Posted by nmudgal on March 17, 2012 at 4:43pm
There should still be some review process after one is given full git access, as there are many people who take it in wrong sense & put modules as which are not may be required.
Just some review process so that same quality which is checked while approving project applications during giving git access can also be maintained later.
Do we really require those projects which may not follow standards to be there ?
Like http://drupal.org/project/contact_profile
No offense, but is it fine to put a module with minimal code [e.g. drupal_alter just] as full project?
This may help in maintaing quality more.
Just a thought!
Thanks.
Comments
imho
We can't even get it managed to review the current applications - so how can we get it managed to review already approved projects? - Would be nice - but It's simply not possible.
Size of a module has nothing to do with it's quality. - if there's a use case for it - why not.
The reason we require a module with enough code to review it seriously is not against that there can't be small modules created by already approved users.
Yeah it's true actually that
Yeah it's true actually that it would be hard to manage review queue of approved projects but won't say not possible, as few months back it used to look same with project application queue.
Also size of a module doesn't really matter, but at least there should be discussion which will let community know regarding any duplicity, following module standards or how can that be done in better way.
Yes there can be small module but what if it is achievable by configuring some other contrib modules?
See "quality" discussion at "What are other communities doing?"
See http://groups.drupal.org/node/178364#comment-592104
And what followed from that, see http://groups.drupal.org/node/180444#comment-599769
In short: drupal.org is a different community and may have different values, but, contributed code is often regarded as positive no matter whether the quality level has been "officially" reviewed in any way. "Unofficial" reviews in the form of votes/stars and text comments can't hurt Drupal.org, and might help.
Help who, now?
I dare say, the people who count aren't self-appointed quality gatekeepers barring "crap" from drupal.org (excepting matters of "crap" related to Security), the people who count are end-users seeking potential solutions on drupal.org. We can't crowdsource code reviews, but I'll bet we can effectively crowdsource that other kind of quality checking (see Community reviews/ratings for Projects).
I did not take a look at the
I did not take a look at the code of that project, but i checked the usage statistics: At times there have been more than 60 REPORTING sites using the module, so i guess there is some kind of demand. Remember that probably majority of CMS users are not able to create custom modules, so even small contributions can make a big difference. We cannot put everything in the core.
However, the quality of the code should be evaluated in all cases, and we can only hope that the community handles that. All we can really do is to make sure that the contributors UNDERSTAND the coding standards and best practices, and that's what the project issue queue is already doing.
Help who -> drupal community
Help who -> drupal community to grow more & better, I suppose.
There are no gatekeepers surely & there should never be. Community review/ratings for projects is one solution, that would help in creating statistics which would help eventually user choose what's best.
We can't crowdsource but I also think there are who would love to help :-)
@bfr As I mentioned earlier, quality of code doesn't depend on factors such as number of lines. Yes as patrickd also mentioned if there is use-case, then why not.
Project application does great job of familiarizing contributors with drupal best code practices, but there could be a platform or something which will get new approved projects in some kind of spotlight so community can review it, find issues out & help in maintain it's quality irrespective of size of module.