- Development - lists.drupal.org

[drupal-devel] Admin help longs to be read by skilled Drupal user
by Kieran Lal 27 May '05

27 May '05

Howdy, 203 of you completed the Drupal documentation survey and told us that the number one thing you wanted was module documentation. The goal of this admin help documentation is: 1) Provide documentation for Drupal modules. 2) Improve the usability of the admin help. 3) Improve the documentation teams ability to work with admin help by single sourcing it from the documentation handbook. Here is the documentation for Drupal core:http://drupal.org/handbook/ modules Here is the documentation for some of Drupal contributions: http:// drupal.org/handbook/modules/contributions The handbook module pages start with executive summaries that are also admin help. Pretty clever, don't you think. The handbooks are currently under development, stay tuned for more improvements. It would be greatly appreciated if knowledgeable Drupal users could read some of them and report back on where they are wrong. If you need a site to confirm this documentation, create an account on: http://demo.civicspacelabs.org/home and send me a link to your profile. I'll make sure you get the access you need to confirm the documentation is accurate. Thanks for you help, Kieran

1 0

[drupal-devel] [bug] New pager code broken taxonomy pagers
by Steven 27 May '05

27 May '05

Issue status update for http://drupal.org/node/23687 Project: Drupal Version: cvs Component: base system Category: bug reports Priority: normal Assigned to: Morbus Iff Reported by: Morbus Iff Updated by: Steven Status: patch Note: if you need to fetch a list of tags ordered by hierarchy in one query, you should look at how comment.module does it. Steven Previous comments: ------------------------------------------------------------------------ May 26, 2005 - 21:44 : Morbus Iff I've not yet looked into it, but I doubt I'll be able to fix it (the original taxonomy/term pager code was a hack that moshe taught me about). In a nutshell, though, the recent pager cleanup in head has broken the term pagers for vocabularies. ------------------------------------------------------------------------ May 26, 2005 - 22:21 : Morbus Iff Attachment: http://drupal.org/files/issues/_p_23687_taxopager.patch (1.88 KB) Patch attached. ------------------------------------------------------------------------ May 27, 2005 - 09:05 : Steven Why can't this code use pager_query() and query the database directly? If free tagging vocabularies can get so big, it seems a very inefficient thing to fetch the entire list.

1 0

[drupal-devel] [bug] New pager code broken taxonomy pagers
by Steven 27 May '05

27 May '05

Issue status update for http://drupal.org/node/23687 Project: Drupal Version: cvs Component: base system Category: bug reports Priority: normal Assigned to: Morbus Iff Reported by: Morbus Iff Updated by: Steven Status: patch Why can't this code use pager_query() and query the database directly? If free tagging vocabularies can get so big, it seems a very inefficient thing to fetch the entire list. Steven Previous comments: ------------------------------------------------------------------------ May 26, 2005 - 21:44 : Morbus Iff I've not yet looked into it, but I doubt I'll be able to fix it (the original taxonomy/term pager code was a hack that moshe taught me about). In a nutshell, though, the recent pager cleanup in head has broken the term pagers for vocabularies. ------------------------------------------------------------------------ May 26, 2005 - 22:21 : Morbus Iff Attachment: http://drupal.org/files/issues/_p_23687_taxopager.patch (1.88 KB) Patch attached.

1 0

[drupal-devel] Drupal.org cache statistics
by Dries Buytaert 27 May '05

27 May '05

Hello world, november last year, I profiled drupal.org's cache observations. Yesterday, Moshe asked to profile it again so we could evaluate the usefulness of Jeremy's "loose caching" mechanism. The past 20 hours, I logged 93.000 unique page requests using the patch at http://buytaert.net/temporary/cache-statistics.patch. Loose caching was enabled. Results: 1. Last year we found that authenticated users were responsible for 15,8 % of all page views. A year later, we see that authenticated users are responsible for 14,9% of all page views. 2. Last year we found that only 27.9% of the page requests actually benefited from the cache system. That is, for more than 2/3th of the page requests, we had to generate a page dynamically. A year later, using "loose caching" rather than "strict caching", we see that 30,7% of the page requests benefit from the cache system. Read: we still have a lot of page cache misses! :( 3. Last year, we found that the cache got flushed once every 207 page requests. A year later, we observe that the cache got flushed once every 190 page requests. We conclude that: 1. Loose caching does not significantly -- or not necessarily -- improve the behavior of drupal.org's page cache (though I'd like to believe that it does when there are sudden traffic spikes/bursts). 2. When writing code, we can NOT assume that a page will benefit from being cached. -- Dries Buytaert :: http://www.buytaert.net/

7 9

[drupal-devel] [feature] Replace $_SERVER["REMOTE_ADDR"] with a function which understands X-Forwarded-For
by degerrit 27 May '05

27 May '05

Issue status update for http://drupal.org/node/20471 Project: Drupal Version: 4.6.0 Component: base system Category: feature requests Priority: normal Assigned to: degerrit Reported by: degerrit Updated by: degerrit Status: patch I will re-write the patch with Dries' suggestions, I haven't had much time this week. I found some interesting examples of scripts doing roughly the same, none of them seem to do input checking (which I hadn't thought of either, but is obviously very needed!). Good thing there are people like Dries around, eh! There's also another possible header: HTTP_CLIENT_IP, I've never seen it but then I don't log all my clients' headers either ;-) http://be2.php.net/getenv (the 2nd comment at this time) http://php.planetmirror.com/manual/en/function.getallheaders.php http://ip-to-country.webhosting.info/node/view/212 http://dev.wp-plugins.org/file/wp-contact-form/trunk/wp-contactform.php?rev… degerrit Previous comments: ------------------------------------------------------------------------ April 14, 2005 - 00:42 : degerrit In several modules, replace $_SERVER["REMOTE_ADDR"] with a function which understands X-Forwarded-For. Proxy servers often (sometimes, always?) insert "X-Forwarded-For" headers which contains the "real" client IP address. I think it is useful to log the client IP address in that situation. This little snippet works for me, and I'd suggest something similar be made into a function? $headers = apache_request_headers(); if (array_key_exists('X-Forwarded-For', $headers)){ $hostname=$headers['X-Forwarded-For'] . ' via ' . $_SERVER["REMOTE_ADDR"]; } else { $hostname=$_SERVER["REMOTE_ADDR"]; } I can make a patch proposal if people like the idea. ------------------------------------------------------------------------ May 21, 2005 - 22:44 : degerrit Attachment: http://drupal.org/files/issues/remote_addr.diff (6.07 KB) Here's my proposal for showing user IP addresses behind proxy servers, especially in log files. It adds a function remote_addr() and replaces all the uses of $_SERVER["REMOTE_ADDR"] with that function. This is my first patch for Drupal, don't be too harsh one me :-) A few notes: - I patched session.inc because $_SERVER["REMOTE_ADDR"] is also called in session.inc and common.inc is not available at that time it seems - maybe I replaced too vigorously in common.inc and sessions.inc, it works for me though - maybe it should be renamed http_remote_addr or apache_remote_addr ------------------------------------------------------------------------ May 22, 2005 - 09:37 : Dries Couple comments: Drupal supports non-Apache webservers: apache_request_headers() might not always be available. Does this patch need input checking? Depends on how/when/where the remote address data is used. Fact is that we don't want people to inject XSS-stuff through HTTP headers. Make sure to double-check this. I'd rename the function to remote_address(). We usually don't abbreviate words like 'address'. The word 'via' can't be translated. We write } else { and not } else {Just thinking up loud. Is it easy to fake or spoof X-Forwarded-For-headers? If it is, it would be easy to trick Drupal (eg. in the statistics.module's log pages we group by remote address). For the watchdog it doesn't matter, but for the statistics module it does. Maybe we need a toggle so remote_address() can be told not to return the 'true IP address'. ------------------------------------------------------------------------ May 25, 2005 - 05:37 : Steven X-Forwarded-For can be spoofed without any trouble. Just connect and send the fake header. I don't think we should pay attention to it. If we do, we need to also store the original IP address. Perhaps we could do "real.ip.#.# (forwarded.ip.#.#)" ------------------------------------------------------------------------ May 25, 2005 - 05:39 : Steven Sorry didn't look at the patch carefully enough. Yes, we need XSS protection: in practice, we always strore remote_addr() in the hostname column in the db. Wherever that column is displayed, it needs to be check_plain()ed.

1 0

[drupal-devel] [bug] broken t()s in common.inc
by killes 27 May '05

27 May '05

Issue status update for http://drupal.org/node/23700 Project: Drupal Version: cvs Component: base 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 Attachment: http://drupal.org/files/issues/broken-t.patch (1.1 KB) there are two broken t() function in common.inc.The patch shoul dalso be applied to 4.6 (will make a new one if it shoul dnot apply). killes(a)www.drop.org

1 0

[drupal-devel] [bug] Forum module's use of if in Postgresql
by patrickslee 26 May '05

26 May '05

Issue status update for http://drupal.org/node/23499 Project: Drupal Version: 4.6.0 Component: forum.module Category: bug reports Priority: normal Assigned to: patrickslee Reported by: cchamilt Updated by: patrickslee Status: patch Attachment: http://drupal.org/files/issues/database.pgsql_3.patch (363 bytes) Here you have it. patrickslee Previous comments: ------------------------------------------------------------------------ May 24, 2005 - 16:32 : cchamilt Postgresql function "if" is defined to return a boolean value, yet forum requests an integer - last_comment_uid. user error: query: SELECT n.nid, l.last_comment_timestamp, IF(l.last_comment_uid, cu.name, l.last_comment_name) as last_comment_name, l.last_comment_uid FROM node n, node_comment_statistics l /*! USE INDEX (node_comment_timestamp) */, users cu, term_node r WHERE n.nid = r.nid AND r.tid = 1 AND n.status = 1 AND n.type = 'forum' AND l.last_comment_uid = cu.uid AND n.nid = l.nid ORDER BY l.last_comment_timestamp DESC LIMIT 1 OFFSET 0 in /var/www/hosts/www.ambigc.com/htdocs/includes/database.pgsql.inc on line 62. warning: pg_query() [function.pg-query]: Query failed: ERROR: function if(integer, character varying, character varying) does not exist HINT: No function matches the given name and argument types. You may need to add explicit type casts. in /var/www/hosts/www.ambigc.com/htdocs/includes/database.pgsql.inc on line 45. ------------------------------------------------------------------------ May 25, 2005 - 10:07 : patrickslee Attachment: http://drupal.org/files/issues/forum.module_3.patch (2.69 KB) I have also encountered this bug and made a possible patch by replacing IF statements to CASE. A patch file is attached. ------------------------------------------------------------------------ May 25, 2005 - 10:28 : adrian There is an IF function defined in the database.pgsql function. On what page is this happening, as what might be happening is that the type isn't being picked up correctly. It's defined as IF(anyelement, anyelement, text) returns anyelement. IIRC. What I have seen happen is that it wants the text field to return integer, or vice versa. ------------------------------------------------------------------------ May 25, 2005 - 10:32 : patrickslee I found the definition in database.pgsql. It is like this: CREATE FUNCTION "if"(boolean, anyelement, anyelement) RETURNS anyelement AS ' SELECT CASE WHEN $1 THEN $2 ELSE $3 END; ' LANGUAGE 'sql'; So it is actually wanting boolean. ------------------------------------------------------------------------ May 25, 2005 - 10:43 : patrickslee So my patch will work, and we can make it easier by replacing the IF function to: IF(l.last_comment_uid 0 AND l.last_comment_uid NULL, cu.name, l.last_comment_name) BTW there is another place in forum.module use IF like this. ------------------------------------------------------------------------ May 25, 2005 - 10:45 : patrickslee Oops... my <>s are stolen. IF(l.last_comment_uid <> 0 AND l.last_comment_uid <> NULL, cu.name, l.last_comment_name) ------------------------------------------------------------------------ May 25, 2005 - 17:20 : cchamilt Yes that line works great. Thanks for your help! ------------------------------------------------------------------------ May 25, 2005 - 17:52 : patrickslee This patch shows another problem. The last post user will always show "Anonymous" or whatever you defined. Looking for solution... ------------------------------------------------------------------------ May 25, 2005 - 18:00 : patrickslee Use this line instead: IF(l.last_comment_uid <> 0 AND l.last_comment_uid <> NULL, cu.name, l.last_comment_name) Don't forget there are two places in forum.module need to be changed. ------------------------------------------------------------------------ May 25, 2005 - 18:10 : patrickslee Oh what have I done! I changed nothing! IF(l.last_comment_uid <> 0 AND l.last_comment_uid IS NOT NULL, cu.name, l.last_comment_name) ------------------------------------------------------------------------ May 25, 2005 - 22:12 : adrian perhaps we could also just add more wrapper functions for the specific types ? create or replace function IF (integer, anyelement, anyelement) returns anyelement as ' SELECT if (($1 0) AND NOT nullvalue($1), $2, $3); ' LANGUAGE 'sql'; Or something. ------------------------------------------------------------------------ May 25, 2005 - 22:41 : plj '' was missing. So what adrian actually means is this: CREATE OR REPLACE FUNCTION IF (integer, anyelement, anyelement) RETURNS anyelement AS ' SELECT if (($1 <> 0) AND NOT nullvalue($1), $2, $3); ' LANGUAGE 'sql'; I applied that and now forums work -- no module patches. PostgreSQL 7.4.7. ------------------------------------------------------------------------ May 25, 2005 - 22:45 : plj Sheesh. I meant that '<>' was missing. Gotta learn to use that preview button... ------------------------------------------------------------------------ May 26, 2005 - 09:53 : patrickslee Is nullvalue() a function? I cannot find it in neither postgres manual nor mysql manual. Use "IS NOT NULL" would be better as it is more standard-compatible. SO we have this: CREATE OR REPLACE FUNCTION IF (integer, anyelement, anyelement) RETURNS anyelement AS ' SELECT IF(($1 <> 0) AND $1 IS NOT NULL, $2, $3); ' LANGUAGE 'sql'; Looks we have a satisfying fix here. Is this fix going to be committed in CVS for next release? ------------------------------------------------------------------------ May 26, 2005 - 12:26 : adrian hostmaster=# \df nullvalue List of functions Result data type | Schema | Name | Argument data types ------------------+------------+-----------+--------------------- boolean | pg_catalog | nullvalue | "any" But it doesn't matter to me =) ------------------------------------------------------------------------ May 27, 2005 - 07:21 : Dries Can you guys provide a patch? I don't have a PostgreSQL setup to test it against. Will commit to DRUPAL-4-6 and HEAD.

1 0

[drupal-devel] [bug] New pager code broken taxonomy pagers
by Morbus Iff 26 May '05

26 May '05

Issue status update for http://drupal.org/node/23687 Project: Drupal Version: cvs Component: base system Category: bug reports Priority: normal Assigned to: Morbus Iff Reported by: Morbus Iff Updated by: Morbus Iff Status: patch Attachment: http://drupal.org/files/issues/_p_23687_taxopager.patch (1.88 KB) Patch attached. Morbus Iff Previous comments: ------------------------------------------------------------------------ May 26, 2005 - 14:44 : Morbus Iff I've not yet looked into it, but I doubt I'll be able to fix it (the original taxonomy/term pager code was a hack that moshe taught me about). In a nutshell, though, the recent pager cleanup in head has broken the term pagers for vocabularies.

1 0

[drupal-devel] [task] Sensical status messages for forum admin
by drumm 26 May '05

26 May '05

Issue status update for http://drupal.org/node/19621 Project: Drupal Version: cvs Component: forum.module Category: tasks Priority: normal Assigned to: drumm Reported by: drumm Updated by: drumm Status: patch Attachment: http://drupal.org/files/issues/forum.module_3.diff (6.73 KB) We are making forums and forum containers here, not terms. This patch returns a status so the calling code can set the appropriate message (or no message at all). This has the happy side effect of preventing any potential problems with calling this from when cron is running and other similar situations. The define()s are done in common.inc to encourage similarly architected code; separate UI from DB manipulation. drumm

7 9

[drupal-devel] [bug] "Create content" menu displays extraneous links
by JonBob 26 May '05

26 May '05

Issue status update for http://drupal.org/node/18275 Project: Drupal Version: cvs Component: node.module Category: bug reports Priority: normal Assigned to: Anonymous Reported by: JonBob Updated by: JonBob Status: patch Attachment: http://drupal.org/files/issues/node_30.patch (509 bytes) In the case where a node module defines no node types, like Flexinode or CCK when no types have yet been defined, the node/add menu displays links that do not work. One-line fix attached. JonBob

4 3