[02:26:09] * Jasper_Deng wonders if Sven_Manguard has checked his Wikidata watchlist notice [02:26:22] not in the last few hours, no [02:26:25] what did I miss? [02:26:51] RfCs [02:26:54] various ones [02:27:17] Oh. Well I've been inactive in general. [02:29:39] It'll be nice when we can do the Commons IW links via wikidata e.g. https://commons.wikimedia.org/w/index.php?title=Category:Rolls-Royce_RB211&action=edit [02:30:21] why aren't we including Commons in this, again, Reedy? [02:30:27] (other than community consensus) [02:30:43] I'm not sure tbh [02:31:09] Presumably commons could just be a "language code" and it would sorta just work..? [02:31:15] It's going to look terrible when I run for Oversighter having failed Crat. However I'm better suited for OS than 'crat. I really only wanted Crat to help with bots. [02:31:31] I've had to get CUs enough to know what is and is not CUable [02:32:00] omfg, no Lydia [02:32:14] For example, one person uploaded their tax return application, complete with EVERYTHING you could think of that would be private [02:32:24] why? Because she didn't have a printer handy! [02:33:05] and these kinds of people (possibly) vote [02:33:13] * Sven_Manguard sighs [02:35:51] * Reedy will ask "Y U NO COMMONS!?" later [02:36:55] Sven_Manguard: to clarify, even if a non-admin ran for OS, he/she would have to go through an RfA-like process [02:37:06] so the foundation would actually be OK w/ that [02:37:18] Jasper_Deng: not really [02:37:27] Have you read the thread on enWP? [02:37:37] I have [02:37:41] Phillipe make it pretty clear that that wouldn't work, IMO [02:37:44] they were OK if an RfA-/like/ process [02:37:58] what's not OK is arbcom appointing them w/o an RfA-like process [02:37:59] Sven_Manguard: philippe commented on the wikidata RFC [02:40:24] I saw [02:45:36] * Jasper_Deng thinks Sven_Manguard is on the fence about voting percentages [02:45:54] hmm? [02:46:03] because I stopped part of the way through the RfC? [02:46:07] yeah [02:46:12] No, I'm doing an important RL email [02:46:36] which reminds me, does anyone here speak Chinese? [02:46:45] I only speak it, not read it [03:04:10] oh, yes, forgot what I came here for. Jasper_Deng, legoktm, others: http://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Wikidata_Phase_2 [03:04:21] * Jasper_Deng saw [03:04:24] I'm undecided [03:04:32] thanks [03:04:38] ill take a look later [03:22:30] stewards have OS, right? [03:22:49] legoktm? [03:22:53] yes [03:22:56] thought so [03:22:58] mhm [03:44:38] hmmm [03:44:56] I could have sworn that Vituzzu was not the only steward we had on the project [03:45:15] am I missing someone? [03:45:18] Jhs is also [03:45:58] ah, yes [03:46:07] were there not more? [03:46:25] sDrewth is too (but not a sysop) [03:46:38] (IRC nick, ofc) [03:48:23] aka billinghurst [03:53:27] We do have a RFA-like process for OS [03:55:54] lol [03:56:06] I entered "I accept these terms for my future edits. Do not show this message again." into the search box [03:56:22] I get "Did you mean: I accept these terms formy future edittools. De no show this mainpage again. [03:56:22] " [03:56:30] lol [04:00:47] I still get that dialogue every time I open up the browser [04:01:08] I never do [04:04:10] I'm getting tired of that message appearing [04:04:29] I can't find the .js file for it [04:05:22] :/ [06:11:51] Hey all. [06:11:57] The "By clicking "save", you agree to the terms of use, and you irrevocably agree to release your contribution under the CC0 license. [06:11:58] I accept these terms for my future edits. Do not show this message again" [06:12:04] message isn't going away for me [06:12:20] I click do not show again and nothing happens [06:12:30] any ideas? [06:15:07] yeah theres a bug filed [06:15:27] * legoktm looks [06:15:41] Sven_Manguard: probably fixed today [06:15:45] https://bugzilla.wikimedia.org/show_bug.cgi?id=47435 [06:15:56] k [06:16:19] there's a patch, but need to find someone to approve it and deploy it :) [06:16:45] Reedy? [06:17:53] Jasper_Deng_busy: probably, but he's not a JS expert so might not be the one to approve it [08:28:20] New review: Daniel Kinzler; "(3 comments)" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59746 [08:29:24] hashar: hi! quick question about gerrit setup... we were discussing that it would be nice of CR+1 was "sticky", like +2 is. [08:29:33] hashar: is this configurable per project? [08:29:42] and, was this already diswcussed for core somewhere? [08:29:55] Guten Tag [08:30:16] Bonjour :P [08:30:28] DanielK_WMDE: OpenStack uses a python script that reapply the votes whenever the change has been detected as being trivial [08:30:36] I am not sure whether it is build in gerrit though [08:30:49] might be a feature request, let me look in our bug tracker [08:31:19] hashar: well, it always re-applies a -2 [08:31:35] yup -2 is meant to be blocking a change regardless of the patchset [08:31:38] hashar: actually, sorry - i was thinking of *-1*, not *+1* [08:31:39] anyway [08:31:48] but the rest of the votes are on the patchset itself [08:31:55] hmhm... [08:32:06] i'd like my -1 to stick around though... [08:32:43] ideally, it could detect whether all the lines that had comments were touched... but even then. [08:32:44] another thing is that when a new patchset is uploaded that will clear its state and shows it bold in the users dashboards [08:33:08] ah https://bugzilla.wikimedia.org/show_bug.cgi?id=41074 "report trivial rebases in Gerrit" (opened by me) [08:33:22] Chad said: Re-applying the votes is already a work-in-progress [08:33:22] yea... but it's easy to miss earlier complains by others. [08:34:07] keeping the votes when rebasing is the minimum, but i'd really like to keep them even in case of non-trivial changes [08:34:16] whoever gave the -1 should have to look and removed it [08:34:51] to me a new patchset need a review again and that make sense to clear the votes [08:34:57] i'd like to try this at least for Wikibase... but I guess it's all hackery, and not easily done per project :/ [08:35:30] anyway, you want to talk about it with ^demon whenever he shows up (usually at 2pm) [08:35:31] hashar: but if someone pointed out an issue two patchsets ago, it's very easy to miss that once a new patch set is up. [08:35:37] I would be happy to participate [08:35:39] yea, will do [08:35:40] thanks [08:35:49] and yeah sometime we can miss a new patchset :( [08:36:11] when I want to make sure I dont miss an update, i do CR-2 [08:36:13] not ideal [08:37:35] hashar: not ideal at all. this is how the issue came up: interpretation of what "CR-2" means. to me it means "bad idea, unfixable, forget it". [08:38:02] but i wals want a sticky "you need to fix this before we can merge it" thingy. [08:38:08] whereas to me it is more like : "that is going to break prod horribly, make sure it does not land in master by mistake" [08:38:09] :D [08:38:20] yea... [08:38:33] bound to cause confusion and ruffled feathers :/ [08:38:39] yup [08:41:11] bah... [09:20:48] Tobi_WMDE: https://gerrit.wikimedia.org/r/#/c/60133/ [09:23:15] hi [09:23:30] hi [09:27:23] 5 million articles left with IW links :D [09:27:28] shoop da wop [09:27:31] *woop [09:30:15] why so many? [09:31:16] would it make sense to go through the wikis, smallest first, and get them cleaned up and empty one by one? then we have milestones and targets, instead of cleaning all of them a bit all the time and finish with all of them at once? (just wondering loud. by the way thanks for the amazing job, addshore) [09:32:23] looking at http://bots.wmflabs.org/~addshore/addbot/ some of them should be too hard to finish ;p [09:32:28] and i think some are already done ;p [09:33:14] looking at that i am still confused :D [09:33:20] have you figured out what the rows mean? [09:33:37] note the bot isnt running yet ;p [09:33:40] Denny_WMDE: if we hit them all at once, we create work lists of conflicts so that no matter what language you speak, you can still help [09:33:43] the collums are now accurate ;p [09:34:14] legoktm: just gonna reorder my bot stuff quickly ;p [09:34:16] legoktm: ok, makes sense, you're right [09:35:07] ok, so what do the columns mean? [09:35:13] Denny_WMDE: Its all in a pretty new db now ;p [09:35:31] "pending" means "still needs to be checked and probably has local iwlinks"? [09:35:41] yes [09:35:53] give me 30 mins and thatll all start updating ;p [09:35:56] and "checked" means "have checked this, and couldn't figure out what to do, left local interwiki links there" [09:36:01] ? [09:36:02] sure :) [09:36:13] yep :) [09:38:56] ok, then i will wait for the list to be updated :) am curious how it looks like [09:39:31] so, ch.wiki e.g. is almost done. there are only 5 pages to be checked, and all others have been cleaned? [09:45:14] New patchset: Tobias Gritschacher; "Make check for ULS language case-insensitive in Selenium tests" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60250 [09:45:35] Change merged: Tobias Gritschacher; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60250 [09:58:59] Change merged: Daniel Kinzler; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60213 [10:06:12] Denny_WMDE: yes [10:06:23] and for example I just took a look at xh which says it has 2 pending [10:06:37] but they are both done :) so as soon as the bot runs that will be removed from the db :) [10:06:51] nearly finished rewriting stuff [10:09:26] will you have a list of all wikipedias with pending 0, checked 0? [10:09:49] or simply a list at the end of "Wikipedias without language links" [10:19:13] liangent: hi [10:19:52] Denny_WMDE: hi [10:20:02] i am now here for questions [10:20:56] New review: Daniel Kinzler; "(3 comments)" [mediawiki/extensions/Wikibase] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/60214 [10:20:56] Change merged: Daniel Kinzler; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60214 [10:21:06] Denny_WMDE: well I have to leave soon later, though it's not that urgent [10:21:29] one question here: do you think data transcluded into articles should be in page language or user language? [10:22:13] it might look strange if a word in a sentence is in another language than the whole setence [10:23:06] blah [10:23:07] even when a word means something different in two languages [10:23:15] enwiki's taxonomy is messed up [10:23:29] it can cause users to misunderstand the info [10:26:08] liangent: it should be in the page language [10:26:24] liangent: or be very clearly marked up to be in a different language [10:26:27] Denny_WMDE: responded on enwiki [10:27:00] rschen7754: responed back already :) [10:27:12] Denny_WMDE: so there shouldn't many fallback issues in this aspect I guess? [10:27:33] New review: Daniel Kinzler; "(2 comments)" [mediawiki/extensions/Wikibase] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/60215 [10:27:33] Change merged: Daniel Kinzler; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60215 [10:28:00] liangent: on the client we should probably never use any user fallback, because this would mess with caching [10:28:05] only pages on wikidata site are affected by possible cache split I assume? [10:28:21] liangent: on the client, if at all, only global fallback would be ok, and should be marked explicitly [10:29:02] how "explicit"? language name in parentheses after the data, or just is enough [10:29:14] liangent: yes, only Wikidata pages would be allowed to mess with the cache. one possibility could be to implement the user language fallback only in javascript. this would mean that caches would still work. a little. [10:30:15] liangent: hmm. some class and lang would be needed, i guess, so that we can add appropriate css [10:30:44] yep [10:30:54] Change merged: Daniel Kinzler; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60216 [10:31:41] ok I gotta go now [10:32:11] btw, Denny_WMDE, what's your expected online time? [10:32:57] the next 5 hours or so [10:33:05] and probably then later again [10:33:08] for deployment [10:33:29] I mean in the future [10:33:37] at what time in a day [10:33:54] New review: Daniel Kinzler; "Seems like it would sense to have constants for the data type IDs. Magic values in literals are erro..." [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60217 [10:35:28] usually between 11am an 6pm Berlin time [10:36:03] Denny_WMDE: ok thanks. afk [10:53:30] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [10:53:30] New patchset: Tobias Gritschacher; "Selenium tests for autocomments/autosummaries" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60252 [10:56:25] Denny_WMDE: bot done again [10:56:37] infact, got 2 more bugs to squish [10:57:02] addshore: still no wikis completely finished? [10:57:07] or do they get removed from the list? [10:57:16] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [10:57:34] they get removed from the list [10:57:44] New patchset: Tobias Gritschacher; "Selenium tests for autocomments/autosummaries" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60252 [10:57:46] as they have no entries in the db ;p [10:58:04] * aude guesses those with interwikis like page#section are still there? [10:58:18] wonder how many? [10:59:33] aude: yes they are [10:59:52] just saw one here: https://ch.wikipedia.org/wiki/T%C3%A5si [10:59:58] New review: Jeroen De Dauw; "(1 comment)" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60214 [11:00:02] they are counted in addshore's db? [11:00:20] hmm, actually they have been on the "checked" side of things [11:00:25] they were although they are now slowly being removed so finding the others will be easier [11:00:26] ok :) [11:00:44] got 1 more bug to squish [11:00:54] addshore: would it be hard to list all wikis that are not in the db anymore? [11:01:09] hmm Denny_WMDE shouldnt be, ill have a go after this bug :) [11:01:29] something like: "23 Wikipedias are done: yadda.wikipedia, bla.wikipedia, blabla.wikipedia, … " [11:01:39] would be awesome :) thanks! [11:02:24] looking to clean up the croatian links… :) [11:05:04] Denny_WMDE: whats the lang code foe croationawiki? [11:05:07] *for [11:05:18] hr [11:05:39] running the bot against the remaining articles now [11:05:55] oh, if i can ask for wishes, also uz.wiki would be awesome [11:06:10] ;p [11:06:21] hehe :) [11:06:29] it's just wishlist :) [11:06:37] they should now be updating ;p [11:06:43] yay [11:07:21] * addshore has just turned off the importing of iwlinks to wd for addbot as there are just too many conflicts now.. [11:08:13] gonna run for all langs now [11:08:19] get ready for all of the numbers to update ;p [11:09:04] it's going to take a while to get used to the edit link on enwiki doing not what i expect now [11:09:26] * aude click edit source to see the links [11:17:06] hmm Denny_WMDE [11:17:06] stalgian|aagcztrt-smg-x-old|bm|cbk-zamboryhbu-vroprgnkkwf|ho|hz|iiab|kji|krhdejg|map-bms|mhn|moj|musvhp-nlw|ngvgmphya-rupa-taraehnomple|sqnqt|udmcpr-classical-min-nan-yue| [11:17:20] but I know any language with a - in is a lie [11:19:04] and thats a mess xD [11:19:10] |nostalgia|aa|bat-smg|bm|cho|ho|hz|ii|kj|kr|mh|mo|mus|nds-nl|ng|sq|udm| [11:19:19] ignoring bat-smg [11:19:24] and nds-nl [11:19:54] (for your reference it should now appear at the bottom of http://bots.wmflabs.org/~addshore/addbot/ [11:20:59] sweet :) [11:21:01] thx :) [11:22:17] awww the list of articles that need fixing just grew a lot… [11:22:27] ;p [11:22:38] it only lists ones the bot has confirmed ;p [11:30:18] rschen7754: I updated http://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Wikidata_Phase_2#In_infoboxes and would like someone to check if I maybe changed the meaning of the options? [11:30:30] was actually looking at that right now [11:30:58] thx [11:30:59] yeah, 3 and 4 are fine [11:31:17] addshore: hehe, i guess there are still plenty of iwconflicts lurking ... [11:31:26] Denny_WMDE: yep [11:31:49] rschen7754: thx. I hope risker will agree with the changes too. do you know when we can expect her to show up? [11:32:16] Denny_WMDE: usually evenings my time (Los Angeles) [11:32:40] ok, that's… middle in the night my time :P [11:32:49] (Berlin) [11:32:51] but she only frequents #wikipedia-en-admins (private) [11:33:23] #wikipedia-en is inhabited by too many trolls lol [11:33:49] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [11:34:12] well, discussion can happen on-wiki as well. or on mailinglist. or wherever. [11:47:11] heh, still removing thousands of iw links every min [11:48:18] there have been more than 240 million language links in the main namespace before phase 1 went live [11:48:55] about 5 GB of text [11:49:20] i am looking forward to do the analysis again at some point, to see how wikidata changed that :) [11:50:09] New review: Daniel Kinzler; "(1 comment)" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60214 [11:50:22] haha, how much text got removed from the pedias? :P [11:50:49] well, i guess about these 5 GB of text :) also the edit rate will probably go down for some wikipedias [11:51:12] right all setup, at 10am my time the bot will start a process for every wiki and re run over the articles :) [11:51:29] what's your time zone? [11:51:54] uk so right now bst which is gmt+1 ;p [11:52:45] ok, so every day at noon my time there are new numbers :) [11:53:05] hehe :) maybe I should start it a bit sooner [11:53:10] maybe 1am [11:53:34] 1am it is :) [11:53:54] also just realise the server runs on gmt so its 01:00GMT [11:55:14] :) [11:55:40] as long as it runs daily, i will get new numbers if i check it once daily anyway ;) [11:57:08] yep :P [11:59:12] and Denny_WMDE one other thing that might interest you is https://www.stathat.com/stats/90o4i [12:02:06] that is a bit minimalistic [12:02:15] what does it mean? :) [12:02:36] how many times people are viewing the page and therefore probably fixing conflicts on there ;p [12:03:50] hmm, i wonder if I could add something to track link clicks on there [13:25:41] New review: Daniel Werner; "I don't quite get why all the functions in the RepoApi still use the datamodel objects. Only the Abs..." [mediawiki/extensions/Wikibase] (master) C: -1; - https://gerrit.wikimedia.org/r/58223 [13:28:48] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [13:35:42] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [14:01:33] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [14:06:46] New patchset: Tobias Gritschacher; "Autosummary for setClaim" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58880 [14:12:08] rschen7754|away: Susan: please poke when around [14:18:23] Reedy: aude: if you have time, any idea why the creation of wikidatawiki-20130417-pages-logging.xml.gz takes so extremely long? it's running for four days now… it feels like it is running longer than the creation of wikidatawiki-20130417-pages-meta-history.xml.bz2 (but I don't have numbers for this claim) [14:19:39] Denny_WMDE: no idea [14:19:45] ariel might know [14:20:13] i know the log table is huge, but still, more than 4 days looks weird [14:20:16] to be found in wikitech as apergos [14:20:24] thx [14:30:19] hello Denny_WMDE [14:31:18] hi liangent [14:31:36] New patchset: Daniel Werner; "(bug 47435) Fix problem with dismissing copyright notice" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60277 [14:31:37] about global fallback + user fallback: how do you think about chromium's language selector ui? I guess it fits our need too [14:32:28] regarding the language selector UI, we already have two of those, I think that is sufficient :) [14:32:29] specifically, we only maintain one fallback chain per user. when a user try to add a language to it, all its fallback targets from global fallback chain are added too [14:32:39] i don't know chromium's ui, also, you have a link? [14:32:50] New review: Tobias Gritschacher; "Selenium statements-tests & QUnit tests OK." [mediawiki/extensions/Wikibase] (master); V: 2 - https://gerrit.wikimedia.org/r/59599 [14:32:55] New review: Daniel Werner; "see I8d2517568b394f9da8e2ad753e5c8d9fe9595a45 for a proper fix which will also work in IE8" [mediawiki/extensions/Wikibase] (master) C: -2; - https://gerrit.wikimedia.org/r/60133 [14:33:10] it's a software so there can't be a link other than "chrome://settings/" to work in chromium... [14:33:34] aude: https://gerrit.wikimedia.org/r/#/c/60277/ [14:33:44] Tobi_WMDE: can you try with selenium and IE [14:33:56] * aude can see if it works in chrome and firefox :) [14:34:14] continuing... after a language (together with its global fallbacks) is added to a user's "language set" as they want [14:34:26] * aude has no way of testing with IE [14:35:08] us neither [14:35:15] however an entry with any other entries falling back to it can't be removed from the list (in chromium, when I remove it, it's added back again automatically) [14:35:19] but I promise that does not break ie8 [14:35:19] * aude rages [14:35:23] ok [14:35:31] Denny_WMDE: ^ [14:35:35] I merge it, can you backport it? [14:35:38] i use chrome and hope it's the same as chromium [14:35:43] Tobi_WMDE: ok [14:35:54] Denny_WMDE: yeah I think so [14:36:34] hmm, actually I think ULS is more useful in selecting a language [14:36:53] Change merged: Tobias Gritschacher; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60277 [14:36:53] that's for users to manage their own language fallback chain [14:36:54] and whatever is stated in the preference in chrome can be stated with babel on the user page too [14:37:05] any reason not to use babel for that? [14:37:12] and well mix global & user fallback [14:37:44] global fallback. when a user add zh-cn to their babel, they may not realize they should add zh-hans and zh too [14:39:26] also we can't quietly place zh-hans and zh (and en, based on our system fallback) just after zh-cn, because they may prefer other languages over zh (or en) [14:40:46] hm. we could always use the user fallback first [14:41:02] and only then go to the global fallback. [14:41:54] in that case users could explicate their order as much as they like, and only after that we try to guess [14:43:21] nope. if a user says place babel zh-cn-N then es-2, it's bad to feed they es instead of zh-hans [14:43:36] because zh-hans is almost identical to zh-cn [14:43:56] in system messages we just write messages for zh-hans and let zh-cn fall back to it [14:44:39] chromium's UI is good because it gives users some feedback about what they'll finally get [14:44:58] yes, but it is so much of an external system [14:45:10] it's like nothing we currently have, which means a lot of new code, etc. [14:45:29] yeah [14:45:58] I'd prefer something that is more aligned with what we already have. [14:46:18] I could imagine a special page which helps you with setting up your babel, and explains the fallback chain etc. [14:46:36] and make such suggestions like "you probably want zh-hans here" [14:46:55] but otherwise, use the explicit list, fall back to the global afterwards [14:47:11] or else we need two kind of fallbacks: "almost the same" and "normal fallback" [14:49:59] Denny_WMDE: I can't well imagine the whole workload but it time permits let me try the first :) [14:52:07] Denny_WMDE: do we already have some widget to allow users to set up a language list? [14:53:58] liangent: no, we only have a widget that lets a user choose a language [14:53:59] ULS [14:54:10] currently only enabled for logged in users [14:56:17] Denny_WMDE: so the "use the explicit list" approach is also not doable without much new code currently.. [14:56:56] why not? [14:57:07] ah, sorry, misunderstood [14:57:09] there is babel [14:57:15] it is just not a widget [14:57:49] Denny_WMDE: you mean {{#babel: ? [14:58:18] yep, as in here: http://www.wikidata.org/w/index.php?title=User:Denny&action=edit [14:59:32] looks like it's being overused [15:00:04] it's too implicit for users to configure user settings by editing their user pages [15:03:21] it would save us from needing to create a completely new preference, and the data is already there [15:10:55] and if the special page explains it, it would not be too implicit [15:11:05] it could be still regarded as overused though [15:13:09] New patchset: Daniel Werner; "info about approved copyright message in one language will remain" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/60281 [15:13:26] Denny_WMDE: which special page? [15:14:11] a new one that tells you "here's how your fallback chain looks like." and where you can figure out what you need to do in order to change it. [15:16:23] Denny_WMDE: having it configurable just on that special page looks more user friendly (that's the other approach mentioned above) but of course it needs more coding work [15:17:07] it's not just more work, it's also more code to maintain later [15:17:10] and we already have babel [15:17:17] i do not understand why not use it? [15:19:53] what about users who want a clean user page? [15:20:52] well, they can't use babel obviously [15:20:57] why would they want that? [15:21:56] you mean an empty user page, or one without the babel boxes? [15:21:57] I don't know but we have some users on zhwiki even insisting keeping their user page red [15:22:07] red as in red link [15:22:13] can babel boxes be put on a user sub page? [15:22:36] that's just annoying. it always means some extra hassle to get to the user contributions etc. [15:22:36] they could be transcluded onto the user page that way [15:23:21] Denny_WMDE: it's not more difficult. red user pages are HTTP 200 too and have links to contribs pages too [15:23:53] http://www.mediawiki.org/wiki/GlobalProfile/design si coming someday maybe :o [15:24:17] one more reason not to put effort into solving this problem with more code [15:24:20] i imagine the idea of user pages changing someday.... [15:24:24] on the global profile this makes much more sense [15:25:35] I don't really know it, but what's its progress? [15:26:16] liangent: seems something very much in mind at the foundaiton, but no idea details, schedule, etc. [15:26:44] more at the idea stage, but the idea of a global watchlist, etc. has been talked about a bit over the years [15:27:02] i can imagine things going in this direction, sooner or later [15:27:23] http://www.mediawiki.org/wiki/GlobalProfile ( no timeline) [15:27:50] * liangent is readingit [15:28:12] anyway, about babel boxes, it seems like a simple approach for now [15:29:12] babel still requires per site setup... [15:29:32] I assume GlobalProfile is tied to global accounts right? [15:29:40] liangent: i think it would be [15:30:22] liangent: that's how i understood it. it would be global [15:33:16] ok then what about make it a hook (or any other way to implement interface) to accept external source about users' exact language perference? [15:33:43] and should system messages benefit from this? [15:36:47] if you let the system messages benefit from this, it might have all kind of hassles with the caches [15:36:57] New patchset: Aude; "(bug 47435) Fix problem with dismissing copyright notice" [mediawiki/extensions/Wikibase] (mw1.22-wmf2) - https://gerrit.wikimedia.org/r/60284 [15:37:34] i think it is currently implemented like this, or at least it should - i.e., if babel is not available, it uses it in different ways [15:37:36] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59130 [15:37:46] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59408 [15:37:57] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59409 [15:38:11] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59410 [15:40:39] Denny_WMDE1: maybe - I've never dived deeply into MessageCache [15:41:23] liangent: i think there are more layers of caching involved for wmf [15:41:30] like squid, and someday varnish [15:41:34] plus parser cache, etc. [15:41:47] squid is the main one we've had issues with, re: universal language selector for anons [15:41:59] that's why it's enabled for logged-in users only [15:42:17] aude: it's already there for long days on zhwiki, about variant support [15:42:26] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59422 [15:42:35] Change merged: Daniel Werner; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59599 [15:42:52]