IRC office hours/Office hours 2013-09-25

Time: 17:00-18:23 UTC (22:30-23:33 IST)
Channel: #wikimedia-office
Timestamps are in IST (UTC+5:30).

[22:30] <arrbee> === Wikimedia Language Engineering Office Hour - Start ===
[22:31] <arrbee> Hello, Welcome to the office hour of the Wikimedia Language Engineering Team.
[22:31] <arrbee> My name is Runa and I am the Outreach and QA coordinator for our team.
[22:31] Nikerabbit Nischayn22
[22:32] <arrbee> The other members of our team present today are: kart_ Nikerabbit pginer siebrand and sucheta
[22:32] * arrbee looks if she missed anyone
[22:33] <arrbee> Our last office hour was held on July 10 2013 and the logs are at: https://meta.wikimedia.org/wiki/IRC_office_hours/Office_hours_2013-07-10
[22:33] <arrbee> Today we will briefly talk about some of our ongoing projects
[22:34] <arrbee> Also we'll talk about our participation in Google Summer of Code (GSoC)
[22:34] <arrbee> After that we are open for questions
[22:34] <arrbee> If you'd like to send in your questions before the Q&A, please send them to me as a private message
[22:35] * gwicke is now known as gwicke_away
[22:35] <arrbee> As I mentioned before, we are the Wikimedia Language Engineering Team
[22:36] <arrbee> Since our last office hour, we have one new person joining in - sucheta . She was an OPW intern last year at the WMF.
[22:37] <arrbee> Our team works on enabling language tools to make MediaWiki and Wikimedia projects support all languages of the world.
[22:37] <arrbee> At present there are ~300 languages within the Wikimedia projects, which we support
[22:37] <sucheta> :-)
[22:38] <arrbee> Presently our focus is on:
[22:38] <arrbee> 1. creating a new landing page for translatewiki.net
[22:38] * marktraceur applauds wildly for sucheta
[22:38] <arrbee> 2. better support of input methods on Visual Editor for languages that use non-latin scripts
[22:39] <arrbee> 3. preparations to support content translation of wiki pages
[22:39] * gwicke_away is now known as gwicke
[22:39] * gwicke is now known as gwicke_away
[22:39] <arrbee> === Announcement ===
[22:39] <arrbee> We begin with a quick announcement
[22:40] <arrbee> The MediaWiki Language Extension Bundle for September (2013.09) was released today.
[22:40] <arrbee> It provides easy way to bring ultimate language support to your MediaWiki.
[22:41] <arrbee> More information is available at https://www.mediawiki.org/wiki/MLEB
[22:41] <arrbee> The release announcement from kart_ can be found here: http://lists.wikimedia.org/pipermail/mediawiki-i18n/2013-September/000751.html
[22:41] <arrbee> === On-going projects ===
[22:42] <arrbee> 1. translatewiki.net Main Page
[22:42] <arrbee> A new main page for translatewiki.net was in the development plan for sometime.
[22:42] <arrbee> The page has now been enabled for logged-in users.
[22:42] <arrbee> You can see it at https://translatewiki.net/wiki/Special:Mainpage
[22:43] <arrbee> The new layout adds some information tiles with statistics for projects, languages and users
[22:43] <arrbee> Now you can view your position in the translation ranking for the last month
[22:44] <Nikerabbit> repeating: you need to login to see it
[22:44] <arrbee> and also easily start translating or proofreading your project from the project cards
[22:44] <arrbee> Next Project update
[22:44] <arrbee> 2. Input methods for Visual Editor
[22:44] * gwicke_away is now known as gwicke
[22:45] <arrbee> At present there are some significant shortcomings when a user tries to write in a non-latin script in the Visual Editor
[22:45] <arrbee> The VE and Language team have started to work together to fix these problems
[22:45] <arrbee> All input method related bugs for VE are now being tracked through this bug
[22:46] <arrbee> https://bugzilla.wikimedia.org/show_bug.cgi?id=33077
[22:46] <arrbee> Final project update for today: 3. Content Translation of wiki pages
[22:46] <arrbee> Development for this project has not started yet
[22:47] <arrbee> However, initial designs are available at: https://www.mediawiki.org/wiki/Content_Translation
[22:47] <marktraceur> arrbee: I'll look at the page and feel free to say "It's on the page", but will that be built in a WMF-specific way?
[22:48] <siebrand> What do you mean by "built in a WMF specific way"?
[22:49] <Ironholds> dependencies on our architecture, maybe?
[22:49] <marktraceur> siebrand: I mean, built in a way that only the WMF's sites will be able to use - we've been doing a better job of supporting third parties and I like encouraging that
[22:49] <siebrand> I can make assumptions, but I choose not to, Ironholds.
[22:49] <marktraceur> siebrand: I'll assume you don't make assumptions from now on, then
[22:50] <siebrand> marktraceur: WMF has a specific project called Wikipedia. THe feature is primarily aimed at facilitating translation of pages in the main namespace of Wikipedia language versions.
[22:50] <marktraceur> 'kay
[22:50] <siebrand> No use cases outside of that are being taken into account at the moment.
[22:50] <marktraceur> siebrand: That page could be cleaned up to make that more obvious, or even the name of the project :)
[22:51] Maeby manybubbles marktraceur MaxSem
[22:51] <siebrand> I'll have a look, and make that clear. Thanks for the tip.
[22:52] <arrbee> Recently the Language Engineering team also had a meeting with the Analytics team to identify metrics for the project.
[22:52] <arrbee> This will allow us to understand how the tool will be used and whether the users can achieve their goals
[22:52] <siebrand> marktraceur: done
[22:52] <marktraceur> Thanks!
[22:52] <arrbee> Slides used during that meeting are at https://commons.wikimedia.org/wiki/File:Content_translation_analytics_overview.pdf
[22:53] <arrbee> More information to follow
[22:53] <arrbee> Meanwhile, if you would like to participate in a usability test that is being conducted by pginer (Pau), please sign up using this form:
[22:53] <arrbee> https://docs.google.com/a/wikimedia.org/forms/d/1yCvPS65eWk9S8uXkksAbDbLsbZQd0ISQKBDFfJnSSo0/viewform
[22:54] <arrbee> === Google Summer of Code (GSoC) ===
[22:55] <arrbee> 4 students had participated for the Language Engineering projects for GSoC this year: harshkothari, mooeypoo, orsagi and prageck
[22:55] <arrbee> We had published a mid-term round up in one of our earlier blog-posts:
[22:55] <arrbee> http://blog.wikimedia.org/2013/07/23/updates-from-the-language-engineering-google-summer-of-code-projects/
[22:55] <arrbee> The official period for GSoC ends in 2 days time and the students are now finishing up with the final tasks
[22:56] <arrbee> I'll provide a quick round-up for each of the projects
[22:56] <arrbee> 1. Mobile app for Translate - orsa
[22:56] <arrbee> The app is now available on the Google Playstore: https://play.google.com/store/apps/details?id=net.translatewiki.app
[22:56] <arrbee> Grab it from there!
[22:57] <arrbee> 2. Browser plugins for Input methods - prageck
[22:57] <siebrand> Or Sagi is not here at the moment. I'm very thankful for his contribution.
[22:57] <arrbee> Browser plugins for jquery.ime were published
[22:57] <arrbee> Firefox extension https://addons.mozilla.org/en-US/firefox/addon/wikimedia-input-tools/
[22:57] <arrbee> Chrome extension https://chrome.google.com/webstore/detail/wikimedia-input-tools/fjnfifedbeeeibikgpggddmfbaeccaoh
[22:58] <siebrand> prageck has taken jQuery.IME and made it into a Firefox and Chrome plugin. This allows a browser user to use the input methods in every web site.
[22:58] <siebrand> Pretty cool.
[22:58] <arrbee> prageck also worked on adding onscreen keyboards to jquery.ime and the basic features are already working
[22:58] <arrbee> http://praveensingh.in/projects/jquery.ime/examples/osk.html
[22:58] <prageck> siebrand: :)
[22:59] <arrbee> 3. RTL support for VE - mooeypoo
[22:59] <kart_> prageck: nice work!
[22:59] <prageck> kart_: Thanks :)
[22:59] <arrbee> Thanks to this project, the crucial RTL bugs in VE were fixed, and VE is usable in RTL wikis
[22:59] * siebrand cheers mooeypoo on!
[22:59] <mooeypoo> \o/
[23:00] <arrbee> The project also included the development of a language selector to state the langauge in which parts of a page are written.
[23:00] <arrbee> And an extension for template data editing
[23:00] <marktraceur> Woo!
[23:00] * marktraceur <3s TemplateData
[23:00] <arrbee> :)
[23:00] <arrbee> 4. Language Dashboard - harshkothari
[23:01] <mooeypoo> Now you can <3 then right-to-left, marktraceur.
[23:01] <marktraceur> >3?
[23:01] <arrbee> A dashboard to query the language coverage on Wikimedia projects was created through this project.
[23:01] <marktraceur> mooeypoo: https://bugzilla.wikimedia.org/show_bug.cgi?id=54140 also.
[23:01] <arrbee> The demo instance is here: http://tools.wmflabs.org/lcm-dashboard/lcmd/
[23:01] <arrbee> (also includes link to the API console)
[23:02] <arrbee> The UI has been optimized to be viewed (list and graphical views) on web and mobile browsers
[23:02] <aharoni> mooeypoo, marktraceur : Ɛ>
[23:02] <arrbee> All the students shall be also following up with more reports about their projects through their blogs/reports. :)
[23:02] <arrbee> That ends the updates.
[23:02] <siebrand> Harsh Kothari isn't here at the moment. He's worked hard towards getting a final product that provides insight in language support in MediaWIki. sucheta has also helped out.
[23:02] <sankarshan> The "internationalization library" check box has a spelling error.
[23:03] <arrbee> sankarshan: noted. Thanks.
[23:03] <arrbee> === Q and A ===
[23:03] <siebrand> Many, many thanks to all the GSoC students for their efforts. It's been really great this year (again!).
[23:04] <mooeypoo> It's been an absolutely awesome experience, thank you all for the support and guidance and to aharoni for the great mentorship!
[23:05] <aharoni> thanks mooeypoo for your great work, and also to all the VisualEditor team for the indispensable help - InezK , James_F|Away , catrope , trevor , rmoen|away and the rest of the gang
[23:06] * rmoen|away is now known as rmoen
[23:06] <aharoni> a little update from me:
[23:07] <aharoni> Nikerabbit added support for translating MathJax to translatewiki.net
[23:07] <aharoni> https://translatewiki.net/wiki/Special:Translate/out-mathjax-0-all
[23:07] <aharoni> This brings us much closer to far nicer display of Mathematical formulas on Wikipedia.
[23:07] <aharoni> The inability to translate it was one of the major blockers until now, and now it's removed.
[23:08] <aharoni> So you are very, very welcome to translate it.
[23:09] <aharoni> questions, anybody?
[23:09] <Scott_WUaS> Hi Everybody ...
[23:09] * cscott pops up
[23:09] <Pavanaja> Any dates for full integration of ULS with VE?
[23:09] <cscott> sorry, i just saw the office hour announcement
[23:10] <siebrand> What do you mean by "full integration", Pavanaja ?
[23:10] <cscott> i'd like to talk about languageconverter, in general, if there's a moment.
[23:10] <aharoni> Pavanaja: mooeypoo integrated the language selection part of ULS into VE.
[23:11] <Pavanaja> siebrand: at present templates and other dialog boxes in VE dont support ULS
[23:11] <aharoni> Santhosh, I and the VE team, are working on improving the typing support.
[23:11] <arrbee> cscott: sure..we just started the Q&A
[23:11] <siebrand> We think we know what needs to be done at the moment, but we do not know when (a) jQuery.IME input methods are supported, and (b) many classes of other input method types.
[23:11] <Scott_WUaS> Are there new/others Wikidatas outside of Wikimedia which are engaging Wikipedia's 285 languages via this Language Engineering team or otherwise? E.g. Google Translate as a related example?
[23:12] <siebrand> We will make jQuery.IME input methods work first, because we think that will increase our understanding of the complexities for the other input methods.
[23:12] <aharoni> Pavanaja: what do you mean by "templates and other dialog boxes in VE dont support ULS"? In what way they don't support it? Are you talking about typing?
[23:12] <siebrand> So those will be second.
[23:12] <Pavanaja> aharoni: Yes
[23:12] <siebrand> It's a high priority for both VisualEditor as well as Language Engineering team, and we have a dedicated developer working on it.
[23:13] <siebrand> As soon as there is progress, it will be made available to users.
[23:15] <aharoni> Pavanaja: Thank you for noting this. Is this reported in Bugzilla?
[23:15] <arrbee> cscott: would you like to go ahead with your question?
[23:15] <Pavanaja> aharoni: Yes
[23:16] <siebrand> cscott: Not sure if this is the right venue for language converter discussion. We should probably plan a proper meeting for this.
[23:16] <arrbee> ahh ok. Lets move to Scott_WUaS' question then
[23:17] <Scott_WUaS> Thanks
[23:17] <siebrand> Scott_WUaS: I'm not aware of other instances of wikidata aside from wikidata.org, which is hosted by Wikimedia.
[23:17] <aharoni> Scott_WUaS: what do you mean by "Wikidatas"? Other installations of Wikidata? The Wikibase extensions?
[23:17] <siebrand> Scott_WUaS: How is Google Translate related to this question, or is that a separate question?
[23:18] <Scott_WUaS> Perhaps non-Wikimedia organizations using Wikidata, which I assume is C.C. licensed and open?
[23:19] <cscott> siebrand: i was thinking i might scoot in at the end after proper users' questions are answered, since a good # of the language team is already assembled?
[23:19] <cscott> but i'm certainly planning to wait until other conversations are done here. ;)
[23:19] <Scott_WUaS> Google Translate is an example of an organization that could fruitfully engage Wikipedia's languages (albeit Google Translate is not wiki)
[23:19] <siebrand> Scott_WUaS: I'm not aware of any such efforts.
[23:20] <Scott_WUaS> Thanks.
[23:20] <siebrand> Scott_WUaS: We are actively seeking contact with Google Translate product management to cooperate on translation related efforts, like translatewiki.net, page translation on Meta-Wiki and mediawiki.org, and other Wikimedai wikis using the Translate extension, and for the future Content Translation feature.
[23:21] <Scott_WUaS> In what ways could a wiki-organization independently engage Wikidata as well as Wikipedia's languages that this Language Engineering team focuses on.
[23:21] <siebrand> cscott: Yeah, shoot. No other open questions, I believe.
[23:21] <Scott_WUaS> siebrand ... Great
[23:22] <siebrand> Scott_WUaS: I think that question is more related to Wikidata than i18n in Wikimedia. Let me look up office hours opportunities for Wikidata.
[23:23] <cscott> so i hope you've heard that erik is having a meeting next week with liangent, James_F|Away, and i to discuss language converter on zhwiki (and VE, presumably)
[23:23] <Scott_WUaS> I'm developing an online, C.C., wiki university project with plans to be in all 7,105 + languages (and is like Wikipedia with MIT OCW)
[23:23] <siebrand> cscott: No.
[23:23] <cscott> siebrand: i was afraid of that
[23:23] <Scott_WUaS> Thanks.
[23:24] <cscott> siebrand: so i'd like either better language team representation in that meeting, or else i need to learn a lot more about language converter beforehand to effectively argue your case.
[23:24] <siebrand> cscott: I think liangent is the leading authority on the language converter.
[23:24] <siebrand> cscott: sad but true.
[23:24] <arrbee> Scott_WUaS: I don't see any wikidata office hour coming up. You keep an eye out on this page for the next one: http://meta.wikimedia.org/wiki/IRC_office_hours#Upcoming_office_hours
[23:24] <arrbee> Scott_WUaS: Lydia_WMDE may be able to help you too
[23:25] <cscott> my understanding of erik's position is that language converter looks like a lot of complexity for a small wiki, and he's hoping to avoid supporting it
[23:25] <Scott_WUaS> arrbee: thanks
[23:25] * bd808 is now known as bd808|LUNCH
[23:25] <siebrand> Scott_WUaS: You can also join #wikimedia-wikidata to engage directly with the Wikidata people.
[23:25] <cscott> liangent is an authority on language converter, but the underlying language problem is more broad
[23:25] * Ironholds is now known as IH|away
[23:25] <Scott_WUaS> siebrand: thanks
[23:26] * rmoen is now known as rmoen|away
[23:26] <siebrand> cscott: Where that could be true for Serbian, I think it's hard to argue that for zh-hans/zh-hant.
[23:26] <cscott> i've been trying to make the case that language converter should be broadly utilized across all our projects, eg to support en-gb/en-us, because if more editors understood the problems it solved, it would be better supported (and the UI for related tasks would improve)
[23:26] <siebrand> (s/Serbian/Serbian and some other languages)
[23:26] <cscott> well, from my knowledge language converter could be used for es-*, pt-pt/pt-br, sr/sr-ec (at least) as well as the zh-* wikis
[23:27] * James_F|Away is now known as James_F
[23:27] <cscott> i think there are lots of arabic dialects/topolects as well?
[23:27] <siebrand> cscott: LanguageConverter is a pretty hard to understand feature, and I think it's probably a barrier to entry. Especially if editors have to be made aware of its existance.
[23:27] <arrbee> Time check - we have <5 minutes in this room. We are on #mediawiki-i18n though to continue the conversations :)
[23:27] <siebrand> cscott: If it would be possible to have it be completely transparent to editors (and possible also readers), I see more value for it.
[23:27] <cscott> language converter has not been documented in anything other than mandarin for most of its life, afaict
[23:28] <siebrand> I think that's most probably correct.
[23:28] <siebrand> It's a very esoteric feature.
[23:28] <cscott> siebrand: right. so i'd argue that we should elevate 'language converter functionality' (if not the actual codebase) to the point where the UX is intuitive even in english.
[23:28] * InezK is now known as InezK_away
[23:28] <cscott> this is a hard sell for erik
[23:29] <siebrand> The problem for me from a priority point of view, is that language conversion is not at a visible priority level.
[23:29] <Nikerabbit> I'm unsure how well language converter would work for handling language variation instead of mainly script conversion
[23:29] <cscott> but the alternatives (as i see them) are: (a) forcibly fork wikis (zh-hant/zh-hans as a start), or (b) keep zh/pt/sr/en-gb in a ghetto.
[23:29] <cscott> Nikerabbit: it works fine. that is, it would work fine if we wanted to make it work. ;)
[23:30] <siebrand> So either I misunderstand the value compared to other things Language Engineering can do, or it's indeed not something that can be called a priority.
[23:30] <siebrand> I'm sorry, I must leave for another meeting now.
[23:30] <Nikerabbit> I'd argue that the current implementation is based on the assumption that word order doesn't change
[23:30] <cscott> Language Converter really has two parts: one is writing-system conversion, and the other is word conversion. they are a bit tangled together for hans/hant because you need flexible word conversion in order to do that particular script conversion. but the tasks are orthogonal.
[23:31] <cscott> Nikerabbit: not really, they use use larger -{ ... }- spans to accomodate.
[23:31] <arrbee> I hate to interrupt, but we have run out of time. The office hour conversation has to be stopped here. Don't want anyone held up from using this channel
[23:31] Nikerabbit Nischayn22
[23:31] <cscott> siebrand: can we find another time -- and or, can you invite yourself to erik's meeting next week?
[23:32] <cscott> i'm going to move over to #mediawiki-i18n
[23:32] <arrbee> cscott: Nikerabbit : siebrand : #mediawiki-i18n is open to continue
[23:32] <siebrand> cscott: I cannot invite myself to that meeting, but I'd like to be invited :). Can you make an effort to make that happen?
[23:32] <arrbee> cscott: Thanks :)
[23:32] <cscott> siebrand: i've suggested it once already. <sigh>
[23:32] <arrbee> If nothing changes, we shall be back with our next office hour 9th October
[23:32] <arrbee> Our office hours are held every 2nd Wednesday of the month. (Today is an exception)
[23:33] <arrbee> Thanks everyone for coming!
[23:33] * aharoni_ is now known as aharoni
[23:33] <arrbee> === Wikimedia Language Engineering Office Hour - End===