Development

Download

development@drupal.org

August 2005

73 participants
637 discussions

[drupal-devel] [bug] Put revisions in their own table
by Cvbge 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/7582 Post a follow up: http://drupal.org/project/comments/add/7582 Project: Drupal Version: cvs Component: node system Category: bug reports Priority: normal Assigned to: killes(a)www.drop.org Reported by: killes(a)www.drop.org Updated by: Cvbge Status: patch (ready to be committed) Attachment: http://drupal.org/files/issues/revisions_41.patch.patch (1.14 KB) Hi, tested the patch a bit: 1. in database.pgsql in the table files a) after removing PRIMARY KEY (fid) you need to remove also "," from the line above b) you CREATE INDEX files_fid_idx ON node(fid); CREATE INDEX files_vid_idx ON node(vid); on wrong tables: not node but files 2. in newly created table node_revisions a) in mysql version there is no DEFAULT, in postgres there is DEFAUL 0, maybe mysql version should be the same as postgres? b) there's no log longtext NOT NULL default '' in pgsql version c) you have translated format int(4) NOT NULL default '0' to format smallint NOT NULL default '0', it also should be integer and not smallint d) you create index: CREATE INDEX node_revisions_title_idx ON node_revisions(title,type);, but there is no type column in the table, and also there is no similar index in mysql version Attached patch should fix the errors Cvbge Previous comments: ------------------------------------------------------------------------ Wed, 05 May 2004 16:25:27 +0000 : killes(a)www.drop.org Currently all node revisions are stored in a serialized field in node.table and retrieved for _each_ page view although they are rarely needed. However, we have agreed that serializing data is bad and that we should try to keep the memory foot print pf Drupal small. Therefore I propose to create a separate revisions table which would be in principle identical to the node table, only that it could have several old copies of the same node. Extra data added by other modules could be added in a serialized field unless we find a better solution. ------------------------------------------------------------------------ Wed, 05 May 2004 17:06:35 +0000 : jhriggs I too think the serialized approach is less than desirable, but here's an alternative. This would likely take some considerable rework in core and contrib, but the following is how we handle similar types of situations in our databases at work. It is more elegant that a separate table, and avoids the (almost exact) duplication of a table. Instead of separate tables, keep all revisions of nodes in the node table as follows: * add field: active (0/1 or Y/N) * add field: revision * every revision of a node is stored in the node table; however, only one revision can be active at any given time * nid can no longer be unique -- primary/unique key becomes (nid, active) * any time a node is loaded, updated (without revision), etc., the active version is used. Thoughts? ------------------------------------------------------------------------ Wed, 05 May 2004 17:57:48 +0000 : killes(a)www.drop.org I am not opposed to your scheme, but I want to stress the following: * Duplicating a table's structure is not bad (IMHO) as long as the content is different. * having two tables will allow us to have a rather small node table. This is (maybe) a performance gain. ------------------------------------------------------------------------ Wed, 05 May 2004 18:37:29 +0000 : jhriggs I don't necessarily think that duplicating a table's structure is _bad_. It just seems to be wasteful and a pain to maintain. (Every change to the node table is made twice...easy to do, but also easy to miss perhaps.) As for performance, as long as nid and the active indicator are indexed, there shouldn't be any performance loss. Also, archiving an old version when making a new revision will be much simpler: just change the active indicator rather than copying an entire node to another table (and ensuring everything gets copied...again a potential maintainance issue). To be honest, I would just like to see the serialized data go away, regarless of what approach is taken. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:49:33 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_07-30-2004.p… (10.47 KB) I'm interested in using Drupal for a large scale wiki-type project. In order to do this, I need revisions to be in their own table. Attached is a patch to do just that. Most of the changes are pretty self explanatory. Spreading out node data across two tables meant that I had to add database functions to do locking/transactions. Without this, race conditions in which the database becomes corrupted are possible. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:54:36 +0000 : Nick Nassar Oh yeah... The patch is a diff against Drupal CVS ------------------------------------------------------------------------ Sat, 31 Jul 2004 00:00:08 +0000 : Anonymous Gerhard speaking. Nick, thanks a lot for your nice patch! It saves me a great deal of labour. I looked through it and immediately liked it. You not only put the old revisions into a new table but also the current one. Do you have an estimate how much more expensive the additional join is? Besides a few minor coding style issues I found a major one: Just a few hours before you uploaded your patch JonBob's node access patch hit core. That means your patch won't apply anymore as all the queries you change have been changed. Can I bug you to update your patch? ------------------------------------------------------------------------ Sat, 31 Jul 2004 01:11:59 +0000 : Anonymous Also I think that your upgrade path loses existing revisions. ------------------------------------------------------------------------ Sat, 31 Jul 2004 02:39:12 +0000 : drumm I think this is the proper way to do things. No columns are duplicated, there is no serialized data, and only the fields that are logically revised are stored. Nothing jumped out at me as a way to have my node modules be able to keep a table of revisions of additional fields. I'm guessing this could be done within the confines of _insert and _update. Assuming the upgrade path works and modules can extend it I give it a +1. ------------------------------------------------------------------------ Sat, 31 Jul 2004 14:40:15 +0000 : Nick Nassar It figures that just as I finish a big patch, another patch comes along and breaks it. Oh well, it should be a pretty easy to fix. I'll work on it. Fixing the upgrade path to keep revisions should be fairly painless. I found another issue that needs to be fixed before this patch gets merged. There format of a node needs to be stored for each revision. Otherwise, for modules that store a format for the nodes, such as page and book, if you write one revision in PHP and the next in HTML, the PHP revision will be displayed as HTML. This is part of a larger issue of how node modules should store revisions of additional fields. I think each module that wants to do this should create another table with (nid, revid) as the primary key. Just as when they want to add fields to a node they create another table with nid as the primary key. As far as performance goes, for sites that make heavy use of revisions, an extra join on primary keys is going to be a lot faster than grabbing all of the revisions from that database everytime. We would need to run benchmarks to determine is the overall difference in speed is for an average site is a gain or a loss. I'm guessing it's very minor either way. ------------------------------------------------------------------------ Mon, 23 Aug 2004 13:55:49 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_08-23-2004.p… (10.92 KB) Here's an updated patch against CVS that puts revisions in their own table, provides an upgrade path, and fixes the format related bugs in the last patch. Hopefully, this can make it into CVS as soon as the freeze is over. ------------------------------------------------------------------------ Mon, 23 Aug 2004 14:10:39 +0000 : moshe weitzman Interesting patch ... drumm's question is still outstanding. how do modules store revisions of their fields? Are they expected to manage this on their own? Thats not how it works today. As an aside, i am seeing profile_ fields in my node.revisions column. One could argue that those need not be saved. They pertain to the node author, not to the node itself. ------------------------------------------------------------------------ Mon, 23 Aug 2004 16:14:39 +0000 : Nick Nassar Having modules be responsible for storing revisions of their own fields is a side-effect of storing revision data in tables. There's really no way around it. However, revisions generally don't make sense for node types that don't have PHP/HTML content, such as polls. I think it's going to be a pretty rare scenario for a new node type to want another field to change per-revision, so it's a pretty good trade-off. Storing fields that shouldn't be part of revisions, such as the profile_ fields, is a side-effect of storing revisions as serialized objects. Applying this patch will free up that wasted space. :) ------------------------------------------------------------------------ Mon, 23 Aug 2004 17:20:57 +0000 : Anonymous There should be a hook that let's the module choose whether it supports history. This way a module author can prevent the user from doing something that may break his module or just cause undefined behavior. If the module doesn't support history then don't let the user/admin choose to add history to nodes of that type. Craig ------------------------------------------------------------------------ Mon, 23 Aug 2004 19:23:29 +0000 : Nick Nassar I agree, there should be an API change to make specifying support for revisions easier. In the interests of keeping patches small and keeping to one change per patch, I think the API change should be a separate issue. A sort of ad-hoc API to decide whether or not a module supports revisions by default already exists. Instead of having a hook, modules set the default value of the "Create new revision" field in the edit form. The admin can change this option in admin/node/configure/defaults. This patch doesn't change that. Revisions are broken for node types that have their own database structure, like polls, even when storing them as serialized objects. This patch doesn't change that, either. ------------------------------------------------------------------------ Tue, 26 Oct 2004 02:35:06 +0000 : moshe weitzman I'm guessing that someone is going to have to demonstrate that this patch performs as well as current drupal before it gets comitted. i think this patch is a few benchmarks from being comitted. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:04:09 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:16 +0000 : Nick Nassar I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:26 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Mon, 15 Nov 2004 05:05:30 +0000 : elias1884 please overthink the revision system default workflow as well. don't look at the revision system as an isolated system but as a part of the whole workflow system! if you combine revisions with the moderation queue the most logic default workflow would be like that: auth user creates node (revision #0) admin approves the node (status = 1, moderation = 0) => node publicly available auth user finds typo and changes node (revision #1, status = 0, moderation = 1) ------- what happens at that point at the moment is, that the node is not accessible anymore at all until the new revision is approved by admin. of course the new revision should not go online until reviewed and approved, this is absolutely correct, but there is no reason to not take the old revision offline, since it was already approved and should therefore be online until the new revision is approved. it is not practical if a node disappears only because the author corrected a typo. ------- admin approves the node (status = 1, moderation = 0) eventhough I first thought a plain boolean active field would not be capable of providing that functionality if finally came to the conclusion, that it can. The only thing to do is to not set that bit, when a new revision is created, but when it is approved (in case moderation is activated under default workflow). Every revision should have its own moderation, status and active field and on approval they are set like this (status=1, moderation=0, active=Y). When you wanna rollback to an old revision, you can chose between all revisions that already have the moderation bit set back to 0 again and the published to 1. There should be an extra permission for rollback! another concern that I have about the default workflow is, that users can't see the content, they have just created, when moderation is enabled. Eventhough, there is a big fat "submission accepted" presented after submissions, unexperienced users tend to question the information those stupid tincans give them, if they can't find their content afterwards. Many users are really lazy bastards and they don't even read the status messages. The best feedback about whether his story was submitted successfully or not of course is, if he can find the story somewhere on the site, maybe with a status message on top of it, mentioning, that the content is currently not publicly available since it has not been approved yet. there should be a my content section under my account, like somebody is trying to do with the workspace module I guess. so my suggestion is to make (status=0, moderation=1) still available for the creator under a my content section somewhere! ------------------------------------------------------------------------ Wed, 24 Nov 2004 04:21:18 +0000 : Nick Nassar I agree. The current workflow for moderation queues and revisions needs to change, but this patch isn't the place for it. The patch is already too big, and it only does the backend stuff. Instead of adding more to this patch and making it take even longer to get into core, would you mind creating a new issue for your UI suggestions, so the those changes can be added as a separate patch? Thanks, Nick ------------------------------------------------------------------------ Sat, 11 Dec 2004 12:26:02 +0000 : Dries This patch is _much_ needed so I'd love to see someone revive it. In order for this patch to be accepted, the following needs to be done: * Update this patch to CVS HEAD. * Rename revid to vid. * Rename node_rev to node_revisions. * Rename node_rev.changed to node_revisions.timestamp. * Rename $rnode to $revision. * Fix the coding style to match Drupal's: proper spacing, single quotes where possible, proper variable names. * Benchmark this patch with a large database with enough revisions. I'd be happy to benchmark this on my local copy of the drupal.org database. * The book.log field should probably move to the node_revisions table. This can be done in a separate patch. * Investigate whether transactions are well-supported. ------------------------------------------------------------------------ Mon, 13 Dec 2004 00:25:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004-r… (11.02 KB) I've worked a bit on the patch (coding style issues as mentioned by Dries). One thing I noticed is that the patch uses REPLACE. IIRC this needs to be chagned to "UPDATE, if fail INSERT" for pgsql compatibility. Nick, are you still interested in working on that patch? I'd like to know how it works on your site and work on getting it into core. ------------------------------------------------------------------------ Mon, 13 Dec 2004 12:33:08 +0000 : Dries Gerhard: your patch does not apply. ------------------------------------------------------------------------ Mon, 13 Dec 2004 17:10:12 +0000 : killes(a)www.drop.org Yes, I know, that was the same version as I mailed to you earlier. ------------------------------------------------------------------------ Mon, 13 Dec 2004 21:02:06 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions.patch (52.96 KB) Ok, upüdated the patch to cvs. ------------------------------------------------------------------------ Tue, 14 Dec 2004 08:58:36 +0000 : Dries Some more comments: * db_begin_transaction() and db_end_transaction() do not belong in database.inc, but in database.mysql.inc and database.pgsql.inc respectively. * The node module calls node_revisionsision_list() which is not defined. (Fxed that on my local copy.) * Do db_begin_transaction() and db_end_transaction() deprecate Jeremy's table locking patch? * The upgrade path assigns the wrong user ID to each revision. * The upgrade path assigns the wrong date to each revision (that or a node's revision page shows the wrong usernames/dates). * The coding style needs a bit of work, but we can worry about that later. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:34:44 +0000 : Nick Nassar If you need any help getting those things fixed, just let me know. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:50:30 +0000 : Nick Nassar How this relates to Jeremy's node locking patch: There was lots of discussion, and node locking was decided against because from an end user point of view you never want a node to be locked. He's now advocating for a much simpler patch that warns users if their changes will overwrite someone elses. That patch still has a race condition, which might be fixed using db_begin_transaction(). http://drupal.org/node/6025 ------------------------------------------------------------------------ Tue, 14 Dec 2004 22:26:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_0.patch (55.96 KB) Here is an updated patch that tries to address Dries concerns. ------------------------------------------------------------------------ Wed, 15 Dec 2004 08:32:50 +0000 : Dries Attachment: http://drupal.org/files/issues/revisions-bug.png (76.06 KB) It didn't fix the aforementioned bugs. See attached screenshot. ------------------------------------------------------------------------ Thu, 06 Jan 2005 20:15:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_1.patch (51.77 KB) Ok, here is a new version. Dries and myself worked hart at it, so please have a look. what is still missing - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. ------------------------------------------------------------------------ Wed, 19 Jan 2005 21:43:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_2.patch (49.49 KB) Here is an updated patch. We discussed to keep the current title in node module and also in the revisiosn table. This is content duplication but will save many joins as many queries only need the title of a node. Discussion is welcome. ------------------------------------------------------------------------ Wed, 19 Jan 2005 23:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_3.patch (29.93 KB) I've implemented the aforementioned solution. This makes the patch much smaller. The patch now also removes taxonomy_node_has_term() which wasn't used anywhere. I'd really apprciate if some people could test drive the patch. It will be another huge improvement for 4.6. ------------------------------------------------------------------------ Thu, 20 Jan 2005 00:05:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_4.patch (30 KB) Another revision. Steven didn't like my literal $node->vid in queries. ------------------------------------------------------------------------ Thu, 20 Jan 2005 01:10:50 +0000 : killes(a)www.drop.org - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. These issues are still open, btw. Especially the first one needs to be tackled. ------------------------------------------------------------------------ Tue, 25 Jan 2005 20:11:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_5.patch (51.13 KB) Here is a patch that has the database tables updated for forum, book, and page module. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:55:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_6.patch (49.18 KB) Yet another update to keep it working with head. The patch now also removes the table definitons for the page table. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:57:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_7.patch (55.69 KB) Sorry, that was the old version, this is the right one. ------------------------------------------------------------------------ Mon, 31 Jan 2005 19:55:03 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_8.patch (55.71 KB) Updated once more. ------------------------------------------------------------------------ Mon, 31 Jan 2005 20:52:08 +0000 : Dries Anyone to help review/test this? ------------------------------------------------------------------------ Mon, 31 Jan 2005 21:22:36 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_9.patch (49.29 KB) Updated again, the update functions occurred twice. Thanks Bart. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:05 +0000 : killes(a)www.drop.org Don't know if the db I am using is corrupted or what. I still do have some didficulties. The latest patch is attached. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_10.patch (49.67 KB) I am probably slowly going mad ... ------------------------------------------------------------------------ Wed, 02 Feb 2005 01:54:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_11.patch (48.95 KB) The update issue still needs investigating. This patch is updated for cvs. ------------------------------------------------------------------------ Wed, 02 Feb 2005 20:20:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_12.patch (49.83 KB) Ok, here is a new version. I've solved my troubles with book.module. There are still some issues with forum module. Possibly due to inconsistent database. ------------------------------------------------------------------------ Wed, 02 Feb 2005 21:31:05 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_13.patch (49.83 KB) Turns out the drupal.org database had indeed some quirks. Please run this query in your oldest db and tell me the result: select nid,type from node where type like '%/%'; If you get a non-zero result we might need to add another security update. The patch could use still more testing, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 01:16:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_14.patch (49 KB) Ok, we are getting somewhere. At a first glance the update is working. There is a problem remaining: the revisions tab will be shown whether the node has revisions or not. Not sure we can/need to fix this. People with a drupal.org account can log in at http://killes.drupaldevs.org/revision/ and poke around. Your permissions will be the same as on drupal.org. Feel free to vreak everything but don't forget to file complaints here. (Note: this is only a pruned version of the drupal.org database with all project nodes and nodes with nids > 7000 dropped). ------------------------------------------------------------------------ Thu, 03 Feb 2005 04:19:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_15.patch (52.39 KB) There was some error in node_save and also the patches to the database.inc files got lost... ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:07:27 +0000 : robertDouglass Submitting book pages doesn't work on your test site. It puts the entire content of the preview inside the body textarea. I wrote a sentence in the body and the log, and pressing preview put several lines of HTML containing both sentences in the body textarea on the preview page, plus the book page wouldn't submit. -R ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:50:59 +0000 : Junyor 0 results here. I started using Drupal with version 4.4, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 23:56:18 +0000 : killes(a)www.drop.org @Junyor: Thanks, that's a good sign. Maybe somebody else has an older db to try. @robertDouglass: The first effect you describe is due to drupaldevs running on PHP 5. I am unsure why the second thing does not work. In node_save() the node object has a nid although there is none in the form. Very strange. I've enabled display of db queries on the testsite. ------------------------------------------------------------------------ Fri, 04 Feb 2005 19:17:55 +0000 : dmjossel No results here on the query: select nid,type from node where type like '%/%'; On a database that was put in place prior to Drupal 4 and is now running on 4.5.2. ------------------------------------------------------------------------ Fri, 04 Feb 2005 20:44:23 +0000 : killes(a)www.drop.org @dmjossel: thanks. @all. The strange problem I reported was apparently php 5 related. After applying Steven's php 5 patch it went away. One error is remaining: If I create a new forum topic it will be shown as part of the book on preview. Hmm, that was due to a db that got corrupted during testing so that is fixed too. Please poke around at the test site and look if you find more errors. ------------------------------------------------------------------------ Sat, 05 Feb 2005 07:16:22 +0000 : Steven By the way, I just remembered that Drupal.org has some blogs lingering on in the database even though blog.module is not enabled. Perhaps this is causing troubles? ------------------------------------------------------------------------ Sat, 05 Feb 2005 11:22:59 +0000 : Anonymous I can't see why it would. Drupal.org will need extra updates for images and project nodes because those have their own tables. GK. ------------------------------------------------------------------------ Sun, 06 Feb 2005 12:49:55 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_16.patch (52.49 KB) Updated to apply to cvs again. ------------------------------------------------------------------------ Tue, 22 Feb 2005 20:15:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_17.patch (49.64 KB) Updated again. All we need is a patch to upload module and an upgrade path for it. ------------------------------------------------------------------------ Fri, 04 Mar 2005 04:22:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_18.patch (52.31 KB) Updated once more. Moved log field from book to node_revisions table as discussed in Antwerp. upload module still missing. We need to decide under which circumstances the log field should be displayed. Should that be added to the workflow? Should it depend on the revisions setting? ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:27:03 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_20.patch (75.52 KB) Ok, here it is: Yet another revision of this grrrrreat patch. Changes from previous versions: - supports versioning for uploaded files. A problem is that if you delete a file, it will be gone for all revisions. - the log field is now in the node_revsions table, but each module has to decide whether to show it or not. I've implemented it for the page and the book type odes. Also, the field can be edited when adding non-book nodes to the book. The log is displayed on the revisions page and if a node is moderated. - the revisions are moved to an old_revisions table to a) get the node table smaller and b) still leave the mavailable for contrib modules that want to retreive old version data. The patch has been applied to killes.drupaldevs.org/revision where it can be tested by anybody (especially people who have "site admin" rights on drupal.org) The database is from drupal.org and you should b able to log in with your pass or simply mail yourself a new one. Gerhard ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:51:56 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_21.patch (59.42 KB) BTW, I marked this a bug because atm the revisions field can grow quite big. Neil has reported problems from some users who were not able to load some nodes due to to many large revisions. Also, som unrelated stuff crept into the patch. New version attached. ------------------------------------------------------------------------ Tue, 08 Mar 2005 05:56:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_22.patch (60.29 KB) Ok, I think I got it. Changes to last version: - uploads are no properly versioned. Missing are still pgsql checks and updates. ------------------------------------------------------------------------ Thu, 10 Mar 2005 16:58:41 +0000 : Anonymous Was able to get http://drupal.org/files/issues/revisions_21.patch to work with drupal-cvs.tar.gz (10 March 2005) by: - includes/database.mysql.inc: Commenting out duplicates for functions function db_begin_transaction and function db_commit_transaction - modules: node.module: Removing "'title' => $node->title," from $node_table_values variable declaration and removing "'title' => "'%s'"," from "$node_table_types" variable declaration. Happy to submit a patch if requested. I'll watch this thread. ------------------------------------------------------------------------ Fri, 11 Mar 2005 23:59:45 +0000 : killes(a)www.drop.org The duplicate function has been removed in rev 22 of this patch. Why do you think the changes in node_save are needed? Titles are saved in both tables for performance reasons. ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:12:21 +0000 : jlerner Hi - I posted comment #62. The changes to node_save appear to be needed because recent patches (both 21 and 22) remove the field 'title' from table 'node'. So without the changes to node_save, node.module is broken and generates errors. Joshua ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:29:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_23.patch (61.17 KB) Thanks, Joshua, for catching this. node:title is there to stay. ------------------------------------------------------------------------ Wed, 13 Apr 2005 16:29:23 +0000 : moshe weitzman since HEAD is open again, perhaps it is a good time to revisit this patch. once this is committed, lets address - http://drupal.org/node/11071 "node_validate does not respect group editing" ------------------------------------------------------------------------ Mon, 18 Apr 2005 15:43:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Mon, 18 Apr 2005 16:16:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24_0.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:19:42 +0000 : Dries I'll commit this patch later this week! If you haven't checked this patch already, I urge you to test/check it out because it will have significant impact on existing code and modules! ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:21:33 +0000 : Dries Also, what do people think about the n.title being duplicated? ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:26:58 +0000 : chx I won't lose any sleep because of duplicated titles... ------------------------------------------------------------------------ Tue, 19 Apr 2005 18:35:58 +0000 : killes(a)www.drop.org Let me explain why I have chosen to duplicate the title (and also the uid): If you look at all the queries in Drupal, you will find that most of them only need the title and th uid of a node. That is, by duplicating it, we save expensive joins on the node_revisions table. Due to this fact, this patch is actually a performance improvement. A note about updating contrib module: Strictly speaking they wouldn't need to be updated. They only need to if their authors decide that their info should be available for revisioning. The upgrade path for forum.module in my update.inc patch (+ the forum patch) should show you what needs to be done. I will write a note for the update page once the patch hits core. ------------------------------------------------------------------------ Sun, 24 Apr 2005 21:21:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_25.patch (60.38 KB) Updated to cvs. Dries: Based on some remarks in #drupal this is the last update I am going to do. Apply it or won't fix it. ------------------------------------------------------------------------ Sat, 30 Apr 2005 03:42:39 +0000 : Jeremy(a)kerneltrap.org Attachment: http://drupal.org/files/issues/revisions_25.patch.patch (528 bytes) That's a big patch. I've only started looking through it. I noticed one little typo, affecting updates. A patch to your last patch is attached. I'm running with the revision patch on my dev server now happily. I like the concept. What happens if you click 'stop' on your browser in the middle of a MySQL "transaction"? I assume that kills the connection to MySQL, and the lock is freed? But this then leaves changes only partially applied? What exactly does locking/unlocking the tables buy us in MySQL? I don't see anywhere that we detect if an apply fails part way through, and thus roll back...? What am I missing? ------------------------------------------------------------------------ Sat, 30 Apr 2005 07:11:28 +0000 : Dries Jeremy: many of us are worried about the performance ramifications this patch introduces. Early experiments showed a small performance improvement (while a performance regression might be expected). More performance reports from large sites like kerneltrap.org will certainly help this patch. Mind to do a quick performance comparision and to report back with some numbers? Thanks. ------------------------------------------------------------------------ Sat, 30 Apr 2005 12:38:02 +0000 : Jeremy(a)kerneltrap.org Dries: I'm not running HEAD on kerneltrap, so this really isn't a possibility. Furthermore, until I understand why we're locking tables, I don't like it. The idea of revisions in their own tables is great. The idea of locking tables to get (without any obvious benefit) there really worries me. ------------------------------------------------------------------------ Sat, 30 Apr 2005 14:16:01 +0000 : killes(a)www.drop.org @Jeremy: Thanks for looking at the patch! Also for catching the typo. :) Did you try to upgrade your database? If yes, how did it go? One of Dries' concerns is the complexity of the upgrade. How many nodes and revisions did the db have? About database locking: This part of the patch was created by Nick and I simply continued to use it. Maybe the code should rather be: if(db_begin_transaction(array('{node}', '{node_revisions}', '{watchdog}', '{sessions}', '{files}'))) { db_query($node_query, $node_table_values); db_query($revisions_query, $revisions_table_values); db_commit_transaction(); ... } The idea is probably to avoid two updates at the same time. I don't think the locking helps if you abort the script at an inconvenient time. Rollbacks aren't implemented in all mysql versions. We could omit the db locking if deemed inappropriate. Maybe Nick can explain his ideas behind this. @Dries: I wonder who the "many of us" are. They certainly haven't spoken to me. Moshe had some reservations about the upgrade path and project module, but the time that project module abused revisions to store issue updates was long ago and his reservations were resolved. Nobody else (besides you of course and now Jeremy) has voiced reservations in a way that was audible to me. If you grep through the patch you will notice that there are only four queries which have a join on the node_revisions table. Two of them are in node_load and in the other cases the join replaced a join on the node table. The two queries in node_load are the only ones that have both a join on the node and the revisions query. Thus, loading of individual nodes might become somewhat slower. All other queries will be faster since the node table is now much smaller. Also, node loading does not have to be slower, it depends on your node table. If you had a lot of revisions and thus a large table, then the new scheme will make your queries actually faster since we do not load the revisions on each and every node load anymore. If you didn't have many revisions your node_load migth be somewhat slower. WRT to the update script Karoly pointed out that we could use multiple insert queries instead one query per revision. This would probably make the update somewhat faster. I am willing to work on this iff you declare that you will commit the patch afterwards. I'd need to know if this will work on pgsql and on all supported mysql versions before I work on it. About locking: Database locking is dog slow, at least on mysql. I was using locks in an earlier version of the upgrade script but had to remove it for (serious!) performance reasons. ------------------------------------------------------------------------ Mon, 09 May 2005 15:07:34 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_26.patch (46.45 KB) Ok, another update, cause I need it myself. I've left out the transaction stuff for now. It is in principle unrelated to this patch and should be discussed elsewhere. This also makes the patch smaller and easier to review (hint, hint). ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:09 +0000 : killes(a)www.drop.org The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_27.patch (39.05 KB) The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 21:23:06 +0000 : Dries Got one error during the upgrade path: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ------------------------------------------------------------------------ Mon, 09 May 2005 21:26:19 +0000 : killes(a)www.drop.org This had happend to me as well, when I tested this patch. The reason is that for some reason the vid is not unique. Most likely there are some entries with vid = 0 in there. Can you check which node types those have? it always was an error in the test database. See: http://drupal.org/node/7582#comment-20678 ------------------------------------------------------------------------ Mon, 09 May 2005 21:27:06 +0000 : Dries Actually, I got 2850 errors during the upgrade. Some of these: sprintf() [function.sprintf]: Too few arguments in drupal-cvs/includes/database.inc on line 154. Some of these: Query was empty query: in drupal-cvs/includes/database.mysql.inc on line 66. And this: Unknown table 'n' in field list query: SELECT n.nid, n.vid FROM node INNER JOIN files f ON n.nid = f.nid in drupal-cvs/includes/database.mysql.inc on line 66. :-) ------------------------------------------------------------------------ Mon, 09 May 2005 21:29:19 +0000 : Dries Or this: user error: Unknown column 'log' in 'field list' query: SELECT parent, weight, log FROM book WHERE nid = 1 in drupal-cvs/includes/database.mysql.inc on line 66 ------------------------------------------------------------------------ Mon, 09 May 2005 21:52:12 +0000 : Dries The time required to generate my main page went from 902 ms (before upgrade) to 2139 ms (after upgrade). The time required to generate a forum listing (?q=forum/x) went from 1872 ms (before upgrade) to 2874 ms (after upgrade). Maybe this is because my database is not consistent as the result of the upgrade errors (yet I don't see any errors on the pages I benchmarked). ------------------------------------------------------------------------ Tue, 10 May 2005 00:24:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_28.patch (53.47 KB) Ok, let me get to this from the bottom to the top: - my test runs indicated a different development wrt timing. If I had gotten your results, I had stopped working on this long ago. So your results are wrong for some reason. - user error: Unknown column 'log' in 'field list' Wasn't my day, the book patch got lost. It is contained now. First -R the old patch, then apply this one. - Unknown table 'n' in field list query: Walkah found this, but I forgot to fix it. Fixed now. - I've no idea where the other queries come from. I am hoping that either your test db is borken or they are follow ups from the other ones. If you let me have your test db, I'll try some debugging. Thanks for wasting your time, too. ------------------------------------------------------------------------ Tue, 10 May 2005 05:07:31 +0000 : Dries I double-checked and the numbers don't seem to lie. I'll test some more after work on another machine to make sure it is not platform-specific. ------------------------------------------------------------------------ Wed, 11 May 2005 03:32:47 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_29.patch (54.83 KB) Ok, here I am again. What I did: 1) Ask Dries to let me have drupal.org database 2) get 400MB of SQL inserts... 3) take 23 minutes to import said data 4) Remove all image and project nodes (don't want to install their modules), 11765 nodes left 5) back up data 6) take tests on non-cached /node page (as anonymous user). ab results: -c 1 -n 25: Requests per second: 1.29 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 663 775 179.7 689 1264 7) Do the same for the tracker page: Requests per second: 0.83 [#/sec] (mean) Total: 1182 1199 7.4 1199 1217 8) Apply my patch (rev. 28). 9) run db update and hold breath 10) update times out... 11) play back backup from 5) 12) wait 13) getting annoyed and removing cache, watchdog, and accesslog before playing in backup. 14) wait again. Understand why Dries doesn't try this patch often. Maybe a smaller DB would do for testing? 15) wait more. get really annoyed. 16) Set time limit to 18000 in update.php 17) try again 18) fails again before the second update is completed. 19) curse. 20) delete search stuff from db. Ooops, sooo much smaller... 21) import again, below 2 minutes... 22) rewrite to use extended insert. Found a bug. 23) still does not complete. Mysql logging to the rescue! 24) tid = 0? Not good. 25) Well, the update works fine till node 10834. 5595 nodes done, 6136 to go. 26) Writing shell based update script. Discovery: 24MB aren't enough. Hopefully 64 are. Nope. extended inserts for revisions are apparently not the brightest idea: Huge memory consumption. Hmm, no, all updates got through. Selecting the revisions to put them into old_revisions table screwed it. Learned about CREATE TABLE old_revisions SELECT syntax. Yay! finally. 24 MB are just not enough the update.php script seems to still break. 27) Benchmarks! /node Requests per second: 1.54 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 632 649 40.5 636 791 /tracker Requests per second: 0.86 [#/sec] (mean) Total: 1119 1165 65.8 1160 1461 Ok: So we get an improvement for many node_loads, but none for simple selects from node. More tests can be done. 28) roll new patch Ain't Drupal fun? ------------------------------------------------------------------------ Wed, 18 May 2005 13:38:05 +0000 : Dries I did another round of tests on _another_ machine and it is not looking good: Before upgrade After upgrade ?q= (main page) 218 ms/request 340 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request ?q=about (book page) 375 ms/request 5400 ms/request The upgrade process itself gave me a number of 'query was empty' and 'sprintf(): too few arguments' reports. Everything seems to work fine though. Looking at the ?q=about page, I see that the following query is executed twice _and_ that each time, it take more than 2 seconds to complete: SELECT n.nid, n.title, b.parent, b.weight FROM node n INNER JOIN book b ON n.vid = b.vid WHERE n.status = 1 AND n.moderate = 0 ORDER BY b.weight, n.title; --8 SHOW INDEX FROM book; +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | book | 1 | book_parent | 1 | parent | A | 92 | NULL | NULL | | BTREE | | | book | 1 | nid | 1 | nid | A | 369 | NULL | NULL | | BTREE | | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ 2 rows in set (0.00 sec) The book module does not appear to have a primary key? Sounds like a bad idea so I added one: mysql> ALTER TABLE book ADD PRIMARY KEY nid (nid); Query OK, 369 rows affected (0.02 sec) Records: 369 Duplicates: 0 Warnings: 0 Next, I wanted to make the vid column a unique key in all node tables: mysql> ALTER TABLE node ADD UNIQUE vid (vid); Query OK, 20392 rows affected (0.81 sec) Records: 20392 Duplicates: 0 Warnings: 0 mysql> ALTER TABLE book ADD UNIQUE vid (vid); ERROR 1062: Duplicate entry '0' for key 2 mysql> ALTER TABLE forum ADD UNIQUE vid (vid); Query OK, 10806 rows affected (0.10 sec) Records: 10806 Duplicates: 0 Warnings: 0 As you can see, it fails for the book table which makes me believe there is some inconsistent data ... I set out to fix that: mysql> SELECT nid, COUNT(nid) AS vids FROM book GROUP BY vid HAVING vids > 1; +-----+------+ | nid | vids | +-----+------+ | 871 | 2 | +-----+------+ 1 row in set (0.00 sec) mysql> SELECT title FROM node WHERE nid = 871; Empty set (0.00 sec) mysql> DELETE FROM book WHERE nid = 871; Query OK, 1 row affected (0.00 sec) mysql> ALTER TABLE book ADD UNIQUE vid (vid); Query OK, 368 rows affected (0.01 sec) Records: 368 Duplicates: 0 Warnings: 0 Looks like everything is well now. Ran some new benchmarks: Before upgrade After upgrade With indices ?q= (main page) 218 ms/request 340 ms/request 336 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request 1531 ms/request ?q=about (book page) 375 ms/request 5400 ms/request 475 ms/request Unfortunately, we're still slower than the original code. ------------------------------------------------------------------------ Wed, 18 May 2005 21:53:31 +0000 : killes(a)www.drop.org Dries, thanks for testing it again. I do think the broken queries you observer have something to do with the bad performance after the update. Please log the queries any I will have a look at them. I've never seen any such queries. My update script also tries to create the appropriate indices, but it will of course fail if the database contains cruft. the indices for the forum are probably missing, too. I am still convinced that the patch is core worthy. ------------------------------------------------------------------------ Thu, 19 May 2005 04:36:09 +0000 : Dries It wouldn't hurt if more people would benchmark this patch. The patch's current performance worries me. Did you check your watchdog messages after upgrading the drupal.org database? Depending on your settings, errors might only be shown in the watchdog. I'll look into the remaining glitches as time permits. Thanks for your persistence in keeping this patch up-to-date. :) ------------------------------------------------------------------------ Thu, 19 May 2005 11:59:22 +0000 : killes(a)www.drop.org Dries: Can you please let me have your updated database? I want to have a look at it and try my own benchmarks with it. And yes, if I did learn something on this project is how to be persistant. ;) ------------------------------------------------------------------------ Fri, 24 Jun 2005 16:25:34 +0000 : killes(a)www.drop.org Here is an idea that occurred to me: The problem with the upgrade process is that keeping the existing revisions requires a lot of work to do. This generates a huge amount of sql queries for a large database and also requires a huge amount of memory. My suggestions is to let update.php only handle the basic upgrade, ie without old revisions. An additional module could be created that would implement a cron based approach to upgrading old revisions one node at a time. it could expose a hook to let contrib modules do their own upgrades. Dries, what do you think? (I am writing "Dries" because he seems to be the only one who is interested in getting this into core...) ------------------------------------------------------------------------ Fri, 24 Jun 2005 22:25:11 +0000 : Junyor Killes: I'm also interested in seeing this hit core. What about adding something to legacy.module to do it? ------------------------------------------------------------------------ Sun, 26 Jun 2005 21:14:54 +0000 : chx This is a sensible approach. Maybe this is the _only_ sensible approach. I have a little problem though: while the conversion is running somehow both revision handlers should be available. ------------------------------------------------------------------------ Sun, 26 Jun 2005 22:16:21 +0000 : killes(a)www.drop.org hehe, one only has to whine and bug enough and one gets some feedback. ;) @junyor: legacy.module would be a good place. my current idea is to auto-enable it in update.php and then disable it again in legacy_cron after all nodes are updated.. ;) @chx: When somebody wants to look at revisions of a node that node could be auto-updated. The only problem are contrib modules: they's need to have some hook in order to update their own data. When somebody looks at the revisions of a node than cannot be updated because the contrib module in question has no such hook, we can optionally let the user discard old revisions I guess. Dries, what do you think? ------------------------------------------------------------------------ Mon, 25 Jul 2005 16:48:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_30.patch (53.13 KB) Sooooo; I've updated this patch once again. Dries didn't like my idea of legacy updates so we will have an option to discard old revisions in case the update should prove difficult. Always keep a backup. ------------------------------------------------------------------------ Mon, 25 Jul 2005 17:31:12 +0000 : Bèr Kessels can we please accept and commit this patch? We can iron out any issues later. This patch is just far too big and complex to be perfect-at-once. ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:22:50 +0000 : Dries @Ber: no, we can't commit this patch blindly. Data loss is a much bigger problem than a syntax error or other code glitch. We can break Drupal but we can't break their data. I've spent quite a bit of time testing the previous version of this patch and noticed significant performance degradation. Tell me, Ber, why can't you test this patch first? ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:40:37 +0000 : Bèr Kessels I was not referring to not testing it. If it is just the upgrade path that proves to be cumbersome, then why can we not fix it afterwards, I.e. when everyone has had a good chance to look at it. Testing such a huge patch requires a lot of work. something no-one just does in his spare hours. We had discussions before to deal in a slightly different way with big changes; to commit them quicker and to leave the ironing out of any left overs for the community. I hooked into that discussion here. For two reasons. One is that Gerhard has spent numerous hours on maintaining this patch. The other one is that the community can be of help with ironing out issues in such a large change, much better than that Gerhard can do on his own. And yes, dataloss is very bad, but no-one should loose data, when he/she followed the instructions (backing up)... ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:00:38 +0000 : Dries Ber: applying this patch takes 15 seconds. Whether I apply this patch for you, or whether you apply it yourself, it will hardly reduce the 'testing cost'. The problem is that data loss can be subtle; it might go unnoticed for a couple days. Make no mistake, I'd like to see this patch committed ASAP, but it warrants some testing. Let's test/benchmark it and commit it. ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:55:03 +0000 : drumm The upgrade script already takes a long time to execute and does not provide feedback to the user about how it is doing. I plan on making the upgrade script able to spread the updates across multiple page views and give user feedback showing that progress is being made. This will hopefully make the speed of this update a moot point. I am working on this for my dayjob, CivicSpace, so it should get done "real soon now", but it should be expected to take awhile (smallish number of weeks). Please make comprimises if necessary. ------------------------------------------------------------------------ Wed, 27 Jul 2005 17:38:57 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_31.patch (48.78 KB) @drumm: the timeout isn't usually the problem. Memory consumption is. I'll look into doing the updates in chunks of 1000 nodes or so. Clousseau found a bug in the patch, updated. ------------------------------------------------------------------------ Thu, 28 Jul 2005 18:28:08 +0000 : moshe weitzman Even if the upgrade issues are resolved, we have to figure out why this patch is slowing down drupal (according to benchmarks posted here). ------------------------------------------------------------------------ Fri, 29 Jul 2005 07:51:45 +0000 : Dries We can't proceed without some additional benchmarking. I already benchmarked it on two machines (see comments above), and I'll benchmark it again after August 10. The performance impact of this patch worries me so if two other people could test and benchmark this patch extensively, that would come a long way. August 10 is close to the code freeze so some help is necessary. If Gerhard/killes reviewed/tested one of your patches, it is time to return the favor ... ------------------------------------------------------------------------ Fri, 29 Jul 2005 15:40:31 +0000 : Jose A Reyero After looking at the patch I'm not really surprised this slows down everything. I thought that the reason why we wanted revisions in their own table was in the first place to have a simpler -and faster- node table. But this patch: - Adds fields and complexity not only to the node table but also to all the main tables for data - Needs additional joins just to retrieve single nodes (which is done many times for a typical main page) - Does away with encapsulation of the revision functionality requiring other modules to handle revissions related data. So please, please, please: - Make revisions table only needed when actually using revisions - Keep that old nice thing of hiding revisions system from other modules - And what's so bad with serializing data anyway? I agree that usually there are better ways to store data. But this is one of the cases where serialization does make sense. Not the current huge serialized field, but only a new table with one serialized node per row. - In general, do it much simpler. This is way too complex and it removes more functionality than it adds. In just four words: keep it simple, please. ------------------------------------------------------------------------ Fri, 29 Jul 2005 21:25:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_32.patch (51.27 KB) Patch updated (v31 was buggy, and there was an update to updates.inc). I've had a discussion with Jose and he will maybe do some testing over the weekend. John is also doing some testing as we speak. ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:06:59 +0000 : Nick Nassar It has been exactly one year since I first submitted this patch. You know what that means: *old patch party in IRC!* Virtual beers on me! w00t! ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:08:33 +0000 : Jose A Reyero Hi again. I had a talk with killes -I have to say he enlightned me about this- and also did some testing. Sorry but I dont have a good set up to provide benchmarks. Here are my conclusions -and also some explanation for the varied results abt performance: - This is faster for simple node listings - This is slower for full node/load ....thus the performance increase/decrease for a given page should depend on the relation between the number of simple node listings (usually blocks) and the number of nodes displayed in the main page. I've also experienced some trouble with attachments -not re-created when creating a new revision- and book relations, same. I'm not sure I like this new "feature" of creating a new revision every time you rollback an old one -in Drupal 4.6 this seems not to happen. Besides these details, I think this patch is good enough -and quite a powerful approach actually- though I still feel it's bit too complex when simply dealing with nodes -without revisions-, which is what we do most of the time, and complexity will be even higher with contrib modules using still more tables, not to mention flexinode -versioning info will have to be scattered everywhere in the db. So this is not a +1 nor -1. On one side I think this is powerful. On the other side I feel this is too complex and really, it's more than I need. I would be happy with a 'revision' table having: nid, vid, data(serialized), that you can forget about when not using revisions. ------------------------------------------------------------------------ Sat, 30 Jul 2005 19:16:41 +0000 : Nick Nassar On a much more serious note that my last comment, unless I'm missing something, you've introduced a subtle bug into this patch since I left it. A and B are both updating revision 5 of a node. A calls db_next_id() and gets a revision id of 6 B calls db_next_id() and gets a revision id of 7 B updates the node's revision to 7 B adds revision 7 to the revision table A updates the node's revision to 6 A adds revision 6 to the revision table Now, you have a situation where the current revision (6) isn't actually the last revision(7). For some uses, that's not a big deal, but in the case of wikis, which is what I originally hoped to use this patch for, it creates problems. ------------------------------------------------------------------------ Mon, 01 Aug 2005 00:04:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_33.patch (51.27 KB) Updated the patch. updates.inc was changed and chx found a typo. @Nick: I don't think the issue is due to this patch. Drupal has currently no mechanism to not let people commit a new version of the same node, whether with or without revisions. drupal_next_id ensures that the revision IDs are unique and the later revisions will be the current one. Which revision is the current one does not depend on order of the version ID, btw. @Jose: what exactly were your problems with uploads? what do I need to do to reproduce? ------------------------------------------------------------------------ Mon, 01 Aug 2005 11:33:56 +0000 : Jose A Reyero Sorry, I dont know what I did the other day but I cannot reproduce it either :-) I applied this new patch. Problems I'm having now (Testing with "stories") - Book outline is gone after creating a new revision - I cannot delete attachments (not sure it is related to your patch or to something else) - Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I? ------------------------------------------------------------------------ Sun, 14 Aug 2005 16:03:41 +0000 : killes(a)www.drop.org Jose, thanks a lot for testing! "- Book outline is gone after creating a new revision" Sounds like a bug, it should not. "- I cannot delete attachments (not sure it is related to your patch or to something else)" Another bug "- Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab" Log messages are possible for nodes of type book and page as well as all nodes that are put into the book. Log messages can be added for any node type, the module author should implement those. "Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I?" No, the db update should be the same. I will try to fix the bugs you found. WRT the db update I've got an announcement to make: I will not migrate old revisions. Why? 1) it is a rather complicated process that requires a lot of memory and is likely to give people on cheap hosting trouble. 2) testing it takes a lot of time. 3) I don't personally need it I will still move old revisions to an old_revisions table from where they can be converted by a later update or a custom script. ------------------------------------------------------------------------ Tue, 16 Aug 2005 21:34:43 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_34.patch (52.92 KB) Ok, here is incarnation No. 34 of this patch. Dries agrees with me that we can put the revision rescue operation into a later update, possible using Drumm's installer. The patch fixes the issue with the file uploads not being deletable, I could not reproduce the problem with the outlines. ------------------------------------------------------------------------ Sat, 20 Aug 2005 13:33:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/scripts.tar.gz (4.39 KB) I've started to do some performance testing. I created a script that creates nodes that have revisions according to the new system. A node has a 50% chance to have a revision and then an equal chance fro between 1 and 25 revisions. I've created a database with 50 nodes and 500 comments, 7 vocabs and 100 total taxo terms. I've also created a script to convert these revisions to the old format (much easier than the other way around). Initial tests didn't show a significant change for /node. I'll try with more nodes and /tracker. Attached are the two scripts in a tarball. They need to be run against the patched install. Also included are the two preliminary results. ------------------------------------------------------------------------ Sat, 20 Aug 2005 15:16:04 +0000 : killes(a)www.drop.org I've now created a databse which has 1000 nodes and each node has 50 or 51 revisions. I have no taxo terms assigned and no comments (those parts didn't change). I am only using stories and pages. I am now fairly confident that this patch is a performance improvement for this kind of setup. for the new patch on the /node page I get Requests per second: 0.27 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 3592 3648 56.9 3640 3985 Waiting: 3573 3626 55.1 3620 3949 Total: 3592 3648 56.9 3640 3985 for the old one I get: Requests per second: 0.23 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 4247 4297 64.0 4286 4657 Waiting: 4156 4263 65.3 4253 4624 Total: 4247 4297 64.0 4286 4657 For the tracker the results are even better: old: Requests per second: 2.00 [#/sec] (mean) new: Requests per second: 3.64 [#/sec] (mean) Anybody interested in obtaining this database can mail me. It is 60MB (gzipped). ------------------------------------------------------------------------ Sat, 20 Aug 2005 16:33:26 +0000 : killes(a)www.drop.org Ok, I've now deleted the revisions and only kept the current one. The initial tests showed that the performance of the patch was miserable. then I remembered that I had shreddered the revisions table and did a mysql optimization. Now the results are as follows: /node: Requests per second: 3.19 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 275 312 44.3 290 499 Waiting: 255 291 44.1 268 479 Total: 275 312 44.3 290 499 Requests per second: 3.18 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 274 314 42.4 292 407 Waiting: 255 294 42.0 273 388 Total: 274 314 42.4 292 407 The lower one is the new system, there is no difference considering statistics. /tracker old: Requests per second: 3.72 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 238 268 33.8 249 332 Waiting: 219 247 33.3 230 313 Total: 238 268 33.8 249 332 new: Requests per second: 3.77 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 240 264 29.8 250 324 Waiting: 221 244 29.0 229 305 Total: 240 264 29.8 250 324 Same applies, no real difference. So, as soon as somebody tests the upgrade path the patch should be committed. ------------------------------------------------------------------------ Sat, 20 Aug 2005 17:18:39 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_35.patch (55.51 KB) Ok, I coudl hear them in advance "1000 nodes without revisions are not sufficient". Well, here are results for 30000 nodes without revisions. /node new: Requests per second: 0.45 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2162 2202 58.9 2188 2485 Waiting: 2143 2182 54.1 2169 2413 Total: 2162 2202 58.9 2188 2485 old: Requests per second: 0.43 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2319 2346 77.6 2332 2761 Waiting: 2223 2324 79.0 2311 2742 Total: 2319 2346 77.6 2332 2761 One could argue that the new approach s slightly faster. I'd like to see mroe results to believe it, but it is pretty clear it isn't slower. /tracker new: Requests per second: 0.41 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2330 2418 116.9 2403 3194 Waiting: 2302 2397 117.7 2382 3175 Total: 2330 2418 116.9 2403 3194 old: Requests per second: 0.39 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2496 2548 50.1 2536 2841 Waiting: 2477 2528 50.0 2518 2822 Total: 2496 2548 50.1 2536 2841 Statistically insignificant differences. All those tests where done with ab and as user No 1. model name : Intel(R) Celeron(R) CPU 2.40GHz 512MB RAM Besides the differences for revisions the change in served pages with different configurations is interesting. A new version is attached. node_save now returns the complete node (not only the nid). ------------------------------------------------------------------------ Mon, 22 Aug 2005 20:13:52 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_36.patch (55.66 KB) Ok, here is version 36 which incorporates today's changes in cvs. It also changes the upgrade path. I was trying my old solution and had to up my php mem limit to 80 MB in order to get it to run through. I hope this one is better but it needs testing. ------------------------------------------------------------------------ Mon, 22 Aug 2005 22:46:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_37.patch (52.29 KB) Ok, I've now moved _all_ of the heavy lifting to mysql. pgsql need to catch up. No problems with either time out or memory limit anymore. Since pgsql plays catch-up all the time, I set it to "ready to be committed". ------------------------------------------------------------------------ Tue, 23 Aug 2005 11:45:59 +0000 : stefan nagtegaal I could apply the patch on my localhost, and did not seen any errors nor disfunctioning drupal features while creating, revising, editing or deleting nodes.. So, +1 for this patch qua implementation it makes sense. Unfortunatly I chaven't got a clue howto benchmark, so that's for another person. @Gerhard: I really think you are a great person by rewriting this patch over and over again, untill Dries decides it is good enough to get in to the trunk. I doubt about the fact if there are more persons in our community who are like you.. People who would revise their patch 37 times, and even after that still open for discussion.. Gerhard, I think you are a great example of how a real coder should do it.. I think we can all learn from you.. ------------------------------------------------------------------------ Tue, 23 Aug 2005 17:15:29 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_38.patch (52.26 KB) Karoly pointed out that node_save(&$node) is nicer. @Steef: nice that you write that, but a _really_ good coder would only have needed one revision. ;p ------------------------------------------------------------------------ Wed, 24 Aug 2005 15:36:03 +0000 : Dries function book_update($node) { - db_query("UPDATE {book} SET parent = %d, weight = %d, log = '%s' WHERE nid = %d", $node->parent, $node->weight, $node->log, $node->nid); + if ($node->is_new || $node->revision) { + db_query("INSERT INTO {book} (nid, vid, parent, weight) VALUES (%d, %d, %d, %d, '%s')", $node->nid, $node->vid, $node->parent, $node->weight); + } + else { + db_query("UPDATE {book} SET parent = %d, weight = %d WHERE vid = %d", $node->parent, $node->weight, $node->vid); + } } Why 'is_new' and 'revision'? Isn't 'is_new' redundant? I loaded a fresh copy of the drupal.org database on my test system. How do I test this? Where is the upgrade script/path? ------------------------------------------------------------------------ Wed, 24 Aug 2005 16:31:09 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_39.patch (52.25 KB) is_new is indeed redundant since it is handeled in _update. New patch attached. The upgrate is as usual in updates.inc and accessed through update.php. ------------------------------------------------------------------------ Thu, 25 Aug 2005 10:35:24 +0000 : jakeg This is probably a little off topic but I can't find info elsewhere and this may affect revisions changes. It appears as though all revisions of a node are accessible to all users (inc. anonymous users) using ?revision=x. Surely only the 'current' revision should be accessible by default, and there should be a permission that can be assigned to roles to view non-current revisions? I don't want regular users looking at non-current revisions! A revision may have been made for various private reasons. Only admin should be able to do that. ------------------------------------------------------------------------ Thu, 25 Aug 2005 13:26:12 +0000 : Dries On the latest snapshopt of the drupal.org database I get this: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {forum} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {node_revisions} ADD KEY nid (nid) FAILED ALTER TABLE {node_revisions} ADD KEY uid (uid) FAILED ALTER TABLE {node_revisions} ADD PRIMARY KEY vid (vid) FAILED CREATE TABLE {old_revisions} SELECT nid, revisions FROM {node} WHERE revisions != '' OK ALTER TABLE {book} DROP log OK ALTER TABLE {node} DROP teaser OK ALTER TABLE {node} DROP body OK ALTER TABLE {node} DROP format OK ALTER TABLE {node} DROP revisions OK ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:30:11 +0000 : killes(a)www.drop.org That's strange, it worked for me with an older version of the same db. Anyway, I've re-rolled the patch and rearranged the update queries a bit. Also everything is now in one update, thanks to Neil for pointing that out. ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:32:17 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_40.patch (52.22 KB) and the patch. ------------------------------------------------------------------------ Sun, 28 Aug 2005 02:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_41.patch (52.4 KB) I've tested the upgrade path on an older copy of the drupal.org db and found and squished two bugs. The test cycle can be shortened a bit by deleting non-essential db tables before importing (cvs, history, project_*, ...)

2 1

[drupal-devel] [bug] Put revisions in their own table
by Cvbge 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/7582 Post a follow up: http://drupal.org/project/comments/add/7582 Project: Drupal Version: cvs Component: node system Category: bug reports Priority: normal Assigned to: killes(a)www.drop.org Reported by: killes(a)www.drop.org Updated by: Cvbge Status: patch (ready to be committed) Attachment: http://drupal.org/files/issues/revisions_41.patch_0.patch (1.24 KB) Forgot to attach the patch for sequence Cvbge Previous comments: ------------------------------------------------------------------------ Wed, 05 May 2004 16:25:27 +0000 : killes(a)www.drop.org Currently all node revisions are stored in a serialized field in node.table and retrieved for _each_ page view although they are rarely needed. However, we have agreed that serializing data is bad and that we should try to keep the memory foot print pf Drupal small. Therefore I propose to create a separate revisions table which would be in principle identical to the node table, only that it could have several old copies of the same node. Extra data added by other modules could be added in a serialized field unless we find a better solution. ------------------------------------------------------------------------ Wed, 05 May 2004 17:06:35 +0000 : jhriggs I too think the serialized approach is less than desirable, but here's an alternative. This would likely take some considerable rework in core and contrib, but the following is how we handle similar types of situations in our databases at work. It is more elegant that a separate table, and avoids the (almost exact) duplication of a table. Instead of separate tables, keep all revisions of nodes in the node table as follows: * add field: active (0/1 or Y/N) * add field: revision * every revision of a node is stored in the node table; however, only one revision can be active at any given time * nid can no longer be unique -- primary/unique key becomes (nid, active) * any time a node is loaded, updated (without revision), etc., the active version is used. Thoughts? ------------------------------------------------------------------------ Wed, 05 May 2004 17:57:48 +0000 : killes(a)www.drop.org I am not opposed to your scheme, but I want to stress the following: * Duplicating a table's structure is not bad (IMHO) as long as the content is different. * having two tables will allow us to have a rather small node table. This is (maybe) a performance gain. ------------------------------------------------------------------------ Wed, 05 May 2004 18:37:29 +0000 : jhriggs I don't necessarily think that duplicating a table's structure is _bad_. It just seems to be wasteful and a pain to maintain. (Every change to the node table is made twice...easy to do, but also easy to miss perhaps.) As for performance, as long as nid and the active indicator are indexed, there shouldn't be any performance loss. Also, archiving an old version when making a new revision will be much simpler: just change the active indicator rather than copying an entire node to another table (and ensuring everything gets copied...again a potential maintainance issue). To be honest, I would just like to see the serialized data go away, regarless of what approach is taken. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:49:33 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_07-30-2004.p… (10.47 KB) I'm interested in using Drupal for a large scale wiki-type project. In order to do this, I need revisions to be in their own table. Attached is a patch to do just that. Most of the changes are pretty self explanatory. Spreading out node data across two tables meant that I had to add database functions to do locking/transactions. Without this, race conditions in which the database becomes corrupted are possible. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:54:36 +0000 : Nick Nassar Oh yeah... The patch is a diff against Drupal CVS ------------------------------------------------------------------------ Sat, 31 Jul 2004 00:00:08 +0000 : Anonymous Gerhard speaking. Nick, thanks a lot for your nice patch! It saves me a great deal of labour. I looked through it and immediately liked it. You not only put the old revisions into a new table but also the current one. Do you have an estimate how much more expensive the additional join is? Besides a few minor coding style issues I found a major one: Just a few hours before you uploaded your patch JonBob's node access patch hit core. That means your patch won't apply anymore as all the queries you change have been changed. Can I bug you to update your patch? ------------------------------------------------------------------------ Sat, 31 Jul 2004 01:11:59 +0000 : Anonymous Also I think that your upgrade path loses existing revisions. ------------------------------------------------------------------------ Sat, 31 Jul 2004 02:39:12 +0000 : drumm I think this is the proper way to do things. No columns are duplicated, there is no serialized data, and only the fields that are logically revised are stored. Nothing jumped out at me as a way to have my node modules be able to keep a table of revisions of additional fields. I'm guessing this could be done within the confines of _insert and _update. Assuming the upgrade path works and modules can extend it I give it a +1. ------------------------------------------------------------------------ Sat, 31 Jul 2004 14:40:15 +0000 : Nick Nassar It figures that just as I finish a big patch, another patch comes along and breaks it. Oh well, it should be a pretty easy to fix. I'll work on it. Fixing the upgrade path to keep revisions should be fairly painless. I found another issue that needs to be fixed before this patch gets merged. There format of a node needs to be stored for each revision. Otherwise, for modules that store a format for the nodes, such as page and book, if you write one revision in PHP and the next in HTML, the PHP revision will be displayed as HTML. This is part of a larger issue of how node modules should store revisions of additional fields. I think each module that wants to do this should create another table with (nid, revid) as the primary key. Just as when they want to add fields to a node they create another table with nid as the primary key. As far as performance goes, for sites that make heavy use of revisions, an extra join on primary keys is going to be a lot faster than grabbing all of the revisions from that database everytime. We would need to run benchmarks to determine is the overall difference in speed is for an average site is a gain or a loss. I'm guessing it's very minor either way. ------------------------------------------------------------------------ Mon, 23 Aug 2004 13:55:49 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_08-23-2004.p… (10.92 KB) Here's an updated patch against CVS that puts revisions in their own table, provides an upgrade path, and fixes the format related bugs in the last patch. Hopefully, this can make it into CVS as soon as the freeze is over. ------------------------------------------------------------------------ Mon, 23 Aug 2004 14:10:39 +0000 : moshe weitzman Interesting patch ... drumm's question is still outstanding. how do modules store revisions of their fields? Are they expected to manage this on their own? Thats not how it works today. As an aside, i am seeing profile_ fields in my node.revisions column. One could argue that those need not be saved. They pertain to the node author, not to the node itself. ------------------------------------------------------------------------ Mon, 23 Aug 2004 16:14:39 +0000 : Nick Nassar Having modules be responsible for storing revisions of their own fields is a side-effect of storing revision data in tables. There's really no way around it. However, revisions generally don't make sense for node types that don't have PHP/HTML content, such as polls. I think it's going to be a pretty rare scenario for a new node type to want another field to change per-revision, so it's a pretty good trade-off. Storing fields that shouldn't be part of revisions, such as the profile_ fields, is a side-effect of storing revisions as serialized objects. Applying this patch will free up that wasted space. :) ------------------------------------------------------------------------ Mon, 23 Aug 2004 17:20:57 +0000 : Anonymous There should be a hook that let's the module choose whether it supports history. This way a module author can prevent the user from doing something that may break his module or just cause undefined behavior. If the module doesn't support history then don't let the user/admin choose to add history to nodes of that type. Craig ------------------------------------------------------------------------ Mon, 23 Aug 2004 19:23:29 +0000 : Nick Nassar I agree, there should be an API change to make specifying support for revisions easier. In the interests of keeping patches small and keeping to one change per patch, I think the API change should be a separate issue. A sort of ad-hoc API to decide whether or not a module supports revisions by default already exists. Instead of having a hook, modules set the default value of the "Create new revision" field in the edit form. The admin can change this option in admin/node/configure/defaults. This patch doesn't change that. Revisions are broken for node types that have their own database structure, like polls, even when storing them as serialized objects. This patch doesn't change that, either. ------------------------------------------------------------------------ Tue, 26 Oct 2004 02:35:06 +0000 : moshe weitzman I'm guessing that someone is going to have to demonstrate that this patch performs as well as current drupal before it gets comitted. i think this patch is a few benchmarks from being comitted. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:04:09 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:16 +0000 : Nick Nassar I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:26 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Mon, 15 Nov 2004 05:05:30 +0000 : elias1884 please overthink the revision system default workflow as well. don't look at the revision system as an isolated system but as a part of the whole workflow system! if you combine revisions with the moderation queue the most logic default workflow would be like that: auth user creates node (revision #0) admin approves the node (status = 1, moderation = 0) => node publicly available auth user finds typo and changes node (revision #1, status = 0, moderation = 1) ------- what happens at that point at the moment is, that the node is not accessible anymore at all until the new revision is approved by admin. of course the new revision should not go online until reviewed and approved, this is absolutely correct, but there is no reason to not take the old revision offline, since it was already approved and should therefore be online until the new revision is approved. it is not practical if a node disappears only because the author corrected a typo. ------- admin approves the node (status = 1, moderation = 0) eventhough I first thought a plain boolean active field would not be capable of providing that functionality if finally came to the conclusion, that it can. The only thing to do is to not set that bit, when a new revision is created, but when it is approved (in case moderation is activated under default workflow). Every revision should have its own moderation, status and active field and on approval they are set like this (status=1, moderation=0, active=Y). When you wanna rollback to an old revision, you can chose between all revisions that already have the moderation bit set back to 0 again and the published to 1. There should be an extra permission for rollback! another concern that I have about the default workflow is, that users can't see the content, they have just created, when moderation is enabled. Eventhough, there is a big fat "submission accepted" presented after submissions, unexperienced users tend to question the information those stupid tincans give them, if they can't find their content afterwards. Many users are really lazy bastards and they don't even read the status messages. The best feedback about whether his story was submitted successfully or not of course is, if he can find the story somewhere on the site, maybe with a status message on top of it, mentioning, that the content is currently not publicly available since it has not been approved yet. there should be a my content section under my account, like somebody is trying to do with the workspace module I guess. so my suggestion is to make (status=0, moderation=1) still available for the creator under a my content section somewhere! ------------------------------------------------------------------------ Wed, 24 Nov 2004 04:21:18 +0000 : Nick Nassar I agree. The current workflow for moderation queues and revisions needs to change, but this patch isn't the place for it. The patch is already too big, and it only does the backend stuff. Instead of adding more to this patch and making it take even longer to get into core, would you mind creating a new issue for your UI suggestions, so the those changes can be added as a separate patch? Thanks, Nick ------------------------------------------------------------------------ Sat, 11 Dec 2004 12:26:02 +0000 : Dries This patch is _much_ needed so I'd love to see someone revive it. In order for this patch to be accepted, the following needs to be done: * Update this patch to CVS HEAD. * Rename revid to vid. * Rename node_rev to node_revisions. * Rename node_rev.changed to node_revisions.timestamp. * Rename $rnode to $revision. * Fix the coding style to match Drupal's: proper spacing, single quotes where possible, proper variable names. * Benchmark this patch with a large database with enough revisions. I'd be happy to benchmark this on my local copy of the drupal.org database. * The book.log field should probably move to the node_revisions table. This can be done in a separate patch. * Investigate whether transactions are well-supported. ------------------------------------------------------------------------ Mon, 13 Dec 2004 00:25:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004-r… (11.02 KB) I've worked a bit on the patch (coding style issues as mentioned by Dries). One thing I noticed is that the patch uses REPLACE. IIRC this needs to be chagned to "UPDATE, if fail INSERT" for pgsql compatibility. Nick, are you still interested in working on that patch? I'd like to know how it works on your site and work on getting it into core. ------------------------------------------------------------------------ Mon, 13 Dec 2004 12:33:08 +0000 : Dries Gerhard: your patch does not apply. ------------------------------------------------------------------------ Mon, 13 Dec 2004 17:10:12 +0000 : killes(a)www.drop.org Yes, I know, that was the same version as I mailed to you earlier. ------------------------------------------------------------------------ Mon, 13 Dec 2004 21:02:06 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions.patch (52.96 KB) Ok, upüdated the patch to cvs. ------------------------------------------------------------------------ Tue, 14 Dec 2004 08:58:36 +0000 : Dries Some more comments: * db_begin_transaction() and db_end_transaction() do not belong in database.inc, but in database.mysql.inc and database.pgsql.inc respectively. * The node module calls node_revisionsision_list() which is not defined. (Fxed that on my local copy.) * Do db_begin_transaction() and db_end_transaction() deprecate Jeremy's table locking patch? * The upgrade path assigns the wrong user ID to each revision. * The upgrade path assigns the wrong date to each revision (that or a node's revision page shows the wrong usernames/dates). * The coding style needs a bit of work, but we can worry about that later. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:34:44 +0000 : Nick Nassar If you need any help getting those things fixed, just let me know. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:50:30 +0000 : Nick Nassar How this relates to Jeremy's node locking patch: There was lots of discussion, and node locking was decided against because from an end user point of view you never want a node to be locked. He's now advocating for a much simpler patch that warns users if their changes will overwrite someone elses. That patch still has a race condition, which might be fixed using db_begin_transaction(). http://drupal.org/node/6025 ------------------------------------------------------------------------ Tue, 14 Dec 2004 22:26:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_0.patch (55.96 KB) Here is an updated patch that tries to address Dries concerns. ------------------------------------------------------------------------ Wed, 15 Dec 2004 08:32:50 +0000 : Dries Attachment: http://drupal.org/files/issues/revisions-bug.png (76.06 KB) It didn't fix the aforementioned bugs. See attached screenshot. ------------------------------------------------------------------------ Thu, 06 Jan 2005 20:15:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_1.patch (51.77 KB) Ok, here is a new version. Dries and myself worked hart at it, so please have a look. what is still missing - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. ------------------------------------------------------------------------ Wed, 19 Jan 2005 21:43:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_2.patch (49.49 KB) Here is an updated patch. We discussed to keep the current title in node module and also in the revisiosn table. This is content duplication but will save many joins as many queries only need the title of a node. Discussion is welcome. ------------------------------------------------------------------------ Wed, 19 Jan 2005 23:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_3.patch (29.93 KB) I've implemented the aforementioned solution. This makes the patch much smaller. The patch now also removes taxonomy_node_has_term() which wasn't used anywhere. I'd really apprciate if some people could test drive the patch. It will be another huge improvement for 4.6. ------------------------------------------------------------------------ Thu, 20 Jan 2005 00:05:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_4.patch (30 KB) Another revision. Steven didn't like my literal $node->vid in queries. ------------------------------------------------------------------------ Thu, 20 Jan 2005 01:10:50 +0000 : killes(a)www.drop.org - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. These issues are still open, btw. Especially the first one needs to be tackled. ------------------------------------------------------------------------ Tue, 25 Jan 2005 20:11:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_5.patch (51.13 KB) Here is a patch that has the database tables updated for forum, book, and page module. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:55:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_6.patch (49.18 KB) Yet another update to keep it working with head. The patch now also removes the table definitons for the page table. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:57:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_7.patch (55.69 KB) Sorry, that was the old version, this is the right one. ------------------------------------------------------------------------ Mon, 31 Jan 2005 19:55:03 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_8.patch (55.71 KB) Updated once more. ------------------------------------------------------------------------ Mon, 31 Jan 2005 20:52:08 +0000 : Dries Anyone to help review/test this? ------------------------------------------------------------------------ Mon, 31 Jan 2005 21:22:36 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_9.patch (49.29 KB) Updated again, the update functions occurred twice. Thanks Bart. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:05 +0000 : killes(a)www.drop.org Don't know if the db I am using is corrupted or what. I still do have some didficulties. The latest patch is attached. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_10.patch (49.67 KB) I am probably slowly going mad ... ------------------------------------------------------------------------ Wed, 02 Feb 2005 01:54:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_11.patch (48.95 KB) The update issue still needs investigating. This patch is updated for cvs. ------------------------------------------------------------------------ Wed, 02 Feb 2005 20:20:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_12.patch (49.83 KB) Ok, here is a new version. I've solved my troubles with book.module. There are still some issues with forum module. Possibly due to inconsistent database. ------------------------------------------------------------------------ Wed, 02 Feb 2005 21:31:05 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_13.patch (49.83 KB) Turns out the drupal.org database had indeed some quirks. Please run this query in your oldest db and tell me the result: select nid,type from node where type like '%/%'; If you get a non-zero result we might need to add another security update. The patch could use still more testing, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 01:16:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_14.patch (49 KB) Ok, we are getting somewhere. At a first glance the update is working. There is a problem remaining: the revisions tab will be shown whether the node has revisions or not. Not sure we can/need to fix this. People with a drupal.org account can log in at http://killes.drupaldevs.org/revision/ and poke around. Your permissions will be the same as on drupal.org. Feel free to vreak everything but don't forget to file complaints here. (Note: this is only a pruned version of the drupal.org database with all project nodes and nodes with nids > 7000 dropped). ------------------------------------------------------------------------ Thu, 03 Feb 2005 04:19:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_15.patch (52.39 KB) There was some error in node_save and also the patches to the database.inc files got lost... ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:07:27 +0000 : robertDouglass Submitting book pages doesn't work on your test site. It puts the entire content of the preview inside the body textarea. I wrote a sentence in the body and the log, and pressing preview put several lines of HTML containing both sentences in the body textarea on the preview page, plus the book page wouldn't submit. -R ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:50:59 +0000 : Junyor 0 results here. I started using Drupal with version 4.4, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 23:56:18 +0000 : killes(a)www.drop.org @Junyor: Thanks, that's a good sign. Maybe somebody else has an older db to try. @robertDouglass: The first effect you describe is due to drupaldevs running on PHP 5. I am unsure why the second thing does not work. In node_save() the node object has a nid although there is none in the form. Very strange. I've enabled display of db queries on the testsite. ------------------------------------------------------------------------ Fri, 04 Feb 2005 19:17:55 +0000 : dmjossel No results here on the query: select nid,type from node where type like '%/%'; On a database that was put in place prior to Drupal 4 and is now running on 4.5.2. ------------------------------------------------------------------------ Fri, 04 Feb 2005 20:44:23 +0000 : killes(a)www.drop.org @dmjossel: thanks. @all. The strange problem I reported was apparently php 5 related. After applying Steven's php 5 patch it went away. One error is remaining: If I create a new forum topic it will be shown as part of the book on preview. Hmm, that was due to a db that got corrupted during testing so that is fixed too. Please poke around at the test site and look if you find more errors. ------------------------------------------------------------------------ Sat, 05 Feb 2005 07:16:22 +0000 : Steven By the way, I just remembered that Drupal.org has some blogs lingering on in the database even though blog.module is not enabled. Perhaps this is causing troubles? ------------------------------------------------------------------------ Sat, 05 Feb 2005 11:22:59 +0000 : Anonymous I can't see why it would. Drupal.org will need extra updates for images and project nodes because those have their own tables. GK. ------------------------------------------------------------------------ Sun, 06 Feb 2005 12:49:55 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_16.patch (52.49 KB) Updated to apply to cvs again. ------------------------------------------------------------------------ Tue, 22 Feb 2005 20:15:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_17.patch (49.64 KB) Updated again. All we need is a patch to upload module and an upgrade path for it. ------------------------------------------------------------------------ Fri, 04 Mar 2005 04:22:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_18.patch (52.31 KB) Updated once more. Moved log field from book to node_revisions table as discussed in Antwerp. upload module still missing. We need to decide under which circumstances the log field should be displayed. Should that be added to the workflow? Should it depend on the revisions setting? ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:27:03 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_20.patch (75.52 KB) Ok, here it is: Yet another revision of this grrrrreat patch. Changes from previous versions: - supports versioning for uploaded files. A problem is that if you delete a file, it will be gone for all revisions. - the log field is now in the node_revsions table, but each module has to decide whether to show it or not. I've implemented it for the page and the book type odes. Also, the field can be edited when adding non-book nodes to the book. The log is displayed on the revisions page and if a node is moderated. - the revisions are moved to an old_revisions table to a) get the node table smaller and b) still leave the mavailable for contrib modules that want to retreive old version data. The patch has been applied to killes.drupaldevs.org/revision where it can be tested by anybody (especially people who have "site admin" rights on drupal.org) The database is from drupal.org and you should b able to log in with your pass or simply mail yourself a new one. Gerhard ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:51:56 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_21.patch (59.42 KB) BTW, I marked this a bug because atm the revisions field can grow quite big. Neil has reported problems from some users who were not able to load some nodes due to to many large revisions. Also, som unrelated stuff crept into the patch. New version attached. ------------------------------------------------------------------------ Tue, 08 Mar 2005 05:56:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_22.patch (60.29 KB) Ok, I think I got it. Changes to last version: - uploads are no properly versioned. Missing are still pgsql checks and updates. ------------------------------------------------------------------------ Thu, 10 Mar 2005 16:58:41 +0000 : Anonymous Was able to get http://drupal.org/files/issues/revisions_21.patch to work with drupal-cvs.tar.gz (10 March 2005) by: - includes/database.mysql.inc: Commenting out duplicates for functions function db_begin_transaction and function db_commit_transaction - modules: node.module: Removing "'title' => $node->title," from $node_table_values variable declaration and removing "'title' => "'%s'"," from "$node_table_types" variable declaration. Happy to submit a patch if requested. I'll watch this thread. ------------------------------------------------------------------------ Fri, 11 Mar 2005 23:59:45 +0000 : killes(a)www.drop.org The duplicate function has been removed in rev 22 of this patch. Why do you think the changes in node_save are needed? Titles are saved in both tables for performance reasons. ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:12:21 +0000 : jlerner Hi - I posted comment #62. The changes to node_save appear to be needed because recent patches (both 21 and 22) remove the field 'title' from table 'node'. So without the changes to node_save, node.module is broken and generates errors. Joshua ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:29:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_23.patch (61.17 KB) Thanks, Joshua, for catching this. node:title is there to stay. ------------------------------------------------------------------------ Wed, 13 Apr 2005 16:29:23 +0000 : moshe weitzman since HEAD is open again, perhaps it is a good time to revisit this patch. once this is committed, lets address - http://drupal.org/node/11071 "node_validate does not respect group editing" ------------------------------------------------------------------------ Mon, 18 Apr 2005 15:43:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Mon, 18 Apr 2005 16:16:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24_0.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:19:42 +0000 : Dries I'll commit this patch later this week! If you haven't checked this patch already, I urge you to test/check it out because it will have significant impact on existing code and modules! ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:21:33 +0000 : Dries Also, what do people think about the n.title being duplicated? ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:26:58 +0000 : chx I won't lose any sleep because of duplicated titles... ------------------------------------------------------------------------ Tue, 19 Apr 2005 18:35:58 +0000 : killes(a)www.drop.org Let me explain why I have chosen to duplicate the title (and also the uid): If you look at all the queries in Drupal, you will find that most of them only need the title and th uid of a node. That is, by duplicating it, we save expensive joins on the node_revisions table. Due to this fact, this patch is actually a performance improvement. A note about updating contrib module: Strictly speaking they wouldn't need to be updated. They only need to if their authors decide that their info should be available for revisioning. The upgrade path for forum.module in my update.inc patch (+ the forum patch) should show you what needs to be done. I will write a note for the update page once the patch hits core. ------------------------------------------------------------------------ Sun, 24 Apr 2005 21:21:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_25.patch (60.38 KB) Updated to cvs. Dries: Based on some remarks in #drupal this is the last update I am going to do. Apply it or won't fix it. ------------------------------------------------------------------------ Sat, 30 Apr 2005 03:42:39 +0000 : Jeremy(a)kerneltrap.org Attachment: http://drupal.org/files/issues/revisions_25.patch.patch (528 bytes) That's a big patch. I've only started looking through it. I noticed one little typo, affecting updates. A patch to your last patch is attached. I'm running with the revision patch on my dev server now happily. I like the concept. What happens if you click 'stop' on your browser in the middle of a MySQL "transaction"? I assume that kills the connection to MySQL, and the lock is freed? But this then leaves changes only partially applied? What exactly does locking/unlocking the tables buy us in MySQL? I don't see anywhere that we detect if an apply fails part way through, and thus roll back...? What am I missing? ------------------------------------------------------------------------ Sat, 30 Apr 2005 07:11:28 +0000 : Dries Jeremy: many of us are worried about the performance ramifications this patch introduces. Early experiments showed a small performance improvement (while a performance regression might be expected). More performance reports from large sites like kerneltrap.org will certainly help this patch. Mind to do a quick performance comparision and to report back with some numbers? Thanks. ------------------------------------------------------------------------ Sat, 30 Apr 2005 12:38:02 +0000 : Jeremy(a)kerneltrap.org Dries: I'm not running HEAD on kerneltrap, so this really isn't a possibility. Furthermore, until I understand why we're locking tables, I don't like it. The idea of revisions in their own tables is great. The idea of locking tables to get (without any obvious benefit) there really worries me. ------------------------------------------------------------------------ Sat, 30 Apr 2005 14:16:01 +0000 : killes(a)www.drop.org @Jeremy: Thanks for looking at the patch! Also for catching the typo. :) Did you try to upgrade your database? If yes, how did it go? One of Dries' concerns is the complexity of the upgrade. How many nodes and revisions did the db have? About database locking: This part of the patch was created by Nick and I simply continued to use it. Maybe the code should rather be: if(db_begin_transaction(array('{node}', '{node_revisions}', '{watchdog}', '{sessions}', '{files}'))) { db_query($node_query, $node_table_values); db_query($revisions_query, $revisions_table_values); db_commit_transaction(); ... } The idea is probably to avoid two updates at the same time. I don't think the locking helps if you abort the script at an inconvenient time. Rollbacks aren't implemented in all mysql versions. We could omit the db locking if deemed inappropriate. Maybe Nick can explain his ideas behind this. @Dries: I wonder who the "many of us" are. They certainly haven't spoken to me. Moshe had some reservations about the upgrade path and project module, but the time that project module abused revisions to store issue updates was long ago and his reservations were resolved. Nobody else (besides you of course and now Jeremy) has voiced reservations in a way that was audible to me. If you grep through the patch you will notice that there are only four queries which have a join on the node_revisions table. Two of them are in node_load and in the other cases the join replaced a join on the node table. The two queries in node_load are the only ones that have both a join on the node and the revisions query. Thus, loading of individual nodes might become somewhat slower. All other queries will be faster since the node table is now much smaller. Also, node loading does not have to be slower, it depends on your node table. If you had a lot of revisions and thus a large table, then the new scheme will make your queries actually faster since we do not load the revisions on each and every node load anymore. If you didn't have many revisions your node_load migth be somewhat slower. WRT to the update script Karoly pointed out that we could use multiple insert queries instead one query per revision. This would probably make the update somewhat faster. I am willing to work on this iff you declare that you will commit the patch afterwards. I'd need to know if this will work on pgsql and on all supported mysql versions before I work on it. About locking: Database locking is dog slow, at least on mysql. I was using locks in an earlier version of the upgrade script but had to remove it for (serious!) performance reasons. ------------------------------------------------------------------------ Mon, 09 May 2005 15:07:34 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_26.patch (46.45 KB) Ok, another update, cause I need it myself. I've left out the transaction stuff for now. It is in principle unrelated to this patch and should be discussed elsewhere. This also makes the patch smaller and easier to review (hint, hint). ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:09 +0000 : killes(a)www.drop.org The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_27.patch (39.05 KB) The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 21:23:06 +0000 : Dries Got one error during the upgrade path: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ------------------------------------------------------------------------ Mon, 09 May 2005 21:26:19 +0000 : killes(a)www.drop.org This had happend to me as well, when I tested this patch. The reason is that for some reason the vid is not unique. Most likely there are some entries with vid = 0 in there. Can you check which node types those have? it always was an error in the test database. See: http://drupal.org/node/7582#comment-20678 ------------------------------------------------------------------------ Mon, 09 May 2005 21:27:06 +0000 : Dries Actually, I got 2850 errors during the upgrade. Some of these: sprintf() [function.sprintf]: Too few arguments in drupal-cvs/includes/database.inc on line 154. Some of these: Query was empty query: in drupal-cvs/includes/database.mysql.inc on line 66. And this: Unknown table 'n' in field list query: SELECT n.nid, n.vid FROM node INNER JOIN files f ON n.nid = f.nid in drupal-cvs/includes/database.mysql.inc on line 66. :-) ------------------------------------------------------------------------ Mon, 09 May 2005 21:29:19 +0000 : Dries Or this: user error: Unknown column 'log' in 'field list' query: SELECT parent, weight, log FROM book WHERE nid = 1 in drupal-cvs/includes/database.mysql.inc on line 66 ------------------------------------------------------------------------ Mon, 09 May 2005 21:52:12 +0000 : Dries The time required to generate my main page went from 902 ms (before upgrade) to 2139 ms (after upgrade). The time required to generate a forum listing (?q=forum/x) went from 1872 ms (before upgrade) to 2874 ms (after upgrade). Maybe this is because my database is not consistent as the result of the upgrade errors (yet I don't see any errors on the pages I benchmarked). ------------------------------------------------------------------------ Tue, 10 May 2005 00:24:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_28.patch (53.47 KB) Ok, let me get to this from the bottom to the top: - my test runs indicated a different development wrt timing. If I had gotten your results, I had stopped working on this long ago. So your results are wrong for some reason. - user error: Unknown column 'log' in 'field list' Wasn't my day, the book patch got lost. It is contained now. First -R the old patch, then apply this one. - Unknown table 'n' in field list query: Walkah found this, but I forgot to fix it. Fixed now. - I've no idea where the other queries come from. I am hoping that either your test db is borken or they are follow ups from the other ones. If you let me have your test db, I'll try some debugging. Thanks for wasting your time, too. ------------------------------------------------------------------------ Tue, 10 May 2005 05:07:31 +0000 : Dries I double-checked and the numbers don't seem to lie. I'll test some more after work on another machine to make sure it is not platform-specific. ------------------------------------------------------------------------ Wed, 11 May 2005 03:32:47 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_29.patch (54.83 KB) Ok, here I am again. What I did: 1) Ask Dries to let me have drupal.org database 2) get 400MB of SQL inserts... 3) take 23 minutes to import said data 4) Remove all image and project nodes (don't want to install their modules), 11765 nodes left 5) back up data 6) take tests on non-cached /node page (as anonymous user). ab results: -c 1 -n 25: Requests per second: 1.29 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 663 775 179.7 689 1264 7) Do the same for the tracker page: Requests per second: 0.83 [#/sec] (mean) Total: 1182 1199 7.4 1199 1217 8) Apply my patch (rev. 28). 9) run db update and hold breath 10) update times out... 11) play back backup from 5) 12) wait 13) getting annoyed and removing cache, watchdog, and accesslog before playing in backup. 14) wait again. Understand why Dries doesn't try this patch often. Maybe a smaller DB would do for testing? 15) wait more. get really annoyed. 16) Set time limit to 18000 in update.php 17) try again 18) fails again before the second update is completed. 19) curse. 20) delete search stuff from db. Ooops, sooo much smaller... 21) import again, below 2 minutes... 22) rewrite to use extended insert. Found a bug. 23) still does not complete. Mysql logging to the rescue! 24) tid = 0? Not good. 25) Well, the update works fine till node 10834. 5595 nodes done, 6136 to go. 26) Writing shell based update script. Discovery: 24MB aren't enough. Hopefully 64 are. Nope. extended inserts for revisions are apparently not the brightest idea: Huge memory consumption. Hmm, no, all updates got through. Selecting the revisions to put them into old_revisions table screwed it. Learned about CREATE TABLE old_revisions SELECT syntax. Yay! finally. 24 MB are just not enough the update.php script seems to still break. 27) Benchmarks! /node Requests per second: 1.54 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 632 649 40.5 636 791 /tracker Requests per second: 0.86 [#/sec] (mean) Total: 1119 1165 65.8 1160 1461 Ok: So we get an improvement for many node_loads, but none for simple selects from node. More tests can be done. 28) roll new patch Ain't Drupal fun? ------------------------------------------------------------------------ Wed, 18 May 2005 13:38:05 +0000 : Dries I did another round of tests on _another_ machine and it is not looking good: Before upgrade After upgrade ?q= (main page) 218 ms/request 340 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request ?q=about (book page) 375 ms/request 5400 ms/request The upgrade process itself gave me a number of 'query was empty' and 'sprintf(): too few arguments' reports. Everything seems to work fine though. Looking at the ?q=about page, I see that the following query is executed twice _and_ that each time, it take more than 2 seconds to complete: SELECT n.nid, n.title, b.parent, b.weight FROM node n INNER JOIN book b ON n.vid = b.vid WHERE n.status = 1 AND n.moderate = 0 ORDER BY b.weight, n.title; --8 SHOW INDEX FROM book; +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | book | 1 | book_parent | 1 | parent | A | 92 | NULL | NULL | | BTREE | | | book | 1 | nid | 1 | nid | A | 369 | NULL | NULL | | BTREE | | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ 2 rows in set (0.00 sec) The book module does not appear to have a primary key? Sounds like a bad idea so I added one: mysql> ALTER TABLE book ADD PRIMARY KEY nid (nid); Query OK, 369 rows affected (0.02 sec) Records: 369 Duplicates: 0 Warnings: 0 Next, I wanted to make the vid column a unique key in all node tables: mysql> ALTER TABLE node ADD UNIQUE vid (vid); Query OK, 20392 rows affected (0.81 sec) Records: 20392 Duplicates: 0 Warnings: 0 mysql> ALTER TABLE book ADD UNIQUE vid (vid); ERROR 1062: Duplicate entry '0' for key 2 mysql> ALTER TABLE forum ADD UNIQUE vid (vid); Query OK, 10806 rows affected (0.10 sec) Records: 10806 Duplicates: 0 Warnings: 0 As you can see, it fails for the book table which makes me believe there is some inconsistent data ... I set out to fix that: mysql> SELECT nid, COUNT(nid) AS vids FROM book GROUP BY vid HAVING vids > 1; +-----+------+ | nid | vids | +-----+------+ | 871 | 2 | +-----+------+ 1 row in set (0.00 sec) mysql> SELECT title FROM node WHERE nid = 871; Empty set (0.00 sec) mysql> DELETE FROM book WHERE nid = 871; Query OK, 1 row affected (0.00 sec) mysql> ALTER TABLE book ADD UNIQUE vid (vid); Query OK, 368 rows affected (0.01 sec) Records: 368 Duplicates: 0 Warnings: 0 Looks like everything is well now. Ran some new benchmarks: Before upgrade After upgrade With indices ?q= (main page) 218 ms/request 340 ms/request 336 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request 1531 ms/request ?q=about (book page) 375 ms/request 5400 ms/request 475 ms/request Unfortunately, we're still slower than the original code. ------------------------------------------------------------------------ Wed, 18 May 2005 21:53:31 +0000 : killes(a)www.drop.org Dries, thanks for testing it again. I do think the broken queries you observer have something to do with the bad performance after the update. Please log the queries any I will have a look at them. I've never seen any such queries. My update script also tries to create the appropriate indices, but it will of course fail if the database contains cruft. the indices for the forum are probably missing, too. I am still convinced that the patch is core worthy. ------------------------------------------------------------------------ Thu, 19 May 2005 04:36:09 +0000 : Dries It wouldn't hurt if more people would benchmark this patch. The patch's current performance worries me. Did you check your watchdog messages after upgrading the drupal.org database? Depending on your settings, errors might only be shown in the watchdog. I'll look into the remaining glitches as time permits. Thanks for your persistence in keeping this patch up-to-date. :) ------------------------------------------------------------------------ Thu, 19 May 2005 11:59:22 +0000 : killes(a)www.drop.org Dries: Can you please let me have your updated database? I want to have a look at it and try my own benchmarks with it. And yes, if I did learn something on this project is how to be persistant. ;) ------------------------------------------------------------------------ Fri, 24 Jun 2005 16:25:34 +0000 : killes(a)www.drop.org Here is an idea that occurred to me: The problem with the upgrade process is that keeping the existing revisions requires a lot of work to do. This generates a huge amount of sql queries for a large database and also requires a huge amount of memory. My suggestions is to let update.php only handle the basic upgrade, ie without old revisions. An additional module could be created that would implement a cron based approach to upgrading old revisions one node at a time. it could expose a hook to let contrib modules do their own upgrades. Dries, what do you think? (I am writing "Dries" because he seems to be the only one who is interested in getting this into core...) ------------------------------------------------------------------------ Fri, 24 Jun 2005 22:25:11 +0000 : Junyor Killes: I'm also interested in seeing this hit core. What about adding something to legacy.module to do it? ------------------------------------------------------------------------ Sun, 26 Jun 2005 21:14:54 +0000 : chx This is a sensible approach. Maybe this is the _only_ sensible approach. I have a little problem though: while the conversion is running somehow both revision handlers should be available. ------------------------------------------------------------------------ Sun, 26 Jun 2005 22:16:21 +0000 : killes(a)www.drop.org hehe, one only has to whine and bug enough and one gets some feedback. ;) @junyor: legacy.module would be a good place. my current idea is to auto-enable it in update.php and then disable it again in legacy_cron after all nodes are updated.. ;) @chx: When somebody wants to look at revisions of a node that node could be auto-updated. The only problem are contrib modules: they's need to have some hook in order to update their own data. When somebody looks at the revisions of a node than cannot be updated because the contrib module in question has no such hook, we can optionally let the user discard old revisions I guess. Dries, what do you think? ------------------------------------------------------------------------ Mon, 25 Jul 2005 16:48:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_30.patch (53.13 KB) Sooooo; I've updated this patch once again. Dries didn't like my idea of legacy updates so we will have an option to discard old revisions in case the update should prove difficult. Always keep a backup. ------------------------------------------------------------------------ Mon, 25 Jul 2005 17:31:12 +0000 : Bèr Kessels can we please accept and commit this patch? We can iron out any issues later. This patch is just far too big and complex to be perfect-at-once. ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:22:50 +0000 : Dries @Ber: no, we can't commit this patch blindly. Data loss is a much bigger problem than a syntax error or other code glitch. We can break Drupal but we can't break their data. I've spent quite a bit of time testing the previous version of this patch and noticed significant performance degradation. Tell me, Ber, why can't you test this patch first? ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:40:37 +0000 : Bèr Kessels I was not referring to not testing it. If it is just the upgrade path that proves to be cumbersome, then why can we not fix it afterwards, I.e. when everyone has had a good chance to look at it. Testing such a huge patch requires a lot of work. something no-one just does in his spare hours. We had discussions before to deal in a slightly different way with big changes; to commit them quicker and to leave the ironing out of any left overs for the community. I hooked into that discussion here. For two reasons. One is that Gerhard has spent numerous hours on maintaining this patch. The other one is that the community can be of help with ironing out issues in such a large change, much better than that Gerhard can do on his own. And yes, dataloss is very bad, but no-one should loose data, when he/she followed the instructions (backing up)... ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:00:38 +0000 : Dries Ber: applying this patch takes 15 seconds. Whether I apply this patch for you, or whether you apply it yourself, it will hardly reduce the 'testing cost'. The problem is that data loss can be subtle; it might go unnoticed for a couple days. Make no mistake, I'd like to see this patch committed ASAP, but it warrants some testing. Let's test/benchmark it and commit it. ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:55:03 +0000 : drumm The upgrade script already takes a long time to execute and does not provide feedback to the user about how it is doing. I plan on making the upgrade script able to spread the updates across multiple page views and give user feedback showing that progress is being made. This will hopefully make the speed of this update a moot point. I am working on this for my dayjob, CivicSpace, so it should get done "real soon now", but it should be expected to take awhile (smallish number of weeks). Please make comprimises if necessary. ------------------------------------------------------------------------ Wed, 27 Jul 2005 17:38:57 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_31.patch (48.78 KB) @drumm: the timeout isn't usually the problem. Memory consumption is. I'll look into doing the updates in chunks of 1000 nodes or so. Clousseau found a bug in the patch, updated. ------------------------------------------------------------------------ Thu, 28 Jul 2005 18:28:08 +0000 : moshe weitzman Even if the upgrade issues are resolved, we have to figure out why this patch is slowing down drupal (according to benchmarks posted here). ------------------------------------------------------------------------ Fri, 29 Jul 2005 07:51:45 +0000 : Dries We can't proceed without some additional benchmarking. I already benchmarked it on two machines (see comments above), and I'll benchmark it again after August 10. The performance impact of this patch worries me so if two other people could test and benchmark this patch extensively, that would come a long way. August 10 is close to the code freeze so some help is necessary. If Gerhard/killes reviewed/tested one of your patches, it is time to return the favor ... ------------------------------------------------------------------------ Fri, 29 Jul 2005 15:40:31 +0000 : Jose A Reyero After looking at the patch I'm not really surprised this slows down everything. I thought that the reason why we wanted revisions in their own table was in the first place to have a simpler -and faster- node table. But this patch: - Adds fields and complexity not only to the node table but also to all the main tables for data - Needs additional joins just to retrieve single nodes (which is done many times for a typical main page) - Does away with encapsulation of the revision functionality requiring other modules to handle revissions related data. So please, please, please: - Make revisions table only needed when actually using revisions - Keep that old nice thing of hiding revisions system from other modules - And what's so bad with serializing data anyway? I agree that usually there are better ways to store data. But this is one of the cases where serialization does make sense. Not the current huge serialized field, but only a new table with one serialized node per row. - In general, do it much simpler. This is way too complex and it removes more functionality than it adds. In just four words: keep it simple, please. ------------------------------------------------------------------------ Fri, 29 Jul 2005 21:25:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_32.patch (51.27 KB) Patch updated (v31 was buggy, and there was an update to updates.inc). I've had a discussion with Jose and he will maybe do some testing over the weekend. John is also doing some testing as we speak. ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:06:59 +0000 : Nick Nassar It has been exactly one year since I first submitted this patch. You know what that means: *old patch party in IRC!* Virtual beers on me! w00t! ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:08:33 +0000 : Jose A Reyero Hi again. I had a talk with killes -I have to say he enlightned me about this- and also did some testing. Sorry but I dont have a good set up to provide benchmarks. Here are my conclusions -and also some explanation for the varied results abt performance: - This is faster for simple node listings - This is slower for full node/load ....thus the performance increase/decrease for a given page should depend on the relation between the number of simple node listings (usually blocks) and the number of nodes displayed in the main page. I've also experienced some trouble with attachments -not re-created when creating a new revision- and book relations, same. I'm not sure I like this new "feature" of creating a new revision every time you rollback an old one -in Drupal 4.6 this seems not to happen. Besides these details, I think this patch is good enough -and quite a powerful approach actually- though I still feel it's bit too complex when simply dealing with nodes -without revisions-, which is what we do most of the time, and complexity will be even higher with contrib modules using still more tables, not to mention flexinode -versioning info will have to be scattered everywhere in the db. So this is not a +1 nor -1. On one side I think this is powerful. On the other side I feel this is too complex and really, it's more than I need. I would be happy with a 'revision' table having: nid, vid, data(serialized), that you can forget about when not using revisions. ------------------------------------------------------------------------ Sat, 30 Jul 2005 19:16:41 +0000 : Nick Nassar On a much more serious note that my last comment, unless I'm missing something, you've introduced a subtle bug into this patch since I left it. A and B are both updating revision 5 of a node. A calls db_next_id() and gets a revision id of 6 B calls db_next_id() and gets a revision id of 7 B updates the node's revision to 7 B adds revision 7 to the revision table A updates the node's revision to 6 A adds revision 6 to the revision table Now, you have a situation where the current revision (6) isn't actually the last revision(7). For some uses, that's not a big deal, but in the case of wikis, which is what I originally hoped to use this patch for, it creates problems. ------------------------------------------------------------------------ Mon, 01 Aug 2005 00:04:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_33.patch (51.27 KB) Updated the patch. updates.inc was changed and chx found a typo. @Nick: I don't think the issue is due to this patch. Drupal has currently no mechanism to not let people commit a new version of the same node, whether with or without revisions. drupal_next_id ensures that the revision IDs are unique and the later revisions will be the current one. Which revision is the current one does not depend on order of the version ID, btw. @Jose: what exactly were your problems with uploads? what do I need to do to reproduce? ------------------------------------------------------------------------ Mon, 01 Aug 2005 11:33:56 +0000 : Jose A Reyero Sorry, I dont know what I did the other day but I cannot reproduce it either :-) I applied this new patch. Problems I'm having now (Testing with "stories") - Book outline is gone after creating a new revision - I cannot delete attachments (not sure it is related to your patch or to something else) - Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I? ------------------------------------------------------------------------ Sun, 14 Aug 2005 16:03:41 +0000 : killes(a)www.drop.org Jose, thanks a lot for testing! "- Book outline is gone after creating a new revision" Sounds like a bug, it should not. "- I cannot delete attachments (not sure it is related to your patch or to something else)" Another bug "- Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab" Log messages are possible for nodes of type book and page as well as all nodes that are put into the book. Log messages can be added for any node type, the module author should implement those. "Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I?" No, the db update should be the same. I will try to fix the bugs you found. WRT the db update I've got an announcement to make: I will not migrate old revisions. Why? 1) it is a rather complicated process that requires a lot of memory and is likely to give people on cheap hosting trouble. 2) testing it takes a lot of time. 3) I don't personally need it I will still move old revisions to an old_revisions table from where they can be converted by a later update or a custom script. ------------------------------------------------------------------------ Tue, 16 Aug 2005 21:34:43 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_34.patch (52.92 KB) Ok, here is incarnation No. 34 of this patch. Dries agrees with me that we can put the revision rescue operation into a later update, possible using Drumm's installer. The patch fixes the issue with the file uploads not being deletable, I could not reproduce the problem with the outlines. ------------------------------------------------------------------------ Sat, 20 Aug 2005 13:33:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/scripts.tar.gz (4.39 KB) I've started to do some performance testing. I created a script that creates nodes that have revisions according to the new system. A node has a 50% chance to have a revision and then an equal chance fro between 1 and 25 revisions. I've created a database with 50 nodes and 500 comments, 7 vocabs and 100 total taxo terms. I've also created a script to convert these revisions to the old format (much easier than the other way around). Initial tests didn't show a significant change for /node. I'll try with more nodes and /tracker. Attached are the two scripts in a tarball. They need to be run against the patched install. Also included are the two preliminary results. ------------------------------------------------------------------------ Sat, 20 Aug 2005 15:16:04 +0000 : killes(a)www.drop.org I've now created a databse which has 1000 nodes and each node has 50 or 51 revisions. I have no taxo terms assigned and no comments (those parts didn't change). I am only using stories and pages. I am now fairly confident that this patch is a performance improvement for this kind of setup. for the new patch on the /node page I get Requests per second: 0.27 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 3592 3648 56.9 3640 3985 Waiting: 3573 3626 55.1 3620 3949 Total: 3592 3648 56.9 3640 3985 for the old one I get: Requests per second: 0.23 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 4247 4297 64.0 4286 4657 Waiting: 4156 4263 65.3 4253 4624 Total: 4247 4297 64.0 4286 4657 For the tracker the results are even better: old: Requests per second: 2.00 [#/sec] (mean) new: Requests per second: 3.64 [#/sec] (mean) Anybody interested in obtaining this database can mail me. It is 60MB (gzipped). ------------------------------------------------------------------------ Sat, 20 Aug 2005 16:33:26 +0000 : killes(a)www.drop.org Ok, I've now deleted the revisions and only kept the current one. The initial tests showed that the performance of the patch was miserable. then I remembered that I had shreddered the revisions table and did a mysql optimization. Now the results are as follows: /node: Requests per second: 3.19 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 275 312 44.3 290 499 Waiting: 255 291 44.1 268 479 Total: 275 312 44.3 290 499 Requests per second: 3.18 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 274 314 42.4 292 407 Waiting: 255 294 42.0 273 388 Total: 274 314 42.4 292 407 The lower one is the new system, there is no difference considering statistics. /tracker old: Requests per second: 3.72 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 238 268 33.8 249 332 Waiting: 219 247 33.3 230 313 Total: 238 268 33.8 249 332 new: Requests per second: 3.77 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 240 264 29.8 250 324 Waiting: 221 244 29.0 229 305 Total: 240 264 29.8 250 324 Same applies, no real difference. So, as soon as somebody tests the upgrade path the patch should be committed. ------------------------------------------------------------------------ Sat, 20 Aug 2005 17:18:39 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_35.patch (55.51 KB) Ok, I coudl hear them in advance "1000 nodes without revisions are not sufficient". Well, here are results for 30000 nodes without revisions. /node new: Requests per second: 0.45 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2162 2202 58.9 2188 2485 Waiting: 2143 2182 54.1 2169 2413 Total: 2162 2202 58.9 2188 2485 old: Requests per second: 0.43 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2319 2346 77.6 2332 2761 Waiting: 2223 2324 79.0 2311 2742 Total: 2319 2346 77.6 2332 2761 One could argue that the new approach s slightly faster. I'd like to see mroe results to believe it, but it is pretty clear it isn't slower. /tracker new: Requests per second: 0.41 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2330 2418 116.9 2403 3194 Waiting: 2302 2397 117.7 2382 3175 Total: 2330 2418 116.9 2403 3194 old: Requests per second: 0.39 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2496 2548 50.1 2536 2841 Waiting: 2477 2528 50.0 2518 2822 Total: 2496 2548 50.1 2536 2841 Statistically insignificant differences. All those tests where done with ab and as user No 1. model name : Intel(R) Celeron(R) CPU 2.40GHz 512MB RAM Besides the differences for revisions the change in served pages with different configurations is interesting. A new version is attached. node_save now returns the complete node (not only the nid). ------------------------------------------------------------------------ Mon, 22 Aug 2005 20:13:52 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_36.patch (55.66 KB) Ok, here is version 36 which incorporates today's changes in cvs. It also changes the upgrade path. I was trying my old solution and had to up my php mem limit to 80 MB in order to get it to run through. I hope this one is better but it needs testing. ------------------------------------------------------------------------ Mon, 22 Aug 2005 22:46:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_37.patch (52.29 KB) Ok, I've now moved _all_ of the heavy lifting to mysql. pgsql need to catch up. No problems with either time out or memory limit anymore. Since pgsql plays catch-up all the time, I set it to "ready to be committed". ------------------------------------------------------------------------ Tue, 23 Aug 2005 11:45:59 +0000 : stefan nagtegaal I could apply the patch on my localhost, and did not seen any errors nor disfunctioning drupal features while creating, revising, editing or deleting nodes.. So, +1 for this patch qua implementation it makes sense. Unfortunatly I chaven't got a clue howto benchmark, so that's for another person. @Gerhard: I really think you are a great person by rewriting this patch over and over again, untill Dries decides it is good enough to get in to the trunk. I doubt about the fact if there are more persons in our community who are like you.. People who would revise their patch 37 times, and even after that still open for discussion.. Gerhard, I think you are a great example of how a real coder should do it.. I think we can all learn from you.. ------------------------------------------------------------------------ Tue, 23 Aug 2005 17:15:29 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_38.patch (52.26 KB) Karoly pointed out that node_save(&$node) is nicer. @Steef: nice that you write that, but a _really_ good coder would only have needed one revision. ;p ------------------------------------------------------------------------ Wed, 24 Aug 2005 15:36:03 +0000 : Dries function book_update($node) { - db_query("UPDATE {book} SET parent = %d, weight = %d, log = '%s' WHERE nid = %d", $node->parent, $node->weight, $node->log, $node->nid); + if ($node->is_new || $node->revision) { + db_query("INSERT INTO {book} (nid, vid, parent, weight) VALUES (%d, %d, %d, %d, '%s')", $node->nid, $node->vid, $node->parent, $node->weight); + } + else { + db_query("UPDATE {book} SET parent = %d, weight = %d WHERE vid = %d", $node->parent, $node->weight, $node->vid); + } } Why 'is_new' and 'revision'? Isn't 'is_new' redundant? I loaded a fresh copy of the drupal.org database on my test system. How do I test this? Where is the upgrade script/path? ------------------------------------------------------------------------ Wed, 24 Aug 2005 16:31:09 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_39.patch (52.25 KB) is_new is indeed redundant since it is handeled in _update. New patch attached. The upgrate is as usual in updates.inc and accessed through update.php. ------------------------------------------------------------------------ Thu, 25 Aug 2005 10:35:24 +0000 : jakeg This is probably a little off topic but I can't find info elsewhere and this may affect revisions changes. It appears as though all revisions of a node are accessible to all users (inc. anonymous users) using ?revision=x. Surely only the 'current' revision should be accessible by default, and there should be a permission that can be assigned to roles to view non-current revisions? I don't want regular users looking at non-current revisions! A revision may have been made for various private reasons. Only admin should be able to do that. ------------------------------------------------------------------------ Thu, 25 Aug 2005 13:26:12 +0000 : Dries On the latest snapshopt of the drupal.org database I get this: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {forum} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {node_revisions} ADD KEY nid (nid) FAILED ALTER TABLE {node_revisions} ADD KEY uid (uid) FAILED ALTER TABLE {node_revisions} ADD PRIMARY KEY vid (vid) FAILED CREATE TABLE {old_revisions} SELECT nid, revisions FROM {node} WHERE revisions != '' OK ALTER TABLE {book} DROP log OK ALTER TABLE {node} DROP teaser OK ALTER TABLE {node} DROP body OK ALTER TABLE {node} DROP format OK ALTER TABLE {node} DROP revisions OK ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:30:11 +0000 : killes(a)www.drop.org That's strange, it worked for me with an older version of the same db. Anyway, I've re-rolled the patch and rearranged the update queries a bit. Also everything is now in one update, thanks to Neil for pointing that out. ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:32:17 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_40.patch (52.22 KB) and the patch. ------------------------------------------------------------------------ Sun, 28 Aug 2005 02:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_41.patch (52.4 KB) I've tested the upgrade path on an older copy of the drupal.org db and found and squished two bugs. The test cycle can be shortened a bit by deleting non-essential db tables before importing (cvs, history, project_*, ...) ------------------------------------------------------------------------ Sun, 28 Aug 2005 09:18:41 +0000 : Cvbge Attachment: http://drupal.org/files/issues/revisions_41.patch.patch (1.14 KB) Hi, tested the patch a bit: 1. in database.pgsql in the table files a) after removing PRIMARY KEY (fid) you need to remove also "," from the line above b) you CREATE INDEX files_fid_idx ON node(fid); CREATE INDEX files_vid_idx ON node(vid); on wrong tables: not node but files 2. in newly created table node_revisions a) in mysql version there is no DEFAULT, in postgres there is DEFAUL 0, maybe mysql version should be the same as postgres? b) there's no log longtext NOT NULL default '' in pgsql version c) you have translated format int(4) NOT NULL default '0' to format smallint NOT NULL default '0', it also should be integer and not smallint d) you create index: CREATE INDEX node_revisions_title_idx ON node_revisions(title,type);, but there is no type column in the table, and also there is no similar index in mysql version Attached patch should fix the errors ------------------------------------------------------------------------ Sun, 28 Aug 2005 09:38:36 +0000 : Cvbge pgsql needs a sequence to be created, fixed in this patch. An error when trying to view an old revision, e.g. in ?q=node/3/revision/1: warning: pg_query(): Query failed: ERROR: column "c.format" must appear in the GROUP BY clause or be used in an aggregate function in /var/www/dt/d/includes/database.pgsql.inc on line 69. user error: query: SELECT c.cid, c.pid, c.nid, c.subject, c.comment, c.format, c.timestamp, c.name, c.mail, c.homepage, u.uid, u.name AS registered_name, u.picture, u.data, c.score, c.users FROM comments c INNER JOIN users u ON c.uid = u.uid WHERE c.cid = 0 AND c.status = 0 GROUP BY c.cid, c.pid, c.nid, c.subject, c.comment, c.timestamp, c.name, c.mail, u.picture, c.homepage, u.uid, u.name, u.picture, u.data, c.score, c.users in /var/www/dt/d/includes/database.pgsql.inc on line 86. It's comment.module line 796, adding c.format to group by statement fixed this problem. That's all for now...

1 0

[drupal-devel] [bug] Put revisions in their own table
by Cvbge 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/7582 Post a follow up: http://drupal.org/project/comments/add/7582 Project: Drupal Version: cvs Component: node system Category: bug reports Priority: normal Assigned to: killes(a)www.drop.org Reported by: killes(a)www.drop.org Updated by: Cvbge Status: patch (ready to be committed) pgsql needs a sequence to be created, fixed in this patch. An error when trying to view an old revision, e.g. in ?q=node/3/revision/1: warning: pg_query(): Query failed: ERROR: column "c.format" must appear in the GROUP BY clause or be used in an aggregate function in /var/www/dt/d/includes/database.pgsql.inc on line 69. user error: query: SELECT c.cid, c.pid, c.nid, c.subject, c.comment, c.format, c.timestamp, c.name, c.mail, c.homepage, u.uid, u.name AS registered_name, u.picture, u.data, c.score, c.users FROM comments c INNER JOIN users u ON c.uid = u.uid WHERE c.cid = 0 AND c.status = 0 GROUP BY c.cid, c.pid, c.nid, c.subject, c.comment, c.timestamp, c.name, c.mail, u.picture, c.homepage, u.uid, u.name, u.picture, u.data, c.score, c.users in /var/www/dt/d/includes/database.pgsql.inc on line 86. It's comment.module line 796, adding c.format to group by statement fixed this problem. That's all for now... Cvbge Previous comments: ------------------------------------------------------------------------ Wed, 05 May 2004 16:25:27 +0000 : killes(a)www.drop.org Currently all node revisions are stored in a serialized field in node.table and retrieved for _each_ page view although they are rarely needed. However, we have agreed that serializing data is bad and that we should try to keep the memory foot print pf Drupal small. Therefore I propose to create a separate revisions table which would be in principle identical to the node table, only that it could have several old copies of the same node. Extra data added by other modules could be added in a serialized field unless we find a better solution. ------------------------------------------------------------------------ Wed, 05 May 2004 17:06:35 +0000 : jhriggs I too think the serialized approach is less than desirable, but here's an alternative. This would likely take some considerable rework in core and contrib, but the following is how we handle similar types of situations in our databases at work. It is more elegant that a separate table, and avoids the (almost exact) duplication of a table. Instead of separate tables, keep all revisions of nodes in the node table as follows: * add field: active (0/1 or Y/N) * add field: revision * every revision of a node is stored in the node table; however, only one revision can be active at any given time * nid can no longer be unique -- primary/unique key becomes (nid, active) * any time a node is loaded, updated (without revision), etc., the active version is used. Thoughts? ------------------------------------------------------------------------ Wed, 05 May 2004 17:57:48 +0000 : killes(a)www.drop.org I am not opposed to your scheme, but I want to stress the following: * Duplicating a table's structure is not bad (IMHO) as long as the content is different. * having two tables will allow us to have a rather small node table. This is (maybe) a performance gain. ------------------------------------------------------------------------ Wed, 05 May 2004 18:37:29 +0000 : jhriggs I don't necessarily think that duplicating a table's structure is _bad_. It just seems to be wasteful and a pain to maintain. (Every change to the node table is made twice...easy to do, but also easy to miss perhaps.) As for performance, as long as nid and the active indicator are indexed, there shouldn't be any performance loss. Also, archiving an old version when making a new revision will be much simpler: just change the active indicator rather than copying an entire node to another table (and ensuring everything gets copied...again a potential maintainance issue). To be honest, I would just like to see the serialized data go away, regarless of what approach is taken. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:49:33 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_07-30-2004.p… (10.47 KB) I'm interested in using Drupal for a large scale wiki-type project. In order to do this, I need revisions to be in their own table. Attached is a patch to do just that. Most of the changes are pretty self explanatory. Spreading out node data across two tables meant that I had to add database functions to do locking/transactions. Without this, race conditions in which the database becomes corrupted are possible. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:54:36 +0000 : Nick Nassar Oh yeah... The patch is a diff against Drupal CVS ------------------------------------------------------------------------ Sat, 31 Jul 2004 00:00:08 +0000 : Anonymous Gerhard speaking. Nick, thanks a lot for your nice patch! It saves me a great deal of labour. I looked through it and immediately liked it. You not only put the old revisions into a new table but also the current one. Do you have an estimate how much more expensive the additional join is? Besides a few minor coding style issues I found a major one: Just a few hours before you uploaded your patch JonBob's node access patch hit core. That means your patch won't apply anymore as all the queries you change have been changed. Can I bug you to update your patch? ------------------------------------------------------------------------ Sat, 31 Jul 2004 01:11:59 +0000 : Anonymous Also I think that your upgrade path loses existing revisions. ------------------------------------------------------------------------ Sat, 31 Jul 2004 02:39:12 +0000 : drumm I think this is the proper way to do things. No columns are duplicated, there is no serialized data, and only the fields that are logically revised are stored. Nothing jumped out at me as a way to have my node modules be able to keep a table of revisions of additional fields. I'm guessing this could be done within the confines of _insert and _update. Assuming the upgrade path works and modules can extend it I give it a +1. ------------------------------------------------------------------------ Sat, 31 Jul 2004 14:40:15 +0000 : Nick Nassar It figures that just as I finish a big patch, another patch comes along and breaks it. Oh well, it should be a pretty easy to fix. I'll work on it. Fixing the upgrade path to keep revisions should be fairly painless. I found another issue that needs to be fixed before this patch gets merged. There format of a node needs to be stored for each revision. Otherwise, for modules that store a format for the nodes, such as page and book, if you write one revision in PHP and the next in HTML, the PHP revision will be displayed as HTML. This is part of a larger issue of how node modules should store revisions of additional fields. I think each module that wants to do this should create another table with (nid, revid) as the primary key. Just as when they want to add fields to a node they create another table with nid as the primary key. As far as performance goes, for sites that make heavy use of revisions, an extra join on primary keys is going to be a lot faster than grabbing all of the revisions from that database everytime. We would need to run benchmarks to determine is the overall difference in speed is for an average site is a gain or a loss. I'm guessing it's very minor either way. ------------------------------------------------------------------------ Mon, 23 Aug 2004 13:55:49 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_08-23-2004.p… (10.92 KB) Here's an updated patch against CVS that puts revisions in their own table, provides an upgrade path, and fixes the format related bugs in the last patch. Hopefully, this can make it into CVS as soon as the freeze is over. ------------------------------------------------------------------------ Mon, 23 Aug 2004 14:10:39 +0000 : moshe weitzman Interesting patch ... drumm's question is still outstanding. how do modules store revisions of their fields? Are they expected to manage this on their own? Thats not how it works today. As an aside, i am seeing profile_ fields in my node.revisions column. One could argue that those need not be saved. They pertain to the node author, not to the node itself. ------------------------------------------------------------------------ Mon, 23 Aug 2004 16:14:39 +0000 : Nick Nassar Having modules be responsible for storing revisions of their own fields is a side-effect of storing revision data in tables. There's really no way around it. However, revisions generally don't make sense for node types that don't have PHP/HTML content, such as polls. I think it's going to be a pretty rare scenario for a new node type to want another field to change per-revision, so it's a pretty good trade-off. Storing fields that shouldn't be part of revisions, such as the profile_ fields, is a side-effect of storing revisions as serialized objects. Applying this patch will free up that wasted space. :) ------------------------------------------------------------------------ Mon, 23 Aug 2004 17:20:57 +0000 : Anonymous There should be a hook that let's the module choose whether it supports history. This way a module author can prevent the user from doing something that may break his module or just cause undefined behavior. If the module doesn't support history then don't let the user/admin choose to add history to nodes of that type. Craig ------------------------------------------------------------------------ Mon, 23 Aug 2004 19:23:29 +0000 : Nick Nassar I agree, there should be an API change to make specifying support for revisions easier. In the interests of keeping patches small and keeping to one change per patch, I think the API change should be a separate issue. A sort of ad-hoc API to decide whether or not a module supports revisions by default already exists. Instead of having a hook, modules set the default value of the "Create new revision" field in the edit form. The admin can change this option in admin/node/configure/defaults. This patch doesn't change that. Revisions are broken for node types that have their own database structure, like polls, even when storing them as serialized objects. This patch doesn't change that, either. ------------------------------------------------------------------------ Tue, 26 Oct 2004 02:35:06 +0000 : moshe weitzman I'm guessing that someone is going to have to demonstrate that this patch performs as well as current drupal before it gets comitted. i think this patch is a few benchmarks from being comitted. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:04:09 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:16 +0000 : Nick Nassar I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:26 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Mon, 15 Nov 2004 05:05:30 +0000 : elias1884 please overthink the revision system default workflow as well. don't look at the revision system as an isolated system but as a part of the whole workflow system! if you combine revisions with the moderation queue the most logic default workflow would be like that: auth user creates node (revision #0) admin approves the node (status = 1, moderation = 0) => node publicly available auth user finds typo and changes node (revision #1, status = 0, moderation = 1) ------- what happens at that point at the moment is, that the node is not accessible anymore at all until the new revision is approved by admin. of course the new revision should not go online until reviewed and approved, this is absolutely correct, but there is no reason to not take the old revision offline, since it was already approved and should therefore be online until the new revision is approved. it is not practical if a node disappears only because the author corrected a typo. ------- admin approves the node (status = 1, moderation = 0) eventhough I first thought a plain boolean active field would not be capable of providing that functionality if finally came to the conclusion, that it can. The only thing to do is to not set that bit, when a new revision is created, but when it is approved (in case moderation is activated under default workflow). Every revision should have its own moderation, status and active field and on approval they are set like this (status=1, moderation=0, active=Y). When you wanna rollback to an old revision, you can chose between all revisions that already have the moderation bit set back to 0 again and the published to 1. There should be an extra permission for rollback! another concern that I have about the default workflow is, that users can't see the content, they have just created, when moderation is enabled. Eventhough, there is a big fat "submission accepted" presented after submissions, unexperienced users tend to question the information those stupid tincans give them, if they can't find their content afterwards. Many users are really lazy bastards and they don't even read the status messages. The best feedback about whether his story was submitted successfully or not of course is, if he can find the story somewhere on the site, maybe with a status message on top of it, mentioning, that the content is currently not publicly available since it has not been approved yet. there should be a my content section under my account, like somebody is trying to do with the workspace module I guess. so my suggestion is to make (status=0, moderation=1) still available for the creator under a my content section somewhere! ------------------------------------------------------------------------ Wed, 24 Nov 2004 04:21:18 +0000 : Nick Nassar I agree. The current workflow for moderation queues and revisions needs to change, but this patch isn't the place for it. The patch is already too big, and it only does the backend stuff. Instead of adding more to this patch and making it take even longer to get into core, would you mind creating a new issue for your UI suggestions, so the those changes can be added as a separate patch? Thanks, Nick ------------------------------------------------------------------------ Sat, 11 Dec 2004 12:26:02 +0000 : Dries This patch is _much_ needed so I'd love to see someone revive it. In order for this patch to be accepted, the following needs to be done: * Update this patch to CVS HEAD. * Rename revid to vid. * Rename node_rev to node_revisions. * Rename node_rev.changed to node_revisions.timestamp. * Rename $rnode to $revision. * Fix the coding style to match Drupal's: proper spacing, single quotes where possible, proper variable names. * Benchmark this patch with a large database with enough revisions. I'd be happy to benchmark this on my local copy of the drupal.org database. * The book.log field should probably move to the node_revisions table. This can be done in a separate patch. * Investigate whether transactions are well-supported. ------------------------------------------------------------------------ Mon, 13 Dec 2004 00:25:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004-r… (11.02 KB) I've worked a bit on the patch (coding style issues as mentioned by Dries). One thing I noticed is that the patch uses REPLACE. IIRC this needs to be chagned to "UPDATE, if fail INSERT" for pgsql compatibility. Nick, are you still interested in working on that patch? I'd like to know how it works on your site and work on getting it into core. ------------------------------------------------------------------------ Mon, 13 Dec 2004 12:33:08 +0000 : Dries Gerhard: your patch does not apply. ------------------------------------------------------------------------ Mon, 13 Dec 2004 17:10:12 +0000 : killes(a)www.drop.org Yes, I know, that was the same version as I mailed to you earlier. ------------------------------------------------------------------------ Mon, 13 Dec 2004 21:02:06 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions.patch (52.96 KB) Ok, upüdated the patch to cvs. ------------------------------------------------------------------------ Tue, 14 Dec 2004 08:58:36 +0000 : Dries Some more comments: * db_begin_transaction() and db_end_transaction() do not belong in database.inc, but in database.mysql.inc and database.pgsql.inc respectively. * The node module calls node_revisionsision_list() which is not defined. (Fxed that on my local copy.) * Do db_begin_transaction() and db_end_transaction() deprecate Jeremy's table locking patch? * The upgrade path assigns the wrong user ID to each revision. * The upgrade path assigns the wrong date to each revision (that or a node's revision page shows the wrong usernames/dates). * The coding style needs a bit of work, but we can worry about that later. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:34:44 +0000 : Nick Nassar If you need any help getting those things fixed, just let me know. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:50:30 +0000 : Nick Nassar How this relates to Jeremy's node locking patch: There was lots of discussion, and node locking was decided against because from an end user point of view you never want a node to be locked. He's now advocating for a much simpler patch that warns users if their changes will overwrite someone elses. That patch still has a race condition, which might be fixed using db_begin_transaction(). http://drupal.org/node/6025 ------------------------------------------------------------------------ Tue, 14 Dec 2004 22:26:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_0.patch (55.96 KB) Here is an updated patch that tries to address Dries concerns. ------------------------------------------------------------------------ Wed, 15 Dec 2004 08:32:50 +0000 : Dries Attachment: http://drupal.org/files/issues/revisions-bug.png (76.06 KB) It didn't fix the aforementioned bugs. See attached screenshot. ------------------------------------------------------------------------ Thu, 06 Jan 2005 20:15:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_1.patch (51.77 KB) Ok, here is a new version. Dries and myself worked hart at it, so please have a look. what is still missing - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. ------------------------------------------------------------------------ Wed, 19 Jan 2005 21:43:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_2.patch (49.49 KB) Here is an updated patch. We discussed to keep the current title in node module and also in the revisiosn table. This is content duplication but will save many joins as many queries only need the title of a node. Discussion is welcome. ------------------------------------------------------------------------ Wed, 19 Jan 2005 23:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_3.patch (29.93 KB) I've implemented the aforementioned solution. This makes the patch much smaller. The patch now also removes taxonomy_node_has_term() which wasn't used anywhere. I'd really apprciate if some people could test drive the patch. It will be another huge improvement for 4.6. ------------------------------------------------------------------------ Thu, 20 Jan 2005 00:05:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_4.patch (30 KB) Another revision. Steven didn't like my literal $node->vid in queries. ------------------------------------------------------------------------ Thu, 20 Jan 2005 01:10:50 +0000 : killes(a)www.drop.org - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. These issues are still open, btw. Especially the first one needs to be tackled. ------------------------------------------------------------------------ Tue, 25 Jan 2005 20:11:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_5.patch (51.13 KB) Here is a patch that has the database tables updated for forum, book, and page module. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:55:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_6.patch (49.18 KB) Yet another update to keep it working with head. The patch now also removes the table definitons for the page table. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:57:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_7.patch (55.69 KB) Sorry, that was the old version, this is the right one. ------------------------------------------------------------------------ Mon, 31 Jan 2005 19:55:03 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_8.patch (55.71 KB) Updated once more. ------------------------------------------------------------------------ Mon, 31 Jan 2005 20:52:08 +0000 : Dries Anyone to help review/test this? ------------------------------------------------------------------------ Mon, 31 Jan 2005 21:22:36 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_9.patch (49.29 KB) Updated again, the update functions occurred twice. Thanks Bart. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:05 +0000 : killes(a)www.drop.org Don't know if the db I am using is corrupted or what. I still do have some didficulties. The latest patch is attached. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_10.patch (49.67 KB) I am probably slowly going mad ... ------------------------------------------------------------------------ Wed, 02 Feb 2005 01:54:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_11.patch (48.95 KB) The update issue still needs investigating. This patch is updated for cvs. ------------------------------------------------------------------------ Wed, 02 Feb 2005 20:20:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_12.patch (49.83 KB) Ok, here is a new version. I've solved my troubles with book.module. There are still some issues with forum module. Possibly due to inconsistent database. ------------------------------------------------------------------------ Wed, 02 Feb 2005 21:31:05 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_13.patch (49.83 KB) Turns out the drupal.org database had indeed some quirks. Please run this query in your oldest db and tell me the result: select nid,type from node where type like '%/%'; If you get a non-zero result we might need to add another security update. The patch could use still more testing, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 01:16:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_14.patch (49 KB) Ok, we are getting somewhere. At a first glance the update is working. There is a problem remaining: the revisions tab will be shown whether the node has revisions or not. Not sure we can/need to fix this. People with a drupal.org account can log in at http://killes.drupaldevs.org/revision/ and poke around. Your permissions will be the same as on drupal.org. Feel free to vreak everything but don't forget to file complaints here. (Note: this is only a pruned version of the drupal.org database with all project nodes and nodes with nids > 7000 dropped). ------------------------------------------------------------------------ Thu, 03 Feb 2005 04:19:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_15.patch (52.39 KB) There was some error in node_save and also the patches to the database.inc files got lost... ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:07:27 +0000 : robertDouglass Submitting book pages doesn't work on your test site. It puts the entire content of the preview inside the body textarea. I wrote a sentence in the body and the log, and pressing preview put several lines of HTML containing both sentences in the body textarea on the preview page, plus the book page wouldn't submit. -R ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:50:59 +0000 : Junyor 0 results here. I started using Drupal with version 4.4, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 23:56:18 +0000 : killes(a)www.drop.org @Junyor: Thanks, that's a good sign. Maybe somebody else has an older db to try. @robertDouglass: The first effect you describe is due to drupaldevs running on PHP 5. I am unsure why the second thing does not work. In node_save() the node object has a nid although there is none in the form. Very strange. I've enabled display of db queries on the testsite. ------------------------------------------------------------------------ Fri, 04 Feb 2005 19:17:55 +0000 : dmjossel No results here on the query: select nid,type from node where type like '%/%'; On a database that was put in place prior to Drupal 4 and is now running on 4.5.2. ------------------------------------------------------------------------ Fri, 04 Feb 2005 20:44:23 +0000 : killes(a)www.drop.org @dmjossel: thanks. @all. The strange problem I reported was apparently php 5 related. After applying Steven's php 5 patch it went away. One error is remaining: If I create a new forum topic it will be shown as part of the book on preview. Hmm, that was due to a db that got corrupted during testing so that is fixed too. Please poke around at the test site and look if you find more errors. ------------------------------------------------------------------------ Sat, 05 Feb 2005 07:16:22 +0000 : Steven By the way, I just remembered that Drupal.org has some blogs lingering on in the database even though blog.module is not enabled. Perhaps this is causing troubles? ------------------------------------------------------------------------ Sat, 05 Feb 2005 11:22:59 +0000 : Anonymous I can't see why it would. Drupal.org will need extra updates for images and project nodes because those have their own tables. GK. ------------------------------------------------------------------------ Sun, 06 Feb 2005 12:49:55 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_16.patch (52.49 KB) Updated to apply to cvs again. ------------------------------------------------------------------------ Tue, 22 Feb 2005 20:15:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_17.patch (49.64 KB) Updated again. All we need is a patch to upload module and an upgrade path for it. ------------------------------------------------------------------------ Fri, 04 Mar 2005 04:22:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_18.patch (52.31 KB) Updated once more. Moved log field from book to node_revisions table as discussed in Antwerp. upload module still missing. We need to decide under which circumstances the log field should be displayed. Should that be added to the workflow? Should it depend on the revisions setting? ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:27:03 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_20.patch (75.52 KB) Ok, here it is: Yet another revision of this grrrrreat patch. Changes from previous versions: - supports versioning for uploaded files. A problem is that if you delete a file, it will be gone for all revisions. - the log field is now in the node_revsions table, but each module has to decide whether to show it or not. I've implemented it for the page and the book type odes. Also, the field can be edited when adding non-book nodes to the book. The log is displayed on the revisions page and if a node is moderated. - the revisions are moved to an old_revisions table to a) get the node table smaller and b) still leave the mavailable for contrib modules that want to retreive old version data. The patch has been applied to killes.drupaldevs.org/revision where it can be tested by anybody (especially people who have "site admin" rights on drupal.org) The database is from drupal.org and you should b able to log in with your pass or simply mail yourself a new one. Gerhard ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:51:56 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_21.patch (59.42 KB) BTW, I marked this a bug because atm the revisions field can grow quite big. Neil has reported problems from some users who were not able to load some nodes due to to many large revisions. Also, som unrelated stuff crept into the patch. New version attached. ------------------------------------------------------------------------ Tue, 08 Mar 2005 05:56:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_22.patch (60.29 KB) Ok, I think I got it. Changes to last version: - uploads are no properly versioned. Missing are still pgsql checks and updates. ------------------------------------------------------------------------ Thu, 10 Mar 2005 16:58:41 +0000 : Anonymous Was able to get http://drupal.org/files/issues/revisions_21.patch to work with drupal-cvs.tar.gz (10 March 2005) by: - includes/database.mysql.inc: Commenting out duplicates for functions function db_begin_transaction and function db_commit_transaction - modules: node.module: Removing "'title' => $node->title," from $node_table_values variable declaration and removing "'title' => "'%s'"," from "$node_table_types" variable declaration. Happy to submit a patch if requested. I'll watch this thread. ------------------------------------------------------------------------ Fri, 11 Mar 2005 23:59:45 +0000 : killes(a)www.drop.org The duplicate function has been removed in rev 22 of this patch. Why do you think the changes in node_save are needed? Titles are saved in both tables for performance reasons. ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:12:21 +0000 : jlerner Hi - I posted comment #62. The changes to node_save appear to be needed because recent patches (both 21 and 22) remove the field 'title' from table 'node'. So without the changes to node_save, node.module is broken and generates errors. Joshua ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:29:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_23.patch (61.17 KB) Thanks, Joshua, for catching this. node:title is there to stay. ------------------------------------------------------------------------ Wed, 13 Apr 2005 16:29:23 +0000 : moshe weitzman since HEAD is open again, perhaps it is a good time to revisit this patch. once this is committed, lets address - http://drupal.org/node/11071 "node_validate does not respect group editing" ------------------------------------------------------------------------ Mon, 18 Apr 2005 15:43:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Mon, 18 Apr 2005 16:16:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24_0.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:19:42 +0000 : Dries I'll commit this patch later this week! If you haven't checked this patch already, I urge you to test/check it out because it will have significant impact on existing code and modules! ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:21:33 +0000 : Dries Also, what do people think about the n.title being duplicated? ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:26:58 +0000 : chx I won't lose any sleep because of duplicated titles... ------------------------------------------------------------------------ Tue, 19 Apr 2005 18:35:58 +0000 : killes(a)www.drop.org Let me explain why I have chosen to duplicate the title (and also the uid): If you look at all the queries in Drupal, you will find that most of them only need the title and th uid of a node. That is, by duplicating it, we save expensive joins on the node_revisions table. Due to this fact, this patch is actually a performance improvement. A note about updating contrib module: Strictly speaking they wouldn't need to be updated. They only need to if their authors decide that their info should be available for revisioning. The upgrade path for forum.module in my update.inc patch (+ the forum patch) should show you what needs to be done. I will write a note for the update page once the patch hits core. ------------------------------------------------------------------------ Sun, 24 Apr 2005 21:21:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_25.patch (60.38 KB) Updated to cvs. Dries: Based on some remarks in #drupal this is the last update I am going to do. Apply it or won't fix it. ------------------------------------------------------------------------ Sat, 30 Apr 2005 03:42:39 +0000 : Jeremy(a)kerneltrap.org Attachment: http://drupal.org/files/issues/revisions_25.patch.patch (528 bytes) That's a big patch. I've only started looking through it. I noticed one little typo, affecting updates. A patch to your last patch is attached. I'm running with the revision patch on my dev server now happily. I like the concept. What happens if you click 'stop' on your browser in the middle of a MySQL "transaction"? I assume that kills the connection to MySQL, and the lock is freed? But this then leaves changes only partially applied? What exactly does locking/unlocking the tables buy us in MySQL? I don't see anywhere that we detect if an apply fails part way through, and thus roll back...? What am I missing? ------------------------------------------------------------------------ Sat, 30 Apr 2005 07:11:28 +0000 : Dries Jeremy: many of us are worried about the performance ramifications this patch introduces. Early experiments showed a small performance improvement (while a performance regression might be expected). More performance reports from large sites like kerneltrap.org will certainly help this patch. Mind to do a quick performance comparision and to report back with some numbers? Thanks. ------------------------------------------------------------------------ Sat, 30 Apr 2005 12:38:02 +0000 : Jeremy(a)kerneltrap.org Dries: I'm not running HEAD on kerneltrap, so this really isn't a possibility. Furthermore, until I understand why we're locking tables, I don't like it. The idea of revisions in their own tables is great. The idea of locking tables to get (without any obvious benefit) there really worries me. ------------------------------------------------------------------------ Sat, 30 Apr 2005 14:16:01 +0000 : killes(a)www.drop.org @Jeremy: Thanks for looking at the patch! Also for catching the typo. :) Did you try to upgrade your database? If yes, how did it go? One of Dries' concerns is the complexity of the upgrade. How many nodes and revisions did the db have? About database locking: This part of the patch was created by Nick and I simply continued to use it. Maybe the code should rather be: if(db_begin_transaction(array('{node}', '{node_revisions}', '{watchdog}', '{sessions}', '{files}'))) { db_query($node_query, $node_table_values); db_query($revisions_query, $revisions_table_values); db_commit_transaction(); ... } The idea is probably to avoid two updates at the same time. I don't think the locking helps if you abort the script at an inconvenient time. Rollbacks aren't implemented in all mysql versions. We could omit the db locking if deemed inappropriate. Maybe Nick can explain his ideas behind this. @Dries: I wonder who the "many of us" are. They certainly haven't spoken to me. Moshe had some reservations about the upgrade path and project module, but the time that project module abused revisions to store issue updates was long ago and his reservations were resolved. Nobody else (besides you of course and now Jeremy) has voiced reservations in a way that was audible to me. If you grep through the patch you will notice that there are only four queries which have a join on the node_revisions table. Two of them are in node_load and in the other cases the join replaced a join on the node table. The two queries in node_load are the only ones that have both a join on the node and the revisions query. Thus, loading of individual nodes might become somewhat slower. All other queries will be faster since the node table is now much smaller. Also, node loading does not have to be slower, it depends on your node table. If you had a lot of revisions and thus a large table, then the new scheme will make your queries actually faster since we do not load the revisions on each and every node load anymore. If you didn't have many revisions your node_load migth be somewhat slower. WRT to the update script Karoly pointed out that we could use multiple insert queries instead one query per revision. This would probably make the update somewhat faster. I am willing to work on this iff you declare that you will commit the patch afterwards. I'd need to know if this will work on pgsql and on all supported mysql versions before I work on it. About locking: Database locking is dog slow, at least on mysql. I was using locks in an earlier version of the upgrade script but had to remove it for (serious!) performance reasons. ------------------------------------------------------------------------ Mon, 09 May 2005 15:07:34 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_26.patch (46.45 KB) Ok, another update, cause I need it myself. I've left out the transaction stuff for now. It is in principle unrelated to this patch and should be discussed elsewhere. This also makes the patch smaller and easier to review (hint, hint). ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:09 +0000 : killes(a)www.drop.org The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_27.patch (39.05 KB) The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 21:23:06 +0000 : Dries Got one error during the upgrade path: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ------------------------------------------------------------------------ Mon, 09 May 2005 21:26:19 +0000 : killes(a)www.drop.org This had happend to me as well, when I tested this patch. The reason is that for some reason the vid is not unique. Most likely there are some entries with vid = 0 in there. Can you check which node types those have? it always was an error in the test database. See: http://drupal.org/node/7582#comment-20678 ------------------------------------------------------------------------ Mon, 09 May 2005 21:27:06 +0000 : Dries Actually, I got 2850 errors during the upgrade. Some of these: sprintf() [function.sprintf]: Too few arguments in drupal-cvs/includes/database.inc on line 154. Some of these: Query was empty query: in drupal-cvs/includes/database.mysql.inc on line 66. And this: Unknown table 'n' in field list query: SELECT n.nid, n.vid FROM node INNER JOIN files f ON n.nid = f.nid in drupal-cvs/includes/database.mysql.inc on line 66. :-) ------------------------------------------------------------------------ Mon, 09 May 2005 21:29:19 +0000 : Dries Or this: user error: Unknown column 'log' in 'field list' query: SELECT parent, weight, log FROM book WHERE nid = 1 in drupal-cvs/includes/database.mysql.inc on line 66 ------------------------------------------------------------------------ Mon, 09 May 2005 21:52:12 +0000 : Dries The time required to generate my main page went from 902 ms (before upgrade) to 2139 ms (after upgrade). The time required to generate a forum listing (?q=forum/x) went from 1872 ms (before upgrade) to 2874 ms (after upgrade). Maybe this is because my database is not consistent as the result of the upgrade errors (yet I don't see any errors on the pages I benchmarked). ------------------------------------------------------------------------ Tue, 10 May 2005 00:24:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_28.patch (53.47 KB) Ok, let me get to this from the bottom to the top: - my test runs indicated a different development wrt timing. If I had gotten your results, I had stopped working on this long ago. So your results are wrong for some reason. - user error: Unknown column 'log' in 'field list' Wasn't my day, the book patch got lost. It is contained now. First -R the old patch, then apply this one. - Unknown table 'n' in field list query: Walkah found this, but I forgot to fix it. Fixed now. - I've no idea where the other queries come from. I am hoping that either your test db is borken or they are follow ups from the other ones. If you let me have your test db, I'll try some debugging. Thanks for wasting your time, too. ------------------------------------------------------------------------ Tue, 10 May 2005 05:07:31 +0000 : Dries I double-checked and the numbers don't seem to lie. I'll test some more after work on another machine to make sure it is not platform-specific. ------------------------------------------------------------------------ Wed, 11 May 2005 03:32:47 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_29.patch (54.83 KB) Ok, here I am again. What I did: 1) Ask Dries to let me have drupal.org database 2) get 400MB of SQL inserts... 3) take 23 minutes to import said data 4) Remove all image and project nodes (don't want to install their modules), 11765 nodes left 5) back up data 6) take tests on non-cached /node page (as anonymous user). ab results: -c 1 -n 25: Requests per second: 1.29 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 663 775 179.7 689 1264 7) Do the same for the tracker page: Requests per second: 0.83 [#/sec] (mean) Total: 1182 1199 7.4 1199 1217 8) Apply my patch (rev. 28). 9) run db update and hold breath 10) update times out... 11) play back backup from 5) 12) wait 13) getting annoyed and removing cache, watchdog, and accesslog before playing in backup. 14) wait again. Understand why Dries doesn't try this patch often. Maybe a smaller DB would do for testing? 15) wait more. get really annoyed. 16) Set time limit to 18000 in update.php 17) try again 18) fails again before the second update is completed. 19) curse. 20) delete search stuff from db. Ooops, sooo much smaller... 21) import again, below 2 minutes... 22) rewrite to use extended insert. Found a bug. 23) still does not complete. Mysql logging to the rescue! 24) tid = 0? Not good. 25) Well, the update works fine till node 10834. 5595 nodes done, 6136 to go. 26) Writing shell based update script. Discovery: 24MB aren't enough. Hopefully 64 are. Nope. extended inserts for revisions are apparently not the brightest idea: Huge memory consumption. Hmm, no, all updates got through. Selecting the revisions to put them into old_revisions table screwed it. Learned about CREATE TABLE old_revisions SELECT syntax. Yay! finally. 24 MB are just not enough the update.php script seems to still break. 27) Benchmarks! /node Requests per second: 1.54 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 632 649 40.5 636 791 /tracker Requests per second: 0.86 [#/sec] (mean) Total: 1119 1165 65.8 1160 1461 Ok: So we get an improvement for many node_loads, but none for simple selects from node. More tests can be done. 28) roll new patch Ain't Drupal fun? ------------------------------------------------------------------------ Wed, 18 May 2005 13:38:05 +0000 : Dries I did another round of tests on _another_ machine and it is not looking good: Before upgrade After upgrade ?q= (main page) 218 ms/request 340 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request ?q=about (book page) 375 ms/request 5400 ms/request The upgrade process itself gave me a number of 'query was empty' and 'sprintf(): too few arguments' reports. Everything seems to work fine though. Looking at the ?q=about page, I see that the following query is executed twice _and_ that each time, it take more than 2 seconds to complete: SELECT n.nid, n.title, b.parent, b.weight FROM node n INNER JOIN book b ON n.vid = b.vid WHERE n.status = 1 AND n.moderate = 0 ORDER BY b.weight, n.title; --8 SHOW INDEX FROM book; +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | book | 1 | book_parent | 1 | parent | A | 92 | NULL | NULL | | BTREE | | | book | 1 | nid | 1 | nid | A | 369 | NULL | NULL | | BTREE | | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ 2 rows in set (0.00 sec) The book module does not appear to have a primary key? Sounds like a bad idea so I added one: mysql> ALTER TABLE book ADD PRIMARY KEY nid (nid); Query OK, 369 rows affected (0.02 sec) Records: 369 Duplicates: 0 Warnings: 0 Next, I wanted to make the vid column a unique key in all node tables: mysql> ALTER TABLE node ADD UNIQUE vid (vid); Query OK, 20392 rows affected (0.81 sec) Records: 20392 Duplicates: 0 Warnings: 0 mysql> ALTER TABLE book ADD UNIQUE vid (vid); ERROR 1062: Duplicate entry '0' for key 2 mysql> ALTER TABLE forum ADD UNIQUE vid (vid); Query OK, 10806 rows affected (0.10 sec) Records: 10806 Duplicates: 0 Warnings: 0 As you can see, it fails for the book table which makes me believe there is some inconsistent data ... I set out to fix that: mysql> SELECT nid, COUNT(nid) AS vids FROM book GROUP BY vid HAVING vids > 1; +-----+------+ | nid | vids | +-----+------+ | 871 | 2 | +-----+------+ 1 row in set (0.00 sec) mysql> SELECT title FROM node WHERE nid = 871; Empty set (0.00 sec) mysql> DELETE FROM book WHERE nid = 871; Query OK, 1 row affected (0.00 sec) mysql> ALTER TABLE book ADD UNIQUE vid (vid); Query OK, 368 rows affected (0.01 sec) Records: 368 Duplicates: 0 Warnings: 0 Looks like everything is well now. Ran some new benchmarks: Before upgrade After upgrade With indices ?q= (main page) 218 ms/request 340 ms/request 336 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request 1531 ms/request ?q=about (book page) 375 ms/request 5400 ms/request 475 ms/request Unfortunately, we're still slower than the original code. ------------------------------------------------------------------------ Wed, 18 May 2005 21:53:31 +0000 : killes(a)www.drop.org Dries, thanks for testing it again. I do think the broken queries you observer have something to do with the bad performance after the update. Please log the queries any I will have a look at them. I've never seen any such queries. My update script also tries to create the appropriate indices, but it will of course fail if the database contains cruft. the indices for the forum are probably missing, too. I am still convinced that the patch is core worthy. ------------------------------------------------------------------------ Thu, 19 May 2005 04:36:09 +0000 : Dries It wouldn't hurt if more people would benchmark this patch. The patch's current performance worries me. Did you check your watchdog messages after upgrading the drupal.org database? Depending on your settings, errors might only be shown in the watchdog. I'll look into the remaining glitches as time permits. Thanks for your persistence in keeping this patch up-to-date. :) ------------------------------------------------------------------------ Thu, 19 May 2005 11:59:22 +0000 : killes(a)www.drop.org Dries: Can you please let me have your updated database? I want to have a look at it and try my own benchmarks with it. And yes, if I did learn something on this project is how to be persistant. ;) ------------------------------------------------------------------------ Fri, 24 Jun 2005 16:25:34 +0000 : killes(a)www.drop.org Here is an idea that occurred to me: The problem with the upgrade process is that keeping the existing revisions requires a lot of work to do. This generates a huge amount of sql queries for a large database and also requires a huge amount of memory. My suggestions is to let update.php only handle the basic upgrade, ie without old revisions. An additional module could be created that would implement a cron based approach to upgrading old revisions one node at a time. it could expose a hook to let contrib modules do their own upgrades. Dries, what do you think? (I am writing "Dries" because he seems to be the only one who is interested in getting this into core...) ------------------------------------------------------------------------ Fri, 24 Jun 2005 22:25:11 +0000 : Junyor Killes: I'm also interested in seeing this hit core. What about adding something to legacy.module to do it? ------------------------------------------------------------------------ Sun, 26 Jun 2005 21:14:54 +0000 : chx This is a sensible approach. Maybe this is the _only_ sensible approach. I have a little problem though: while the conversion is running somehow both revision handlers should be available. ------------------------------------------------------------------------ Sun, 26 Jun 2005 22:16:21 +0000 : killes(a)www.drop.org hehe, one only has to whine and bug enough and one gets some feedback. ;) @junyor: legacy.module would be a good place. my current idea is to auto-enable it in update.php and then disable it again in legacy_cron after all nodes are updated.. ;) @chx: When somebody wants to look at revisions of a node that node could be auto-updated. The only problem are contrib modules: they's need to have some hook in order to update their own data. When somebody looks at the revisions of a node than cannot be updated because the contrib module in question has no such hook, we can optionally let the user discard old revisions I guess. Dries, what do you think? ------------------------------------------------------------------------ Mon, 25 Jul 2005 16:48:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_30.patch (53.13 KB) Sooooo; I've updated this patch once again. Dries didn't like my idea of legacy updates so we will have an option to discard old revisions in case the update should prove difficult. Always keep a backup. ------------------------------------------------------------------------ Mon, 25 Jul 2005 17:31:12 +0000 : Bèr Kessels can we please accept and commit this patch? We can iron out any issues later. This patch is just far too big and complex to be perfect-at-once. ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:22:50 +0000 : Dries @Ber: no, we can't commit this patch blindly. Data loss is a much bigger problem than a syntax error or other code glitch. We can break Drupal but we can't break their data. I've spent quite a bit of time testing the previous version of this patch and noticed significant performance degradation. Tell me, Ber, why can't you test this patch first? ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:40:37 +0000 : Bèr Kessels I was not referring to not testing it. If it is just the upgrade path that proves to be cumbersome, then why can we not fix it afterwards, I.e. when everyone has had a good chance to look at it. Testing such a huge patch requires a lot of work. something no-one just does in his spare hours. We had discussions before to deal in a slightly different way with big changes; to commit them quicker and to leave the ironing out of any left overs for the community. I hooked into that discussion here. For two reasons. One is that Gerhard has spent numerous hours on maintaining this patch. The other one is that the community can be of help with ironing out issues in such a large change, much better than that Gerhard can do on his own. And yes, dataloss is very bad, but no-one should loose data, when he/she followed the instructions (backing up)... ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:00:38 +0000 : Dries Ber: applying this patch takes 15 seconds. Whether I apply this patch for you, or whether you apply it yourself, it will hardly reduce the 'testing cost'. The problem is that data loss can be subtle; it might go unnoticed for a couple days. Make no mistake, I'd like to see this patch committed ASAP, but it warrants some testing. Let's test/benchmark it and commit it. ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:55:03 +0000 : drumm The upgrade script already takes a long time to execute and does not provide feedback to the user about how it is doing. I plan on making the upgrade script able to spread the updates across multiple page views and give user feedback showing that progress is being made. This will hopefully make the speed of this update a moot point. I am working on this for my dayjob, CivicSpace, so it should get done "real soon now", but it should be expected to take awhile (smallish number of weeks). Please make comprimises if necessary. ------------------------------------------------------------------------ Wed, 27 Jul 2005 17:38:57 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_31.patch (48.78 KB) @drumm: the timeout isn't usually the problem. Memory consumption is. I'll look into doing the updates in chunks of 1000 nodes or so. Clousseau found a bug in the patch, updated. ------------------------------------------------------------------------ Thu, 28 Jul 2005 18:28:08 +0000 : moshe weitzman Even if the upgrade issues are resolved, we have to figure out why this patch is slowing down drupal (according to benchmarks posted here). ------------------------------------------------------------------------ Fri, 29 Jul 2005 07:51:45 +0000 : Dries We can't proceed without some additional benchmarking. I already benchmarked it on two machines (see comments above), and I'll benchmark it again after August 10. The performance impact of this patch worries me so if two other people could test and benchmark this patch extensively, that would come a long way. August 10 is close to the code freeze so some help is necessary. If Gerhard/killes reviewed/tested one of your patches, it is time to return the favor ... ------------------------------------------------------------------------ Fri, 29 Jul 2005 15:40:31 +0000 : Jose A Reyero After looking at the patch I'm not really surprised this slows down everything. I thought that the reason why we wanted revisions in their own table was in the first place to have a simpler -and faster- node table. But this patch: - Adds fields and complexity not only to the node table but also to all the main tables for data - Needs additional joins just to retrieve single nodes (which is done many times for a typical main page) - Does away with encapsulation of the revision functionality requiring other modules to handle revissions related data. So please, please, please: - Make revisions table only needed when actually using revisions - Keep that old nice thing of hiding revisions system from other modules - And what's so bad with serializing data anyway? I agree that usually there are better ways to store data. But this is one of the cases where serialization does make sense. Not the current huge serialized field, but only a new table with one serialized node per row. - In general, do it much simpler. This is way too complex and it removes more functionality than it adds. In just four words: keep it simple, please. ------------------------------------------------------------------------ Fri, 29 Jul 2005 21:25:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_32.patch (51.27 KB) Patch updated (v31 was buggy, and there was an update to updates.inc). I've had a discussion with Jose and he will maybe do some testing over the weekend. John is also doing some testing as we speak. ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:06:59 +0000 : Nick Nassar It has been exactly one year since I first submitted this patch. You know what that means: *old patch party in IRC!* Virtual beers on me! w00t! ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:08:33 +0000 : Jose A Reyero Hi again. I had a talk with killes -I have to say he enlightned me about this- and also did some testing. Sorry but I dont have a good set up to provide benchmarks. Here are my conclusions -and also some explanation for the varied results abt performance: - This is faster for simple node listings - This is slower for full node/load ....thus the performance increase/decrease for a given page should depend on the relation between the number of simple node listings (usually blocks) and the number of nodes displayed in the main page. I've also experienced some trouble with attachments -not re-created when creating a new revision- and book relations, same. I'm not sure I like this new "feature" of creating a new revision every time you rollback an old one -in Drupal 4.6 this seems not to happen. Besides these details, I think this patch is good enough -and quite a powerful approach actually- though I still feel it's bit too complex when simply dealing with nodes -without revisions-, which is what we do most of the time, and complexity will be even higher with contrib modules using still more tables, not to mention flexinode -versioning info will have to be scattered everywhere in the db. So this is not a +1 nor -1. On one side I think this is powerful. On the other side I feel this is too complex and really, it's more than I need. I would be happy with a 'revision' table having: nid, vid, data(serialized), that you can forget about when not using revisions. ------------------------------------------------------------------------ Sat, 30 Jul 2005 19:16:41 +0000 : Nick Nassar On a much more serious note that my last comment, unless I'm missing something, you've introduced a subtle bug into this patch since I left it. A and B are both updating revision 5 of a node. A calls db_next_id() and gets a revision id of 6 B calls db_next_id() and gets a revision id of 7 B updates the node's revision to 7 B adds revision 7 to the revision table A updates the node's revision to 6 A adds revision 6 to the revision table Now, you have a situation where the current revision (6) isn't actually the last revision(7). For some uses, that's not a big deal, but in the case of wikis, which is what I originally hoped to use this patch for, it creates problems. ------------------------------------------------------------------------ Mon, 01 Aug 2005 00:04:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_33.patch (51.27 KB) Updated the patch. updates.inc was changed and chx found a typo. @Nick: I don't think the issue is due to this patch. Drupal has currently no mechanism to not let people commit a new version of the same node, whether with or without revisions. drupal_next_id ensures that the revision IDs are unique and the later revisions will be the current one. Which revision is the current one does not depend on order of the version ID, btw. @Jose: what exactly were your problems with uploads? what do I need to do to reproduce? ------------------------------------------------------------------------ Mon, 01 Aug 2005 11:33:56 +0000 : Jose A Reyero Sorry, I dont know what I did the other day but I cannot reproduce it either :-) I applied this new patch. Problems I'm having now (Testing with "stories") - Book outline is gone after creating a new revision - I cannot delete attachments (not sure it is related to your patch or to something else) - Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I? ------------------------------------------------------------------------ Sun, 14 Aug 2005 16:03:41 +0000 : killes(a)www.drop.org Jose, thanks a lot for testing! "- Book outline is gone after creating a new revision" Sounds like a bug, it should not. "- I cannot delete attachments (not sure it is related to your patch or to something else)" Another bug "- Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab" Log messages are possible for nodes of type book and page as well as all nodes that are put into the book. Log messages can be added for any node type, the module author should implement those. "Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I?" No, the db update should be the same. I will try to fix the bugs you found. WRT the db update I've got an announcement to make: I will not migrate old revisions. Why? 1) it is a rather complicated process that requires a lot of memory and is likely to give people on cheap hosting trouble. 2) testing it takes a lot of time. 3) I don't personally need it I will still move old revisions to an old_revisions table from where they can be converted by a later update or a custom script. ------------------------------------------------------------------------ Tue, 16 Aug 2005 21:34:43 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_34.patch (52.92 KB) Ok, here is incarnation No. 34 of this patch. Dries agrees with me that we can put the revision rescue operation into a later update, possible using Drumm's installer. The patch fixes the issue with the file uploads not being deletable, I could not reproduce the problem with the outlines. ------------------------------------------------------------------------ Sat, 20 Aug 2005 13:33:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/scripts.tar.gz (4.39 KB) I've started to do some performance testing. I created a script that creates nodes that have revisions according to the new system. A node has a 50% chance to have a revision and then an equal chance fro between 1 and 25 revisions. I've created a database with 50 nodes and 500 comments, 7 vocabs and 100 total taxo terms. I've also created a script to convert these revisions to the old format (much easier than the other way around). Initial tests didn't show a significant change for /node. I'll try with more nodes and /tracker. Attached are the two scripts in a tarball. They need to be run against the patched install. Also included are the two preliminary results. ------------------------------------------------------------------------ Sat, 20 Aug 2005 15:16:04 +0000 : killes(a)www.drop.org I've now created a databse which has 1000 nodes and each node has 50 or 51 revisions. I have no taxo terms assigned and no comments (those parts didn't change). I am only using stories and pages. I am now fairly confident that this patch is a performance improvement for this kind of setup. for the new patch on the /node page I get Requests per second: 0.27 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 3592 3648 56.9 3640 3985 Waiting: 3573 3626 55.1 3620 3949 Total: 3592 3648 56.9 3640 3985 for the old one I get: Requests per second: 0.23 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 4247 4297 64.0 4286 4657 Waiting: 4156 4263 65.3 4253 4624 Total: 4247 4297 64.0 4286 4657 For the tracker the results are even better: old: Requests per second: 2.00 [#/sec] (mean) new: Requests per second: 3.64 [#/sec] (mean) Anybody interested in obtaining this database can mail me. It is 60MB (gzipped). ------------------------------------------------------------------------ Sat, 20 Aug 2005 16:33:26 +0000 : killes(a)www.drop.org Ok, I've now deleted the revisions and only kept the current one. The initial tests showed that the performance of the patch was miserable. then I remembered that I had shreddered the revisions table and did a mysql optimization. Now the results are as follows: /node: Requests per second: 3.19 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 275 312 44.3 290 499 Waiting: 255 291 44.1 268 479 Total: 275 312 44.3 290 499 Requests per second: 3.18 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 274 314 42.4 292 407 Waiting: 255 294 42.0 273 388 Total: 274 314 42.4 292 407 The lower one is the new system, there is no difference considering statistics. /tracker old: Requests per second: 3.72 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 238 268 33.8 249 332 Waiting: 219 247 33.3 230 313 Total: 238 268 33.8 249 332 new: Requests per second: 3.77 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 240 264 29.8 250 324 Waiting: 221 244 29.0 229 305 Total: 240 264 29.8 250 324 Same applies, no real difference. So, as soon as somebody tests the upgrade path the patch should be committed. ------------------------------------------------------------------------ Sat, 20 Aug 2005 17:18:39 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_35.patch (55.51 KB) Ok, I coudl hear them in advance "1000 nodes without revisions are not sufficient". Well, here are results for 30000 nodes without revisions. /node new: Requests per second: 0.45 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2162 2202 58.9 2188 2485 Waiting: 2143 2182 54.1 2169 2413 Total: 2162 2202 58.9 2188 2485 old: Requests per second: 0.43 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2319 2346 77.6 2332 2761 Waiting: 2223 2324 79.0 2311 2742 Total: 2319 2346 77.6 2332 2761 One could argue that the new approach s slightly faster. I'd like to see mroe results to believe it, but it is pretty clear it isn't slower. /tracker new: Requests per second: 0.41 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2330 2418 116.9 2403 3194 Waiting: 2302 2397 117.7 2382 3175 Total: 2330 2418 116.9 2403 3194 old: Requests per second: 0.39 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2496 2548 50.1 2536 2841 Waiting: 2477 2528 50.0 2518 2822 Total: 2496 2548 50.1 2536 2841 Statistically insignificant differences. All those tests where done with ab and as user No 1. model name : Intel(R) Celeron(R) CPU 2.40GHz 512MB RAM Besides the differences for revisions the change in served pages with different configurations is interesting. A new version is attached. node_save now returns the complete node (not only the nid). ------------------------------------------------------------------------ Mon, 22 Aug 2005 20:13:52 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_36.patch (55.66 KB) Ok, here is version 36 which incorporates today's changes in cvs. It also changes the upgrade path. I was trying my old solution and had to up my php mem limit to 80 MB in order to get it to run through. I hope this one is better but it needs testing. ------------------------------------------------------------------------ Mon, 22 Aug 2005 22:46:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_37.patch (52.29 KB) Ok, I've now moved _all_ of the heavy lifting to mysql. pgsql need to catch up. No problems with either time out or memory limit anymore. Since pgsql plays catch-up all the time, I set it to "ready to be committed". ------------------------------------------------------------------------ Tue, 23 Aug 2005 11:45:59 +0000 : stefan nagtegaal I could apply the patch on my localhost, and did not seen any errors nor disfunctioning drupal features while creating, revising, editing or deleting nodes.. So, +1 for this patch qua implementation it makes sense. Unfortunatly I chaven't got a clue howto benchmark, so that's for another person. @Gerhard: I really think you are a great person by rewriting this patch over and over again, untill Dries decides it is good enough to get in to the trunk. I doubt about the fact if there are more persons in our community who are like you.. People who would revise their patch 37 times, and even after that still open for discussion.. Gerhard, I think you are a great example of how a real coder should do it.. I think we can all learn from you.. ------------------------------------------------------------------------ Tue, 23 Aug 2005 17:15:29 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_38.patch (52.26 KB) Karoly pointed out that node_save(&$node) is nicer. @Steef: nice that you write that, but a _really_ good coder would only have needed one revision. ;p ------------------------------------------------------------------------ Wed, 24 Aug 2005 15:36:03 +0000 : Dries function book_update($node) { - db_query("UPDATE {book} SET parent = %d, weight = %d, log = '%s' WHERE nid = %d", $node->parent, $node->weight, $node->log, $node->nid); + if ($node->is_new || $node->revision) { + db_query("INSERT INTO {book} (nid, vid, parent, weight) VALUES (%d, %d, %d, %d, '%s')", $node->nid, $node->vid, $node->parent, $node->weight); + } + else { + db_query("UPDATE {book} SET parent = %d, weight = %d WHERE vid = %d", $node->parent, $node->weight, $node->vid); + } } Why 'is_new' and 'revision'? Isn't 'is_new' redundant? I loaded a fresh copy of the drupal.org database on my test system. How do I test this? Where is the upgrade script/path? ------------------------------------------------------------------------ Wed, 24 Aug 2005 16:31:09 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_39.patch (52.25 KB) is_new is indeed redundant since it is handeled in _update. New patch attached. The upgrate is as usual in updates.inc and accessed through update.php. ------------------------------------------------------------------------ Thu, 25 Aug 2005 10:35:24 +0000 : jakeg This is probably a little off topic but I can't find info elsewhere and this may affect revisions changes. It appears as though all revisions of a node are accessible to all users (inc. anonymous users) using ?revision=x. Surely only the 'current' revision should be accessible by default, and there should be a permission that can be assigned to roles to view non-current revisions? I don't want regular users looking at non-current revisions! A revision may have been made for various private reasons. Only admin should be able to do that. ------------------------------------------------------------------------ Thu, 25 Aug 2005 13:26:12 +0000 : Dries On the latest snapshopt of the drupal.org database I get this: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {forum} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {node_revisions} ADD KEY nid (nid) FAILED ALTER TABLE {node_revisions} ADD KEY uid (uid) FAILED ALTER TABLE {node_revisions} ADD PRIMARY KEY vid (vid) FAILED CREATE TABLE {old_revisions} SELECT nid, revisions FROM {node} WHERE revisions != '' OK ALTER TABLE {book} DROP log OK ALTER TABLE {node} DROP teaser OK ALTER TABLE {node} DROP body OK ALTER TABLE {node} DROP format OK ALTER TABLE {node} DROP revisions OK ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:30:11 +0000 : killes(a)www.drop.org That's strange, it worked for me with an older version of the same db. Anyway, I've re-rolled the patch and rearranged the update queries a bit. Also everything is now in one update, thanks to Neil for pointing that out. ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:32:17 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_40.patch (52.22 KB) and the patch. ------------------------------------------------------------------------ Sun, 28 Aug 2005 02:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_41.patch (52.4 KB) I've tested the upgrade path on an older copy of the drupal.org db and found and squished two bugs. The test cycle can be shortened a bit by deleting non-essential db tables before importing (cvs, history, project_*, ...) ------------------------------------------------------------------------ Sun, 28 Aug 2005 09:18:41 +0000 : Cvbge Attachment: http://drupal.org/files/issues/revisions_41.patch.patch (1.14 KB) Hi, tested the patch a bit: 1. in database.pgsql in the table files a) after removing PRIMARY KEY (fid) you need to remove also "," from the line above b) you CREATE INDEX files_fid_idx ON node(fid); CREATE INDEX files_vid_idx ON node(vid); on wrong tables: not node but files 2. in newly created table node_revisions a) in mysql version there is no DEFAULT, in postgres there is DEFAUL 0, maybe mysql version should be the same as postgres? b) there's no log longtext NOT NULL default '' in pgsql version c) you have translated format int(4) NOT NULL default '0' to format smallint NOT NULL default '0', it also should be integer and not smallint d) you create index: CREATE INDEX node_revisions_title_idx ON node_revisions(title,type);, but there is no type column in the table, and also there is no similar index in mysql version Attached patch should fix the errors

1 0

[drupal-devel] [bug] Block regions not initiated for newly enabled theme, so theme unusable
by Dries 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/29506 Post a follow up: http://drupal.org/project/comments/add/29506 Project: Drupal Version: cvs Component: system.module Category: bug reports Priority: critical Assigned to: nedjo Reported by: nedjo Updated by: Dries Status: patch (code needs work) There is something wrong with the PHPdoc of system_initialize_theme_blocks(): + * particular region isn't available in the new theme, the block is assigned + * instead to the new theme's default region. (I can fix that for you when I commit the patch.) Dries Previous comments: ------------------------------------------------------------------------ Wed, 24 Aug 2005 03:10:30 +0000 : nedjo Following the introduction of extentendable regions, see this issue [1], several users have reported problems: "A fresh install of Drupal with bluemarine displays blocks correctly, but when I switch the default theme to pushbutton, all the blocks disappear. They do show up in ?q=admin/block, but just with the yellow helper boxes, no blocks are really displayed.--Goba " "Same problem as Goba. The "region" field in the blocks table is sometimes blank. This makes many themes almost unusable in a Drupal default install. I don't know how the fields are blank, but the following SQL helped me get back to theme development: UPDATE blocks SET region="left" WHERE region=""--Thox " "another confirmation. I found out about the exact same sql query as Thox to fix it:) We came across this in an update. The theme had the regions defined, but since it was a non-default theme the blocks dissapeared. A drupal site, without a navigation block is simply useless.--Ber " The issue suggests a bug in the initiation of theme blocks. When a new theme is enabled, it's supposed to be initiated with blocks based on the default theme's blocks. I'll post a fix ASAP. [1] http://www.drupal.org/node/16216 ------------------------------------------------------------------------ Wed, 24 Aug 2005 04:11:57 +0000 : nedjo I can't reproduce the reported issue with a clean install. Are the themes causing these problems "updated" (are they phptemplate-based or, if not, do they have _regions() functions)? The specific problem seems to be that the region field isn't being set when block regions are initiated. Here's the relevant portion of function system_initialize_theme_blocks(): <?php // If the region isn't supported by the theme, assign the block to the theme's default region. if (!array_key_exists($block['region'], $regions)) { $block['region'] = system_default_region($theme); } ?> This is executed if a newly-enabled theme doesn't have blocks already assigned to it. It gets the default region's settings--except the region. This code portion tests whether the default region's region is available in the newly-enabled theme, and, if not, substitutes the theme's default region. This would all fail if the theme wasn't returning any regions--but any phptemplate theme, or the core ones, should be.... Anyone see what's going wrong? ------------------------------------------------------------------------ Wed, 24 Aug 2005 18:02:44 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-region-error-catching.patch (5.73 KB) I'm still not quite sure what the issue is, but clearly better error handling is needed. A particular issue is the scenario where a site admin upgrades a Drupal install without removing non-upgraded themes, or, on a new install, tries to install non-upgraded themes. As things stand, either of these could lead to unusable themes being the default, and all blocks disappearing. The attached (initial, untested) patch adds error handling to both updates.inc and system.module. New for allinstalls: * theme region initiation return FALSE on failure TRUE on success. * on initiation failure, theme is disabled, and, if it was set to default theme, bluemarine is substituted, with message given Updating: * tests if current default theme has been upgraded, and if not sets bluemarine as the default, issuing a notice * tries to initialize blocks for each enabled theme and, on failure, disables the theme. This should help prevent upgrade problems. If anyone has a chance to review and test this, I'd appreciate it, as I don't at present have a test install to upgrade. ------------------------------------------------------------------------ Thu, 25 Aug 2005 14:07:38 +0000 : Thox The problem occurs with all of the core themes except Bluemarine. That means .theme (chameleon) and phptemplate (pushbutton) themes are both affected equally. ------------------------------------------------------------------------ Thu, 25 Aug 2005 15:33:37 +0000 : Thox When I enable a theme, it goes through system_initialize_theme_blocks(), which assigns the system_default_region() to the new theme. Unfortunately, that's blank. There's no themes loaded (disabled by system_listing_save()), so there's no regions available, so there's no default region to use. Your patch stops old themes from working, can we not simply set the default region (if not found) to "left"? This works fine for me to fix this problem. ------------------------------------------------------------------------ Thu, 25 Aug 2005 16:26:03 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-region.patch (6.2 KB) This bug came from a recent change to function list_themes(), which now lists only enabled themes. function system_listing_save() fetched available regions (using, indirectly, list_themes()) *before* a theme was enabled. As well as the error handling (discussed above), designed to prevent non-upgraded themes from being enabled and to give meaningful error messages, the attached patch moves theme region initiation to after themes are enabled. On my testing, this fixes the reported error. Please test and report back. I haven't yet tested the error handling changes to updates.inc. Anyone able to do that? To test it, try to upgrade a site that has some default (and therefore, with an upgrade, updated) themes, and some themes (already enabled) that are not upgraded. The expected behaviour is that the upgraded themes have their regions initiated, while any non-upgraded themes fail, are disabled (i.e., have their status reset to 0), with messages given. ------------------------------------------------------------------------ Thu, 25 Aug 2005 17:57:34 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme_region2.patch (8.02 KB) Theme initiation was still broken for styles (unless the theme they are based on was already initialized). This is a perhaps unintended side-effect of limiting list_themes() to enabled themes--data for a theme on which a style is based are not available if that theme is not enabled. I've changed the code to load theme data directly rather than through list_themes(), so it's now I believe working. (Also fixed a minor bug in error reporting.) ------------------------------------------------------------------------ Thu, 25 Aug 2005 18:41:37 +0000 : stefan nagtegaal Nedjo, unfortunatly your patch isn't conform the drupal guidelines. See some comments here: <?php + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nOK\n"); + } ?> Why is this?? If this is _really_ needed, please use something like: <?php + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\n". t("Success") ."\n"); + } ?> Secondly: <?php + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be upgraded.' . "\nFAILED\n", array('%theme' => $theme))); ?> Should be something like: <?php + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be updated.', array('%theme' => theme('placeholder', $theme)) . "\n". t('Failed') ."\n")); ?> The difference is "Failed" instead of "FAILED". We do not use UPPERCASE errors. And I think if we use sentences as "Failed", we should use "Success" either instead of "OK".. Then you didn't made these messages translatable, and lacked the use of theme('placeholder', $text). Same here: <?php + $ret[] = array(1, t('Blocks initiated for %theme theme.') . "\nOK\n", array('%theme' => $theme))); ?> Do it like this: <?php + $ret[] = array(1, t('Blocks initiated for theme %theme.', array('%theme' => theme('placeholder', $theme)) . "\n". t('Success') ."\n")); ?> So, in short: - Try to minimise HTML in user translatable strings; - Don't use CAPITAL ERRORS, we don't do that atm; - Use theme('placeholder', $string) where needed; Though, I did not test the functionality of the patch (yet).. ------------------------------------------------------------------------ Thu, 25 Aug 2005 19:22:20 +0000 : nedjo "unfortunatly your patch isn't conform the drupal guidelines. " Well, okay. I'm just following what updates.inc already does, see function update_sql(), so that messaging is consistent. Perhaps you'd like to submit a patch to fix the allcaps issue in updates.inc. I'm not going to do that now, as my aim is to get this particular (critical) problem fixed. ------------------------------------------------------------------------ Thu, 25 Aug 2005 22:25:32 +0000 : chx No, no, no! Fix core themes, write something to update.inc if need to be and document changes. Never ever put update code in the running part of Drupal. ------------------------------------------------------------------------ Thu, 25 Aug 2005 23:08:59 +0000 : nedjo I think there's some confusion here. The portion of my patch in question *is* to update.inc--not "the running part of Drupal". ------------------------------------------------------------------------ Thu, 25 Aug 2005 23:34:55 +0000 : chx I stand corrected -- sorry. But please provide a new update_XXX function instead of patching the current the one. Stefan, in update.inc we use FAILED yes and I can't see placeholder calls. ------------------------------------------------------------------------ Fri, 26 Aug 2005 00:01:36 +0000 : drumm Updates should not make the return value themselves. <?php + $result = db_query("UPDATE {blocks} SET theme = '%s'", $default_theme); + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nOK\n"); + } + else { + $ret[] = array('0', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nFAILED\n"); + } ?> I can't see how this is different from using update_sql(). <?php + if (!system_initialize_theme_blocks($theme)) { + update_sql("UPDATE {system} SET status = 0, WHERE type = 'theme' AND name = '%s'", $theme); + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be upgraded.' . "\nFAILED\n", array('%theme' => $theme))); + } + else { + $ret[] = array(1, t('Blocks initiated for %theme theme.' . "\nOK\n", array('%theme' => $theme))); + } ?> Use drupal_set_message(t(...), 'error') instead. Monor issue: sentences should not have double spaces between them. And another update should be written to patch up whatever the last update left behind since many developers have the broken code and should not be running the same update again. ------------------------------------------------------------------------ Fri, 26 Aug 2005 00:02:17 +0000 : drumm Updating the status. ------------------------------------------------------------------------ Fri, 26 Aug 2005 05:24:31 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-regions.patch (5.65 KB) Thanks for the tips, though they kinda miss the point. My inclusion of update enhancements is evidently distracting from the main issue--so I've deleted it from the patch. There is, in any case, no bug or broken code to fix in the past update. The issue here is that the "list_theme() improvements" patch has broken theme block initialization--before anyone had a chance to see the new region functionality in action. The attached patch fixes this. Please review and apply ASAP. Aside: my manual messaging in the updates was an admittedly crude workaround to two limitations: (1) update_sql doesn't allow additional arguments (i.e., string or number query parameters) beyond the sql, and (2) since updates may do more than simply issue queries, there's a need for more flexible reporting. I agree that drupal_set_message() *could* be used for this, but I think it makes more sense to have flexible inline messages. So I've roughed in a patch to do that (and also address some of stefan's requests), see http://drupal.org/node/29002 ------------------------------------------------------------------------ Fri, 26 Aug 2005 08:34:38 +0000 : Thox Attachment: http://drupal.org/files/issues/default-regions.patch (617 bytes) Your patch works, but I don't see why it needs to be so complicated. Can you please explain why something simple such as my attached patch is not enough? Note that it would allow allow all of the old themes to work as usual and fixes the troubles with new themes. ------------------------------------------------------------------------ Fri, 26 Aug 2005 09:23:47 +0000 : robertDouglass Nedjo, I tried your most recent patch and it worked well. I haven't compared approaches, but was happy to get my site up and running with this *awsome* new feature. I had fun popping the login box into the footer, or header or content area. This is a whole new dimension of Drupal excitement. ------------------------------------------------------------------------ Fri, 26 Aug 2005 09:30:15 +0000 : stefan nagtegaal Attachment: http://drupal.org/files/issues/theme-regions_0.patch (5.66 KB) Updated patch to use theme('placeholder', $string) and deleted the double space between a translatable sentence.. Functionality seems to work.. One thing I do not like is the wording you use, though I do not have a better idea (yet).. Can someone who's native language is English look at the drupal_set_messages()'s, please? Stefan ------------------------------------------------------------------------ Fri, 26 Aug 2005 16:28:36 +0000 : nedjo Thanks for the tweaks, Stefan. "Can you please explain why something simple such as my attached patch is not enough?--Thox " It's a good question. My basic reasoning is that we shouldn't make upgrading themes optional, and shouldn't support non-complying themes. It's better to require upgrades, as we do for everything else, because failing to do so encourages lax coding and uneven support. Also, because the suggested workaround doesn't fix the bug, all themes would return simply left and right regions--and we'd be right back where we started! Overall, I'm thinking now I was mistaken to include error handling in this patch. Whether to allow non-complying code to be enabled is a worthy but separate issue. With modules, for example, we don't evaluate whether they work before enabling them--we rely on errors and the good sense of site admins to remove a non-complying module. I was thinking, "Themes are different, since setting a broken/nonupgraded theme as default could make a site unusable, since menus disappear, and novices don't know the right url to use to reset the theme to a supported one." But my approach was awkward and limited (e.g., it tested only one part of required theme functionality before disabling). If we're going to test for compliance before enabling (and I'm not convinced we should), we need to do so in a full, even way--which is to say, not in this patch! (The best, easiest thing would be simply to introduce a way for modules and themes to advertise their version support, as has been suggested and I believe begun elsewhere.) Updates *should* include better error handling; so I'll try to get in a separate update patch before the 4.7 code freeze. At any rate, the key point is to fix the bug introduced by issue #29002 [2]. I'll post a "bug-fix only" patch as soon as I can. [2] http://drupal.org/node/29002 ------------------------------------------------------------------------ Sat, 27 Aug 2005 21:19:12 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme_region3.patch (4.37 KB) Okay, here's a stripped down fix. This should be ready to apply--but I'll wait for someone else to affirm that before setting the status! (Aside: I gave an incorrect link, above, to the issue I created to improve update_sql() and messaging. The issue is actually at: http://drupal.org/node/29701)

1 0

[drupal-devel] [bug] Put revisions in their own table
by killes 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/7582 Post a follow up: http://drupal.org/project/comments/add/7582 Project: Drupal Version: cvs Component: node system Category: bug reports Priority: normal Assigned to: killes(a)www.drop.org Reported by: killes(a)www.drop.org Updated by: killes(a)www.drop.org Status: patch (ready to be committed) Attachment: http://drupal.org/files/issues/revisions_41.patch (52.4 KB) I've tested the upgrade path on an older copy of the drupal.org db and found and squished two bugs. The test cycle can be shortened a bit by deleting non-essential db tables before importing (cvs, history, project_*, ...) killes(a)www.drop.org Previous comments: ------------------------------------------------------------------------ Wed, 05 May 2004 16:25:27 +0000 : killes(a)www.drop.org Currently all node revisions are stored in a serialized field in node.table and retrieved for _each_ page view although they are rarely needed. However, we have agreed that serializing data is bad and that we should try to keep the memory foot print pf Drupal small. Therefore I propose to create a separate revisions table which would be in principle identical to the node table, only that it could have several old copies of the same node. Extra data added by other modules could be added in a serialized field unless we find a better solution. ------------------------------------------------------------------------ Wed, 05 May 2004 17:06:35 +0000 : jhriggs I too think the serialized approach is less than desirable, but here's an alternative. This would likely take some considerable rework in core and contrib, but the following is how we handle similar types of situations in our databases at work. It is more elegant that a separate table, and avoids the (almost exact) duplication of a table. Instead of separate tables, keep all revisions of nodes in the node table as follows: * add field: active (0/1 or Y/N) * add field: revision * every revision of a node is stored in the node table; however, only one revision can be active at any given time * nid can no longer be unique -- primary/unique key becomes (nid, active) * any time a node is loaded, updated (without revision), etc., the active version is used. Thoughts? ------------------------------------------------------------------------ Wed, 05 May 2004 17:57:48 +0000 : killes(a)www.drop.org I am not opposed to your scheme, but I want to stress the following: * Duplicating a table's structure is not bad (IMHO) as long as the content is different. * having two tables will allow us to have a rather small node table. This is (maybe) a performance gain. ------------------------------------------------------------------------ Wed, 05 May 2004 18:37:29 +0000 : jhriggs I don't necessarily think that duplicating a table's structure is _bad_. It just seems to be wasteful and a pain to maintain. (Every change to the node table is made twice...easy to do, but also easy to miss perhaps.) As for performance, as long as nid and the active indicator are indexed, there shouldn't be any performance loss. Also, archiving an old version when making a new revision will be much simpler: just change the active indicator rather than copying an entire node to another table (and ensuring everything gets copied...again a potential maintainance issue). To be honest, I would just like to see the serialized data go away, regarless of what approach is taken. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:49:33 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_07-30-2004.p… (10.47 KB) I'm interested in using Drupal for a large scale wiki-type project. In order to do this, I need revisions to be in their own table. Attached is a patch to do just that. Most of the changes are pretty self explanatory. Spreading out node data across two tables meant that I had to add database functions to do locking/transactions. Without this, race conditions in which the database becomes corrupted are possible. ------------------------------------------------------------------------ Fri, 30 Jul 2004 19:54:36 +0000 : Nick Nassar Oh yeah... The patch is a diff against Drupal CVS ------------------------------------------------------------------------ Sat, 31 Jul 2004 00:00:08 +0000 : Anonymous Gerhard speaking. Nick, thanks a lot for your nice patch! It saves me a great deal of labour. I looked through it and immediately liked it. You not only put the old revisions into a new table but also the current one. Do you have an estimate how much more expensive the additional join is? Besides a few minor coding style issues I found a major one: Just a few hours before you uploaded your patch JonBob's node access patch hit core. That means your patch won't apply anymore as all the queries you change have been changed. Can I bug you to update your patch? ------------------------------------------------------------------------ Sat, 31 Jul 2004 01:11:59 +0000 : Anonymous Also I think that your upgrade path loses existing revisions. ------------------------------------------------------------------------ Sat, 31 Jul 2004 02:39:12 +0000 : drumm I think this is the proper way to do things. No columns are duplicated, there is no serialized data, and only the fields that are logically revised are stored. Nothing jumped out at me as a way to have my node modules be able to keep a table of revisions of additional fields. I'm guessing this could be done within the confines of _insert and _update. Assuming the upgrade path works and modules can extend it I give it a +1. ------------------------------------------------------------------------ Sat, 31 Jul 2004 14:40:15 +0000 : Nick Nassar It figures that just as I finish a big patch, another patch comes along and breaks it. Oh well, it should be a pretty easy to fix. I'll work on it. Fixing the upgrade path to keep revisions should be fairly painless. I found another issue that needs to be fixed before this patch gets merged. There format of a node needs to be stored for each revision. Otherwise, for modules that store a format for the nodes, such as page and book, if you write one revision in PHP and the next in HTML, the PHP revision will be displayed as HTML. This is part of a larger issue of how node modules should store revisions of additional fields. I think each module that wants to do this should create another table with (nid, revid) as the primary key. Just as when they want to add fields to a node they create another table with nid as the primary key. As far as performance goes, for sites that make heavy use of revisions, an extra join on primary keys is going to be a lot faster than grabbing all of the revisions from that database everytime. We would need to run benchmarks to determine is the overall difference in speed is for an average site is a gain or a loss. I'm guessing it's very minor either way. ------------------------------------------------------------------------ Mon, 23 Aug 2004 13:55:49 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_08-23-2004.p… (10.92 KB) Here's an updated patch against CVS that puts revisions in their own table, provides an upgrade path, and fixes the format related bugs in the last patch. Hopefully, this can make it into CVS as soon as the freeze is over. ------------------------------------------------------------------------ Mon, 23 Aug 2004 14:10:39 +0000 : moshe weitzman Interesting patch ... drumm's question is still outstanding. how do modules store revisions of their fields? Are they expected to manage this on their own? Thats not how it works today. As an aside, i am seeing profile_ fields in my node.revisions column. One could argue that those need not be saved. They pertain to the node author, not to the node itself. ------------------------------------------------------------------------ Mon, 23 Aug 2004 16:14:39 +0000 : Nick Nassar Having modules be responsible for storing revisions of their own fields is a side-effect of storing revision data in tables. There's really no way around it. However, revisions generally don't make sense for node types that don't have PHP/HTML content, such as polls. I think it's going to be a pretty rare scenario for a new node type to want another field to change per-revision, so it's a pretty good trade-off. Storing fields that shouldn't be part of revisions, such as the profile_ fields, is a side-effect of storing revisions as serialized objects. Applying this patch will free up that wasted space. :) ------------------------------------------------------------------------ Mon, 23 Aug 2004 17:20:57 +0000 : Anonymous There should be a hook that let's the module choose whether it supports history. This way a module author can prevent the user from doing something that may break his module or just cause undefined behavior. If the module doesn't support history then don't let the user/admin choose to add history to nodes of that type. Craig ------------------------------------------------------------------------ Mon, 23 Aug 2004 19:23:29 +0000 : Nick Nassar I agree, there should be an API change to make specifying support for revisions easier. In the interests of keeping patches small and keeping to one change per patch, I think the API change should be a separate issue. A sort of ad-hoc API to decide whether or not a module supports revisions by default already exists. Instead of having a hook, modules set the default value of the "Create new revision" field in the edit form. The admin can change this option in admin/node/configure/defaults. This patch doesn't change that. Revisions are broken for node types that have their own database structure, like polls, even when storing them as serialized objects. This patch doesn't change that, either. ------------------------------------------------------------------------ Tue, 26 Oct 2004 02:35:06 +0000 : moshe weitzman I'm guessing that someone is going to have to demonstrate that this patch performs as well as current drupal before it gets comitted. i think this patch is a few benchmarks from being comitted. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:04:09 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:16 +0000 : Nick Nassar I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Wed, 27 Oct 2004 01:05:26 +0000 : Nick Nassar Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004.p… (11 KB) I ran some really unscientific benchmarks, and it looks like this patch has a negligible affect on performance. I used apache bench and the database from theregular.org, which doesn't contain any revisions (worst case scenario for this patch) and contains several hundred nodes. Both the patched and unpatched versions hovered between 2.36 and 2.38 requests per second. The command I used was: ab -n50 -C 'PHPSESSID=b01a9f92880ef215b0ed6f1314a5eba2' http://192.168.0.100/ An updated patch that should apply to CVS is attached. ------------------------------------------------------------------------ Mon, 15 Nov 2004 05:05:30 +0000 : elias1884 please overthink the revision system default workflow as well. don't look at the revision system as an isolated system but as a part of the whole workflow system! if you combine revisions with the moderation queue the most logic default workflow would be like that: auth user creates node (revision #0) admin approves the node (status = 1, moderation = 0) => node publicly available auth user finds typo and changes node (revision #1, status = 0, moderation = 1) ------- what happens at that point at the moment is, that the node is not accessible anymore at all until the new revision is approved by admin. of course the new revision should not go online until reviewed and approved, this is absolutely correct, but there is no reason to not take the old revision offline, since it was already approved and should therefore be online until the new revision is approved. it is not practical if a node disappears only because the author corrected a typo. ------- admin approves the node (status = 1, moderation = 0) eventhough I first thought a plain boolean active field would not be capable of providing that functionality if finally came to the conclusion, that it can. The only thing to do is to not set that bit, when a new revision is created, but when it is approved (in case moderation is activated under default workflow). Every revision should have its own moderation, status and active field and on approval they are set like this (status=1, moderation=0, active=Y). When you wanna rollback to an old revision, you can chose between all revisions that already have the moderation bit set back to 0 again and the published to 1. There should be an extra permission for rollback! another concern that I have about the default workflow is, that users can't see the content, they have just created, when moderation is enabled. Eventhough, there is a big fat "submission accepted" presented after submissions, unexperienced users tend to question the information those stupid tincans give them, if they can't find their content afterwards. Many users are really lazy bastards and they don't even read the status messages. The best feedback about whether his story was submitted successfully or not of course is, if he can find the story somewhere on the site, maybe with a status message on top of it, mentioning, that the content is currently not publicly available since it has not been approved yet. there should be a my content section under my account, like somebody is trying to do with the workspace module I guess. so my suggestion is to make (status=0, moderation=1) still available for the creator under a my content section somewhere! ------------------------------------------------------------------------ Wed, 24 Nov 2004 04:21:18 +0000 : Nick Nassar I agree. The current workflow for moderation queues and revisions needs to change, but this patch isn't the place for it. The patch is already too big, and it only does the backend stuff. Instead of adding more to this patch and making it take even longer to get into core, would you mind creating a new issue for your UI suggestions, so the those changes can be added as a separate patch? Thanks, Nick ------------------------------------------------------------------------ Sat, 11 Dec 2004 12:26:02 +0000 : Dries This patch is _much_ needed so I'd love to see someone revive it. In order for this patch to be accepted, the following needs to be done: * Update this patch to CVS HEAD. * Rename revid to vid. * Rename node_rev to node_revisions. * Rename node_rev.changed to node_revisions.timestamp. * Rename $rnode to $revision. * Fix the coding style to match Drupal's: proper spacing, single quotes where possible, proper variable names. * Benchmark this patch with a large database with enough revisions. I'd be happy to benchmark this on my local copy of the drupal.org database. * The book.log field should probably move to the node_revisions table. This can be done in a separate patch. * Investigate whether transactions are well-supported. ------------------------------------------------------------------------ Mon, 13 Dec 2004 00:25:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/Drupal-Improved_Revision_Schema_10-26-2004-r… (11.02 KB) I've worked a bit on the patch (coding style issues as mentioned by Dries). One thing I noticed is that the patch uses REPLACE. IIRC this needs to be chagned to "UPDATE, if fail INSERT" for pgsql compatibility. Nick, are you still interested in working on that patch? I'd like to know how it works on your site and work on getting it into core. ------------------------------------------------------------------------ Mon, 13 Dec 2004 12:33:08 +0000 : Dries Gerhard: your patch does not apply. ------------------------------------------------------------------------ Mon, 13 Dec 2004 17:10:12 +0000 : killes(a)www.drop.org Yes, I know, that was the same version as I mailed to you earlier. ------------------------------------------------------------------------ Mon, 13 Dec 2004 21:02:06 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions.patch (52.96 KB) Ok, upüdated the patch to cvs. ------------------------------------------------------------------------ Tue, 14 Dec 2004 08:58:36 +0000 : Dries Some more comments: * db_begin_transaction() and db_end_transaction() do not belong in database.inc, but in database.mysql.inc and database.pgsql.inc respectively. * The node module calls node_revisionsision_list() which is not defined. (Fxed that on my local copy.) * Do db_begin_transaction() and db_end_transaction() deprecate Jeremy's table locking patch? * The upgrade path assigns the wrong user ID to each revision. * The upgrade path assigns the wrong date to each revision (that or a node's revision page shows the wrong usernames/dates). * The coding style needs a bit of work, but we can worry about that later. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:34:44 +0000 : Nick Nassar If you need any help getting those things fixed, just let me know. ------------------------------------------------------------------------ Tue, 14 Dec 2004 17:50:30 +0000 : Nick Nassar How this relates to Jeremy's node locking patch: There was lots of discussion, and node locking was decided against because from an end user point of view you never want a node to be locked. He's now advocating for a much simpler patch that warns users if their changes will overwrite someone elses. That patch still has a race condition, which might be fixed using db_begin_transaction(). http://drupal.org/node/6025 ------------------------------------------------------------------------ Tue, 14 Dec 2004 22:26:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_0.patch (55.96 KB) Here is an updated patch that tries to address Dries concerns. ------------------------------------------------------------------------ Wed, 15 Dec 2004 08:32:50 +0000 : Dries Attachment: http://drupal.org/files/issues/revisions-bug.png (76.06 KB) It didn't fix the aforementioned bugs. See attached screenshot. ------------------------------------------------------------------------ Thu, 06 Jan 2005 20:15:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_1.patch (51.77 KB) Ok, here is a new version. Dries and myself worked hart at it, so please have a look. what is still missing - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. ------------------------------------------------------------------------ Wed, 19 Jan 2005 21:43:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_2.patch (49.49 KB) Here is an updated patch. We discussed to keep the current title in node module and also in the revisiosn table. This is content duplication but will save many joins as many queries only need the title of a node. Discussion is welcome. ------------------------------------------------------------------------ Wed, 19 Jan 2005 23:33:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_3.patch (29.93 KB) I've implemented the aforementioned solution. This makes the patch much smaller. The patch now also removes taxonomy_node_has_term() which wasn't used anywhere. I'd really apprciate if some people could test drive the patch. It will be another huge improvement for 4.6. ------------------------------------------------------------------------ Thu, 20 Jan 2005 00:05:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_4.patch (30 KB) Another revision. Steven didn't like my literal $node->vid in queries. ------------------------------------------------------------------------ Thu, 20 Jan 2005 01:10:50 +0000 : killes(a)www.drop.org - database upgrades for the core modules with an own table - contrib modules need an upgrade too. - do we need nid and vid in both the node and the node_revisions table? - the amount of sql queries means a good stress testing for large databases. These issues are still open, btw. Especially the first one needs to be tackled. ------------------------------------------------------------------------ Tue, 25 Jan 2005 20:11:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_5.patch (51.13 KB) Here is a patch that has the database tables updated for forum, book, and page module. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:55:59 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_6.patch (49.18 KB) Yet another update to keep it working with head. The patch now also removes the table definitons for the page table. ------------------------------------------------------------------------ Sat, 29 Jan 2005 22:57:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_7.patch (55.69 KB) Sorry, that was the old version, this is the right one. ------------------------------------------------------------------------ Mon, 31 Jan 2005 19:55:03 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_8.patch (55.71 KB) Updated once more. ------------------------------------------------------------------------ Mon, 31 Jan 2005 20:52:08 +0000 : Dries Anyone to help review/test this? ------------------------------------------------------------------------ Mon, 31 Jan 2005 21:22:36 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_9.patch (49.29 KB) Updated again, the update functions occurred twice. Thanks Bart. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:05 +0000 : killes(a)www.drop.org Don't know if the db I am using is corrupted or what. I still do have some didficulties. The latest patch is attached. ------------------------------------------------------------------------ Wed, 02 Feb 2005 00:27:49 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_10.patch (49.67 KB) I am probably slowly going mad ... ------------------------------------------------------------------------ Wed, 02 Feb 2005 01:54:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_11.patch (48.95 KB) The update issue still needs investigating. This patch is updated for cvs. ------------------------------------------------------------------------ Wed, 02 Feb 2005 20:20:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_12.patch (49.83 KB) Ok, here is a new version. I've solved my troubles with book.module. There are still some issues with forum module. Possibly due to inconsistent database. ------------------------------------------------------------------------ Wed, 02 Feb 2005 21:31:05 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_13.patch (49.83 KB) Turns out the drupal.org database had indeed some quirks. Please run this query in your oldest db and tell me the result: select nid,type from node where type like '%/%'; If you get a non-zero result we might need to add another security update. The patch could use still more testing, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 01:16:54 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_14.patch (49 KB) Ok, we are getting somewhere. At a first glance the update is working. There is a problem remaining: the revisions tab will be shown whether the node has revisions or not. Not sure we can/need to fix this. People with a drupal.org account can log in at http://killes.drupaldevs.org/revision/ and poke around. Your permissions will be the same as on drupal.org. Feel free to vreak everything but don't forget to file complaints here. (Note: this is only a pruned version of the drupal.org database with all project nodes and nodes with nids > 7000 dropped). ------------------------------------------------------------------------ Thu, 03 Feb 2005 04:19:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_15.patch (52.39 KB) There was some error in node_save and also the patches to the database.inc files got lost... ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:07:27 +0000 : robertDouglass Submitting book pages doesn't work on your test site. It puts the entire content of the preview inside the body textarea. I wrote a sentence in the body and the log, and pressing preview put several lines of HTML containing both sentences in the body textarea on the preview page, plus the book page wouldn't submit. -R ------------------------------------------------------------------------ Thu, 03 Feb 2005 07:50:59 +0000 : Junyor 0 results here. I started using Drupal with version 4.4, though. ------------------------------------------------------------------------ Thu, 03 Feb 2005 23:56:18 +0000 : killes(a)www.drop.org @Junyor: Thanks, that's a good sign. Maybe somebody else has an older db to try. @robertDouglass: The first effect you describe is due to drupaldevs running on PHP 5. I am unsure why the second thing does not work. In node_save() the node object has a nid although there is none in the form. Very strange. I've enabled display of db queries on the testsite. ------------------------------------------------------------------------ Fri, 04 Feb 2005 19:17:55 +0000 : dmjossel No results here on the query: select nid,type from node where type like '%/%'; On a database that was put in place prior to Drupal 4 and is now running on 4.5.2. ------------------------------------------------------------------------ Fri, 04 Feb 2005 20:44:23 +0000 : killes(a)www.drop.org @dmjossel: thanks. @all. The strange problem I reported was apparently php 5 related. After applying Steven's php 5 patch it went away. One error is remaining: If I create a new forum topic it will be shown as part of the book on preview. Hmm, that was due to a db that got corrupted during testing so that is fixed too. Please poke around at the test site and look if you find more errors. ------------------------------------------------------------------------ Sat, 05 Feb 2005 07:16:22 +0000 : Steven By the way, I just remembered that Drupal.org has some blogs lingering on in the database even though blog.module is not enabled. Perhaps this is causing troubles? ------------------------------------------------------------------------ Sat, 05 Feb 2005 11:22:59 +0000 : Anonymous I can't see why it would. Drupal.org will need extra updates for images and project nodes because those have their own tables. GK. ------------------------------------------------------------------------ Sun, 06 Feb 2005 12:49:55 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_16.patch (52.49 KB) Updated to apply to cvs again. ------------------------------------------------------------------------ Tue, 22 Feb 2005 20:15:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_17.patch (49.64 KB) Updated again. All we need is a patch to upload module and an upgrade path for it. ------------------------------------------------------------------------ Fri, 04 Mar 2005 04:22:58 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_18.patch (52.31 KB) Updated once more. Moved log field from book to node_revisions table as discussed in Antwerp. upload module still missing. We need to decide under which circumstances the log field should be displayed. Should that be added to the workflow? Should it depend on the revisions setting? ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:27:03 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_20.patch (75.52 KB) Ok, here it is: Yet another revision of this grrrrreat patch. Changes from previous versions: - supports versioning for uploaded files. A problem is that if you delete a file, it will be gone for all revisions. - the log field is now in the node_revsions table, but each module has to decide whether to show it or not. I've implemented it for the page and the book type odes. Also, the field can be edited when adding non-book nodes to the book. The log is displayed on the revisions page and if a node is moderated. - the revisions are moved to an old_revisions table to a) get the node table smaller and b) still leave the mavailable for contrib modules that want to retreive old version data. The patch has been applied to killes.drupaldevs.org/revision where it can be tested by anybody (especially people who have "site admin" rights on drupal.org) The database is from drupal.org and you should b able to log in with your pass or simply mail yourself a new one. Gerhard ------------------------------------------------------------------------ Sat, 05 Mar 2005 19:51:56 +0000 : Anonymous Attachment: http://drupal.org/files/issues/revisions_21.patch (59.42 KB) BTW, I marked this a bug because atm the revisions field can grow quite big. Neil has reported problems from some users who were not able to load some nodes due to to many large revisions. Also, som unrelated stuff crept into the patch. New version attached. ------------------------------------------------------------------------ Tue, 08 Mar 2005 05:56:01 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_22.patch (60.29 KB) Ok, I think I got it. Changes to last version: - uploads are no properly versioned. Missing are still pgsql checks and updates. ------------------------------------------------------------------------ Thu, 10 Mar 2005 16:58:41 +0000 : Anonymous Was able to get http://drupal.org/files/issues/revisions_21.patch to work with drupal-cvs.tar.gz (10 March 2005) by: - includes/database.mysql.inc: Commenting out duplicates for functions function db_begin_transaction and function db_commit_transaction - modules: node.module: Removing "'title' => $node->title," from $node_table_values variable declaration and removing "'title' => "'%s'"," from "$node_table_types" variable declaration. Happy to submit a patch if requested. I'll watch this thread. ------------------------------------------------------------------------ Fri, 11 Mar 2005 23:59:45 +0000 : killes(a)www.drop.org The duplicate function has been removed in rev 22 of this patch. Why do you think the changes in node_save are needed? Titles are saved in both tables for performance reasons. ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:12:21 +0000 : jlerner Hi - I posted comment #62. The changes to node_save appear to be needed because recent patches (both 21 and 22) remove the field 'title' from table 'node'. So without the changes to node_save, node.module is broken and generates errors. Joshua ------------------------------------------------------------------------ Sun, 13 Mar 2005 16:29:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_23.patch (61.17 KB) Thanks, Joshua, for catching this. node:title is there to stay. ------------------------------------------------------------------------ Wed, 13 Apr 2005 16:29:23 +0000 : moshe weitzman since HEAD is open again, perhaps it is a good time to revisit this patch. once this is committed, lets address - http://drupal.org/node/11071 "node_validate does not respect group editing" ------------------------------------------------------------------------ Mon, 18 Apr 2005 15:43:42 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Mon, 18 Apr 2005 16:16:32 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_24_0.patch (60.39 KB) Updated. ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:19:42 +0000 : Dries I'll commit this patch later this week! If you haven't checked this patch already, I urge you to test/check it out because it will have significant impact on existing code and modules! ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:21:33 +0000 : Dries Also, what do people think about the n.title being duplicated? ------------------------------------------------------------------------ Tue, 19 Apr 2005 05:26:58 +0000 : chx I won't lose any sleep because of duplicated titles... ------------------------------------------------------------------------ Tue, 19 Apr 2005 18:35:58 +0000 : killes(a)www.drop.org Let me explain why I have chosen to duplicate the title (and also the uid): If you look at all the queries in Drupal, you will find that most of them only need the title and th uid of a node. That is, by duplicating it, we save expensive joins on the node_revisions table. Due to this fact, this patch is actually a performance improvement. A note about updating contrib module: Strictly speaking they wouldn't need to be updated. They only need to if their authors decide that their info should be available for revisioning. The upgrade path for forum.module in my update.inc patch (+ the forum patch) should show you what needs to be done. I will write a note for the update page once the patch hits core. ------------------------------------------------------------------------ Sun, 24 Apr 2005 21:21:19 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_25.patch (60.38 KB) Updated to cvs. Dries: Based on some remarks in #drupal this is the last update I am going to do. Apply it or won't fix it. ------------------------------------------------------------------------ Sat, 30 Apr 2005 03:42:39 +0000 : Jeremy(a)kerneltrap.org Attachment: http://drupal.org/files/issues/revisions_25.patch.patch (528 bytes) That's a big patch. I've only started looking through it. I noticed one little typo, affecting updates. A patch to your last patch is attached. I'm running with the revision patch on my dev server now happily. I like the concept. What happens if you click 'stop' on your browser in the middle of a MySQL "transaction"? I assume that kills the connection to MySQL, and the lock is freed? But this then leaves changes only partially applied? What exactly does locking/unlocking the tables buy us in MySQL? I don't see anywhere that we detect if an apply fails part way through, and thus roll back...? What am I missing? ------------------------------------------------------------------------ Sat, 30 Apr 2005 07:11:28 +0000 : Dries Jeremy: many of us are worried about the performance ramifications this patch introduces. Early experiments showed a small performance improvement (while a performance regression might be expected). More performance reports from large sites like kerneltrap.org will certainly help this patch. Mind to do a quick performance comparision and to report back with some numbers? Thanks. ------------------------------------------------------------------------ Sat, 30 Apr 2005 12:38:02 +0000 : Jeremy(a)kerneltrap.org Dries: I'm not running HEAD on kerneltrap, so this really isn't a possibility. Furthermore, until I understand why we're locking tables, I don't like it. The idea of revisions in their own tables is great. The idea of locking tables to get (without any obvious benefit) there really worries me. ------------------------------------------------------------------------ Sat, 30 Apr 2005 14:16:01 +0000 : killes(a)www.drop.org @Jeremy: Thanks for looking at the patch! Also for catching the typo. :) Did you try to upgrade your database? If yes, how did it go? One of Dries' concerns is the complexity of the upgrade. How many nodes and revisions did the db have? About database locking: This part of the patch was created by Nick and I simply continued to use it. Maybe the code should rather be: if(db_begin_transaction(array('{node}', '{node_revisions}', '{watchdog}', '{sessions}', '{files}'))) { db_query($node_query, $node_table_values); db_query($revisions_query, $revisions_table_values); db_commit_transaction(); ... } The idea is probably to avoid two updates at the same time. I don't think the locking helps if you abort the script at an inconvenient time. Rollbacks aren't implemented in all mysql versions. We could omit the db locking if deemed inappropriate. Maybe Nick can explain his ideas behind this. @Dries: I wonder who the "many of us" are. They certainly haven't spoken to me. Moshe had some reservations about the upgrade path and project module, but the time that project module abused revisions to store issue updates was long ago and his reservations were resolved. Nobody else (besides you of course and now Jeremy) has voiced reservations in a way that was audible to me. If you grep through the patch you will notice that there are only four queries which have a join on the node_revisions table. Two of them are in node_load and in the other cases the join replaced a join on the node table. The two queries in node_load are the only ones that have both a join on the node and the revisions query. Thus, loading of individual nodes might become somewhat slower. All other queries will be faster since the node table is now much smaller. Also, node loading does not have to be slower, it depends on your node table. If you had a lot of revisions and thus a large table, then the new scheme will make your queries actually faster since we do not load the revisions on each and every node load anymore. If you didn't have many revisions your node_load migth be somewhat slower. WRT to the update script Karoly pointed out that we could use multiple insert queries instead one query per revision. This would probably make the update somewhat faster. I am willing to work on this iff you declare that you will commit the patch afterwards. I'd need to know if this will work on pgsql and on all supported mysql versions before I work on it. About locking: Database locking is dog slow, at least on mysql. I was using locks in an earlier version of the upgrade script but had to remove it for (serious!) performance reasons. ------------------------------------------------------------------------ Mon, 09 May 2005 15:07:34 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_26.patch (46.45 KB) Ok, another update, cause I need it myself. I've left out the transaction stuff for now. It is in principle unrelated to this patch and should be discussed elsewhere. This also makes the patch smaller and easier to review (hint, hint). ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:09 +0000 : killes(a)www.drop.org The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 20:32:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_27.patch (39.05 KB) The patch contained the update functions twice. ------------------------------------------------------------------------ Mon, 09 May 2005 21:23:06 +0000 : Dries Got one error during the upgrade path: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ------------------------------------------------------------------------ Mon, 09 May 2005 21:26:19 +0000 : killes(a)www.drop.org This had happend to me as well, when I tested this patch. The reason is that for some reason the vid is not unique. Most likely there are some entries with vid = 0 in there. Can you check which node types those have? it always was an error in the test database. See: http://drupal.org/node/7582#comment-20678 ------------------------------------------------------------------------ Mon, 09 May 2005 21:27:06 +0000 : Dries Actually, I got 2850 errors during the upgrade. Some of these: sprintf() [function.sprintf]: Too few arguments in drupal-cvs/includes/database.inc on line 154. Some of these: Query was empty query: in drupal-cvs/includes/database.mysql.inc on line 66. And this: Unknown table 'n' in field list query: SELECT n.nid, n.vid FROM node INNER JOIN files f ON n.nid = f.nid in drupal-cvs/includes/database.mysql.inc on line 66. :-) ------------------------------------------------------------------------ Mon, 09 May 2005 21:29:19 +0000 : Dries Or this: user error: Unknown column 'log' in 'field list' query: SELECT parent, weight, log FROM book WHERE nid = 1 in drupal-cvs/includes/database.mysql.inc on line 66 ------------------------------------------------------------------------ Mon, 09 May 2005 21:52:12 +0000 : Dries The time required to generate my main page went from 902 ms (before upgrade) to 2139 ms (after upgrade). The time required to generate a forum listing (?q=forum/x) went from 1872 ms (before upgrade) to 2874 ms (after upgrade). Maybe this is because my database is not consistent as the result of the upgrade errors (yet I don't see any errors on the pages I benchmarked). ------------------------------------------------------------------------ Tue, 10 May 2005 00:24:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_28.patch (53.47 KB) Ok, let me get to this from the bottom to the top: - my test runs indicated a different development wrt timing. If I had gotten your results, I had stopped working on this long ago. So your results are wrong for some reason. - user error: Unknown column 'log' in 'field list' Wasn't my day, the book patch got lost. It is contained now. First -R the old patch, then apply this one. - Unknown table 'n' in field list query: Walkah found this, but I forgot to fix it. Fixed now. - I've no idea where the other queries come from. I am hoping that either your test db is borken or they are follow ups from the other ones. If you let me have your test db, I'll try some debugging. Thanks for wasting your time, too. ------------------------------------------------------------------------ Tue, 10 May 2005 05:07:31 +0000 : Dries I double-checked and the numbers don't seem to lie. I'll test some more after work on another machine to make sure it is not platform-specific. ------------------------------------------------------------------------ Wed, 11 May 2005 03:32:47 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_29.patch (54.83 KB) Ok, here I am again. What I did: 1) Ask Dries to let me have drupal.org database 2) get 400MB of SQL inserts... 3) take 23 minutes to import said data 4) Remove all image and project nodes (don't want to install their modules), 11765 nodes left 5) back up data 6) take tests on non-cached /node page (as anonymous user). ab results: -c 1 -n 25: Requests per second: 1.29 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 663 775 179.7 689 1264 7) Do the same for the tracker page: Requests per second: 0.83 [#/sec] (mean) Total: 1182 1199 7.4 1199 1217 8) Apply my patch (rev. 28). 9) run db update and hold breath 10) update times out... 11) play back backup from 5) 12) wait 13) getting annoyed and removing cache, watchdog, and accesslog before playing in backup. 14) wait again. Understand why Dries doesn't try this patch often. Maybe a smaller DB would do for testing? 15) wait more. get really annoyed. 16) Set time limit to 18000 in update.php 17) try again 18) fails again before the second update is completed. 19) curse. 20) delete search stuff from db. Ooops, sooo much smaller... 21) import again, below 2 minutes... 22) rewrite to use extended insert. Found a bug. 23) still does not complete. Mysql logging to the rescue! 24) tid = 0? Not good. 25) Well, the update works fine till node 10834. 5595 nodes done, 6136 to go. 26) Writing shell based update script. Discovery: 24MB aren't enough. Hopefully 64 are. Nope. extended inserts for revisions are apparently not the brightest idea: Huge memory consumption. Hmm, no, all updates got through. Selecting the revisions to put them into old_revisions table screwed it. Learned about CREATE TABLE old_revisions SELECT syntax. Yay! finally. 24 MB are just not enough the update.php script seems to still break. 27) Benchmarks! /node Requests per second: 1.54 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Total: 632 649 40.5 636 791 /tracker Requests per second: 0.86 [#/sec] (mean) Total: 1119 1165 65.8 1160 1461 Ok: So we get an improvement for many node_loads, but none for simple selects from node. More tests can be done. 28) roll new patch Ain't Drupal fun? ------------------------------------------------------------------------ Wed, 18 May 2005 13:38:05 +0000 : Dries I did another round of tests on _another_ machine and it is not looking good: Before upgrade After upgrade ?q= (main page) 218 ms/request 340 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request ?q=about (book page) 375 ms/request 5400 ms/request The upgrade process itself gave me a number of 'query was empty' and 'sprintf(): too few arguments' reports. Everything seems to work fine though. Looking at the ?q=about page, I see that the following query is executed twice _and_ that each time, it take more than 2 seconds to complete: SELECT n.nid, n.title, b.parent, b.weight FROM node n INNER JOIN book b ON n.vid = b.vid WHERE n.status = 1 AND n.moderate = 0 ORDER BY b.weight, n.title; --8 SHOW INDEX FROM book; +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ | book | 1 | book_parent | 1 | parent | A | 92 | NULL | NULL | | BTREE | | | book | 1 | nid | 1 | nid | A | 369 | NULL | NULL | | BTREE | | +-------+------------+-------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+ 2 rows in set (0.00 sec) The book module does not appear to have a primary key? Sounds like a bad idea so I added one: mysql> ALTER TABLE book ADD PRIMARY KEY nid (nid); Query OK, 369 rows affected (0.02 sec) Records: 369 Duplicates: 0 Warnings: 0 Next, I wanted to make the vid column a unique key in all node tables: mysql> ALTER TABLE node ADD UNIQUE vid (vid); Query OK, 20392 rows affected (0.81 sec) Records: 20392 Duplicates: 0 Warnings: 0 mysql> ALTER TABLE book ADD UNIQUE vid (vid); ERROR 1062: Duplicate entry '0' for key 2 mysql> ALTER TABLE forum ADD UNIQUE vid (vid); Query OK, 10806 rows affected (0.10 sec) Records: 10806 Duplicates: 0 Warnings: 0 As you can see, it fails for the book table which makes me believe there is some inconsistent data ... I set out to fix that: mysql> SELECT nid, COUNT(nid) AS vids FROM book GROUP BY vid HAVING vids > 1; +-----+------+ | nid | vids | +-----+------+ | 871 | 2 | +-----+------+ 1 row in set (0.00 sec) mysql> SELECT title FROM node WHERE nid = 871; Empty set (0.00 sec) mysql> DELETE FROM book WHERE nid = 871; Query OK, 1 row affected (0.00 sec) mysql> ALTER TABLE book ADD UNIQUE vid (vid); Query OK, 368 rows affected (0.01 sec) Records: 368 Duplicates: 0 Warnings: 0 Looks like everything is well now. Ran some new benchmarks: Before upgrade After upgrade With indices ?q= (main page) 218 ms/request 340 ms/request 336 ms/request ?q=forum (forum overview) 754 ms/request 1520 ms/request 1531 ms/request ?q=about (book page) 375 ms/request 5400 ms/request 475 ms/request Unfortunately, we're still slower than the original code. ------------------------------------------------------------------------ Wed, 18 May 2005 21:53:31 +0000 : killes(a)www.drop.org Dries, thanks for testing it again. I do think the broken queries you observer have something to do with the bad performance after the update. Please log the queries any I will have a look at them. I've never seen any such queries. My update script also tries to create the appropriate indices, but it will of course fail if the database contains cruft. the indices for the forum are probably missing, too. I am still convinced that the patch is core worthy. ------------------------------------------------------------------------ Thu, 19 May 2005 04:36:09 +0000 : Dries It wouldn't hurt if more people would benchmark this patch. The patch's current performance worries me. Did you check your watchdog messages after upgrading the drupal.org database? Depending on your settings, errors might only be shown in the watchdog. I'll look into the remaining glitches as time permits. Thanks for your persistence in keeping this patch up-to-date. :) ------------------------------------------------------------------------ Thu, 19 May 2005 11:59:22 +0000 : killes(a)www.drop.org Dries: Can you please let me have your updated database? I want to have a look at it and try my own benchmarks with it. And yes, if I did learn something on this project is how to be persistant. ;) ------------------------------------------------------------------------ Fri, 24 Jun 2005 16:25:34 +0000 : killes(a)www.drop.org Here is an idea that occurred to me: The problem with the upgrade process is that keeping the existing revisions requires a lot of work to do. This generates a huge amount of sql queries for a large database and also requires a huge amount of memory. My suggestions is to let update.php only handle the basic upgrade, ie without old revisions. An additional module could be created that would implement a cron based approach to upgrading old revisions one node at a time. it could expose a hook to let contrib modules do their own upgrades. Dries, what do you think? (I am writing "Dries" because he seems to be the only one who is interested in getting this into core...) ------------------------------------------------------------------------ Fri, 24 Jun 2005 22:25:11 +0000 : Junyor Killes: I'm also interested in seeing this hit core. What about adding something to legacy.module to do it? ------------------------------------------------------------------------ Sun, 26 Jun 2005 21:14:54 +0000 : chx This is a sensible approach. Maybe this is the _only_ sensible approach. I have a little problem though: while the conversion is running somehow both revision handlers should be available. ------------------------------------------------------------------------ Sun, 26 Jun 2005 22:16:21 +0000 : killes(a)www.drop.org hehe, one only has to whine and bug enough and one gets some feedback. ;) @junyor: legacy.module would be a good place. my current idea is to auto-enable it in update.php and then disable it again in legacy_cron after all nodes are updated.. ;) @chx: When somebody wants to look at revisions of a node that node could be auto-updated. The only problem are contrib modules: they's need to have some hook in order to update their own data. When somebody looks at the revisions of a node than cannot be updated because the contrib module in question has no such hook, we can optionally let the user discard old revisions I guess. Dries, what do you think? ------------------------------------------------------------------------ Mon, 25 Jul 2005 16:48:51 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_30.patch (53.13 KB) Sooooo; I've updated this patch once again. Dries didn't like my idea of legacy updates so we will have an option to discard old revisions in case the update should prove difficult. Always keep a backup. ------------------------------------------------------------------------ Mon, 25 Jul 2005 17:31:12 +0000 : Bèr Kessels can we please accept and commit this patch? We can iron out any issues later. This patch is just far too big and complex to be perfect-at-once. ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:22:50 +0000 : Dries @Ber: no, we can't commit this patch blindly. Data loss is a much bigger problem than a syntax error or other code glitch. We can break Drupal but we can't break their data. I've spent quite a bit of time testing the previous version of this patch and noticed significant performance degradation. Tell me, Ber, why can't you test this patch first? ------------------------------------------------------------------------ Mon, 25 Jul 2005 19:40:37 +0000 : Bèr Kessels I was not referring to not testing it. If it is just the upgrade path that proves to be cumbersome, then why can we not fix it afterwards, I.e. when everyone has had a good chance to look at it. Testing such a huge patch requires a lot of work. something no-one just does in his spare hours. We had discussions before to deal in a slightly different way with big changes; to commit them quicker and to leave the ironing out of any left overs for the community. I hooked into that discussion here. For two reasons. One is that Gerhard has spent numerous hours on maintaining this patch. The other one is that the community can be of help with ironing out issues in such a large change, much better than that Gerhard can do on his own. And yes, dataloss is very bad, but no-one should loose data, when he/she followed the instructions (backing up)... ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:00:38 +0000 : Dries Ber: applying this patch takes 15 seconds. Whether I apply this patch for you, or whether you apply it yourself, it will hardly reduce the 'testing cost'. The problem is that data loss can be subtle; it might go unnoticed for a couple days. Make no mistake, I'd like to see this patch committed ASAP, but it warrants some testing. Let's test/benchmark it and commit it. ------------------------------------------------------------------------ Mon, 25 Jul 2005 20:55:03 +0000 : drumm The upgrade script already takes a long time to execute and does not provide feedback to the user about how it is doing. I plan on making the upgrade script able to spread the updates across multiple page views and give user feedback showing that progress is being made. This will hopefully make the speed of this update a moot point. I am working on this for my dayjob, CivicSpace, so it should get done "real soon now", but it should be expected to take awhile (smallish number of weeks). Please make comprimises if necessary. ------------------------------------------------------------------------ Wed, 27 Jul 2005 17:38:57 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_31.patch (48.78 KB) @drumm: the timeout isn't usually the problem. Memory consumption is. I'll look into doing the updates in chunks of 1000 nodes or so. Clousseau found a bug in the patch, updated. ------------------------------------------------------------------------ Thu, 28 Jul 2005 18:28:08 +0000 : moshe weitzman Even if the upgrade issues are resolved, we have to figure out why this patch is slowing down drupal (according to benchmarks posted here). ------------------------------------------------------------------------ Fri, 29 Jul 2005 07:51:45 +0000 : Dries We can't proceed without some additional benchmarking. I already benchmarked it on two machines (see comments above), and I'll benchmark it again after August 10. The performance impact of this patch worries me so if two other people could test and benchmark this patch extensively, that would come a long way. August 10 is close to the code freeze so some help is necessary. If Gerhard/killes reviewed/tested one of your patches, it is time to return the favor ... ------------------------------------------------------------------------ Fri, 29 Jul 2005 15:40:31 +0000 : Jose A Reyero After looking at the patch I'm not really surprised this slows down everything. I thought that the reason why we wanted revisions in their own table was in the first place to have a simpler -and faster- node table. But this patch: - Adds fields and complexity not only to the node table but also to all the main tables for data - Needs additional joins just to retrieve single nodes (which is done many times for a typical main page) - Does away with encapsulation of the revision functionality requiring other modules to handle revissions related data. So please, please, please: - Make revisions table only needed when actually using revisions - Keep that old nice thing of hiding revisions system from other modules - And what's so bad with serializing data anyway? I agree that usually there are better ways to store data. But this is one of the cases where serialization does make sense. Not the current huge serialized field, but only a new table with one serialized node per row. - In general, do it much simpler. This is way too complex and it removes more functionality than it adds. In just four words: keep it simple, please. ------------------------------------------------------------------------ Fri, 29 Jul 2005 21:25:26 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_32.patch (51.27 KB) Patch updated (v31 was buggy, and there was an update to updates.inc). I've had a discussion with Jose and he will maybe do some testing over the weekend. John is also doing some testing as we speak. ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:06:59 +0000 : Nick Nassar It has been exactly one year since I first submitted this patch. You know what that means: *old patch party in IRC!* Virtual beers on me! w00t! ------------------------------------------------------------------------ Sat, 30 Jul 2005 16:08:33 +0000 : Jose A Reyero Hi again. I had a talk with killes -I have to say he enlightned me about this- and also did some testing. Sorry but I dont have a good set up to provide benchmarks. Here are my conclusions -and also some explanation for the varied results abt performance: - This is faster for simple node listings - This is slower for full node/load ....thus the performance increase/decrease for a given page should depend on the relation between the number of simple node listings (usually blocks) and the number of nodes displayed in the main page. I've also experienced some trouble with attachments -not re-created when creating a new revision- and book relations, same. I'm not sure I like this new "feature" of creating a new revision every time you rollback an old one -in Drupal 4.6 this seems not to happen. Besides these details, I think this patch is good enough -and quite a powerful approach actually- though I still feel it's bit too complex when simply dealing with nodes -without revisions-, which is what we do most of the time, and complexity will be even higher with contrib modules using still more tables, not to mention flexinode -versioning info will have to be scattered everywhere in the db. So this is not a +1 nor -1. On one side I think this is powerful. On the other side I feel this is too complex and really, it's more than I need. I would be happy with a 'revision' table having: nid, vid, data(serialized), that you can forget about when not using revisions. ------------------------------------------------------------------------ Sat, 30 Jul 2005 19:16:41 +0000 : Nick Nassar On a much more serious note that my last comment, unless I'm missing something, you've introduced a subtle bug into this patch since I left it. A and B are both updating revision 5 of a node. A calls db_next_id() and gets a revision id of 6 B calls db_next_id() and gets a revision id of 7 B updates the node's revision to 7 B adds revision 7 to the revision table A updates the node's revision to 6 A adds revision 6 to the revision table Now, you have a situation where the current revision (6) isn't actually the last revision(7). For some uses, that's not a big deal, but in the case of wikis, which is what I originally hoped to use this patch for, it creates problems. ------------------------------------------------------------------------ Mon, 01 Aug 2005 00:04:14 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_33.patch (51.27 KB) Updated the patch. updates.inc was changed and chx found a typo. @Nick: I don't think the issue is due to this patch. Drupal has currently no mechanism to not let people commit a new version of the same node, whether with or without revisions. drupal_next_id ensures that the revision IDs are unique and the later revisions will be the current one. Which revision is the current one does not depend on order of the version ID, btw. @Jose: what exactly were your problems with uploads? what do I need to do to reproduce? ------------------------------------------------------------------------ Mon, 01 Aug 2005 11:33:56 +0000 : Jose A Reyero Sorry, I dont know what I did the other day but I cannot reproduce it either :-) I applied this new patch. Problems I'm having now (Testing with "stories") - Book outline is gone after creating a new revision - I cannot delete attachments (not sure it is related to your patch or to something else) - Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I? ------------------------------------------------------------------------ Sun, 14 Aug 2005 16:03:41 +0000 : killes(a)www.drop.org Jose, thanks a lot for testing! "- Book outline is gone after creating a new revision" Sounds like a bug, it should not. "- I cannot delete attachments (not sure it is related to your patch or to something else)" Another bug "- Usage of the log message: Is that mensage intended for revisions or only for book outlines? Currently the field is displayed only for outlines but shown in the revision tab" Log messages are possible for nodes of type book and page as well as all nodes that are put into the book. Log messages can be added for any node type, the module author should implement those. "Btw, the db update failed for me in patch 32, so I did it manually. I didnt do a db update again after rolling back v32 and applying v33 -should I?" No, the db update should be the same. I will try to fix the bugs you found. WRT the db update I've got an announcement to make: I will not migrate old revisions. Why? 1) it is a rather complicated process that requires a lot of memory and is likely to give people on cheap hosting trouble. 2) testing it takes a lot of time. 3) I don't personally need it I will still move old revisions to an old_revisions table from where they can be converted by a later update or a custom script. ------------------------------------------------------------------------ Tue, 16 Aug 2005 21:34:43 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_34.patch (52.92 KB) Ok, here is incarnation No. 34 of this patch. Dries agrees with me that we can put the revision rescue operation into a later update, possible using Drumm's installer. The patch fixes the issue with the file uploads not being deletable, I could not reproduce the problem with the outlines. ------------------------------------------------------------------------ Sat, 20 Aug 2005 13:33:40 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/scripts.tar.gz (4.39 KB) I've started to do some performance testing. I created a script that creates nodes that have revisions according to the new system. A node has a 50% chance to have a revision and then an equal chance fro between 1 and 25 revisions. I've created a database with 50 nodes and 500 comments, 7 vocabs and 100 total taxo terms. I've also created a script to convert these revisions to the old format (much easier than the other way around). Initial tests didn't show a significant change for /node. I'll try with more nodes and /tracker. Attached are the two scripts in a tarball. They need to be run against the patched install. Also included are the two preliminary results. ------------------------------------------------------------------------ Sat, 20 Aug 2005 15:16:04 +0000 : killes(a)www.drop.org I've now created a databse which has 1000 nodes and each node has 50 or 51 revisions. I have no taxo terms assigned and no comments (those parts didn't change). I am only using stories and pages. I am now fairly confident that this patch is a performance improvement for this kind of setup. for the new patch on the /node page I get Requests per second: 0.27 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 3592 3648 56.9 3640 3985 Waiting: 3573 3626 55.1 3620 3949 Total: 3592 3648 56.9 3640 3985 for the old one I get: Requests per second: 0.23 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 4247 4297 64.0 4286 4657 Waiting: 4156 4263 65.3 4253 4624 Total: 4247 4297 64.0 4286 4657 For the tracker the results are even better: old: Requests per second: 2.00 [#/sec] (mean) new: Requests per second: 3.64 [#/sec] (mean) Anybody interested in obtaining this database can mail me. It is 60MB (gzipped). ------------------------------------------------------------------------ Sat, 20 Aug 2005 16:33:26 +0000 : killes(a)www.drop.org Ok, I've now deleted the revisions and only kept the current one. The initial tests showed that the performance of the patch was miserable. then I remembered that I had shreddered the revisions table and did a mysql optimization. Now the results are as follows: /node: Requests per second: 3.19 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 275 312 44.3 290 499 Waiting: 255 291 44.1 268 479 Total: 275 312 44.3 290 499 Requests per second: 3.18 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 274 314 42.4 292 407 Waiting: 255 294 42.0 273 388 Total: 274 314 42.4 292 407 The lower one is the new system, there is no difference considering statistics. /tracker old: Requests per second: 3.72 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 238 268 33.8 249 332 Waiting: 219 247 33.3 230 313 Total: 238 268 33.8 249 332 new: Requests per second: 3.77 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 240 264 29.8 250 324 Waiting: 221 244 29.0 229 305 Total: 240 264 29.8 250 324 Same applies, no real difference. So, as soon as somebody tests the upgrade path the patch should be committed. ------------------------------------------------------------------------ Sat, 20 Aug 2005 17:18:39 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_35.patch (55.51 KB) Ok, I coudl hear them in advance "1000 nodes without revisions are not sufficient". Well, here are results for 30000 nodes without revisions. /node new: Requests per second: 0.45 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2162 2202 58.9 2188 2485 Waiting: 2143 2182 54.1 2169 2413 Total: 2162 2202 58.9 2188 2485 old: Requests per second: 0.43 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2319 2346 77.6 2332 2761 Waiting: 2223 2324 79.0 2311 2742 Total: 2319 2346 77.6 2332 2761 One could argue that the new approach s slightly faster. I'd like to see mroe results to believe it, but it is pretty clear it isn't slower. /tracker new: Requests per second: 0.41 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2330 2418 116.9 2403 3194 Waiting: 2302 2397 117.7 2382 3175 Total: 2330 2418 116.9 2403 3194 old: Requests per second: 0.39 [#/sec] (mean) Connection Times (ms) min mean[+/-sd] median max Connect: 0 0 0.0 0 0 Processing: 2496 2548 50.1 2536 2841 Waiting: 2477 2528 50.0 2518 2822 Total: 2496 2548 50.1 2536 2841 Statistically insignificant differences. All those tests where done with ab and as user No 1. model name : Intel(R) Celeron(R) CPU 2.40GHz 512MB RAM Besides the differences for revisions the change in served pages with different configurations is interesting. A new version is attached. node_save now returns the complete node (not only the nid). ------------------------------------------------------------------------ Mon, 22 Aug 2005 20:13:52 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_36.patch (55.66 KB) Ok, here is version 36 which incorporates today's changes in cvs. It also changes the upgrade path. I was trying my old solution and had to up my php mem limit to 80 MB in order to get it to run through. I hope this one is better but it needs testing. ------------------------------------------------------------------------ Mon, 22 Aug 2005 22:46:38 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_37.patch (52.29 KB) Ok, I've now moved _all_ of the heavy lifting to mysql. pgsql need to catch up. No problems with either time out or memory limit anymore. Since pgsql plays catch-up all the time, I set it to "ready to be committed". ------------------------------------------------------------------------ Tue, 23 Aug 2005 11:45:59 +0000 : stefan nagtegaal I could apply the patch on my localhost, and did not seen any errors nor disfunctioning drupal features while creating, revising, editing or deleting nodes.. So, +1 for this patch qua implementation it makes sense. Unfortunatly I chaven't got a clue howto benchmark, so that's for another person. @Gerhard: I really think you are a great person by rewriting this patch over and over again, untill Dries decides it is good enough to get in to the trunk. I doubt about the fact if there are more persons in our community who are like you.. People who would revise their patch 37 times, and even after that still open for discussion.. Gerhard, I think you are a great example of how a real coder should do it.. I think we can all learn from you.. ------------------------------------------------------------------------ Tue, 23 Aug 2005 17:15:29 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_38.patch (52.26 KB) Karoly pointed out that node_save(&$node) is nicer. @Steef: nice that you write that, but a _really_ good coder would only have needed one revision. ;p ------------------------------------------------------------------------ Wed, 24 Aug 2005 15:36:03 +0000 : Dries function book_update($node) { - db_query("UPDATE {book} SET parent = %d, weight = %d, log = '%s' WHERE nid = %d", $node->parent, $node->weight, $node->log, $node->nid); + if ($node->is_new || $node->revision) { + db_query("INSERT INTO {book} (nid, vid, parent, weight) VALUES (%d, %d, %d, %d, '%s')", $node->nid, $node->vid, $node->parent, $node->weight); + } + else { + db_query("UPDATE {book} SET parent = %d, weight = %d WHERE vid = %d", $node->parent, $node->weight, $node->vid); + } } Why 'is_new' and 'revision'? Isn't 'is_new' redundant? I loaded a fresh copy of the drupal.org database on my test system. How do I test this? Where is the upgrade script/path? ------------------------------------------------------------------------ Wed, 24 Aug 2005 16:31:09 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_39.patch (52.25 KB) is_new is indeed redundant since it is handeled in _update. New patch attached. The upgrate is as usual in updates.inc and accessed through update.php. ------------------------------------------------------------------------ Thu, 25 Aug 2005 10:35:24 +0000 : jakeg This is probably a little off topic but I can't find info elsewhere and this may affect revisions changes. It appears as though all revisions of a node are accessible to all users (inc. anonymous users) using ?revision=x. Surely only the 'current' revision should be accessible by default, and there should be a permission that can be assigned to roles to view non-current revisions? I don't want regular users looking at non-current revisions! A revision may have been made for various private reasons. Only admin should be able to do that. ------------------------------------------------------------------------ Thu, 25 Aug 2005 13:26:12 +0000 : Dries On the latest snapshopt of the drupal.org database I get this: ALTER TABLE {book} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {forum} ADD PRIMARY KEY vid (vid) FAILED ALTER TABLE {node_revisions} ADD KEY nid (nid) FAILED ALTER TABLE {node_revisions} ADD KEY uid (uid) FAILED ALTER TABLE {node_revisions} ADD PRIMARY KEY vid (vid) FAILED CREATE TABLE {old_revisions} SELECT nid, revisions FROM {node} WHERE revisions != '' OK ALTER TABLE {book} DROP log OK ALTER TABLE {node} DROP teaser OK ALTER TABLE {node} DROP body OK ALTER TABLE {node} DROP format OK ALTER TABLE {node} DROP revisions OK ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:30:11 +0000 : killes(a)www.drop.org That's strange, it worked for me with an older version of the same db. Anyway, I've re-rolled the patch and rearranged the update queries a bit. Also everything is now in one update, thanks to Neil for pointing that out. ------------------------------------------------------------------------ Fri, 26 Aug 2005 01:32:17 +0000 : killes(a)www.drop.org Attachment: http://drupal.org/files/issues/revisions_40.patch (52.22 KB) and the patch.

1 0

[drupal-devel] jsbug module
by Gordon Heydon 28 Aug '05

28 Aug '05

Hi, I have been experimenting with the new js stuff in drupal and using them for some clients where I can see it will make a difference. I have however found a few bugs and to make life a little easier I have created a small module which can be used to demonstrate these bugs in a standard drupal environment which should make things easier for someone to fix. If there are any other js bugs that people know about then please let me know and I will include them into this module to make it easier to fix the js bugs. At the moment there is only one bug outstanding with I have just created an issue for ( http://drupal.org/node/29839 ) which is that under IE a select list will show through the dropdown list. The second is a solved bug but I included it so it can be easily tested on older cvs versions to prove it is fixed. I hope this helps. Gordon

1 0

[drupal-devel] [bug] Disabled Picture support still tries to create a directory
by Uwe Hermann 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/18252 Post a follow up: http://drupal.org/project/comments/add/18252 Project: Drupal Version: cvs Component: user.module Category: bug reports Priority: minor Assigned to: Anonymous Reported by: budda Updated by: Uwe Hermann -Status: active +Status: patch (ready to be committed) +1. It's a nice usability improvement, applies cleanly to HEAD, and works as expected. Uwe Hermann Previous comments: ------------------------------------------------------------------------ Wed, 02 Mar 2005 13:42:08 +0000 : budda On ?q=admin/user/configure it tries to create the directory %files%/pictures even when I have picture support disabled. Surely this folder should only be created when I have the picture support enabled? Error in full: " warning: mkdir(files/xxxxxx/pictures): Permission denied in /var/www/html/drupal-4.5.2/includes/file.inc on line 77. " ------------------------------------------------------------------------ Mon, 22 Aug 2005 02:57:16 +0000 : Uwe Hermann Still applies to HEAD. ------------------------------------------------------------------------ Thu, 25 Aug 2005 22:48:58 +0000 : tostinni Attachment: http://drupal.org/files/issues/user_create_image_dir.patch (1.35 KB) The path in HEAD is admin/settings/user Here is a little patch that check first if picture support is enable and then create the path. Is it so important to check this ? And I think that your files directory should get write permissions, this is a normal behaviour of Drupal to store its files.

1 0

[drupal-devel] [bug] conf_init should strip out port numbers in bootstrap.inc
by Uwe Hermann 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/19934 Post a follow up: http://drupal.org/project/comments/add/19934 Project: Drupal Version: cvs Component: base system Category: bug reports Priority: normal Assigned to: chx Reported by: bwooster47 Updated by: Uwe Hermann Status: patch (code needs review) Attachment: http://drupal.org/files/issues/ports.patch (2.14 KB) Patch didn't apply anymore. Updated. I didn't test it, but +1 for the idea. Uwe Hermann Previous comments: ------------------------------------------------------------------------ Sun, 03 Apr 2005 23:07:58 +0000 : bwooster47 The latest Drupal from CVS (4.6.0) offers multisite configuration files, but currently the conf_init function in includes/bootstrap.inc will also include the :portnum URL specification in the directory name, which does not seem right (and the colon character may cause problems on Mac/Windows). Snippet from bootstrap.inc: * Example for a fictitious site installed at * http://www.drupal.org/mysite/test/ the 'settings.php' is * searched in the following directories: ..... If the site is: http://www.drupal.org:8080/mysite/test/ the same list of directories should be searched as listed in the example in the file, and the :8080 should not be added to the directory name (I think...) I don't know enough PHP to submit a patch, but this is probably easy for those who know this stuff :-) ------------------------------------------------------------------------ Fri, 08 Apr 2005 07:08:53 +0000 : chx Attachment: http://drupal.org/files/issues/conf_init_ports.patch (768 bytes) Here you are. ------------------------------------------------------------------------ Fri, 08 Apr 2005 11:16:52 +0000 : Dries Won't commit. Removing the port seems like a bad idea (I could have two drupal sites at two different ports but otherwise identical URLs) and adding regex' all over the map isn't advised. If ":" is an invalid character, then there might be more such characters. I suggest we replace them with '.' and that we document this behavior much like we do with $db_url. ------------------------------------------------------------------------ Mon, 11 Apr 2005 23:25:51 +0000 : chx Attachment: http://drupal.org/files/issues/conf_init_ports_0.patch (1.35 KB) ------------------------------------------------------------------------ Tue, 19 Apr 2005 17:03:46 +0000 : bwooster47 Regarding update #3 - does this mean the code has made it in, and will be available in the next release? Secondly: I follow the logic for the update, and it makes sense for non-standard HTTP port numbers, but not sure if the logic is valid when port 80 is used. For example, http://yourdomain.com/ and http://yourdomain.com:80/ both will go to the exact same page, but looks like the search for the conf.php will be different. Not sure if this matters... ------------------------------------------------------------------------ Mon, 01 Aug 2005 01:23:28 +0000 : killes(a)www.drop.org patch still applies, my testsite kept working (is not on a non-standard port, though). ------------------------------------------------------------------------ Mon, 01 Aug 2005 05:31:56 +0000 : Dries I think the PHPdoc is confusing. The second example URL is non-existing/invalid. Did you meant showing a configuration file? I agree with bwooster that this gets tricky with the default port '80'. Do we need to discuss this some more? ------------------------------------------------------------------------ Sat, 06 Aug 2005 07:20:03 +0000 : Boris Mann I like the period (".") character as the delimiter for ports. If people really do tack on :80, then site admins can symlink domain.com.80 to domain.com (much as you would need to symlink www.domain.com to domain.com as well). I think this is largely a documentation issue. ------------------------------------------------------------------------ Fri, 19 Aug 2005 01:27:39 +0000 : Morbus Iff Attachment: http://drupal.org/files/issues/_ports.patch (2.25 KB) +1. This patch is absolutely required for my home development, which is behind an ISP that doesn't like webservers on port 80. I've attached a new patch that fixes the doc mistake from the last patch, as well as adds a note to the INSTALL.txt too. I'm with Boris on the :80 issue - if a developer is using it knowingly, he should be smart enough to put 2 and 2 together. ------------------------------------------------------------------------ Fri, 19 Aug 2005 01:32:10 +0000 : Morbus Iff Attachment: http://drupal.org/files/issues/_ports_0.patch (2.25 KB) Grr. Parens in the wrong place. Right patch attached.

1 0

[drupal-devel] [feature] Multiple select functionality in profile module
by mathias 28 Aug '05

28 Aug '05

Issue status update for http://drupal.org/node/19356 Project: Drupal Version: cvs Component: profile.module Category: feature requests Priority: normal Assigned to: mathias Reported by: mathias Updated by: mathias Status: patch Attachment: http://drupal.org/files/issues/profile-multiple-select.patch (7.48 KB) This patch adds a 'multiple select' checkbox to the 'list selection' field type interface, allowing users to select multiple items from a single list. It looks the same as the 'multiple select' checkbox used during vocabulary creation. Field values are stored separately in the profile_values table and thus don't need to be serialized, keeping search functionality in tact. Note: this patch is not intended for 4.6. mathias

3 5

[drupal-devel] [bug] Block regions not initiated for newly enabled theme, so theme unusable
by nedjo 27 Aug '05

27 Aug '05

Issue status update for http://drupal.org/node/29506 Post a follow up: http://drupal.org/project/comments/add/29506 Project: Drupal Version: cvs Component: system.module Category: bug reports Priority: critical Assigned to: nedjo Reported by: nedjo Updated by: nedjo Status: patch (code needs work) Attachment: http://drupal.org/files/issues/theme_region3.patch (4.37 KB) Okay, here's a stripped down fix. This should be ready to apply--but I'll wait for someone else to affirm that before setting the status! (Aside: I gave an incorrect link, above, to the issue I created to improve update_sql() and messaging. The issue is actually at: http://drupal.org/node/29701) nedjo Previous comments: ------------------------------------------------------------------------ Wed, 24 Aug 2005 03:10:30 +0000 : nedjo Following the introduction of extentendable regions, see this issue [1], several users have reported problems: "A fresh install of Drupal with bluemarine displays blocks correctly, but when I switch the default theme to pushbutton, all the blocks disappear. They do show up in ?q=admin/block, but just with the yellow helper boxes, no blocks are really displayed.--Goba " "Same problem as Goba. The "region" field in the blocks table is sometimes blank. This makes many themes almost unusable in a Drupal default install. I don't know how the fields are blank, but the following SQL helped me get back to theme development: UPDATE blocks SET region="left" WHERE region=""--Thox " "another confirmation. I found out about the exact same sql query as Thox to fix it:) We came across this in an update. The theme had the regions defined, but since it was a non-default theme the blocks dissapeared. A drupal site, without a navigation block is simply useless.--Ber " The issue suggests a bug in the initiation of theme blocks. When a new theme is enabled, it's supposed to be initiated with blocks based on the default theme's blocks. I'll post a fix ASAP. [1] http://www.drupal.org/node/16216 ------------------------------------------------------------------------ Wed, 24 Aug 2005 04:11:57 +0000 : nedjo I can't reproduce the reported issue with a clean install. Are the themes causing these problems "updated" (are they phptemplate-based or, if not, do they have _regions() functions)? The specific problem seems to be that the region field isn't being set when block regions are initiated. Here's the relevant portion of function system_initialize_theme_blocks(): <?php // If the region isn't supported by the theme, assign the block to the theme's default region. if (!array_key_exists($block['region'], $regions)) { $block['region'] = system_default_region($theme); } ?> This is executed if a newly-enabled theme doesn't have blocks already assigned to it. It gets the default region's settings--except the region. This code portion tests whether the default region's region is available in the newly-enabled theme, and, if not, substitutes the theme's default region. This would all fail if the theme wasn't returning any regions--but any phptemplate theme, or the core ones, should be.... Anyone see what's going wrong? ------------------------------------------------------------------------ Wed, 24 Aug 2005 18:02:44 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-region-error-catching.patch (5.73 KB) I'm still not quite sure what the issue is, but clearly better error handling is needed. A particular issue is the scenario where a site admin upgrades a Drupal install without removing non-upgraded themes, or, on a new install, tries to install non-upgraded themes. As things stand, either of these could lead to unusable themes being the default, and all blocks disappearing. The attached (initial, untested) patch adds error handling to both updates.inc and system.module. New for allinstalls: * theme region initiation return FALSE on failure TRUE on success. * on initiation failure, theme is disabled, and, if it was set to default theme, bluemarine is substituted, with message given Updating: * tests if current default theme has been upgraded, and if not sets bluemarine as the default, issuing a notice * tries to initialize blocks for each enabled theme and, on failure, disables the theme. This should help prevent upgrade problems. If anyone has a chance to review and test this, I'd appreciate it, as I don't at present have a test install to upgrade. ------------------------------------------------------------------------ Thu, 25 Aug 2005 14:07:38 +0000 : Thox The problem occurs with all of the core themes except Bluemarine. That means .theme (chameleon) and phptemplate (pushbutton) themes are both affected equally. ------------------------------------------------------------------------ Thu, 25 Aug 2005 15:33:37 +0000 : Thox When I enable a theme, it goes through system_initialize_theme_blocks(), which assigns the system_default_region() to the new theme. Unfortunately, that's blank. There's no themes loaded (disabled by system_listing_save()), so there's no regions available, so there's no default region to use. Your patch stops old themes from working, can we not simply set the default region (if not found) to "left"? This works fine for me to fix this problem. ------------------------------------------------------------------------ Thu, 25 Aug 2005 16:26:03 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-region.patch (6.2 KB) This bug came from a recent change to function list_themes(), which now lists only enabled themes. function system_listing_save() fetched available regions (using, indirectly, list_themes()) *before* a theme was enabled. As well as the error handling (discussed above), designed to prevent non-upgraded themes from being enabled and to give meaningful error messages, the attached patch moves theme region initiation to after themes are enabled. On my testing, this fixes the reported error. Please test and report back. I haven't yet tested the error handling changes to updates.inc. Anyone able to do that? To test it, try to upgrade a site that has some default (and therefore, with an upgrade, updated) themes, and some themes (already enabled) that are not upgraded. The expected behaviour is that the upgraded themes have their regions initiated, while any non-upgraded themes fail, are disabled (i.e., have their status reset to 0), with messages given. ------------------------------------------------------------------------ Thu, 25 Aug 2005 17:57:34 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme_region2.patch (8.02 KB) Theme initiation was still broken for styles (unless the theme they are based on was already initialized). This is a perhaps unintended side-effect of limiting list_themes() to enabled themes--data for a theme on which a style is based are not available if that theme is not enabled. I've changed the code to load theme data directly rather than through list_themes(), so it's now I believe working. (Also fixed a minor bug in error reporting.) ------------------------------------------------------------------------ Thu, 25 Aug 2005 18:41:37 +0000 : stefan nagtegaal Nedjo, unfortunatly your patch isn't conform the drupal guidelines. See some comments here: <?php + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nOK\n"); + } ?> Why is this?? If this is _really_ needed, please use something like: <?php + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\n". t("Success") ."\n"); + } ?> Secondly: <?php + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be upgraded.' . "\nFAILED\n", array('%theme' => $theme))); ?> Should be something like: <?php + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be updated.', array('%theme' => theme('placeholder', $theme)) . "\n". t('Failed') ."\n")); ?> The difference is "Failed" instead of "FAILED". We do not use UPPERCASE errors. And I think if we use sentences as "Failed", we should use "Success" either instead of "OK".. Then you didn't made these messages translatable, and lacked the use of theme('placeholder', $text). Same here: <?php + $ret[] = array(1, t('Blocks initiated for %theme theme.') . "\nOK\n", array('%theme' => $theme))); ?> Do it like this: <?php + $ret[] = array(1, t('Blocks initiated for theme %theme.', array('%theme' => theme('placeholder', $theme)) . "\n". t('Success') ."\n")); ?> So, in short: - Try to minimise HTML in user translatable strings; - Don't use CAPITAL ERRORS, we don't do that atm; - Use theme('placeholder', $string) where needed; Though, I did not test the functionality of the patch (yet).. ------------------------------------------------------------------------ Thu, 25 Aug 2005 19:22:20 +0000 : nedjo "unfortunatly your patch isn't conform the drupal guidelines. " Well, okay. I'm just following what updates.inc already does, see function update_sql(), so that messaging is consistent. Perhaps you'd like to submit a patch to fix the allcaps issue in updates.inc. I'm not going to do that now, as my aim is to get this particular (critical) problem fixed. ------------------------------------------------------------------------ Thu, 25 Aug 2005 22:25:32 +0000 : chx No, no, no! Fix core themes, write something to update.inc if need to be and document changes. Never ever put update code in the running part of Drupal. ------------------------------------------------------------------------ Thu, 25 Aug 2005 23:08:59 +0000 : nedjo I think there's some confusion here. The portion of my patch in question *is* to update.inc--not "the running part of Drupal". ------------------------------------------------------------------------ Thu, 25 Aug 2005 23:34:55 +0000 : chx I stand corrected -- sorry. But please provide a new update_XXX function instead of patching the current the one. Stefan, in update.inc we use FAILED yes and I can't see placeholder calls. ------------------------------------------------------------------------ Fri, 26 Aug 2005 00:01:36 +0000 : drumm Updates should not make the return value themselves. <?php + $result = db_query("UPDATE {blocks} SET theme = '%s'", $default_theme); + if ($result) { + $ret[] = array('1', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nOK\n"); + } + else { + $ret[] = array('0', check_plain("UPDATE {blocks} SET theme = '$default_theme'") ."\nFAILED\n"); + } ?> I can't see how this is different from using update_sql(). <?php + if (!system_initialize_theme_blocks($theme)) { + update_sql("UPDATE {system} SET status = 0, WHERE type = 'theme' AND name = '%s'", $theme); + $ret[] = array(0, t('The theme %theme failed to initialize its blocks and so has been disabled. The theme may need to be upgraded.' . "\nFAILED\n", array('%theme' => $theme))); + } + else { + $ret[] = array(1, t('Blocks initiated for %theme theme.' . "\nOK\n", array('%theme' => $theme))); + } ?> Use drupal_set_message(t(...), 'error') instead. Monor issue: sentences should not have double spaces between them. And another update should be written to patch up whatever the last update left behind since many developers have the broken code and should not be running the same update again. ------------------------------------------------------------------------ Fri, 26 Aug 2005 00:02:17 +0000 : drumm Updating the status. ------------------------------------------------------------------------ Fri, 26 Aug 2005 05:24:31 +0000 : nedjo Attachment: http://drupal.org/files/issues/theme-regions.patch (5.65 KB) Thanks for the tips, though they kinda miss the point. My inclusion of update enhancements is evidently distracting from the main issue--so I've deleted it from the patch. There is, in any case, no bug or broken code to fix in the past update. The issue here is that the "list_theme() improvements" patch has broken theme block initialization--before anyone had a chance to see the new region functionality in action. The attached patch fixes this. Please review and apply ASAP. Aside: my manual messaging in the updates was an admittedly crude workaround to two limitations: (1) update_sql doesn't allow additional arguments (i.e., string or number query parameters) beyond the sql, and (2) since updates may do more than simply issue queries, there's a need for more flexible reporting. I agree that drupal_set_message() *could* be used for this, but I think it makes more sense to have flexible inline messages. So I've roughed in a patch to do that (and also address some of stefan's requests), see http://drupal.org/node/29002 ------------------------------------------------------------------------ Fri, 26 Aug 2005 08:34:38 +0000 : Thox Attachment: http://drupal.org/files/issues/default-regions.patch (617 bytes) Your patch works, but I don't see why it needs to be so complicated. Can you please explain why something simple such as my attached patch is not enough? Note that it would allow allow all of the old themes to work as usual and fixes the troubles with new themes. ------------------------------------------------------------------------ Fri, 26 Aug 2005 09:23:47 +0000 : robertDouglass Nedjo, I tried your most recent patch and it worked well. I haven't compared approaches, but was happy to get my site up and running with this *awsome* new feature. I had fun popping the login box into the footer, or header or content area. This is a whole new dimension of Drupal excitement. ------------------------------------------------------------------------ Fri, 26 Aug 2005 09:30:15 +0000 : stefan nagtegaal Attachment: http://drupal.org/files/issues/theme-regions_0.patch (5.66 KB) Updated patch to use theme('placeholder', $string) and deleted the double space between a translatable sentence.. Functionality seems to work.. One thing I do not like is the wording you use, though I do not have a better idea (yet).. Can someone who's native language is English look at the drupal_set_messages()'s, please? Stefan ------------------------------------------------------------------------ Fri, 26 Aug 2005 16:28:36 +0000 : nedjo Thanks for the tweaks, Stefan. "Can you please explain why something simple such as my attached patch is not enough?--Thox " It's a good question. My basic reasoning is that we shouldn't make upgrading themes optional, and shouldn't support non-complying themes. It's better to require upgrades, as we do for everything else, because failing to do so encourages lax coding and uneven support. Also, because the suggested workaround doesn't fix the bug, all themes would return simply left and right regions--and we'd be right back where we started! Overall, I'm thinking now I was mistaken to include error handling in this patch. Whether to allow non-complying code to be enabled is a worthy but separate issue. With modules, for example, we don't evaluate whether they work before enabling them--we rely on errors and the good sense of site admins to remove a non-complying module. I was thinking, "Themes are different, since setting a broken/nonupgraded theme as default could make a site unusable, since menus disappear, and novices don't know the right url to use to reset the theme to a supported one." But my approach was awkward and limited (e.g., it tested only one part of required theme functionality before disabling). If we're going to test for compliance before enabling (and I'm not convinced we should), we need to do so in a full, even way--which is to say, not in this patch! (The best, easiest thing would be simply to introduce a way for modules and themes to advertise their version support, as has been suggested and I believe begun elsewhere.) Updates *should* include better error handling; so I'll try to get in a separate update patch before the 4.7 code freeze. At any rate, the key point is to fix the bug introduced by issue #29002 [2]. I'll post a "bug-fix only" patch as soon as I can. [2] http://drupal.org/node/29002

1 0

← Newer
1
...
14
15
16
17
18
19
20
...
64
Older →

Jump to page: