On Tuesday 07 August 2007, Neil Drumm wrote:

On 8/6/07, Larry Garfield <larry@garfieldtech.com> wrote:

...

It doesn't seem to help, though, since just today I saw a new module[1] cross the wire that is a duplicate of one I submitted a few months ago[2].

[1] http://drupal.org/project/default_filter [2] http://drupal.org/project/filterbynodetype

I actually win this one with 11 months: http://cvs.drupal.org/viewvc.py/drupal/contributions/modules/input_format_r estrictions/

Hm, touche. I don't see it in the project list[1], though. Methinks some merging may be in order. Or perhaps it's a sign that this functionality has no business not being in core, so we should make sure it gets in for Drupal 7. :-) [1] http://drupal.org/project/Modules -- Larry Garfield AIM: LOLG42 larry@garfieldtech.com ICQ: 6817012 "If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of every one, and the receiver cannot dispossess himself of it." -- Thomas Jefferson

Here's the problem. We are creating yet another 'voluntary' procedure that new people will not know about, old time contributors who aren't paying attention to list right now won't know about and I suspect will be abandoned and unmaintained like many similar such initiatives in the past. About the only real solution is to have several people who are familiar with the existing modules review the ideas and point to existing modules. Of course this would really have to be a group of people that aren't doing stuff now. They would have to build the proceedures and habits into the community to make it really last beyond this current surge of interest. -sp On 8/7/07, Augustin (Beginner) <drupal.beginner@wechange.org> wrote:

On Tuesday 07 August 2007 22:18, Larry Garfield wrote:

...

...

...

It doesn't seem to help, though, since just today I saw a new module[1] cross the wire that is a duplicate of one I submitted a few months ago[2].

[1] http://drupal.org/project/default_filter [2] http://drupal.org/project/filterbynodetype

I actually win this one with 11 months: http://cvs.drupal.org/viewvc.py/drupal/contributions/modules/input_format _r estrictions/

Hm, touche. I don't see it in the project list[1], though.

And here is a fourth one: http://drupal.org/project/filter_default

4 modules to have different default input format by role AND/OR by node types, one of which has never been published on d.o. (and never tagged for any particular core branch).

You might want to get in touch with the maintainer of the fourth module, decide on which of the 4 projects becomes the semi-official one, and abandon the other three...

A.

-- http://www.wechange.org/ Because we and the world need to change.

http://www.reuniting.info/ Intimate Relationships, peace and harmony in the couple.

Didn't we just have 4 existing modules doing almost the same thing by long time contributors who already know this in theory? There was a time until 4.7 when I did in fact know every module in the contrib repository. I don't have the time or inclination anymore. Something like this is really only solved by a human filter. Anything that involves a static list being checked has consistently failed in the past. -sp On 8/7/07, Chris Kennedy <chrisken@mail.utexas.edu> wrote:

What about this: before you make a new module, check if it already exists. If true, abort.

Steven Peck wrote (8/7/2007 7:39 PM):

...

Here's the problem.

We are creating yet another 'voluntary' procedure that new people will not know about, old time contributors who aren't paying attention to list right now won't know about and I suspect will be abandoned and unmaintained like many similar such initiatives in the past.

About the only real solution is to have several people who are familiar with the existing modules review the ideas and point to existing modules. Of course this would really have to be a group of people that aren't doing stuff now. They would have to build the proceedures and habits into the community to make it really last beyond this current surge of interest.

-sp

On 8/7/07, Augustin (Beginner) <drupal.beginner@wechange.org> wrote:

...

On Tuesday 07 August 2007 22:18, Larry Garfield wrote:

...

...

...

It doesn't seem to help, though, since just today I saw a new module[1] cross the wire that is a duplicate of one I submitted a few months ago[2].

[1] http://drupal.org/project/default_filter [2] http://drupal.org/project/filterbynodetype

I actually win this one with 11 months: http://cvs.drupal.org/viewvc.py/drupal/contributions/modules/input_format _r estrictions/

Hm, touche. I don't see it in the project list[1], though.

And here is a fourth one: http://drupal.org/project/filter_default

4 modules to have different default input format by role AND/OR by node types, one of which has never been published on d.o. (and never tagged for any particular core branch).

You might want to get in touch with the maintainer of the fourth module, decide on which of the 4 projects becomes the semi-official one, and abandon the other three...

Personally, I don't think duplicated modules are that big of a problem and it is certainly something we should _not_ punish people for. I'm a firm believer of the fact that "deliberate duplication" is a good. "Unfortunate duplication", i.e. duplication as the result of the fact that the author didn't know an almost identical module already existed, is something we have to reduce though.

I have a pretty good idea of existing modules as they stand now, not all, but a fair proportion, and would be interested in helping with this. When I read the first e-mail in this thread I immediately thought of a "duplicated module hall of shame" (possibly with a nicer title) wiki in the contributed module status group or similar, to serve as a reference point for people to avoid yet another duplicate. I can think of at least 3 candidate sets of module for this already.

-- Dries Buytaert :: http://www.buytaert.net/

On 09 Aug 2007, at 07:54, Neil Drumm wrote:

...

Personally, I don't think duplicated modules are that big of a problem and it is certainly something we should _not_ punish people for. I'm a firm believer of the fact that "deliberate duplication" is a good. "Unfortunate duplication", i.e. duplication as the result of the fact that the author didn't know an almost identical module already existed, is something we have to reduce though.

Is the problem not looking, or looking and not finding? I doubt we can make a lot of progress telling people to look if they are not motivated to do it. Improving module findability would be a bigger help. The module creators who are motivated would have an easier job. And the much larger group of module users would be able to find modules easier.

I agree. A lot of these groups, like the contributed-module-ideas group (to some extend), are quick ugly hacks to work around an underlying problem. I'd prefer to see people spend their time on improving the underlying problem (i.e. module findability). It keeps amazing me why people aren't a little bit more critic towards these kind of hacks. I guess other people will fix the module findability problem then ... -- Dries Buytaert :: http://www.buytaert.net/

Such a search block should also have a drop down for the version, IMO. Larry Garfield wrote:

On Thu, 9 Aug 2007 14:06:51 +0200, Dries Buytaert <dries.buytaert@gmail.com> wrote:

...

I agree. A lot of these groups, like the contributed-module-ideas group (to some extend), are quick ugly hacks to work around an underlying problem. I'd prefer to see people spend their time on improving the underlying problem (i.e. module findability). It keeps amazing me why people aren't a little bit more critic towards these kind of hacks. I guess other people will fix the module findability problem then ...

My apologies to dww before I even say anything... :-)

Would it be difficult, once the search module is back on, to offer somewhere a search box that is pre-set to only search project nodes? I have found in the past that doing an Advanced search for my keyword(s) restricted to "project" nodes (shouldn't that be Project now?) is a surprisingly good way to track down a module. It's a PITA to have to do, though, especially when d.o is slow. Could we have some sort of optional-default-off "project search" block that just has that restriction pre-set? It doesn't need to be fancy, just forward to the normal search system for results.

And while we're on the subject, I'll reiterate my previous point that for 99% of all users, including those with CVS access, something in a person's sandbox does not exist.

--Larry Garfield

-- Sean Robertson Web Developer NGP Software, Inc. seanr@ngpsoftware.com (202) 686-9330 http://www.ngpsoftware.com

On Aug 9, 2007, at 5:06 AM, Dries Buytaert wrote:

A lot of these groups, like the contributed-module-ideas group (to some extend), are quick ugly hacks to work around an underlying problem. I'd prefer to see people spend their time on improving the underlying problem (i.e. module findability).

Yes and no. I agree module "findability" (nice word) is currently not ideal on d.o. Yes, I'd love to see other people help with this. However, this isn't the only problem here, which is why I support some kind of way for people say "I've got an idea for a module to do X, what do y'all think?". Many modules in contrib are so powerful and flexible now, that lots of things can be solved by creatively bolting of existing things together. For example, the answer to "I think I need a module for X" might be "no, all you need is to setup a node_queue for that, and use a view that does blah with a filter on node_queue exclusion". I'd have no friggin' clue to search for "node_queue" or even to know about node_queue exclusion views, but that might just be the answer to my problem (this is a real example from a conversation I had with Earl not too long ago). I'm sure I could come up with countless other examples of problems solved by non-obvious combinations of existing powerful contribs, but I think everyone will grant that this is true without more evidence. Therefore, while I'm fully in favor of people spending cycles on making modules easier to find on d.o, I'm not opposed to a more structured place for people to say "I think I need to write a module to allow X, should I do it?" and get informed answers from people more familiar with the existing state of the art. Cheers, -Derek

Quoting Neil Drumm <drumm@delocalizedham.com>:

On 8/8/07, Dries Buytaert <dries.buytaert@gmail.com> wrote:

...

Personally, I don't think duplicated modules are that big of a problem and it is certainly something we should _not_ punish people for. I'm a firm believer of the fact that "deliberate duplication" is a good. "Unfortunate duplication", i.e. duplication as the result of the fact that the author didn't know an almost identical module already existed, is something we have to reduce though.

Is the problem not looking, or looking and not finding? I doubt we can make a lot of progress telling people to look if they are not motivated to do it. Improving module findability would be a bigger help. The module creators who are motivated would have an easier job. And the much larger group of module users would be able to find modules easier.

I think it is more looking and not finding that is the issue. The Modules, Themes and Translations pages are clearly highlighted at d.o. but only those that are published. New works and redesigns need some attention. I'm inclined to say that the default view of the modules page should be by name rather than category but then there are hundreds of modules so this probably isn't feasible. I see more categories being added recently probably in hopes of modules being more easily found but now we're approaching category duplication with different names; e.g. "Commerce/Advertising" vs "e-Commerce module" or "Content" vs "Content display". IMO, we need a sub categorization to give further depth or meaning to what someone might be looking for to allow the user to drill down to few modules. 210 modules in one category is asking a lot for someone to review when that someone is probably clueless anyway. I might as well review by name and use the browser find function to maybe find one. Earnie

RE: "As to the "voluntary" aspect of this proposal, I have no authority to make this mandatory." Really, none of this is mandatory. When it comes to contributing back to the project, we are all, effectively, volunteers. If there is another mechanism (in addition to research before starting to code, communicating with maintainers of similar modules, posting plans to the dev list, etc) for reducing duplication of effort, great. If people use it, even better. Thanks for starting up the group. I'm hitting send, and then subscribing at g.d.o -- Cheers, Bill John Wilkins wrote:

On Aug 7, 2007, at 4:39 PM, Steven Peck wrote:

...

Here's the problem.

We are creating yet another 'voluntary' procedure that new people will not know about, old time contributors who aren't paying attention to list right now won't know about and I suspect will be abandoned and unmaintained like many similar such initiatives in the past.

The reason I created the group is I have an honest interest in knowing about new modules; I regularly track 5.x releases at http://drupal.org/taxonomy/term/78/0/feed . And I suspect that there are many others with the same interest; if we can get several in one group, we can keep the group active. And I absolutely agree that it's important to educate existing developers who aren't watching the list right now.

If there are no objections to the idea, I will be looking for appropriate docs to update to point at the group, so that it does become part of the new module process.

As to the "voluntary" aspect of this proposal, I have no authority to make this mandatory. And I suspect that those that do have that authority would like to see proof that it is working before making it mandatory.

...

About the only real solution is to have several people who are familiar with the existing modules review the ideas and point to existing modules.

I agree. We need a group of people with knowledge of existing modules. I'm hoping to attract such people to the Contributed Module Ideas group.

- John

John Wilkins | Albin.Net | nickname: JohnAlbin

-- Bill Fitzgerald http://www.funnymonkey.com Tools for Teachers 503.897.7160

I think that that is the point that John is trying to make. People are duplicating work and code and polluting the namespace. If you could post to somewhere, saying: 'I've got this idea', then someone could let you down gently, and say they'd already done it. Then you'd be free to extend their work, and improve upon it, collaboratively. Otherwise d.o is going to end up having a ton of modules, that all do roughly the same thing, (RSS aggregation anyone?) On 07/08/07, Larry Garfield <larry@garfieldtech.com> wrote:

I forget where exactly I found it on the site, but there's already an RSS feed for new modules:

http://drupal.org/taxonomy/term/14/0/feed

It doesn't seem to help, though, since just today I saw a new module[1] cross the wire that is a duplicate of one I submitted a few months ago[2].

[1] http://drupal.org/project/default_filter [2] http://drupal.org/project/filterbynodetype

On Monday 06 August 2007, John wrote:

...

Hi all,

When a developer has an idea for a module, they are often not the only one with that idea. Currently, new modules are mostly invisible until they show up as a project. And IDEAS for new modules are / completely/ invisible unless they are discussed in the ephemeral IRC. As I'm sure you all know, this leads to code dupe.

If we had a central location for announcing modules, we could: * Reduce module duplication. We can prevent two independent developers from writing separate, duplicate modules. * Increase developer collaboration. We can get interested developers working together at an early stage of development. * Improve module quality. Others can shape and improve the ideas, leading to a better module from the beginning (or, at the least, a better road map for future development.)

Rather than cluttering up the drupal-devel list with module announcements, I created a Drupal Group: http://groups.drupal.org/ contributed-module-ideas (the name was chosen to be in-line with the contributed-module-status group.)

If we can get enough buy-in to this idea, I think it would be a good idea to add a note to the "Maintaining a project on drupal.org" page ( http://drupal.org/node/7765 ) that modules should be announced on the Contributed Module Ideas group, in addition to requesting a CVS account. This might also help out CVS Administrators if more eyeballs were reviewing the modules associated with new CVS account requests.

What does everyone think? If you have a specific criticisms or enhancements, reply to the list. If you simply have "+1" comment, why not just subscribe to the group?

- John

John Wilkins | Albin.Net | nickname: JohnAlbin

-- Larry Garfield AIM: LOLG42 larry@garfieldtech.com ICQ: 6817012

"If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of every one, and the receiver cannot dispossess himself of it." -- Thomas Jefferson

-- Regards Steven Jones