Services dev team
| Project: | Services |
| Version: | 7.x-1.x-dev |
| Component: | Miscellaneous |
| Category: | feature request |
| Priority: | normal |
| Assigned: | Unassigned |
| Status: | active |
Jump to:
Hi, once of the things I'm currently missing for this module is the activity in the issue queue. It is being quite slow I'd say. Before considering this a blame, let me explain. I'm really interested in having a funky version of the services for Drupal 7, and seriously I started to contribute and talk to the people I've found on IRC, or email. For the purpose of having services for D7 working, I also contacted other maintainers (cweagans, e.g.) to see if other services could fit in the D7 version of the services module (in this case, blogapi in on the way ;) ). I went to irc again, and asked about being involved, and the re-design ideas for the module.
After two weeks, the issues being active are only mine, the patches are only mine, and I'm focusing the whole discussion. This could not be good, and would require more activity from the current project maintainers. Heyrocker and Robloach also provided some feedback, and are at least available on irc.
So, just want to remember the D7CX pledge, and the clock tick-tacking smoothly. In somehow we need to make the discussion active, alive, and reviewable :)
The issue queue is slow, and conversation are moving on IRC from time to time. What if we do obey our selfs and participate in real time discussions on IRC? what if we do have a channel for that (the topic is important enough to do so) so the bot can log the conversations for people not being connected (or even awake)?
I promissed I'd participate, now I, don't blame me for doing so!
Cheers,
ilo

#1
I have made #drupal-services on IRC, and Morbus has put the bot in there. I think this is a good idea, both for giving us a place to discuss issues, and also because more and more people are asking services-related questions in channel and this gives them a place to go. If we get lucky maybe we can find some regular attendees besides the maintainers. At least this way the conversation can become a bit more focused.
As far as speed, I'm dedicated to working on these patches but it is slow work for me. Small focused patches make this easier. For instance there are three patches in the queue right now and one depends on the other two (because the tests for services.module don't run properly without the patches for user and system services.) However I don't really feel I can test the user and system service patches without having the main module running. This makes the reviews a lot harder. I would, again, much prefer we simply port the existing modules and get them to the point where they mostly run before adding new stuff, including tests. It will be smaller steps but we will actually move faster because the patches are less daunting.
I have more notes and some updates to the main module patch I just need to get them together. There's a patch in the D6 queue I really want to get into the D7 branch as quickly as possible as well (the hash-marks one).
Anyone else got any suggestions on keeping things moving?
#2
Heyrocker++!
I think this is a very important step that will make the issue queue more active and push the wheel. Probably setting up regular irc meetings is a good option, but for now people in the channel may discuss the issue queue or new issues on real time.
Other idea about services that may catch end users interest are focused services. It is a good approach to have system base services (node, taxonomy, user.. ), but looks like a very vertical approach for specific purposes, that can also be used for services application (here is where node service and permissions fight a real war). In fact, I would rather recommend replace with a general Entity services, now that exists an entity controller in D7.
Having horizontal services (Moveable Type API, bugzilla API.. ) also, brings the opportunity where users may share ideas and necessities. Yesterday, cweagans (blogapi maintainer) and I discussed a path to get blogapi service integrated in services module for d7 (considering services D7CX pledge), and IMO it will provide services is a very good base for oriented web services, not just a simple system, node, users, taxonomy, .. interface. RDF went his own way in this, but still are a lot of services that could be integrated.
Lastly, about testing, I was working on a services test class earlier, that I didn't include (there is an issue for simpletest in D7 about make the setUp() argument an array) in the patches I submitted. I'll clean and provide as soon as possible so we can start designing the test cases, but the real problem right now is that 5 modules are required to simple login into the webservices: services, services_keyauth, user_service, system_service and a whatever_server, so the core of services right now are these 5 services, and this is why I submitted also patches for user and system services.
Thanks heyrocker for reserving the space in the irc, going there ASAP!