Posted by jpetso on May 23, 2009 at 6:53am
Our first all-inclusive meetup, featuring both students (marvil07 and chrono325, now known as dhax) and two of three mentors (jpetso and sdboyer-laptop). Covered topics include development goals, workflow issues (using "gsoc2009" as well as "gsoc2009-marvil07" and "gsoc2009-chrono325" tags for issues), student development repositories (pushing everything onto Github for the time being), and how to manage further communication (there's going to be a weekly IRC meeting, every Sunday at 18:30 UTC). All subsequent logs will be tracked in the d.o issue http://drupal.org/node/470722.
[Sun May 17 2009] [20:25:35] Join You have joined the channel #drupal-vcs (n=jpetso@93.83.249.126). [20:25:35] Topic The channel topic is "http://github.com/sdboyer/drupal-git-scripts/tree/master". [20:25:36] Topic The topic was set by sirkitree on 05/13/2009 06:40:28 PM. [20:25:51] CTCP Received Version request from freenode-connect. [20:26:09] Mode Channel modes: secret, no messages from outside [20:26:09] Created This channel was created on 05/08/2009 07:36:39 AM. [20:28:38] <marvil07> hi jpetso and sdboyer-laptop [20:28:56] <jpetso> hi all! just made it in time :) [20:29:58] <jpetso> so it seems we've got everyone here except skiquel, which is great, thanks for passing by :) [20:30:45] <chrono325> did he say he was coming? [20:30:46] <Druplicon> chrono325: 3 weeks 4 days ago <skiquel> tell chrono325 Hey. [20:31:26] <jpetso> chrono325: no, but more people are here than said they would [20:31:48] <chrono325> okay, he may show up then [20:33:22] <jpetso> the bad news is that i'm pretty much unprepared, time passes more quickly than i thought it would (even now that I should have free time all around) [20:34:13] <jpetso> the good news is that sdboyer always saves the party when he's around, because he's so cool. :D [20:35:05] <jpetso> so, let's start with your questions, in case you got any [20:35:29] <jpetso> chrono325: i see you created your wiki page, that's cool [20:35:52] <chrono325> jpetso: yup, thought I would get prepared before this :) [20:36:08] <jpetso> chrono325: there's not much in there yet, but that can of course change over time :) [20:36:52] <chrono325> jpetso: that's what I figured. there isn't too much to say at this point [20:37:05] <jpetso> chrono325: that's probably true, yes [20:37:41] <jpetso> let me look up the bullet points from the welcome post [20:38:12] <chrono325> I did have a question, actually. How much work (if any) needs to be done on the project* modules to get them integrated with the new versioncontrol API? [20:38:39] <jpetso> chrono325: define "new" Version Control API [20:39:04] <chrono325> jpetso: the one I will be working on [20:39:29] <chrono325> jpetso: since it used to be hardcoded for CVS, if I'm not mistaken [20:40:00] <jpetso> chrono325: as far as i can see, project* tries its best to stay out of the way, and should hardly interfere with these changes [20:40:54] <jpetso> chrono325: essentially, we're going to need our own, different kind of maintainership definition with our own implementation, but that doesn't matter all that much to project* [20:41:28] <chrono325> jpetso: so it is VCS-agnostic? I'm just wondering whether I will need to perform additional changes to project* once the git backend (for example) is done. [20:42:06] <jpetso> chrono325: afaik, dww and the project* team are trying to split out the notion of maintainership from project* to cvs.module [20:42:56] <jpetso> chrono325: or did they already do that? anyways, versioncontrol_project has its own maintainer table anyways, so yeah, this stuff can work without specific integration or support of project [20:43:01] <chrono325> jpetso: I see. My main concern is how much time I would need to devote to working on project before the changes can be deployed to d.o [20:43:38] <jpetso> chrono325: ah right, that might be a more difficult issue [20:44:23] <chrono325> jpetso: yeah, so what might be involved? I've been telling all my friends how cool I am that I could be deploying to d.o, but I feared that it might actually be more work. [20:46:53] <jpetso> chrono325: i think as long as there's some possibility to provide "backwards compatibility" (maintainers done in a remotely similar way to what versioncontrol_project or cvs.module do now) then additional features should be no issue for deployment, as long as it's implemented in a clean way [20:47:23] <jpetso> chrono325: project* pretty much places an emphasis on evolutionary changes [20:48:10] <jpetso> chrono325: in any case, we really need to separate a) the capability to implement those changes, and b) the actual deployment [20:48:45] <jpetso> chrono325: there's a good chance those two won't be coupled. [20:49:39] <jpetso> chrono325: for initial d.o deployment (my issue, still not done, sorry), the most important thing is to support CVS as good as it's supported now. everything else can be done at a later time. [20:50:06] <jpetso> chrono325: that's the only approach that'll work for version control on d.o, imho [20:50:50] <chrono325> jpetso: makes sense. [20:51:34] <jpetso> chrono325: does that answer your question, or did you want to know something different? [20:52:31] <chrono325> jpetso: that helps, I mainly just wanted to have a rough idea of what would be needed on top of simply writing the git and subversion integration to get deployed to d.o [20:52:59] <chrono325> jpetso: and whether I should budget time for it into the schedule [20:54:04] <jpetso> ok. so, our premier goal is to simply make Version Control API as good as possible, and if d.o wants that kind of functionality (they should want it) then they can have it [20:54:48] <jpetso> occasional feedback from dww is critical too, i'd like to add [20:55:12] <jpetso> because he decides what gets deployed and what doesn't [20:55:52] <jpetso> so make sure to rather write clean, secure & readable code instead of implementing every possible feature [20:56:40] <jpetso> of course, if we can have both, great. :] [20:57:54] <chrono325> and it should all be simpletested, right? [20:59:49] <jpetso> if feasible, yes. personally, i'm not the developer best known for consistently equipping his own modules with simpletests, so i don't quite have the street credibility to demand every feature being tested... [21:00:12] <chrono325> heh :) [21:00:32] <jpetso> ...but it's a good thing, and helps the module in the end [21:01:35] <chrono325> fair enough. [21:01:59] <jpetso> i think in Version Control API, simpletests would be most useful at the "lower end", where the backends interface with the VCS binaries [21:03:00] <jpetso> i wish i had written some for the SVN backend, and because i didn't, it's even harder to avoid regressions in sdboyer-laptop's rewrite [21:04:15] <jpetso> not sure how to best test Version Control API itself, most probably we would need to make the FakeVCS backend into an actually functioning module. [21:04:43] <jpetso> (because it's hard to test most stuff without working backends) [21:05:02] <chrono325> I see [21:06:37] <jpetso> i guess marvil07 will be more concerned with this stuff though, chrono325's project hardly involves the actual API module [21:07:47] <jpetso> chrono325: on the other hand... er. hm. whatever, i need to be better prepared to bring that up. some other time. [21:08:22] <jpetso> chrono325: have you followed corni's progress with the Git backend? [21:11:06] <chrono325> jpetso: not really, no. I've been spending more time learning drupal module development [21:11:21] <jpetso> chrono325: cool, how far have you got? [21:12:08] <chrono325> jpetso: about halfway through "Learning Drupal 6 Module Development" [21:12:27] <chrono325> jpetso: still have to do some coding of my own, though [21:14:11] <jpetso> chrono325: mmkay. in case you haven't yet, you really need to get comfortable with the Version Control API source code too [21:15:08] <jpetso> chrono325: but well, i trust that'll work out [21:15:25] <chrono325> jpetso: okay. Is there any specific way I should attack it or just go in and read through the whole durn thing? [21:16:12] <jpetso> chrono325: on the project page - http://drupal.org/project/versioncontrol - there's a link to the handbook [21:16:36] <jpetso> chrono325: i wrote some introductory stuff in there, should be helpful for the birds-eye view [21:17:40] <chrono325> jpetso: I'll get on that. My "oodles of free time" has seemed to dissolve pretty quickly, too. [21:17:53] <jpetso> chrono325: although in the end, inspecting at least versioncontrol.module itself won't be avoidable [21:18:36] <jpetso> chrono325: you should be able to get by with the per-function API docs, though, i spent quite a bit of effort to make them cool [21:20:28] <chrono325> jpetso: sounds good. Now that I have my summer residence plans squared away, I can focus the bulk of my time on this. [21:20:58] <jpetso> chrono325: cool. [21:22:29] <jpetso> marvil07: still here? [21:24:11] <jpetso> ok, so... let's see how we go from here [21:24:57] <jpetso> i'll stick to the welcome post for now... [21:25:17] <jpetso> "2. Obtain a CVS account." [21:26:01] <jpetso> marvil07 already has a CVS account, as he already contributed to the Git backend [21:26:43] <jpetso> also, "3. Create a 'project' on drupal.org for your project" [21:27:19] <jpetso> both of those bullet points are concerned with committing stuff [21:28:43] <jpetso> we still didn't work out the process how your code will be developed, where it will be committed and how it'll get incorporated into upstream [21:29:36] <chrono325> I talked with skiquel, and he suggested gitHub for my stuff [21:29:50] <jpetso> right. [21:30:46] <chrono325> obviously, that would be outside the normal drupal development line, but it would be a bit nicer on the devel side, as we would be using git and not CVS [21:31:13] <jpetso> that's not totally in line with the general Drupal SoC policies, but as we've got a bunch of Git people here i think that's ok. [21:32:12] <jpetso> i suggested to move Version Control API development to Git for the duration of your SoC projects, that would also go to Github [21:32:26] <marvil07> sorry.. comming back.. and reading [21:33:21] <jpetso> sdboyer-laptop: ping? [21:34:17] <chrono325> moving to git for SoC projects would be nice. CVS isn't too difficult, but I'm much more comfortable on Git [21:36:30] <jpetso> i think that's true for everyone but me, and i can well live with it (in fact, i'm looking forward to some practical multi-developer experience with Git) [21:37:57] <jpetso> marvil07: do you think we should split the various projects into separate Github projects, or should we have a single repository with all relevant modules in it? [21:38:19] <jpetso> sdboyer-laptop: ^ [21:38:52] <jpetso> maybe Sam isn't here after all? hm. [21:42:45] Quit marvil07 has left this server ("Abandonando"). [21:42:57] Join marvil07 has joined this channel (n=marvil07@64.76.110.180). [21:43:05] <jpetso> marvil07: hi! [21:43:08] <jpetso> :P [21:44:18] <marvil07> hi jpetso [21:44:26] <jpetso> marvil07, chrono325: on issue tracking and weekly status reports, this year's SoC admins have been somewhat unclear. [21:44:33] <marvil07> really sorry because of being and not being... but now I'm reading :p [21:46:01] <jpetso> marvil07, chrono325: as you could read in the welcome post, the suggested solution for this stuff was an own d.o project for each SoC project [21:46:05] <jpetso> marvil07, chrono325: initially. [21:47:57] <marvil07> aswering question about separate projects on git hub.. welll.. [21:47:59] <jpetso> marvil07, chrono325: the discussion in the comments, where it has been decided that we'll use the "gsoc2009" tag, has still not been folded into the post. [21:48:31] <marvil07> is better maintaint individual logs for gitcvsexportcommit funcionallity [21:49:01] <marvil07> but t would be kind of difficult to follow lots of projects [21:49:01] <jpetso> marvil07: yes, that makes sense [21:49:30] <marvil07> but, not so much. I think it's ok one for each one [21:50:50] <marvil07> jpetso, I do not understand something: are we going to maintain syncronized cvs repos on d.o? [21:51:09] <jpetso> marvil07: yeah, that'll be my task [21:51:22] <marvil07> :o [21:51:44] <jpetso> marvil07: not cool? [21:52:18] <marvil07> taht means you commit our changes to cvs on d.o and we wont commit to cvs, but yes to github [21:52:23] <marvil07> if thatt, you rock! [21:53:41] <jpetso> marvil07: kinda... i thought we'd all maintain our own trees, and i pull from you whenever a mergeable milestone is reached [21:54:24] <marvil07> that's ok jpetso [21:54:25] <jpetso> marvil07: then commit those changes to d.o. or rather the other way round? [21:56:24] <jpetso> so my next task is to create Github projects for all the involved modules [21:57:06] <marvil07> there is one for versioncontrol_git :D [21:57:23] <marvil07> are u jpetso on github? [21:58:26] <jpetso> marvil07: versioncontrol_git is on gitorious though... skiquel has suggested github because it's more mature, i'm pretty indifferent on that really [21:58:40] <jpetso> marvil07: any jpetso you'll find is always me :) [21:58:49] <marvil07> jeje [21:59:15] <marvil07> well.. you are right.. is on gitorious :p, I forgot it [21:59:31] <marvil07> ok, I'm marvil07 everywhere too [22:00:42] <jpetso> to finish the above point, we're not going to have d.o projects for your SoC projects, per decree from me. [22:01:36] <marvil07> that's fine, but I think wwe should use issue tracker on d.o [22:01:59] <jpetso> marvil07, chrono325: i'm still waiting on Dmitry's SoC site to go live (he promised it, but it's pretty close to the start already), and overall, let's focus on the d.o/g.d.o infrastructure wherever possible [22:02:20] <marvil07> ok [22:02:22] <jpetso> marvil07, chrono325: unless AlexUA or Dmitry get this done, we're all issue queue and g.d.o discussions [22:02:46] <marvil07> ok [22:02:48] <chrono325> sounds good [22:02:49] <marvil07> chrono325, [22:02:54] <jpetso> marvil07, chrono325: please pick the tag for your d.o issues [22:02:56] <jpetso> marvil07, chrono325: please pick the tag for your d.o issues [22:02:58] <jpetso> er. [22:03:22] <jpetso> marvil07, chrono325: i suggest "gsoc2009-marvil07" and "gsoc2009-chrono325" [22:04:01] <jpetso> chrono325: plus "gsoc2009" for both of you [22:04:11] <jpetso> er. [22:04:17] <jpetso> anyways. [22:04:21] <marvil07> well, no problems [22:04:45] * sdboyer-laptop laughs [22:04:50] <sdboyer-laptop> looks like i'm awfully late to saving the party [22:05:04] <jpetso> sdboyer-laptop: well, you made it :D [22:05:16] <sdboyer-laptop> sorry, some crazy stuff happening in the street here [22:05:19] <marvil07> sdboyer-laptop, I felt the same some minutes ago [22:05:26] <jpetso> makes two tags on each issue you work on for the SoC, that way we can group all the issues on one page. [22:05:28] <sdboyer-laptop> couple friends pulled me out, i only just got back in :) [22:05:34] * sdboyer-laptop reads backscroll [22:06:08] <jpetso> sdboyer-laptop: nothing like a good time with socializing :] [22:06:40] <marvil07> chrono325, I'm a little worry about how do we are going to work togheter, I think it's no really defined the line between our projects [22:09:07] * sdboyer-laptop is nearly done [22:09:32] <chrono325> marvil07: yeah, I've been thinking about that too. [22:09:36] <marvil07> I think we should make one line project timeline [22:09:53] <sdboyer-laptop> ok. heh, several big points here [22:09:54] <sdboyer-laptop> first [22:10:11] <sdboyer-laptop> see the channel topic for the git/d.o cvs integration scripts i've been working on for quite some time [22:10:16] <sdboyer-laptop> and neclimdul has recently stepped up to help quite a bit with [22:10:31] <sdboyer-laptop> it's rapidly approaching the point where git integration with d.o cvs in contrib is pretty much trivial [22:10:39] <chrono325> marvil07: could you point me to some more info about your project, like a detailed proposal (I'm looking at your abstract on socghop.appspot.com [22:10:57] <sdboyer-laptop> next... [22:11:03] <jpetso> oh damn. major fail on my staying on top of things. [22:11:19] <jpetso> s/my/me/ [22:11:58] <sdboyer-laptop> chrono325: though i'd deLIGHT in seeing a new vcs on d.o, there are a lot, lot of steps between here and there, only some of which are technical (though most of those technical ones are under the purview of this group here) [22:12:01] <sdboyer-laptop> ...more or less... [22:12:18] <sdboyer-laptop> chrono325: dww had a great, brief lullabot podcast recently in which he highlighted the basic things necessary to actually make a migration [22:12:25] <sdboyer-laptop> if you haven't listened to it already, it's worth bending an ear to [22:12:51] <marvil07> omg [22:12:56] <sdboyer-laptop> THAT all said, i'd LOVE to migrate asap, and have been kicking around the idea of starting the simultaneous ball-rolling in the various places we'll need to to get it movin [22:13:18] <sdboyer-laptop> with the intention of having most of the ducks in a row by the time of DC Paris [22:13:23] <sdboyer-laptop> ...we'll see on that :) [22:13:38] <jpetso> sdboyer-laptop: chrono325 didn't refer to the initial migration but to his versioncontrol_project/DVCS changes being deployed [22:14:01] <jpetso> sdboyer-laptop: after Version Control API is already deployed on d.o, which i need to take care of asap [22:14:02] <sdboyer-laptop> jpetso chrono325: ahh, ok, sorry, thanks for the clarification [22:14:27] <sdboyer-laptop> chrono325: side note, Matt Butcher is my cubemate :) [22:14:44] <jpetso> aka mfb [22:14:50] <sdboyer-laptop> mbutcher [22:14:53] <sdboyer-laptop> mfb's someone else [22:14:58] <jpetso> oh. sorry. [22:15:03] <sdboyer-laptop> nooooo worries [22:15:11] <sdboyer-laptop> jpetso: re: the new SVN backend, it's actually very, very nearly ready for prime time [22:16:11] <sdboyer-laptop> latest stats indicate it a) runs two to three times faster than the old one, and b) cleans up memory efficiently, which means that whereas memory usage increased linearly over time, it now stays pretty much flat [22:16:23] <sdboyer-laptop> s/whereas/whereas before/ [22:16:26] <jpetso> awesome [22:16:31] <sdboyer-laptop> yeah, i'm a happy camper :) [22:16:46] <sdboyer-laptop> i've got to write phpunit tests for svnlib just in general, still [22:16:49] <chrono325> sdboyer-laptop: where is that podcast for the migration? [22:16:50] <sdboyer-laptop> and it's not 1.4-compatible [22:16:55] <sdboyer-laptop> chrono325: lemme find it... [22:17:04] <marvil07> it should be hard to propose a _git backend that kicks that :p [22:17:14] <chrono325> sdboyer-laptop: thanks [22:17:46] <sdboyer-laptop> HAH, look at that, earl did another lullacast! [22:17:52] <marvil07> sdboyer-laptop, I did not find it neither [22:18:02] <sdboyer-laptop> it's a little older, nearly there [22:18:11] <sdboyer-laptop> shows how out of the loop i am, i didn't even know he did another panels thing :P [22:18:17] <jpetso> sdboyer-laptop: they're spilling podcasts all over the place, i just can't keep up anymore [22:18:25] <jpetso> sdboyer-laptop: i did listen to dww's, though [22:18:34] <sdboyer-laptop> ROFL [22:18:39] <sdboyer-laptop> and two of my colleagues [22:18:40] <sdboyer-laptop> OH [22:18:41] <sdboyer-laptop> right [22:18:43] <sdboyer-laptop> i forgot [22:18:50] <sdboyer-laptop> these drupal voices things are brief interviews they did at DCDC [22:19:01] <sdboyer-laptop> http://www.lullabot.com/drupal-voices/drupal-voices-23-derek-wright-proj... [22:19:03] <sdboyer-laptop> that's dww's [22:19:35] <sdboyer-laptop> making the git backend that efficient is just a question of taking the approach i used for svn and applying it to git :) [22:19:58] * sdboyer-laptop suspects that'll be FAR faster, though [22:20:10] <sdboyer-laptop> since svn is just ridiculously slow at retrieving large amounts of data [22:20:20] <jpetso> sdboyer-laptop: would you consider the SVN backend "fast enough" now? [22:20:23] * sdboyer-laptop hates `svn info` [22:20:26] <sdboyer-laptop> jpetso: yup. [22:20:37] <sdboyer-laptop> it's not gonna get any faster without rewriting subversion :) [22:20:50] <sdboyer-laptop> or, possibly, SWIG bindings [22:20:58] <sdboyer-laptop> but i sorta doubt even those would make much of a differences [22:21:00] <jpetso> or Version Control API requirements [22:21:14] <sdboyer-laptop> hmm? [22:21:48] <jpetso> like, not requiring source items from the start, and fetching them on the fly instead [22:22:06] <jpetso> but ok, that shouldn't make a big difference for the backend itself i guess [22:22:16] <sdboyer-laptop> i think most of the optmizations that can be done have been done [22:22:21] * sdboyer-laptop may have missed something big, but [22:22:23] <sdboyer-laptop> the way it works now [22:22:31] <sdboyer-laptop> it never stores any big arrays [22:22:37] <sdboyer-laptop> which is what eats memory [22:22:47] <sdboyer-laptop> in fact, it'll even query a url multiple times, on occasion, because it turns out that's faster [22:22:50] <sdboyer-laptop> (when batched correctly) [22:23:07] <jpetso> sdboyer-laptop: ok, i'll check it out soon (tm) [22:23:42] <sdboyer-laptop> it just grabs exactly what it needs as it needs it for each revision, makes liberal use of xpath to find additional bits and do extended metadata logic checking, then fires out the same array as we used in the previous version [22:23:44] <sdboyer-laptop> jpetso: ;) [22:24:02] <jpetso> sdboyer-laptop: this time, i have to, because chrono325's first task involves versioncontrol_svn (at least, hook scripts for it) [22:24:15] <jpetso> sdboyer-laptop: i hope you didn't already write those either? :P [22:24:38] <jpetso> s/either/too/ [22:24:38] * marvil07 rearlly di not notice the new lullabot podcast feed, grhh(I was just listening at main podcast feed) [22:25:56] <sdboyer-laptop> jpetso: i haven't, nope. that bit i haven't done [22:26:42] <sdboyer-laptop> my most recent commits changed the other invocations to the backend from within versioncontrol_svn, but i haven't done anything with the hook scripts, or anything designed to be run by cron, iirc [22:27:05] <jpetso> sdboyer-laptop: nice, because if everyone is doing this work then we need to reshuffle tasks for chrono325 ;) [22:27:18] <sdboyer-laptop> jpetso: f'sho :) [22:27:23] <sdboyer-laptop> well i'm happy to stand back on some pieces here [22:27:23] <jpetso> :D [22:27:32] <sdboyer-laptop> in fact, that's probably optimal anyway [22:27:39] <sdboyer-laptop> with the end of my quarter coming and a thesis to fire out [22:28:40] <jpetso> sdboyer-laptop: comments on the Github issue? (single repository or separate ones) [22:28:54] <sdboyer-laptop> jpetso: ahh right. sorry, forgot to respond on that bit [22:29:13] <sdboyer-laptop> jpetso: i'd tend to say separate [22:30:07] <sdboyer-laptop> especially if there's ever to be interoperation with drupal cvs [22:30:07] <jpetso> sdboyer-laptop: k, then you and marvil07 are on the same page - two people who know Git and the code, that's plenty to convince me [22:30:15] <sdboyer-laptop> because that's what my scripts entail [22:30:18] <sdboyer-laptop> jpetso: cool :) [22:31:16] <marvil07> well, it seem like we are all in the same direction :D [22:31:24] <sdboyer-laptop> always a good thing! [22:31:41] <jpetso> sdboyer-laptop: ah. but your scripts don't have anything to do with the actual Version Control API hook script thingy, right? [22:31:55] <jpetso> (for the log: http://github.com/sdboyer/drupal-git-scripts/tree/master) [22:33:28] <marvil07> :o, yesterday I see that project, but I did not try it, I was working on a python git-drupal script, but no enought time for that [22:34:05] <marvil07> jpetso, it seem like they are scripts ofr development [22:34:37] <sdboyer-laptop> jpetso: no, sorry. theyr'e purely for managing a local workflow where you interoperate between keeping and doing all your work in git, and firing things back into drupal cvs [22:34:53] <sdboyer-laptop> i.e., via structured wrappers around git-cvsexportcommit [22:35:05] <marvil07> yeah, like http://drupal.org/node/288873 :D [22:35:06] <Druplicon> http://drupal.org/node/288873 => Maintaining a Drupal module using Git => 5 IRC mentions [22:35:28] <jpetso> right. then i revoke my previous statement on not being on top of things, i knew that you had *those* lying around somewhere and misunderstood your pointer above [22:35:40] <sdboyer-laptop> yup, these scripts systematize what's there to make it headache-free [22:35:44] <sdboyer-laptop> and multiple-branch friendly [22:35:58] <marvil07> :o [22:36:09] * marvil07 should try it [22:36:13] <sdboyer-laptop> yeah, believe me, i was one of the first people to use those instructions and realize their limitations :) [22:36:47] <sdboyer-laptop> ooh, but lookie, you were the one who first wrote em :P [22:36:48] <jpetso> sdboyer-laptop: anything that i need to take special care about when importing the projects into Github? [22:37:32] <sdboyer-laptop> hmm, july 29...yup, that's about a week before i started kickin around with my local git/cvs environment :) [22:37:36] <sdboyer-laptop> jpetso: nothin i can think of [22:37:46] <jpetso> k. [22:37:59] <marvil07> :o [22:42:34] <jpetso> marvil07, chrono325: your next steps? [22:42:56] <marvil07> well, coming back to the main topic :p, we have now two wikis: http://groups.drupal.org/node/22301 and http://groups.drupal.org/node/21624 [22:42:58] <Druplicon> http://groups.drupal.org/node/22301 => Completion of Version Control Integration and Deployment to Drupal.org => 1 IRC mention [22:42:59] <Druplicon> http://groups.drupal.org/node/21624 => Version Control API and family changes => 2 IRC mentions [22:43:39] <chrono325> dive into the versioncontrol api, look at all of the existing SVN work, figure out what needs to be done to complete it [22:44:15] <jpetso> chrono325: sounds good [22:45:06] <marvil07> looking at those wikis, I see the only conflicting thing is the git backend [22:45:54] <marvil07> so, there is no really too much intersection :p, so branching on github do the work :D [22:46:52] <marvil07> :o, and the "web-based support for creating branches of projects, similar to Launchpad." [22:47:07] <jpetso> :-o [22:47:46] <jpetso> well as i said, we won't have that for both of you probably [22:48:54] <jpetso> except if we can work out a way to cooperate on this, but SoC policies pretty much forbid cooperation (bah) [22:49:30] <jpetso> so this task needs to be replaced with another task for either marvil07 or chrono325. [22:50:01] <chrono325> what should it be? [22:50:23] <jpetso> chrono325: "it" like "another task"? [22:50:30] <chrono325> yeah :) [22:51:46] <jpetso> chrono325: i thought that improvements in Repoview would be cool, and if not that then we'll come up with something - there's still plenty to do, don't worry :P [22:52:27] <jpetso> maybe take another look at the ideas page from back then. [22:52:35] <chrono325> jpetso: I don't have any particular preference one way or the other, I'd like to work on anything that is useful and interesting [22:53:18] <marvil07> well, while thinking on that, at the on of the day I should be posting to versioncontrol d.o issue queue a detailed class diagram about where to put each function(now methods) [22:53:43] <jpetso> chrono325: anyways, let's save that for later - first, i want you to get through the mid-term evaluation with me being able to hand out excellent ratings [22:54:28] <chrono325> jpetso: me too :) [22:55:00] <jpetso> marvil07: looking forward to that! [22:55:45] <jpetso> chrono325: common goals always make things easier :] [22:56:23] <chrono325> so do we want to figure out a new common timeline? [22:56:51] <chrono325> sorting out the overlap and taking into account how much has already been done on the SVN and Git modules. [22:58:05] <jpetso> chrono325: your SVN hook script stuff is very much separate from the work that sdboyer-laptop did [22:58:20] <jpetso> chrono325: no adaptations needed there. [22:58:41] <chrono325> jpetso: ah, so that part of the timeline is unchanged. [22:58:49] <jpetso> chrono325: yep. [22:59:02] <jpetso> chrono325: for the Git backend, hook scripts are also still needed. [22:59:53] <jpetso> chrono325: it's just that corni did a great job of revamping the backend, with a little help from marvil07 as far as i could see [23:00:23] <jpetso> chrono325: so we can drop the "porting it to the new API and Drupal 6" point [23:00:55] <chrono325> jpetso: sounds good [23:02:09] <jpetso> chrono325: if you want, we can replace it with "transform the Git interfacing code to an sdboyer like OOP library" [23:02:24] <jpetso> chrono325: but let's see how well things go with the hook scripts, then look further [23:03:13] <marvil07> that's right, corni rocks [23:03:19] <marvil07> jpetso, where was "porting it to the new API and Drupal 6" point [23:03:35] <marvil07> that a good idea, let make it step by step now for both of us [23:04:07] <jpetso> marvil07: that was part of the "Git backend and hooks" point on chrono325's wiki page [23:04:46] <jpetso> chrono325: how about you re-post your application on g.d.o so that marvil07 can take a look at it? [23:05:11] <marvil07> well, I'm really sorry, but I have to go now, and before that I want to propose to have this metting at least all sundays at same time(maybe a little later, I'm on -0500 :p) [23:05:35] <chrono325> jpetso: sure. [23:06:26] <jpetso> marvil07: -0500 is like, Mountain Time or such? [23:06:49] <marvil07> it 16:06 here [23:06:57] <jpetso> marvil07: or are they named differently in South America? :] [23:06:57] <marvil07> s/it/its/ [23:07:14] * jpetso really needs to learn all the American timezones by heart sometime soon [23:07:22] <jpetso> chrono325: what's your timezone? [23:07:51] <marvil07> and yours jpetso and sdboyer-laptop ? [23:08:27] <chrono325> jpetso: US eastern [23:08:33] <jpetso> anyways, it's now 23:07 for me, so much later than 19:00 UTC (21:00 in Austria, +0200 with daylight saving time) might not be ideal for me [23:08:38] <chrono325> I'm not sure what that is relative to UTC [23:08:49] <chrono325> it's 5pm here [23:09:29] <jpetso> chrono325: UTC is currently 21:09 [23:09:50] <jpetso> chrono325: 21-17 = 4, so you're -0400 [23:09:52] <chrono325> ack, my tiny American brain can't handle 24-hour time! :o [23:10:16] <chrono325> got it [23:10:37] <jpetso> :] [23:10:53] <chrono325> I'm going to have to get better at learning timezone stuff :) [23:11:36] <jpetso> marvil07, chrono325: so Sunday, starting from something between 18:00-19:00 UTC would be ok for you? [23:12:03] <chrono325> yeah, that would work [23:12:25] <jpetso> marvil07, chrono325: if it doesn't work out sometimes, we can reschedule [23:13:17] <marvil07> that's ok [23:13:24] <jpetso> marvil07: if it's too early for you, you can come a bit later maybe [23:13:43] <jpetso> but on the whole, i can't do it much later. [23:14:02] <marvil07> ok jpetso [23:14:06] <jpetso> k, then let's stick to 18:30 UTC, weekly, every Sunday [23:14:15] <jpetso> is weekly enough? [23:15:21] <marvil07> I'll posting at night and iddling here all the time I can, issues on d.o can make the non-syncronized communication [23:15:40] <jpetso> right, if issues work out then that'll be fine [23:15:48] <marvil07> yep [23:15:58] <marvil07> well, thanks all for comming! [23:16:11] <marvil07> gtg now [23:16:15] <jpetso> anyways, i'd like to hear from you at least every two or three days, even if no progress is made for some reason [23:16:21] <jpetso> marvil07: thanks, see you! [23:16:28] <marvil07> :D [23:16:29] Quit marvil07 has left this server ("Abandonando"). [23:17:43] <jpetso> chrono325: same goes for you - be active in the issue queue(s) so that we can follow what you're thinking, coding and whatnot [23:18:55] <chrono325> jpetso: will do. At this point, it's mainly going to be catching up to the current state of affairs, but I'll be ready to contribute by the time coding starts. [23:19:52] <jpetso> chrono325: IRC or other means are also ok. i (and skiquel) just need to know what's going on, so that we can make the best of it [23:21:03] <chrono325> jpetso: sounds good. [23:21:51] <jpetso> chrono325: you'll be required by the Drupal SoC team to write weekly reports, apart from that it should be enough to keep me up to date on what you're working on and whether there are any issues that you need help with [23:22:19] <chrono325> jpetso: where should I post those weekly reports? [23:22:39] <chrono325> jpetso: wiki page on g.d.o? Or my own blog? [23:24:02] <jpetso> chrono325: i would assume that Dmitry gets his Drupal SoC website done in time. if not then probably g.d.o. in any case, that'll be announced by the org team unless they don't keep up with their tasks [23:25:04] <jpetso> chrono325: we've got an internal mailing list for stuff like that - not all too frequent posts, but if something crops up then we'll discuss it there first [23:25:55] <chrono325> jpetso: and they will make a post to the "important annoucements" tag on the gsoc g.d.o group? [23:26:34] <jpetso> chrono325: i really hope they will. i might get involved with that myself if nothing else happens. [23:27:38] <jpetso> chrono325: that's not to say that the org team is lazy or anything, Alex UA does a good job on the whole [23:28:34] <jpetso> chrono325: it's just that i'm spoiled with webchick running the SoC, and she just handled everything incredibly well and timely and all, that's hard to beat by any successor :D [23:30:45] <chrono325> jpetso: I see. I've heard great things about webchick, she seems to be very on top of things. [23:30:52] <chrono325> jpetso: a tough act to follow [23:31:34] <jpetso> chrono325: most awesome open source contributor you'll ever meet. without a doubt. [23:32:01] <chrono325> jpetso: good to have her working with drupal then. [23:32:12] <jpetso> chrono325: indeed :D [23:32:44] <chrono325> Well, if there isn't anything else that needs discussing now, I need to get some dinner. Is there anything else? [23:33:15] <jpetso> chrono325: i was just thinking and looking around, but i don't remember anything pressing. [23:33:36] <jpetso> chrono325: in fact, i'd like to go to sleep soon, it's been an exhausting day :] [23:34:05] <chrono325> jpetso: okay, I'll be checking my email and lurking on IRC often, so let me know if anything comes up [23:34:19] <jpetso> i'll notify you when anything of importance is happening, and i'll be on IRC too [23:34:44] <jpetso> if you need any help on some Drupal or Version Control API stuff, just ask [23:36:34] <jpetso> chrono325: well then, i'm going to go offline in case we're done [23:36:47] <jpetso> chrono325: we're done? [23:36:59] <chrono325> jpetso: yup, get some sleep :) [23:37:09] Quit chrono325 has left this server (Remote closed the connection). [23:37:24] <jpetso> sdboyer-laptop: thanks for coming, too!