[02:16:46] mwalker: if you're around I could chat for a little bit. [02:22:05] mvolz, ah; hello! [02:25:21] mvolz, somewhat late across the pond! to your email -- OCLC operates worldcat; and it provides information about more than just ISBNs; basically if it's printed media they supposedly have some way of querying for the metadata [02:25:52] Yup. I'm going to be a zombie tomorrow. [02:25:57] Ok excellent [02:25:58] but when we were scheming here in the office about sustainable ways to start, we figured that the OCLC data would probably be a good test case for data [02:26:15] otoh; it's up to you on if you want to start there, or with the zotero stuff [02:26:59] personally I think just getting the isbn data would be a huge win so I'd advocate for that; but I dont know a huge amount of where the pain points are (isbn's might be well covered) [02:27:15] I think I'll probably do a quite POC app and then try to hook in a few of the zotero handlers. and then go more general [02:27:28] if that seems to make sense [02:27:38] translaters* not handlers [02:28:09] so then just trying to run a web accessible zotero installation is pretty much out? [02:28:22] I couldn't find much documentation on that TBH [02:28:36] clearly zotero.org is doing it though [02:28:40] neither could we; it seems like zotero assumes that everyone is their own individual user [02:28:42] which we could do [02:28:49] but it didn't seem to have the most utility [02:28:57] I agree [02:28:59] and it requires that people run the zotero app [02:29:05] yeah too clunky [02:29:23] but... somewhat cool! [02:29:27] for those who already use it [02:29:52] but I think that's out of scope unless you really want to do it [02:30:03] definitely don't want to! [02:30:05] :) [02:30:11] k [02:30:23] do you have any sense of the architecture you're planning on writing? [02:30:34] and have you written node applications before? [02:30:42] nope [02:30:44] total n00b. [02:31:03] so we might be going a little slow at first! [02:31:35] I had a look at parsoid today and will probably try to model off that. I'll be asking gabriel a lot of questions I suspect. [02:32:12] k; another two applications you can look at are the mathoid (in extensions/math) and offlinecontentgenerator [02:32:26] ok [02:32:32] but basically parsoid is the model you want -- because you're going to be doing real time requests [02:33:13] we run node 0.8 on the cluster right now; but the plan is to be running 0.10 hopefully soon [02:33:17] so I would advice developing with 0.10 [02:33:23] ok [02:33:44] so I noticed parsoid uses the cluster module [02:34:05] yes; node is cothreaded by default which means it only actually uses 1 thread [02:34:11] not the best usage of CPU cores [02:34:29] so we have a controlling process which spawns off a bunch of child threads to consume all resources [02:34:47] I was thinking I'd start out with 1 thread whilst getting my feet wet... [02:34:48] it also, not sure if parsoid does this yet, allows us to SIGHUP the process for a graceful restart [02:35:09] kk [02:35:34] you should look into the express docs as well [02:36:00] we're using that for HTTP handling in everything but ocg (because I had to be special due to legacy reasons) [02:36:01] for api aspects [02:36:06] ok [02:36:35] have you looked at how VE is going to talk to your service? [02:36:41] NOPE [02:36:56] I was thinking I'd actually start with a pretty barebones backend [02:37:05] and then try to figure that out relatively early on [02:37:07] on local [02:37:21] to make sure I'm not going down some totally crazy path [02:37:48] will need to talk to trevor a lot when he gets back [02:38:07] for the hooks in VE certainly; but you can spec out the API youre going to expose now [02:38:20] right [02:39:55] I know next to nothing about DOI so I'll have to learn about that [02:40:37] but my understanding is that it's a URI type of thing like : with rules on identifiers based on the prefix? [02:42:24] mvolz, mwalker: parsoid is shutting down gracefully on SIGTERM [02:42:54] ah Gabriel! I figured you'd gone home [02:43:15] restart is handled by systemd (or an old-fashioned init script) [02:44:39] I think I might have missed something :) did it not shut down gracefully before? [02:45:03] at some point it didn't [02:45:14] you need to use disassociate for that IIRC [02:45:49] disconnect: http://nodejs.org/api/cluster.html#cluster_worker_disconnect [02:46:18] I would not worry much about that for now though -- it's easy to add later [02:46:54] request rates will be low, and the only thing that could use some real cpu cycles would be DOM parsing [02:47:03] Ok. [02:48:21] Alright; I think I may actually need to go to sleep [02:48:46] where are you located? [02:49:57] mvolz, kk -- if you're going to be up by 9; we can talk more when you're awake, otherwise I'm going to fall asleep -- but I'll be online at 3PM your time (I think that's 7AM PDT) to talk more [02:50:02] London UTC +1 [02:50:32] good morning to you then ;) [02:51:38] * gwicke just arrived from UTC-1 last night [02:51:53] Good morning and good night! I'll probably be available 9:30ish provided sufficient amounts of caffeine [02:52:33] gwicke: nice, jetlag :) [02:52:53] ok night [02:52:59] night [02:53:03] tty tomorrow [02:53:58] goodnight! [02:55:09] gwicke, think you'll be able to join us for a 7AM brainstorm? [02:55:37] I presume you're going to be asleep at 1:30 this morning [02:55:37] 7AM PDT? [02:55:40] aye [02:55:55] maybe, supported by jetlag [02:55:55] that's 3PM her time [02:55:58] :D [02:56:08] normally I get up between 7 and 730 [02:56:34] I'm making an exception for bootstrapping this project; I usually roll out of bed at 8 [02:56:57] I may have to get up earlier though since she wants to work between 9-5 / 10-6 [02:57:03] *start getting up [02:57:37] or use more async comms [02:57:40] *nods* [07:02:56] Hello Everyone [07:03:24] I need asstiance with publishing a wikipedia value [08:30:36] mwalker: I assume you're asleep, but if not, good morning :) [08:30:47] good morning :) [08:34:45] mvolz, I'm presently writing some very janky python :p [08:35:01] Excellent! My favourite kind [08:35:22] quite! it is a popular choice [08:36:00] ok -- for this meeting I mostly just wanted to know if you have something to do today? [08:36:45] ... and if you had any questions for me about node at the foundation [08:37:10] Well I think I did yesterday but now I'm not so sure. I was originally just going to do a POC scraper because I really don't know node , more of a learning thing than production ready [08:37:36] but now I'm wondering if should go directly into handling a zotero translator. [08:38:19] so everyone seems to use cheerio for scraping but zotero translators are mostly using xpath? [08:38:28] there seems to be an xpath node module [08:38:55] going back a step; POC? [08:39:03] sorry, proof of concept :) [08:40:24] ah; ok... um... I would say use xpath or jsdom (but that's because I'm familiar with it where I'm not familiar with cheerio) [08:40:50] if zotero are using xpath then that's probably something good to start with [08:41:41] ok [08:41:42] alternatively; since the plan is to integrate with the zotero scrapers + oclc data; I would actually start with the oclc api [08:41:51] that way you dont end up trashing code [08:42:28] ok, I wasn't sure of oclc was a definite thing or not [08:42:55] unless you have some DOI proxy that can do it better than they can; I think we're stuck with them [08:43:10] given that they are sort of the canonical source of all things ISBN [08:43:25] *better == at all [08:43:33] ok [08:44:00] * mwalker reiterates that I know next to nothing about DOI and that there really could be awesome things out there [08:44:07] I guess a good question is, what are we using DOI internally for exactly anyway [08:44:13] I dont know! [08:44:18] :D [08:44:30] that would probably be a question better directed at James_F or another long term editor [08:44:44] well the current way it's used on wikipedia is sort of the reverse of the way it was intended to be used. [08:45:11] they're supposed to permanently point to a digital objects, whose URL may change [08:45:58] but the way we've been using it is to use DOI to get the citation and it's really not optimized for that. It's like trying to find the key by the value [08:46:37] the citation itself isn't really part of the relationship, [08:46:51] which is primarily a DOI -> URL/owner relationship [08:47:00] heh; ok [08:47:22] anyway this is getting very philosophical [08:47:54] so in that case, from a position of nativity, I would say you should have a goal of generating DOI citations rather than reading them [08:48:47] and if your service is given a DOI link, parse it out into it's provider and id form and bootstrap the process that way [08:49:24] but the reality is that DOIs aren't used by humans, which is who we're serving- URLs are. so... [08:49:55] anyway I think for today I'll give xpath a go. [08:50:03] hokay [08:51:13] Thanks! [08:55:52] (http://www.crossref.org/02publishers/doi_display_guidelines.html originally doi: was supposed to be a browser implemented protocol) [08:57:14] that was ambitious of them :p [12:40:23] mwalker: did you guys see this already? https://github.com/zotero/zotero-node [12:52:42] So it appears there is a zotero server installation that they implemented in xpc-shell instead of node https://github.com/zotero/translation-server but not sure if that's helpful [14:58:39] Morning [14:59:04] afternoon [15:49:31] GSoC and OPW Q&A at the ECT meeting? [15:51:16] I think so! [15:52:25] * muninn-project hopes so [15:52:41] mvolz, thanks [15:53:00] mvolz, thanks [15:53:23] muninn-project, thanks ^^ [15:55:12] Hi there [15:55:34] We will start https://www.mediawiki.org/wiki/Engineering_Community_Team/Meetings/2014-05-20 in 5 minutes. [15:56:34] * guillom waves [15:57:04] * muninn-project nods. [15:58:11] * andre__ waves too [15:58:53] * tonythomas says hi [15:59:06] o/ [15:59:26] andre__: Nice blog post btw :) [15:59:41] guillom, thanks. I felt that need to summarize it for myself :) [15:59:46] * Jeff_Green says hi [16:00:11] 9h UTC [16:00:22] Are you ready to start? [16:00:29] Hi sumanah guillom andre__ [16:00:32] yes [16:00:34] hi [16:00:34] yes [16:00:42] yup [16:00:47] Alright [16:00:48] jlschatz: Morning. [16:00:57] Welcome to the Engineering Community Team monthly meeting [16:01:00] hello [16:01:06] In fact we meet every Tuesday, but once a month we do it here on IRC. [16:01:08] qgil, thanks [16:01:13] The team: https://www.mediawiki.org/wiki/Engineering_Community_Team [16:01:13] just to say congratz on tech news milestone [16:01:23] matanya: thanks :) [16:01:29] andre__, guillom, sumanah and myself are here to discuss and answer questions in our best capacity. [16:01:37] The open agenda can be found at https://www.mediawiki.org/wiki/Engineering_Community_Team/Meetings/2014-05-20 [16:01:42] muninn-project: good morn/afternoon [16:01:47] The only point so far is Questions & Answers about Google Summer of Code 2014 and FOSS Outreach Program for Women/Round 8. [16:01:56] If you have any other topics, just let us know and we will find time for them at the end after discussing GSoC/OPW [16:02:03] IF YOU ARE HERE FOR THE ECT MEETING RISE YOUR HAND. EOOOO! o/ [16:02:09] _o/ [16:02:14] o/ [16:02:15] _o/ [16:02:20] _o/ [16:02:21] o/ [16:02:26] \o/ [16:02:29] Hi [16:02:30] o/ [16:02:31] o/ [16:02:35] o/ [16:02:42] Oh my god, it's full of people! [16:02:43] o/ [16:02:45] \o/ I think. [16:02:45] o/ [16:02:46] o/ [16:02:46] o/ [16:02:56] o/ [16:02:59] o/ [16:03:01] =jlschatz my nick is taken o_0 [16:03:02] o/ [16:03:07] o/ [16:03:12] * Krenair is lurking IRC, as always [16:03:27] o/ [16:03:31] 139 people basically [16:03:35] o/ [16:03:51] * Raylton feels a IRC party  [16:03:55] great ! [16:04:07] I think this is a record number of people actually here, not just idling, for an ECT meeting :) [16:04:22] This is indeed great. We seem to have a critical mass of GSoC and OPW interns and mentors. [16:04:41] For those that couldn't make it, logs will be available at https://www.mediawiki.org/wiki/Engineering_Community_Team/Meetings/2014-05-20 [16:04:47] Yeah, I'm happy to see so many people here! Welcome & thanks for your help in making Wikimedia better for the world! [16:04:59] :-) [16:05:31] Good. Anybody landing later can just say hi. Let's move on. [16:05:41] About GSoC and OPW [16:05:47] https://www.mediawiki.org/wiki/Google_Summer_of_Code_2014 [16:06:00] https://www.mediawiki.org/wiki/FOSS_Outreach_Program_for_Women/Round_8 [16:06:12] I'm the org admin for both programs, hoping that most of my work has been already done, willing to just sit back and see how you complete 23 projects in the next three months. :) [16:06:50] Since yesterday, all interns are supposed to be working in their projects, well in sync with mentors, after a successful community bonding period. [16:07:02] From a program coordination point of view, all you are expected to do is regular reporting, as explained in the wiki pages. [16:07:10] So the first question is for you: how was the community bonding period? Is anybody still having misc problems impeding them to start working? [16:08:20] https://www.mediawiki.org/wiki/Evaluating_and_Improving_MediaWiki_web_API_client_libraries I have already started enjoying mentoring the project "Evaluating and Improving MediaWiki web API client libraries" by fhocutt - but I think she has had a bit of trouble with Linux wireless drivers! :/ [16:08:28] (this is a biased question, since those having lots of problem perhaps didn't even know about this IRC meeting...) [16:08:29] It was great for me. No official community-bonding anyway, but was around for some time [16:08:38] I'm fighting IRC quite a bit (as you can see). [16:08:47] it slowed her down but she still got nearly everything done that she planned on doing during her community bonding period [16:09:05] I'm also still getting used to IRC [16:09:05] well i have bonding with one of my mentors(Tpt) but others(aubrey and Yann) not so quite. [16:09:05] I've had a lot of communication already today and yesterday [16:09:15] I'm happy with how things are going with jlschatz_intern up to this point, the support network is pretty solid. [16:09:17] qgil, The bonding actually started 3 months ago ;) But myself, Nikerabbit and Nemo_bis are having daily 15 minute meetups on #mediawiki-i18n for the past three weeks [16:09:40] I was a little surprised to discover yesterday I had two new mentors :) [16:09:45] and we have something up on gerrit too :) [16:09:45] I'm actually not starting until the first week of June (my mentor is aware and this was outlined in my proposal) but have had communication and sorted out some reporting tools already [16:10:06] which I'm very happy for as my project has changed technologies. [16:10:22] I'm in the same building as one of my mentors at the moment, so communication is good :) [16:10:26] The community bonding period for me went awesome. I am working on 'Wikidata Outreach' project mentored by Lydia Pintscher. She's extremely helpful and friendly.....I bonded well to community with her help :) [16:10:31] qgil we have made progress https://meta.wikimedia.org/wiki/Book_management_2014, but I feel that the time zones prevents us from communicating as much as we should [16:10:33] ali_king_intern: nice! [16:10:47] I had problems with slow SSH into the bastion servers, but Jeff and I got it fixed with mosh ! great alternative to those away from the wikitech labs servers and experiencing slow SSH [16:10:58] i started the frontend for my project. But the core i'll be starting after a week / university exams.. [16:11:23] Raylton, timezone differences are the norm more than the exception, for all of us. Managng asynchronous communication is essential in free software projects. [16:11:25] I'm in touch with my mentors via IRC and I've been looking at the code and making minor patches. [16:11:33] rohit-dua: university exams - same here ! gets over in two days for me. [16:11:35] I am okay with the project so far and the communication is good with my mentor andrewbogott [16:11:43] * andrewbogott waves [16:11:54] * Micru waves [16:12:03] tonythomas: lucky you. ;-) [16:12:03] I started with some basic idea for the project and exploring various features in mediaiwiki. Submitted two reports for work. [16:12:04] tonythomas, toolserver was much faster for European users; but mosh works, yes [16:12:19] hi qgil, thanks for your answer! [16:12:30] rillke: specially if you are in India - mosh is great ! [16:12:35] I'm glad to have 2 people assisting mentors so we can cover more languages (esp ones I'm not familiar with yet) [16:12:39] Spent most of the time documenting the requirements and reading. Apart from daily chats, I had a meeting with my mentor. afk. [16:12:43] no wonder, since toolserver hardware is in Amsterdam [16:12:44] qgil, i know ^^ [16:12:48] I am in good pace with my work since the start of May ..... planned to few outreach ideas along with my mentor and shall soon be finalizing and publishing them ! [16:12:55] qgil, thanks [16:13:01] qgil, reports should be submitted weekly or monthly? [16:13:09] mutante: It used to be almost 1 sec delay between keystrokes :\ [16:13:24] rohit-dua, we will talk about reports as soon as we are done with the community bonding discussion [16:13:48] currently preparing a presentation for SMWCon on Thursday - also a great opportunity to talk to useful people [16:14:06] fyi, Micru is referring to this reply, which is relevant to GSoC/OPW: http://lists.wikimedia.org/pipermail/wikimedia-l/2014-May/071856.html [16:14:56] pretty good for the bonding time. we discuss in mailing list wikizh-l, not only the mentor, but also other guys [16:15:09] qgil, do you think it is convenient for all GsoC projects to use Phabricator? [16:15:24] Micru, letś talk about this after the reports. :) [16:15:31] The first report is about the community bonding period. [16:15:34] ok, ok, qgil :) [16:15:49] You have all learned something, you have made first decisions as a team, you have found some expected and unexpected obstacles... [16:16:14] Please reflect the most important aspects in a short report. This will give you some practice reportin. :) [16:16:21] Any questions about this first report? [16:16:48] what would be the approximate word count of this? [16:16:49] Copying from https://www.mediawiki.org/wiki/Google_Summer_of_Code_2014#Status : [16:16:56] community bonding report including links to [16:16:56] minimum viable product and goals [16:16:56] communication plan [16:16:56] lessons learned since 21 April [16:17:04] where should we post them? [16:17:18] sandaru_, reports are as short as possible, providing important details and links [16:17:21] Any preferred format to be followed for the report ? [16:17:47] Every time you write your report you have to think hat there are 22 more interns doing the same, and this is only two programs among the many we have around [16:18:16] roger [16:18:23] Also in the wiki we are identifying good practices (well, just one so far, but mentors and other vets are welcome to add more examples) [16:18:28] The example provided: https://www.mediawiki.org/wiki/User:Mariapacana/OPW_Progress_Report [16:18:44] Thanks qgil [16:18:45] Ok so FOSS OPW interns should follow the same format as GSoc i.e. weekly reports? [16:18:47] Create a subpage under your project wiki page and start reporting there, creating new sections for new reports [16:19:06] qgil: ok [16:19:20] We have seen all kinds of frequency used, from monthly to daily. It is a good idea to default to weekly reports. [16:19:51] mvolz, https://www.mediawiki.org/wiki/FOSS_Outreach_Program_for_Women/Round_8#Status contains exactly the same text, transcluded [16:20:07] We want to align GSoC and OPW processes and standards as much as possible [16:20:47] Why weekly? Because you can miss one week and itś not a big problem, but if you miss a second report it has been "only "14 days missing. [16:21:17] Previously we were defaulting to monthly reports, but then we found about serious problems really late (for a project that is supposed to last about 3 months only) [16:21:37] More questions about reporting? More opinions from mentors, ECT colleagues...? [16:21:55] how should this relate to the (minimum) fortnightly blog posts, if at all? [16:21:56] I think twice a month is a good rhythm, more is welcome [16:21:56] e.g watch list grouping project [16:22:22] qgil. I think wikimedia should require reports "every few days" [16:22:41] matanya2: what's a watch list grouping project? [16:22:54] i do that with deepali and molly... good results [16:23:00] previous gsoc project that died [16:23:11] ali_king_intern, OPW have this requirement for fortnightly blog posts. At least from the Wikimedia point of view, you could paste your two weekly reports preceded by an intro, and be good. [16:23:12] matanya2: thanks [16:23:19] Raylton: can you point to a sample report? am I right in inferring that each one was maybe a paragraph long? [16:23:50] guillom, Raylton "Wikimedia" is suggesting weekly, but every team can decide the way of woerking that fits best for them. [16:24:07] sounds good to me :) [16:24:17] We only want a simple communication plan agreed with the mentors, no sttress, and no ugly surprises. :) [16:24:33] sumanah, https://meta.wikimedia.org/wiki/Book_management_2013/Progress#Weekly_reports [16:25:18] Raylton, sumanah, other, you are encouraged to link to more god examples of reports next to the example of Maria Pacana at the GSoC page. [16:25:40] Anything else about reporting? Have I missed any questions? [16:25:49] qgil: I have an OT Q re: ugly surprises, will there be time later? [16:26:00] gnome gsoc reports are good [16:26:39] qgil, yes... I have problems with google-melange [16:26:46] Hi everyone, where can we host our server side scripts, wikimedia labs? [16:26:53] jlschatz, the ugliest surprise (imho) is when everybody thinks that an intern is doing ok, just having some temporary difficulties, until you realize that the project has been stalled and about to fail [16:27:16] all because having too much patience with silence, or with misc arguments avoiding some core problem [16:27:26] qgil, I sent an email to you but got no answer [16:27:31] apsdehal, gerrit or github ? [16:27:39] Regular reporting is not the solution for this problem, but at least helps detecting the symptoms earlier and better. [16:27:47] (gerrit-wm) [16:27:54] No I want to run, actually I have to create an api for wikimedia login [16:27:56] qgil: yes, that would be bad :). Very frequent reporting sounds really helpful for that. My Q was about the "if funding comes in" portion of our OPW contracts. [16:28:05] Raylton, the one about you not being officially a mentor? In my ToDo list, no worries. [16:28:09] apsdehal: yes [16:28:21] To which I can redirect my client side app towards that api for login [16:28:38] what about phabricator. shall it be used for all projects? [16:29:01] jlschatz, I'm not sure I understand the link between ugly surprises and funding. Can you explain, please? [16:29:05] jlschatz: [16:29:11] whoops. [16:29:16] Actually software I am building upon already has openid login [16:29:29] But wikimedia doesn't supports it [16:29:37] apsdehal, can you leave discussions about your project for later, please? [16:29:47] jlschatz: are you referring to gnomes issues with cash flow problems for the last round? [16:29:55] qgil: the contract says that if funding falls through, interns won't receive payments even if they're successful in their projects [16:30:00] qgil: Yes, sorry for that [16:30:10] apsdehal: spontaneously I'd say: Wikimedia Labs. [16:30:18] qgil, thanks. please answer before the midterm. [16:30:39] mvolz: i think the change in the contract language was a result of their cash flow issue (i read about it online, didn't hear from Gnome specifically on it) [16:30:41] jlschatz, Wikimedia pays its bills as much as Google. :) You shouldn't worry about funding. [16:30:56] qgil: :) thanks [16:31:19] Ok, anything else about reporting before moving to the question about Phabricator? [16:31:45] yes - when should community reports be submitted? My guess is as soon as possible... [16:32:17] a deadline is always welcome [16:33:02] rillke: you just need to update it as soon as you complete a milestone right ? why deadlines ? [16:33:33] thepwnco, the community bonding report can be submitted... soon. Let's say this week, since you are required to submit your first actual report about the project by 30 May. [16:33:42] tonythomas, deadlines help some of us focus [16:34:07] qgil, a blog post is ok for that? [16:34:46] We have just a few dates defined during the program, some of them set by the program organizers aka Google & GNOME. They are useful as sync points. [16:34:47] qgil: submitting == updating your report wiki page right ? [16:34:57] Raylton, for what? [16:35:06] tonythomas, yes. [16:35:15] qgil: cool then [16:35:28] qgil, community bonding report [16:35:54] We recommend you to submit brief reports in a wiki subpage (see the example above). If you want to write blog post, send updates to mailing lists, etc, you can also do it. [16:35:59] should we touch https://www.mediawiki.org/wiki/Google_Summer_of_Code_2014 again or leave it as it is ? [16:36:16] - for listing reports etc. [16:36:27] rillke, it is useful to keep your row in that table up to date. It gives a good view of the status of the whole program. [16:36:54] rillke, no need to list every weekly report, but main milestones. [16:36:54] e.g. mid-term evaluations [16:37:33] Reporting ok? :) [16:37:49] :) [16:38:03] (I will try to capture the usual questions and document the answers in the GSoC / OPW wiki pages. [16:38:14] The question about Phabricator [16:38:30] The bottom line here is that you need to define the way you want to work. [16:39:07] The wiki project page is a must, it needs to be updated reflecting always the current plan + the subpage with reporting. [16:39:55] Then, about the tasks, we have been using Bugzilla as a way to define tasks, blockers, etc, for coding projects. [16:40:09] Is it broadly known here what "Phabricator" is, when it comes to our infrastructure? (I wonder if I missed an email I'm not aware of) [16:40:20] (In short: It will replace Bugzilla at some point.) [16:40:21] Most of your projects have a central report in Bugzilla, so it is easy to create blockers for specific tasks, bugs, etc. [16:40:31] andre__, I'm getting there. :) [16:41:09] And then most of your projects also have a related code repository, by default in Gerrit. [16:41:47] We are just in the middle of a transition to go from these various tools to Phabricator (keeping MediaWiki for documentation) [16:41:55] https://www.mediawiki.org/wiki/Phabricator [16:42:35] In this context, you may want to consider the possibility of using/testing Phabricator for your project. [16:42:40] ... which is the advice I gave to Micru earlier today at http://lists.wikimedia.org/pipermail/wikimedia-l/2014-May/071856.html [16:43:05] Phabricator is useful to organize tasks, and you could use our instance at http://fab.wmflabs.org/ already now [16:43:18] ok [16:43:22] And it reliably works :) [16:43:34] For instance, this is what rillke is doing at http://fab.wmflabs.org/project/view/26/ [16:44:17] For code review... it might be premature, although it depends on how brave and how much are you willing to be an alphatester. [16:44:53] So you don't have to use Phabricator, and the default keeps being Bugzilla, but if you want to try and help testing/learning about this tool, you are welcome. [16:45:00] Is it clear? [16:45:20] qgil i need to leave. i'll see the logs and ask some question by email. ok? [16:45:40] ok Raylton [16:45:57] thanks [16:46:08] qgil, where should i host server side scripts(like python-bot scripts)? [16:46:51] rohit-dua, your mentors should know, but I guess the answer is Wikimedia Labs [16:46:59] rohit-dua: would Wikimedia Labs work for this? [16:47:53] sndre__, yes. i meant where should i publish the code for the bot for review process [16:47:53] https://www.mediawiki.org/wiki/Wikimedia_Labs [16:48:14] gerrit [16:48:14] rohit-dua, please agree this with your mentors [16:48:25] qgil, ok :) [16:48:31] Gerrit is the default option, yes. [16:48:52] qgil: let me make sure: https://www.mediawiki.org/wiki/Google_Summer_of_Code_2014#Status "community bonding report including links to: * minimum viable product and goals * communication plan * lessons learned since 21 April" -- can you please explain what you mean by "minimum viable product"? Because I think you are saying "DESCRIPTION of what the mentee WILL achieve, at minimum, within the next 3 months" [16:48:52] More questions? [16:49:19] About minimum viable product [16:49:26] +1 for sumanah's Q [16:49:33] 1 min... searching [16:49:46] honestly I think we should either remove that jargon entirely and leave "goal" (which everyone understands) or explain better. [16:50:18] See https://www.mediawiki.org/wiki/Mentorship_programs/Selection_process [16:50:34] also for sumanah's question, if that's for the entire project is there not a section for "things achieved since 21 April"? [16:50:41] I find the MVP concept very useful although it's true that everyone might not understand it. [16:50:46] or is that just "lessons learned"? [16:50:48] "A good team must work on a good plan. A good plan is expected to define a minimum viable product (the content of your first testable release) and build the rest of features on top of it." [16:50:51] in most cases, the goals/mvp description that would belong that initial "community bonding report" is what the mentee wrote up in their Deliverables section of their initial proposal, I think [16:51:13] qgil: that's still unclear and confusing, because that sentence you just quoted conflates the planning and the building. [16:51:18] ^ that's what I thought as well. I'd pretty much just be saying "see [[section]] in proposal" [16:51:31] mvolz: it's also less relevant to projects like mine that focus on standards and documentation [16:51:36] sumanah, yes yes, letś clarify. Still searching references.http://en.wikipedia.org/wiki/Minimum_viable_product [16:51:42] That's true; although my project has changed pretty substantially since then. [16:51:45] fhocutt: good point [16:51:46] The problem we want to solve is: [16:51:59] projects that at the end of the program cannot deliver a single thing [16:52:02] I understand the terminology, I'm just not going to have a testable release [16:52:06] because they aimed to do everything but finished nothing [16:52:15] Right [16:52:43] this is why already during the selection process we have encouraged students and mentors to define a minimum viable product, a first minimal release to build upon [16:52:58] again, qgil, when you say "to build upon" it [16:53:00] it gets confusing [16:53:01] is there a sample community bonding report? [16:53:11] this is new to this round, correct? [16:53:19] Even in the case of documentation you may want to define a focus in one areas as opposed to try to push all "chapters"at once [16:53:29] yes [16:53:54] which my mentee has already done, as she worked out the list of deliverables she would be delivering [16:53:54] sumanah, "release soon, release often" [16:54:00] qgil: yes. I am objecting to your wording as confusing. [16:54:14] get something out that someone else can look at (a minimum viable product, a prototype...) and then continue improving it [16:54:23] this makes sense as a concept but I still don't know what I'm supposed to put in the report. [16:55:04] fhocutt: My advice: just copy your list of deliverables (a.k.a. goals) that you aim to achieve by the end of the summer. [16:55:06] do we need to repeat what our minimum viable product is if we've already defined it in our proposal? [16:55:06] please propose a better wording [16:55:44] sumanah, I still think it is useful to define your first "publication", that should come before the mid-term evaluation [16:55:50] fhocutt: If we find that you have time and ability, then you'll be able to do more stuff beyond that -- file more bugs, polish more documents, and so on. [16:55:50] your initial focus [16:56:30] ok, then, fhocutt, copy the list of the things you'll be achieving in the first 6 weeks. you have your deliverables set up by week already so that should be fairly straightforward [16:56:43] Looking at previosu projects, those that didn't have anything to show before the mid-term struggled during the second half. [16:56:47] thanks, sumanah, will do [16:56:49] cool [16:56:57] Have to go. Thanks for this meeting, qgil. Very useful IMO. [16:57:29] wctaiwan, you don't need to details you plans in the community bonding report, just link to the relevant places. [16:57:40] okay, thanks. [16:57:42] siebrand, thanks! [16:57:43] also have to run. Thanks for this. Happy interning all! [16:58:13] sumanah, I will try to clarify the wording. Nobody had complained until now. :) [16:58:32] 2 minutes [16:58:35] I changed https://www.mediawiki.org/wiki/Google_Summer_of_Code_2014#Status to ask for "goals for the first half of the internship" [16:58:43] sumanah, good. [16:59:36] Ok, it looks that this is it. Thank you everybody! [16:59:47] If you have more questions, you can find us at wikitech-l and the usual places. [16:59:54] thanks qgil [17:00:02] thank you qgil [17:00:05] thanks qgil, sumanah [17:00:21] Nice meeting qgil . Thanks :) [17:00:28] Thank you for this informative discussion qgil ! :) [17:00:32] thanks qgil [17:00:37] thanks everyone [17:00:38] Thanks, qgil! [17:01:02] ^^ and everyone! sumanah, too - you read my mind! [17:01:10] :) [17:01:30] bye! [17:01:39] * tonythomas waves bye-bye [17:02:42] Thanks qgil [17:04:42] Hm. [17:09:03] qgil: andre__: BTW, our current and past projects are now listed automatically in our navigation template (courtesy of a few improvements to my Lua module): https://www.mediawiki.org/wiki/Template:Engineering_Community_Team [17:09:55] qgil: is there another meeting at 1600 UTC? https://m.mediawiki.org/wiki/Engineering_Community_Team/Meetings/2014-05-20 (or were my TMZs out of whack?) [17:10:33] 16:00 UTC was an hour ago [17:10:53] guillom: so then it's the latter :D [17:10:57] :) [17:29:39] guillom, automatically? I just added it manually. Sorry. [17:30:30] jlschatz, according to http://www.timeanddate.com/worldclock/fixedtime.html?iso=20140520T09&p1=224&ah=1 16:00 UTC was 90 minutes ago. :) [17:30:39] qgil: projects are listed automatically; meeting notes still need to be added manually for the moment [17:31:28] qgil: yes. I have a meeting *next* Tuesday at 11am my time and I still had that window open. Bah. Rookie. [17:32:11] guillom, ah ok. [18:50:27] plop