[03:37:07] so qualifiers today? [03:39:01] still no sourcing? [07:10:11] Lydia_WMDE, aude: Hi. What's the status of the link to add interlanguage links to an article that doesn't have any? [07:12:45] aharoni: you just need to create the item [07:13:47] aharoni: that's a question aude needs to answer [07:13:54] i would like to know the answer too [07:14:15] Jasper_Deng: sourcing is possible [07:14:26] people are just figuring out how to use it exactly [07:14:41] rschen7754: yes that's the plan [07:14:48] :D [07:15:45] legoktm: I don't think so. See https://en.wikipedia.org/wiki/Reading_revolution and https://www.wikidata.org/wiki/Q7300603 . [07:16:31] aharoni: oh you mean having the "edit links" show up if there are no other links? [07:16:49] Yes. [07:17:10] I remember seeing some code that does it a few weeks ago. By Hooman if I'm not mistaken. [07:17:25] So I'm wondering what became of it. [07:17:34] It's one of the most requested features. [07:57:40] Lydia_WMDE: not sure if you saw in the backlog but i got around to publishing some graphs of wikidata bot activity, that are updated every hour: http://inkdroid.org/wikidata-bots/ [07:58:10] Lydia_WMDE: I'm still interested in calling together a wikidata bot meeting, if you still think there is a need [08:10:45] aharoni: think the widget can come out of "experimental" next deployment [08:10:58] edit links also, if item is connected but has no interwiki links [08:11:22] aude: thanks. what do you mean by experimental? [08:11:30] is it deployed in any way now? [08:11:41] aharoni: it's on test2 (i think) [08:11:49] and when is the next deployment? [08:11:56] it wasn't ready for production and we have "experimental" for such stuff [08:11:59] should be next week [08:12:05] which day? [08:12:19] i think monday is enwiki and others are wednesday, normally [08:12:33] should be normal schedule [08:12:41] OK. [08:12:53] https://www.mediawiki.org/wiki/MediaWiki_1.22/Roadmap [08:13:03] we follow same schedule as core [08:13:15] I haven't maintained my own Wikibase installation in a while... If I just install it on my own local test wiki, am I supposed to see it already? [08:13:52] there's a setting for it [08:14:04] 'enableSiteLinkWidget' = true (i think) [08:14:16] cool, thanks. [08:14:28] let me find a page on test2 that has it [08:14:35] test2 is not the latest code, though [08:18:31] https://test2.wikipedia.org/wiki/Achille_Campo (with add links) [08:18:34] you need to log in [08:19:00] https://test2.wikipedia.org/wiki/Arthur_Oliver_Wheeler (connected only to english), has edit links [08:19:14] aharoni: ^ [08:20:03] aude: and if there's no item at all? [08:20:21] oh, the first one. [08:20:25] aharoni: then it makes a new item [08:20:34] or if you pick itwiki Achille Campo [08:20:55] then it will add enwiki to the existing item (which has just itwiki, in this case) [08:21:13] if what you choose in the widget is used in the item, then the connection is made to the item [08:21:25] And why do I have to be logged in? [08:21:58] we're doing cross wiki edits so requires to be logged in on repo [08:22:11] if you login on wikipedia, you get logged in on wikidata [08:22:40] I'm not sure that I understand it. [08:22:51] we're editing the repo via the api [08:23:07] OK, and why isn't it possible to do it anonymously? [08:24:14] it might be but gets a little mor ecomplicated [08:25:43] Should be possible. The obligation to log in must not creep in. [08:26:26] we can think about it.... [08:52:24] New patchset: Henning Snater; "Introduced setError in claimview and referenceview widget" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59599 [09:13:14] edsu: seen it - thanks! that is very useful [09:13:39] edsu: from the dev team side i think we don't need a bot meeting anymore but that shouldn't stop you from having one anyway of course [09:14:18] Lydia_WMDE: ok :) [09:42:57] New review: Siebrand; "L10n/i18n LGTM" [mediawiki/extensions/Wikibase] (master) C: 1; - https://gerrit.wikimedia.org/r/58880 [09:43:43] New review: Siebrand; "If you do that, Daniel, I can merge the change. Deal? :)" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59030 [09:44:17] New review: Siebrand; "Per Daniel." [mediawiki/extensions/Wikibase] (master) C: -1; - https://gerrit.wikimedia.org/r/59030 [09:49:06] New patchset: Siebrand; "Add overview of dispatch states to DispatchStats." [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58464 [09:49:35] New review: Siebrand; "PS3 adds full message keys as comment for generated message keys." [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/58464 [09:51:30] New review: Siebrand; "i18n/L10n reviewed (one update made in PS3)." [mediawiki/extensions/Wikibase] (master) C: 1; - https://gerrit.wikimedia.org/r/58464 [10:10:30] * Moe_Epsilon waves and departs [12:57:40] btw addshore i set this up for my bot https://www.stathat.com/stats/K5Vs [12:58:06] i can easily add any other bots to it :) [12:58:20] er, create a separate graph for them* [13:51:45] who's not in a meeting? [13:51:54] * legoktm raises his hand [13:51:55] * jeremyb_ pokes Denny_WMDE [13:54:07] i've commented on 45005 [13:54:08] !b 45005 [13:54:09] https://bugzilla.wikimedia.org/45005 [13:54:44] Silke_WMDE: Lydia_WMDE: aude: Denny_WMDE: ^ [13:54:56] Silke_WMDE: can i has a webcam on the meeting room so i know when they're done? ;-) [13:55:22] hehe [13:55:26] sure [13:55:40] that's part of your new job, riiiiight? [13:57:20] jeremyb_: ideally de.wikidata.org/wiki/Berlin should redirect to www.wikidata.org/wiki/Special:ItemByTitle/dewiki/Berlin [13:57:27] but that's not part of that change [13:57:40] Denny_WMDE: why not? should it be 301 or 302? [13:57:43] right now I would just be happy if wikidata.org would finally reliably go to www.wikidata.org [13:57:51] i am not sure about the difference [13:58:01] can you explain quickly, or do you have a suggestion already? [13:58:35] i will check my own notes [13:59:27] Denny_WMDE: i'm just asking what you want in the end [13:59:47] 302 looks good [13:59:52] if i could do them all right now (or an hour from now) should i do lang codes or not? [14:00:03] (not that i can...) [14:00:19] lang codes would be awesome [14:00:59] as said: {langcode}.wikidata.org/wiki/{Title} to www.wikidata.org/wiki/Special:ItemByTitle/{langcode}wiki/{Title} [14:01:24] ohhh, forgot about that part :P [14:01:31] ok, lang codes will wait :) [14:01:43] ok :) [14:02:44] i put it in the bug [14:03:10] great [14:31:39] jeremyb_: I wrote down how we did the redirects for our dev system, just in case https://wikitech.wikimedia.org/wiki/Nova_Resource:Wikidata-dev [14:31:42] Abraham_WMDE: "2013-04-17" [14:31:59] thx [14:32:28] Silke_WMDE: danke [14:34:03] Abraham_WMDE: you're missing a club mate? [14:34:11] Tobi_WMDE: yes [16:33:42] New patchset: Daniel Kinzler; "Expand legend on Special:DispatchStats" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59030 [16:34:31] New review: Daniel Kinzler; "there you go, then :)" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59030 [16:35:13] Reedy: deployment starts in 1.5 hours, right? [16:35:19] after db migration? [16:35:20] Nope [16:35:25] huh? [16:35:29] when? [16:35:55] Wednesday 17th 18:00-20:00 11 a.m. - 1 p.m. Complete migration of enwiki, dewiki, wikidata to mariadb with master swaps [Asher, Peter] [16:35:55] Wednesday 17th 20:00-22:00 1 p.m. - 3 p.m. MediaWiki general deployment window (1.22) [Sam / Aaron / RobLa] [16:36:01] k [16:36:02] 2.5 hours [16:36:12] good :) [16:36:17] enough time to get food [16:53:12] aude: Reedy: I'll go to the cinema and then post announcements when i am back - please keep me posted here [16:53:24] k [16:53:37] * aude must skip the cinema :( [16:54:16] New review: Daniel Werner; "(1 comment)" [mediawiki/extensions/Wikibase] (master) C: -1; - https://gerrit.wikimedia.org/r/51164 [17:28:47] New patchset: Daniel Kinzler; "Use getUserPermissionsErrors in EditEntityAction::showPermissionError" [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/56606 [17:29:07] Change merged: Daniel Kinzler; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/56606 [17:59:08] * hoo slaps himself for linking the wrong bug [18:40:13] me here [18:40:43] hi me! :) [18:41:24] * hoo can't think of a structure for the RepoApi which doesn't plain suck :/ [18:42:03] hmmm [18:42:29] hey is there already a new dispatcher? [18:43:20] Sk1d: no, although the dispatcher is on new hardware, new data center now [18:43:22] If I do it like Daniel suggested we end up with either a lot of useless wrapping functions or have a mix of wb.RepoApi and wb.AbstractedRepoApi all over the place -.- [18:44:06] aude: do you think we can start increasing the bot edits slowly? [18:45:10] Sk1d: probably not yet but a question for denny [18:45:43] * aude was hoping to have dispatch stats available via the api but not quite yet for that either [18:45:54] Danny_WMDE: ping [18:46:08] Denny_WMDE: ping [18:46:17] if something like that could be used by bots to adjust editing rate, that could be nice [18:46:29] yes like maxlag [18:46:39] and we're still considering some throttle at a technical / config level for the wiki [18:46:46] There are hooks in the API... we even could hook up onto maxlag [18:46:49] don't know what the limit would be [18:46:59] Sk1d: one trick for you: type de [18:47:12] hoo: probably, although the api could provide more stats about the dispatching [18:47:19] more info.... [18:47:28] hi Danny_B :) [18:47:34] hello kate [18:47:40] +i [18:47:50] Yeah... I'm not sure how much resources it costs to get those stats from the DB everytime (haven't looked at the code) [18:48:02] Danny_B: thx [18:48:07] hoo: not sure [18:48:31] perhaps they could be cached like every minute or something [18:48:34] that woudl be [18:48:42] would be enough granularity [18:48:53] Memcached to the rescue :) [18:49:09] * aude screenscraping them to make charts [18:49:24] grab them every few minutes and that's more than good enough [18:49:30] aude: did you see my link last time: http://tools.wmflabs.org/legobot/dispatchstats.log [18:49:35] yeah [18:50:11] * legoktm is here too :P [18:50:18] I've been thinking about it [18:50:36] * hoo postpones the refactor of wikibase store and does some code review instead [18:50:52] I was thinking of just having a web page that bots can check, and simply says how long they should sleep for in between edits [18:51:09] And it would simply base that number off of dispatch lag, some fraction of it [18:51:12] Sounds nasty, tbh [18:51:30] hoo: got a better idea? :P [18:51:54] and if dispatch lag rose above some #, the bot just goes into some kind of holding pattern [18:52:04] I doubt adding smth. like maxlag for the dispatch lag is impossible [18:52:12] or even hard to implement [18:52:36] hoo: it's not [18:52:49] my interwiki import bot is now doing imports at abou 100 edits/ min as the dispatch lag is still on zero and i dont think this schould increas the dispatch lag [18:52:51] it will just take 2 weeks minimum to be deployed [18:53:12] Well... first of all someone has to implement it :P [18:53:15] daniel was workign on something and need to see what's needed [18:53:17] to get it done [18:53:31] Sk1d: o.O [18:54:39] legoktm: I think we should increase the limit for bots to 70 or 100 edits/min and see if this is ok for the system [18:55:03] Yeah [18:55:22] My bots auto-kill themselves if the lag goes over 1000 [18:56:08] aude: does creating new items get the dispatchlag up? I am not sure if you can see this in the watchlist. [18:56:17] Sk1d: It does. [18:56:31] My bot went on a creation spree, and thats what started all of this in the first place :/ [18:56:59] Sk1d: it might [18:57:11] I've established that patrolling edits and deleting pages do not increase lag. [18:57:19] the bigger the diff (e.g. hundreds of site links), it might be slightly slower to dispatch [18:57:38] shouldn't be a huge difference and don't have numbers to back that up [19:15:49] Brief Database Maintenance in progress, please try again in 3 minutes ... [19:17:02] aww my bot crashed :( [19:17:03] benestar: switching over wikidata to use mariadb instead of mysql (differences between the two should not be noticable, however) [19:17:04] pywikibot.data.api.APIError: readonly: The wiki is currently in read-only mode [19:17:30] hehe [19:17:33] http://lists.wikimedia.org/pipermail/wikitech-l/2012-December/064994.html [19:17:36] aude: Lydia_WMDE wrote on pc [19:17:43] [02:17:20 PM] <+logmsgbot> !log asher synchronized wmf-config/db-eqiad.php 's1/5 writeable' [19:17:51] benestar: ah, ok [19:19:24] :P [19:19:41] this was really fast... xD [19:20:08] :) [19:23:17] Should this filter: https://www.wikidata.org/wiki/Special:AbuseFilter/18 use the same tag as this filter: https://www.wikidata.org/wiki/Special:AbuseFilter/11 ? [19:23:56] benestar: sure [19:24:05] ok :) [19:28:49] New review: Aude; "(1 comment)" [mediawiki/extensions/Wikibase] (master) C: -1; - https://gerrit.wikimedia.org/r/59388 [19:37:33] Yo guys [19:47:00] are qualifiers being deployed tonight? [19:47:54] DroidClient when the db migration is done [19:48:20] and it's done, so soon.... [19:48:22] thx. I thought it already was [19:48:43] 19:43 < binasher> !log enwiki, dewiki, wikidata are fully on mariadb 5.5.30 [19:48:48] done 5 min ago [19:50:10] aude: there should be no more readonly time tonight right? [19:51:08] legoktm: right [19:51:53] aude: what means "soon.." ? [19:52:18] benestar: next few weeks ;) [19:52:28] ah, thanks [19:52:45] heh [19:53:04] i was joking >.< [19:53:09] should be within next hour, hopefully sooner :) [19:53:10] :P [19:53:34] when Reedy is available to do deployments [19:53:55] New patchset: Daniel Kinzler; "Factor ClientStoreFactory into WikibaseClinet." [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59711 [19:54:04] https://gerrit.wikimedia.org/r/#/c/59667/ is ready [19:57:40] let me tidy up my desk a little bit, then we should be good to kick the other wikis over [19:57:45] k [19:57:53] no hurry [20:09:06] legoktm: do you want to try https://www.wikidata.org/wiki/User:Bene*/recentchanges ? [20:09:25] I've created it today and it helps to fight vandalism [20:09:53] does it auto refresh? [20:09:59] have you seen what krinkle wrote? [20:09:59] yes [20:10:03] no [20:10:04] it has a ton more features [20:10:10] :O [20:10:19] why doesn't it appear at WD:TOOLS? [20:10:23] *WD:Tools [20:10:32] https://meta.wikimedia.org/wiki/User:Krinkle/Tools/Real-Time_Recent_Changes [20:11:00] New patchset: Daniel Kinzler; "Factor ClientStoreFactory into WikibaseClinet." [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59711 [20:18:58] New review: Aude; "only slight performance improvement, but still is improvement" [mediawiki/extensions/Wikibase] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/59188 [20:19:00] Change merged: Aude; [mediawiki/extensions/Wikibase] (master) - https://gerrit.wikimedia.org/r/59188 [20:48:55] me me again [20:50:32] Denny_WMDE: DanielK_WMDE so we have an issue with the change notification https://bugzilla.wikimedia.org/47334 [20:50:55] assume it's new repo + old client and trying to debug/reproduce it [20:51:02] sounds familiar [20:51:35] probably unrelated but they were also seeing some core warning for DatabaseMysql [20:52:48] looks like sth the tests should have caught if reproducible [20:53:10] oh, maybe we never tested it with a 22wmf1 client, only with 21wmf12, could be. [20:54:37] i can't imagine what changed there [20:54:51] except there are qualifiers [20:55:12] ok, got wmf2 repo and wmf1 client [20:56:26] Ummm, how much until qualifiers are live? [20:59:15] Hahc21: you missed it, its being delayed for another 2 weeks [20:59:35] what [20:59:36] omg [20:59:38] T.T [20:59:42] :P [20:59:42] jk [20:59:51] they're working on it [21:00:05] oh my [21:00:09] you scared me u.u [21:00:23] http://p.defau.lt/?jUnh3ng0cDb_p3APQPEmAg [21:01:49] :( [21:02:11] so it's wmf2 repo? [21:02:19] what is that? [21:02:37] Hahc21: the previous version is fine [21:02:40] broken stuff is broken [21:02:47] new version apparently has a bug :( [21:03:00] i am tryign to reproduce [21:03:03] the bug(s) [21:03:51] that would explain https://bugzilla.wikimedia.org/47334 though [21:04:13] * aude thinks so [21:05:04] aude: Reedy: hey :) back at home now - how are things going? [21:05:39] Lydia_WMDE: bugs [21:05:41] bug [21:05:50] ok [21:16:08] PHP Notice: Uninitialized string offset: 0 in /var/www/common/php-1.21wmf6/extensions/Wikibase/lib/includes/changes/ChangeRow.php on line 231 [21:16:28] wmf6? ;) [21:16:47] http://pastebin.com/dSsg0H6z [21:16:58] Reedy: heh, it's 1.22wmf2 [21:17:22] * aude just updates my branches now [21:17:37] Should I jdeploy the Wikidata, DataValues and diff updates in the mean time? [21:17:59] i am not convinced [21:18:08] not yet [21:18:19] * aude wants to understand why it's broken [21:19:01] after the above notice, i get [21:19:02] PHP Warning: array_key_exists() expects parameter 2 to be array, boolean given in /var/www/common/php-1.21wmf6/extensions/Wikibase/lib/includes/changes/DiffChange.php on line 107 [21:19:41] aude: ChangeRow.php on line 231?! [21:19:42] http://pastebin.com/z5G2gGb7 [21:19:48] That line is in a comment?! [21:20:20] uh, it's wmf2 core with the same version of extensions as before [21:20:25] * aude goes to look at the lines [21:20:41] that's what would've been deployed before reedy updates everything [21:20:43] aude: new version of what? [21:20:53] * aude wants to double check the update also [21:20:53] aude: oh, so it's not mw1.22-wmf2 of wikidata? [21:21:00] right [21:21:05] now i'm confused [21:21:06] * aude checks like 231 [21:21:09] line [21:21:12] what should i patch against for the clients? [21:21:27] what's currently deployed there?` [21:21:56] line 231 is " if ( $str[0] === '{' ) { // json " [21:22:09] in unserializeInfo( $str ) { [21:23:03] DanielK_WMDE: client is easy.... mw1.21-wmf12b [21:23:16] what we've had there and clients have not changed [21:25:30] * Hahc21 can't wait to see the qualifiers working xD [21:28:03] you can see them on http://wikidata-test-repo.wikimedia.de/wiki/Testwiki:Main_Page :) [21:28:10] I already tested them [21:28:11] ^^ [21:28:19] you like them? [21:28:25] A lot :P [21:32:37] DanielK_WMDE: ChangeRow::unserializeInfo( $str ) is returning false for me [21:33:07] * aude not quite sure how it fits with the production error but seems related or at least not good [21:34:31] doesn't matter if changesAsJson is true or false [21:36:26] aude: changesAsJson changes the serialization, not unserialization [21:36:35] can yopu eyeball the serialized data? [21:36:41] my guess is that it'S getting truncated [21:36:45] but, err... [21:36:53] it's returning false for you when exactly? [21:36:58] always [21:36:59] ? [21:37:06] with what combination of client/repo versions? [21:38:12] string = '' [21:38:15] apparently [21:38:25] uhu [21:38:40] that would imply the field isn't even initialized [21:38:41] DanielK_WMDE: mw1.21wmf12b on both [21:38:49] err. [21:38:56] but that worked for several weeks now, right? [21:38:58] * aude verifying my string [21:39:01] and the error just popped up? [21:39:07] yeah, but with new version of core [21:39:22] uh... right... how would that mess with the serialization? [21:39:24] * aude wonders what's different [21:39:24] well, whatever [21:39:26] which version? [21:39:37] 1.22wmf1? [21:39:37] 1.22wmf2 [21:39:50] but that's not live?.... [21:39:56] it was [21:40:01] where is the error happening on the live systems, then? [21:40:02] it's rolledback [21:40:14] oh, it was and got rolled back? because of this issue? [21:40:16] not exactly sure [21:40:19] hrm [21:40:23] * DanielK_WMDE is a bit confused [21:40:26] http://p.defau.lt/?jUnh3ng0cDb_p3APQPEmAg is the live error [21:40:28] It was creating a huge spam of warnings in the apache logs [21:40:33] and rolling that back fixes the issue? interesting [21:40:44] * aude can't reproduce that cause i get the other one which seems related perhaps [21:40:56] Reedy: it was likely also creatring invalid/broken (or at best missing) entriers in recentchanges [21:40:57] so much so that within a minute or 2, 90%+ of the last 1000 lines are that error [21:41:06] * aude tries switching my core [21:41:39] Reedy: i can add defensive code to the client, to make the error look nicer, but that doesn't help with the root cause... [21:42:01] right [21:42:13] if the client is getting bad change data, then well no good [21:42:23] indeed [21:42:36] and ignore it there, then we probably get different errors or no changes in the clients [21:42:39] we can think about how to handle this more nicely, but if the data isn't there.... what can we do? [21:42:50] ignore the change.... nasty [21:43:21] well, we already do. we ignore the change and all the following changes in the batch [21:43:24] the process just dies [21:43:30] so, better ignore the one change [21:44:11] actually, the job runner may re-try this. but won't succeed the next time either. at some point, it'll give up. [21:44:41] aude, Reedy: coming up with defensive code on the client, will try to reproduce the issue after that. [21:45:14] k [21:46:47] hm, i can't really see how getMetadata can return null :/ [21:47:24] with 1.22wmf1 core, my repo is happy [21:49:09] switch to 1.22wmf2, not happy [21:49:16] same version of the extensions [21:49:44] arrggg! [21:49:48] just a second. [21:49:54] this might actually be my fault [21:50:04] * aude tries the new branch of our extension [21:50:12] huh? [21:51:58] aude: there was a core change to ORMTable between wmf1 and wmf2. Wikibase needs to be patched to be compatible. I suppose mw1.21wmf12b doesn't have the patch. [21:52:13] aude, Reedy: breaking core change: https://gerrit.wikimedia.org/r/#/c/49707/ [21:52:25] aude, Reedy: required fix: https://gerrit.wikimedia.org/r/#/c/56384/ [21:52:41] aude, Reedy: i have not verified this, but it seems consistent with the problem [21:52:43] ugh [21:52:46] and it's a known incompatibility [21:52:46] but ok [21:53:14] i should have warned you about it and taken care the core change didn't get roleld out before our fix is live :/ [21:54:27] ok, trying [21:57:18] DanielK_WMDE: well, jttps://gerrit.wikimedia.org/r/#/c/56384/ [21:57:31] been merged a while ago and should be in new branch [21:57:40] so i will just try that and double check things [21:58:01] for sanity... :) [21:58:20] aude: to verify on the repo, use: php phpunit.php --group Wikibase --filter Change [21:58:27] fails horribly without the patch, passes with the patch [21:58:53] but it has been merged more than 2 weeks ago [21:59:23] Denny_WMDE: yes, but client wikis are running an version older than that [21:59:30] we skipped a cycle, remember? [21:59:44] but that's also the reason i forgot about it :/ [21:59:55] DanielK_WMDE: i do see an issue and wonder if it's just me [21:59:57] aude: backporting... [22:00:05] aude: hm?= [22:00:10] ok [22:00:28] DanielK_WMDE: new core and new repo branch should be fine [22:00:36] yes [22:00:53] new core and old repo code will break [22:01:02] new core and old client code will also break, i think [22:01:05] obviously [22:01:22] old core and new wikibase code should work [22:01:25] not convinced as it already would have broken [22:01:28] i tested for that [22:01:50] aude: hm? what would have broken? [22:01:52] anyway..... [22:01:56] different issue [22:02:01] Special:RecentChanges [22:02:11] PHP Fatal error: Class 'ValueParsers\\Result' not found in /var/www/common/php-1.21wmf6/extensions/Wikibase/lib/includes/parsers/EntityIdParser.php on line 90 [22:02:21] i can load the wiki fine, but not recent changes [22:02:25] New patchset: Daniel Kinzler; "Make ChangeRow work with new ORMTable." [mediawiki/extensions/Wikibase] (mw1.21-wmf12b) - https://gerrit.wikimedia.org/r/59740 [22:02:46] DanielK_WMDE: we can do the backport to be safe [22:02:55] no issue with that, although i want to test it [22:02:56] aude: make sou [22:03:09] aude: make sure your branch of DataValues is in sync with your branch of Wikibase [22:03:15] ValueParser\Result got moved [22:03:18] DanielK_WMDE: it is [22:03:24] mw1.22-wmf2 [22:03:26] of all [22:03:37] aude: wbsetaliases with set= (empty) is a valid thing to do, no? [22:03:40] * aude checked a few times but can check again [22:04:01] hoo: apprently, but not sure [22:04:15] set aliases is a bit strange, how it works [22:04:34]