[00:29:38] New patchset: Bsitu; "New primary and secondary link behavior for Echo" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/68817 [00:29:38] New patchset: Bsitu; "Add HTML email support to Echo notification" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/70112 [00:31:11] New review: Bsitu; "nothing changed in patch set #10, it submits this when I submit another patch that depends on this" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/68817 [00:55:12] New patchset: Wikinaut; "Combine text and ID update fields" [mediawiki/extensions/UserMerge] (master) - https://gerrit.wikimedia.org/r/69017 [00:55:52] New patchset: Wikinaut; "Combine text and ID update fields" [mediawiki/extensions/UserMerge] (master) - https://gerrit.wikimedia.org/r/69017 [00:56:59] New review: Wikinaut; "Tyler: I added code to deduplicate watchlist entries of old- and newuser before they will get merged..." [mediawiki/extensions/UserMerge] (master) C: 1; - https://gerrit.wikimedia.org/r/69017 [00:58:36] New review: Parent5446; "Left comments on PS10." [mediawiki/extensions/UserMerge] (master) C: -1; - https://gerrit.wikimedia.org/r/69017 [01:06:34] New review: Wikinaut; "ok, got your comments. " [mediawiki/extensions/UserMerge] (master) - https://gerrit.wikimedia.org/r/69017 [01:16:15] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70037 [01:16:36] New patchset: Bsitu; "Fix the error in html email when pear mail_mime is missing" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70114 [01:24:57] New review: AlephNull; "Bump." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/65301 [01:32:19] New patchset: Rfaulk; "mod. simplify _create_lock, more logging." [sartoris] (master) - https://gerrit.wikimedia.org/r/70116 [01:32:41] Change merged: Rfaulk; [sartoris] (master) - https://gerrit.wikimedia.org/r/70116 [01:40:14] New review: Physikerwelt; "I'm not sure if the way you implemented this is efficient. I think Hashar, who raised that comment w..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/65301 [01:58:04] New patchset: Physikerwelt; "made isvalidxml non static" [mediawiki/extensions/Math] (LaTeXML) - https://gerrit.wikimedia.org/r/70118 [03:32:51] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7 build #69: FAILURE in 8 min 1 sec: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7/69/ [03:35:11] Yippie, build fixed! [03:35:12] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8 build #107: FIXED in 7 min 49 sec: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8/107/ [03:37:06] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/70105 [03:42:44] New review: Ori.livneh; "My comment above still holds true." [mediawiki/vagrant] (master) C: -1; - https://gerrit.wikimedia.org/r/63783 [04:20:08] Change abandoned: Santhosh; "alternate patch at https://gerrit.wikimedia.org/r/69467" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/69456 [04:28:10] New review: Nischayn22; "Can you try running a SMW_refresh?" [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/69668 [04:46:53] New review: FastLizard4; "Code review looks good, some minor whitespacing issues I may or may not fix later." [mediawiki/extensions/CommunityTwitter] (master) C: 2; - https://gerrit.wikimedia.org/r/70029 [04:49:30] New review: FastLizard4; "Seems to work; merging" [mediawiki/extensions/CommunityTwitter] (master); V: 2 - https://gerrit.wikimedia.org/r/70029 [04:49:31] Change merged: FastLizard4; [mediawiki/extensions/CommunityTwitter] (master) - https://gerrit.wikimedia.org/r/70029 [05:13:43] New patchset: Pastakhov; "Add while loops (version 0.4.5)" [mediawiki/extensions/Foxway] (master) - https://gerrit.wikimedia.org/r/69836 [05:50:47] New patchset: Santhosh; "Update jquery.ime from upstream" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/70122 [06:08:23] New review: KartikMistry; "LGTM." [mediawiki/extensions/UniversalLanguageSelector] (master) C: 1; - https://gerrit.wikimedia.org/r/70122 [06:43:34] Change merged: jenkins-bot; [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/70122 [07:17:36] New review: Wikinaut; "(1 comment)" [mediawiki/extensions/UserMerge] (master) - https://gerrit.wikimedia.org/r/69017 [07:41:16] good morning [07:54:11] New patchset: Matthias Mullie; "(bug 37477) Feedback details text on "View feedback page" cut off for very long string without whitespace" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/61568 [07:59:29] New patchset: Nikerabbit; "Implement sandboxed user logic for main page" [mediawiki/extensions/TwnMainPage] (master) - https://gerrit.wikimedia.org/r/69064 [08:04:38] ori-l: I have seen your patch regarding some fluoride demuxer ( https://gerrit.wikimedia.org/r/#/c/69818/ ). Seems to be handling udp2log fatal and exceptions isn't it ? [08:04:43] to have them logged in a db? [08:05:26] yeah, it doesn't do anything with them yet other than print them to standard out [08:05:49] bawolf asked about getting involved in this project on wikitech-l so i wanted to have something for someone to hack on [08:06:34] so you have the locomotive on some rails and bawolf can extend from there [08:06:35] nice [08:07:01] that might one day replace our fatalmonitor script that basically tails the apache log [08:26:54] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69953 [08:31:35] New patchset: Hashar; "jshint now votes on mw/ext/EventLogging" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70123 [08:31:59] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70123 [08:33:07] ori-l: jshint is now voting on EventLogging extension, no qunit tests yet though ( https://bugzilla.wikimedia.org/show_bug.cgi?id=49513 ) [08:39:00] wooo, thanks, hashar! [08:39:22] do you know what it'd take to get qunit too? [08:39:32] I am not sure how it works [08:39:37] well have a rough idea [08:39:44] and Timo has enabled it for VisualEditor [08:40:11] oh and GuidedTour [08:40:30] yes, this seems to be the guidedtour change: https://gerrit.wikimedia.org/r/#/c/69786/ [08:40:38] i'll try to follow it exactly and submit a patch [08:40:48] or do you think i should just wait for timo to do it? [08:40:59] I can try :)) [08:41:03] maybe it will just work eheh [08:41:36] cool, let's try it [08:41:53] generating the job [08:46:34] and enabling tests on submission [08:47:30] New patchset: Hashar; "test pipeline for mw/ext/EventLogging" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70124 [08:47:30] New patchset: Hashar; "qunit on mw/ext/EventLogging" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70125 [08:48:00] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70124 [08:48:26] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70125 [08:49:40] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70126 [08:50:05] mwext-EventLogging-testextensions-master: SUCCESS [08:50:05] yeah [08:50:11] mwext-EventLogging-qunit: FAILURE [08:50:13] oh no :( [08:53:13] hashar: well, maybe those are actual bugs [08:53:15] I should investigate [08:53:19] possibly [08:53:20] New patchset: Hashar; "mwext-EventLogging-qunit non voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70127 [08:53:30] note that EventLogging is loaded without any documentation [08:53:46] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70127 [08:53:47] what do you mean? [08:56:10] New patchset: Hashar; "qunit job for mw/ext/EventLogging" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70128 [08:58:55] ori-l: yeah hmm sorry [08:59:02] I meant, EventLogging extension is just included [08:59:07] and no specific configuration is provided [08:59:14] that might be the cause of some tests errors [08:59:35] I have copy pasted the console log to the bug https://bugzilla.wikimedia.org/show_bug.cgi?id=49513 [08:59:44] Matthew can have a look I guess [09:00:48] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70128 [09:04:53] ori-l: thoughts on moving the JsonSchema classes (the validator) into core? [09:07:52] YuviPanda: dunno; what's the rationale? [09:08:08] ori-l: with UW, that's 3 extensions using it... [09:09:38] """ What’s interesting is the proportion of people without any college education at Google has increased over time as well. So we have teams where you have 14 percent of the team made up of people who’ve never gone to college.""" [09:09:39] :-D [09:09:48] came on one internal list : http://www.nytimes.com/2013/06/20/business/in-head-hunting-big-data-may-not-be-such-a-big-deal.html [09:09:58] about Big Data and head hunting [09:10:19] YuviPanda: what's the third? [09:10:27] Zero doesn't use it? [09:11:05] YuviPanda: no (to its detriment, IMO) -- yurik decided to have his own set of light-weight conventions [09:11:13] grr. [09:11:21] if you could write a quick proposal somewhere about including JsonSchema in core, we can bring up the subject in our mediawiki core team meeting [09:11:29] it is today at 9pm utc [09:11:44] or we can bring it next week ;-) [09:12:08] I am not sure whether it is worth it though [09:12:18] YuviPanda: up to you, really. I don't have a strong feeling either way, to be honest. The scope of what belongs in core is not something I've thought about a lot. [09:12:27] hmm, okay [09:13:04] I am just not sure if we should include it by itself now, or wait to include something like a 'ValidatedJsonContentHandler' [09:13:10] you could ask robla about it too, it's his code [09:13:10] which would be more useful [09:13:18] hmm, true. [09:13:34] thinking about it I think actually moving it as a ContentHandler would be more useufl [09:13:36] *useful [09:14:29] well, I *definitely* agree that it doesn't make sense for it to be so tightly coupled to EventLogging [09:14:59] I wouldn't have done that if I were starting over, and if I had lots of time to spare I'd decouple the namespace / content from the rest of the extension [09:15:13] I don't know that I'd put it in core, but a separate extension certainly [09:15:27] hmm, I think the Content stuff should go into core [09:15:37] moving just JsonSchema into core would not make too much sense [09:15:42] so... I guess it's fine for now [09:15:56] hashar: so nothing this week :) [09:15:59] ori-l: EL is deployed on all wikis, right? [09:16:09] yep [09:16:18] if there is any use to have it decoupled, go for it [09:16:27] might even just be a standalone extension [09:16:46] hashar: yeah, i think doing that would be uncontroversially good [09:16:52] just a matter of time [09:17:39] robla has a 'mediawiki-jsondata' extension but it's pre-contenthandler and so not very salvageable [09:17:53] but it's here if you're interested, YuviPanda: http://jsonwidget.org/wiki/JsonData [09:18:14] hmm [09:18:22] i think it'll be easier to just extract it out of EL / UW [09:18:37] probably, yes [09:19:23] andre__: sorry for the cranky email! [09:19:45] hope returning to your inbox after travelling was not too awful [09:20:49] hashar: it must be some setup issue; those tests work on my dev instance and on test2.wikipedia.org, and it's too strange for *all* tests in that module to fail [09:21:49] must be :/ [09:21:53] or sqlite related [09:22:24] ori-l: from a very quick read this morning I think it was a totally okay, didn't see anything personal or so. I still need to re-read that thread though when I'm more awake and reply :) [09:23:09] New patchset: Hashar; "mwext-Echo-testextension.* are now voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70130 [09:23:16] andre__: :) [09:23:31] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70130 [09:23:36] my jetlag somehow turned into the wrong direction - instead of staying -9h I ended up with +9h and woke up way earlier than way later today :) [09:24:23] ori-l: I have a pretty thick skin and assume that people mean well, so you need to try way harder next time ;)) [09:25:09] challenge accepted! [09:25:10] j/k [09:26:04] hehe [09:39:29] Change abandoned: Hashar; "(no reason)" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70126 [10:12:13] New patchset: Hashar; "python jobs for pywikibot/*" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70132 [10:12:34] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70132 [10:30:46] New patchset: Hashar; "python-lint template" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70135 [10:30:46] New patchset: Hashar; "python jobs are now voting +2 and use a template" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70136 [10:30:46] New patchset: Hashar; "pyflakes on labs/nagios-builder" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70137 [10:31:05] New patchset: Hashar; "labs-nagios-builder-pyflakes" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70138 [10:31:38] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70138 [10:32:05] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70135 [10:32:36] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70136 [10:32:57] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70137 [11:14:48] New patchset: Nikerabbit; "Introducing TranslationStash" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/66540 [11:14:48] New patchset: Nikerabbit; "Message suggester for sandboxed users" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/70142 [11:26:45] New patchset: Rtdwivedi; "Separated data fetching from manipulation for page." [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69802 [11:28:36] New review: Aklapper; "+2 - thanks for the patch, works perfectly fine (tested locally) by hiding the bullet, and we don't ..." [wikimedia/bugzilla/modifications] (master) C: 2; - https://gerrit.wikimedia.org/r/69983 [11:29:45] New review: Se4598; "breaks for the other cases (e.g. space, digit, special chars), see my previous comment." [mediawiki/extensions/WikiEditor] (master) C: -1; - https://gerrit.wikimedia.org/r/69896 [11:49:01] New review: Se4598; "in addition: I don't know how it is outside of [a-z] like lowercase umlaut ö with breaks not at dew..." [mediawiki/extensions/WikiEditor] (master) - https://gerrit.wikimedia.org/r/69896 [12:01:13] New review: Tpt; "Some problems. I hope I have found everything and the next version will be the good one." [mediawiki/extensions/ProofreadPage] (master) C: -1; - https://gerrit.wikimedia.org/r/69802 [12:01:27] New patchset: Hashar; "translatewiki: get rid of job-group" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70145 [12:06:30] could use some english help, should i name variable: hasScripts or haveScripts ? [12:07:09] New patchset: Faidon; "Remove gate-and-submit from puppet template" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70146 [12:08:07] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70146 [12:19:12] hashar: i assume it's a bool so: it hasScripts, they haveScripts, so depending on context/who has sth. ? [12:19:35] se4598: the context is a single server [12:19:40] checking whether it has the scripts [12:19:46] so yeah hasScripts I guess :) [12:19:49] se4598: thank you! [12:22:22] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70145 [12:23:09] New review: J; "looks good but needs rebase against master." [mediawiki/extensions/TimedMediaHandler] (master) C: -1; - https://gerrit.wikimedia.org/r/69952 [12:31:38] YuviPanda: mogge [12:31:43] Change merged: J; [mediawiki/extensions/TimedMediaHandler] (master) - https://gerrit.wikimedia.org/r/69869 [12:31:50] Nikerabbit: nugge [12:35:29] YuviPanda: wana try sending the invites this week, since I managed to use the wrong time last week? [12:35:41] mm doing it [12:35:47] 9pm [12:35:48] right [12:36:47] Nikerabbit: did i get the time right? [12:37:44] YuviPanda: no idea, it looks like the same as last week (which was hour too late) according to or [12:37:58] wait that was an hour too late? [12:38:28] I... am confused? I think the time last time was right... [12:38:49] damn [12:38:51] [17:40:21] [freenode] [orsa(~orsa@IGLD-84-229-13-103.inter.net.il)] hi, I see we moved this 1 hour later [12:39:01] looks like we don't have consensus about the time ;) [12:40:01] New patchset: Hashar; "jobs needing android SDK are labeled hasAndroidSdk" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70152 [12:41:30] New patchset: Hashar; "jobs needing android SDK are labeled hasAndroidSdk" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70152 [12:41:50] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70152 [12:45:02] New patchset: J; "Tune Special:TimedMediaHandler" [mediawiki/extensions/TimedMediaHandler] (master) - https://gerrit.wikimedia.org/r/68029 [12:58:46] New patchset: Hashar; "DataValues & Wikibase jslint are now voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70156 [12:59:27] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70156 [13:10:52] New patchset: Rtdwivedi; "Separated data fetching from manipulation for page." [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69802 [13:26:05] hashar: Cutting out volunteers doesn't make something more professional. [13:26:26] Elsie: what do you mean ? [13:26:49] I think the whole point is that we are trying to make things a bit more professional, which need a ton of cleanup beforehadn [13:26:50] it is not like 99% of my work is publicly available and often reviewed by various volunteers :-] [13:26:56] ahh from a few hours ago [13:27:17] Elsie: I was not really targeting volunteers when talking about bugzilla admin rights [13:27:19] I had them [13:27:26] as well as a bunch of staff [13:27:33] so some of us would sometime change components / rights / whatever [13:27:43] without informing other people [13:27:50] which some time was a bit messy [13:28:00] people asking "who did that change" ? :-) [13:28:08] Imagine if you had a site where anyone could edit. [13:28:13] by restricting who can do the modification and have them log their actions somewhere, we can track what is happening [13:28:14] It'd be chaos. [13:28:29] but there is nothing preventing a volunteer from being granted some administrative rights in bugzilla [13:28:47] Except that they have to sign an indefinite non-disclosure agreement. [13:29:02] for some part of bugzilla they should indeed [13:29:23] The silliness and bureaucracy increases at a worrying pace. [13:29:45] that is exactly what we did when we migrated from subversion to git [13:30:02] we basically removed write access to everyone but staff Engineering [13:30:10] established a process with the volunteers [13:30:19] and start giving write rights to other [13:30:49] that surely has been a bit of pain for a few weeks, but in the end we have a nicer process which is easier for volunteers to get enrolled and ultimately get write access [13:30:57] I guess andre__ is doing the same for bugzilla [13:31:13] then you will have identified volunteers with privileges rights and anyone can refer to them [13:31:28] Relegating volunteers to second-class citizens is great if you're in first-class. [13:31:35] * Elsie shrugs. [13:31:42] I am not feeding trolls sorry [13:31:56] like more than half of my interactions are with volunteers and I LOVE it [13:32:09] Elsie, where were volunteers cut out? [13:32:12] * andre__ tries to get the context [13:32:29] if you have a perception that volunteers are second class volunteers, we need to improve our communication probably and you can help us change that feeling. [13:32:54] Elsie, where were volunteers relegated to second-class citizens? [13:32:54] andre__: this morning we were talking about a wikitech-l thread regarding bugzilla admin rights that got removed from most people [13:33:01] you probably have seen Rob reply already [13:33:13] hashar, yeah. I don't get the connection to "cutting out volunteers". [13:35:03] ahh [13:35:03] I don't see bureaucracy here yet. But when 30 people can happily change everything in Bugzilla and don't tell each other and changes are contradicting themselves and duplicating efforts, it does suck [13:35:13] just found out Elsie is MZMcBride :-] [13:35:32] ah, yet another IRC nickname? Yay. I don't get that part. :) [13:35:36] andre__: yup I fully agree with you [13:35:38] andre__: How do you duplicate effort in Bugzilla? [13:35:45] You make the same component twice? [13:35:53] You mark the same bug resolved? [13:35:57] Elsie: by creating keywords when there are already tracking bugs, for example. Been there, done that. [13:36:05] The horror. [13:36:11] Elsie, exactly!! [13:36:26] andre__: now we want to make sure that the very helpful people have the proper rights to keep helping [13:36:56] yeah. And I still need to reply to that thread that help is always welcome. I think the term "admin" is just misinterpreted. [13:37:09] Elsie: are you missing any rights in Bugzilla ? [13:37:37] "admin" sounds really awesome. but it's not needed for 95% of the tasks. [13:38:06] hashar: Probably not. [13:38:08] Elsie, so what's the example for "cutting out volunteers" exactly, in case you're up for a serious discussion and not trolling only? [13:38:23] It's a general trend. [13:38:32] How many volunteer Bugzilla admins are there now? [13:38:34] One? [13:38:39] 1.5 if you count former staff. [13:38:41] 2. same as months ago. [13:38:44] yeah, 1.5 [13:38:48] how is it relevant? [13:39:05] There used to be a lot more. [13:39:10] They've been cut out over time. [13:39:10] Elsie, when? [13:39:12] Professionalization! [13:39:27] Maybe, but to check that I'd need to know which timeframe we talk about. [13:39:39] I can only assure you that there have always been two since I started here (which is 8 months ago) [13:39:57] and again I'm not sure what that number is meant to tell. [13:40:28] add me [13:40:35] I got my volunteer bugzilla admin rights removed [13:40:44] (but I am a contractor now, so maybe I don't count) [13:44:31] Sue used "professionalization" as the rationale for removing wikimediafoundation.org admins. When I saw you used it as well, I was annoyed. [13:44:43] who is "you"? [13:44:52] hashar. [13:45:13] * Elsie afks. [13:45:36] Elsie: ah sorry :( [13:45:53] New patchset: Rtdwivedi; "Editing system for pages separated." [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69973 [13:45:53] Elsie: will use another term next time, like "avoid the crazy plate of spaghetti" [13:53:13] manybubbles: hey there :-] [13:53:26] manybubbles: I was wondering if you had any puppet classes yet to install solr [13:53:57] hashar: nope - I'm talking this morning with faidon about packaging it. can't install it worth anything without packages. [13:54:48] manybubbles: whenever there is something more or less working, I would like to add a solr instance in beta [13:55:03] and maybe start configuring CirrusSearch in beta to replace the current lucene search extension [13:55:37] hashar: I'd like that. [13:55:45] is beta on the labs ganglion? [13:55:57] that will also provides you a cluster that more or less look like production [13:56:10] and let you do the CirrusSearch configuration using the oeprations/mediawiki-config.git repo :) [13:56:33] cool. [13:56:51] New review: Tpt; "Some issues remaining. Most features are now working well!" [mediawiki/extensions/ProofreadPage] (master) C: -1; - https://gerrit.wikimedia.org/r/69802 [14:06:39] New patchset: Hashar; "use hasSlaveScripts label when requiring scripts" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70167 [14:07:14] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70167 [14:11:17] New patchset: Rtdwivedi; "Separated data fetching from manipulation for page." [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69802 [14:26:02] New patchset: Hashar; "erb and pp lint requires slaves scripts" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70168 [14:26:44] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70168 [14:36:17] New patchset: Tpt; "Beggining of refactoring of Page pages [DO NOT SUBMIT] Change-Id: I48ed31bdd01e89046ca2b3f86a1cb394ff18eb29" [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/70169 [14:41:08] Nikerabbit: sweet. Do send out the email about orsa's availability tho [14:48:11] New patchset: Hashar; "jenkins validation please ignore" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/70174 [14:48:45] Change abandoned: Hashar; "(no reason)" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/70174 [14:52:15] New review: Manybubbles; "I think it would be worth pulling the prefix and full text search handling of namespaces into a sing..." [mediawiki/extensions/CirrusSearch] (master) C: -1; - https://gerrit.wikimedia.org/r/69944 [14:56:53] New patchset: Siebrand; "Localisation updates from http://translatewiki.net." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70176 [14:58:30] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/70177 [14:58:51] Change abandoned: Hashar; "(no reason)" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/70177 [14:59:18] New patchset: Hashar; "use label hasContintPackages for pep8 and pyflakes jobs" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70178 [14:59:18] New patchset: Hashar; "phplint jobs require slave scripts" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70179 [14:59:33] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70178 [14:59:45] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70179 [15:00:47] New review: Tpt; "Two last details and everything will be good!" [mediawiki/extensions/ProofreadPage] (master) C: -1; - https://gerrit.wikimedia.org/r/69802 [15:27:38] New patchset: Rtdwivedi; "Separated data fetching from manipulation for page." [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69802 [15:28:05] New review: Anomie; "I don't want to +2 this myself, since I originated this bit of the code." [mediawiki/core] (master) C: 1; - https://gerrit.wikimedia.org/r/69950 [15:33:56] New review: Tpt; "All looks good!" [mediawiki/extensions/ProofreadPage] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/69802 [15:34:05] Change merged: Tpt; [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/69802 [15:34:23] New patchset: Hashar; "clean up php extensions definitions" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70184 [15:34:26] New patchset: Hashar; "PHP extensions jobs marked with hasContintPackages" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70185 [15:35:26] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)" [mediawiki/php/luasandbox] (master) - https://gerrit.wikimedia.org/r/70186 [15:36:12] Change abandoned: Hashar; "(no reason)" [mediawiki/php/luasandbox] (master) - https://gerrit.wikimedia.org/r/70186 [15:36:35] New review: Hashar; "worked for luasandbox https://gerrit.wikimedia.org/r/70186" [integration/jenkins-job-builder-config] (master) C: 2; - https://gerrit.wikimedia.org/r/70184 [15:36:36] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70184 [15:36:50] New review: Hashar; "worked for luasandbox https://gerrit.wikimedia.org/r/70186" [integration/jenkins-job-builder-config] (master) C: 2; - https://gerrit.wikimedia.org/r/70185 [15:36:51] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70185 [15:40:56] New patchset: Reedy; "Fix parameter order to match actual in sendMail.php" [mediawiki/extensions/SecurePoll] (master) - https://gerrit.wikimedia.org/r/68847 [15:41:03] Change merged: Reedy; [mediawiki/extensions/SecurePoll] (master) - https://gerrit.wikimedia.org/r/68847 [15:41:34] Change abandoned: Reedy; "Guess this wasn't needed" [mediawiki/core] (wmf/1.22wmf7) - https://gerrit.wikimedia.org/r/68844 [15:42:32] New patchset: Reedy; "Cater for different license for testwikidatawiki too" [mediawiki/extensions/WikimediaMessages] (master) - https://gerrit.wikimedia.org/r/67301 [15:43:42] New patchset: Reedy; "Fixup truncated license header" [mediawiki/extensions/TorBlock] (master) - https://gerrit.wikimedia.org/r/60439 [15:48:37] ^demon: Hmm, Gerrit is giving me 500s for all page views after taking a long time to load [15:48:52] I have to do a deployment soon :( [15:49:29] nm it's back up [15:49:38] <^demon> !log restarting gerrit [15:49:45] <^demon> Whoops wrong chan [15:59:53] heh [16:02:35] New review: Anomie; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67875 [16:03:36] New patchset: Reedy; "Initial version of script to automatically upload changelogs to mw.org" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/67416 [16:04:05] New patchset: Reedy; "Initial version of script to automatically upload changelogs to mw.org" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/67416 [16:04:28] ^demon: found another extension that does not self update :/ Vector . Do you want a bug report ? [16:04:39] <^demon> No, I'll look. [16:04:51] last commit referenced is Date: Fri Mar 29 22:32:03 2013 -0300 [16:05:40] Change merged: Reedy; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/67416 [16:10:48] Reedy: greg-g so there are no jenkins jobs on mediawiki/tools/relaease, don't you want some ? :-] [16:11:46] <^demon> hashar: No, it's up to date. [16:11:50] :) [16:11:56] <^demon> That's just the commit date of the last commit. The merge date is 5d ago [16:12:20] New patchset: Hashar; "mw-tools-release-phplint" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70194 [16:12:24] <^demon> HEAD~1 is "Mon Jun 10 07:32:19 2013 +0000" [16:12:25] ^demon: ooops sorry :( [16:12:39] <^demon> np [16:13:06] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70194 [16:13:37] New review: CSteipp; "(1 comment)" [mediawiki/extensions/TorBlock] (master) - https://gerrit.wikimedia.org/r/60439 [16:13:50] New review: Anomie; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69876 [16:13:57] New patchset: Anomie; "API: Fix imageinfo iiurlheight on audio files" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69876 [16:14:14] New review: Anomie; "(1 comment)" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [16:15:45] New patchset: Reedy; "Comment INSIDE php tags" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70195 [16:16:05] Change merged: Reedy; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70195 [16:16:45] New review: Cmcmahon; "update docs re: altered build times" [qa/browsertests] (master) C: 2; - https://gerrit.wikimedia.org/r/69865 [16:16:45] Change merged: Cmcmahon; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69865 [16:21:39] New patchset: Demon; "Respect namespaces when searching. Also fix prefix search output" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/69944 [16:21:59] New patchset: Daniel Kinzler; "Assert write perms when opening a sqlite file." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70196 [16:22:17] New review: Demon; "Still untested :)" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/69944 [16:25:03] New patchset: Daniel Kinzler; "Don't catch errors thrown during DB setup." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70197 [16:26:11] Reedy: Do you know if https://bugzilla.wikimedia.org/show_bug.cgi?id=48873 is still an issue? [16:29:05] Yup [16:29:10] New review: Manybubbles; "Makes sense to me now. Can you handle the same namespace issue when building titles for full text s..." [mediawiki/extensions/CirrusSearch] (master) C: 1; - https://gerrit.wikimedia.org/r/69944 [16:29:44] ^demon: do you want me to very that change as well by checking it against my side? [16:30:10] <^demon> If you want. I just hadn't played with it yet. [16:32:58] New patchset: Hashar; "trigger mw-tools-release-phplint" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70199 [16:33:53] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70199 [16:34:45] New patchset: Demon; "Respect namespaces when searching. Also fix prefix search output" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/69944 [16:35:42] New patchset: Demon; "Respect namespaces when searching" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/69944 [16:35:48] New patchset: Hashar; "jenkins validation (php linting)" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70200 [16:36:24] New patchset: Hashar; "jenkins validation (php linting).." [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70200 [16:36:57] Change abandoned: Hashar; "(no reason)" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70200 [16:37:00] New patchset: Krinkle; "Update VisualEditor submodule to latest master" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70201 [16:38:00] New patchset: Krinkle; "Update VisualEditor submodule to latest master" [mediawiki/core] (wmf/1.22wmf7) - https://gerrit.wikimedia.org/r/70202 [16:39:34] New patchset: Hashar; "jobs for mw/ext/CirrusSearch" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70203 [16:41:21] hashar: Sorry for cancelling the ci meeting so last minute. I woke up a few minutes before the meeting would start (I'm in San Francisco) and on my way to the office for a VE deployment (doing that right now) [16:41:32] New patchset: Hashar; "triggers for mw/ext/CirrusSearch" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70204 [16:41:34] If you're available in 1-2 hours we could do it then, or skip it for now. [16:41:41] Krinkle: yeah that is what I thought :-] [16:42:06] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70203 [16:42:53] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf7) - https://gerrit.wikimedia.org/r/70202 [16:42:57] Krinkle: can't really do it tonight though. Promised my wife I will not touch the computer until my first conf call [16:43:30] Krinkle: maybe tomorrow ? Not sure when you head out for lunch [16:44:06] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70204 [16:44:18] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70201 [16:44:21] * Platonides sells hashar a touch-less keyboard [16:45:15] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/70205 [16:46:14] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)." [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/70205 [16:48:09] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT).." [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/70205 [16:48:30] Change abandoned: Hashar; "(no reason)" [mediawiki/extensions/CirrusSearch] (master) - https://gerrit.wikimedia.org/r/70205 [16:48:51] ^demon: CirrusSearch has some basic linting jobs now. [16:48:58] ^demon: whenever you get unit tests, I can add them in :-] [16:49:03] <^demon> Oh cool, thanks. [16:49:40] I am off for now, will be back later for some conf calls thus will barely be on IRC [16:49:44] *wave* [16:50:43] New patchset: Reedy; "Commit from backup..." [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70207 [16:51:39] New review: Platonides; "If it's not writable, I would instead make MediaWiki read-only. That's much nicer than completely bl..." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/70196 [16:57:47] New patchset: Matmarex; "Refactor watchlist token handling" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64565 [16:58:24] New review: Matmarex; "Daniel, replied and done. Thanks!" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64565 [17:02:53] New patchset: Matmarex; "Refactor watchlist token handling" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64565 [17:04:26] New patchset: Subramanya Sastry; "Added several new tests for preventing regressions in Parsoid" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70209 [17:06:48] New patchset: Cmcmahon; "Working code for June 27 tech demo meetup" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/68283 [17:07:24] Change merged: Cmcmahon; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/68283 [17:13:25] New review: Aaron Schulz; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69876 [17:15:44] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69876 [17:18:36] Change abandoned: Reedy; "(no reason)" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70207 [17:25:17] New patchset: Reedy; "Minor updates as to how I have it currently" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70213 [17:30:04] New patchset: Reedy; "Minor updates as to how I have it currently" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70213 [17:31:09] Change merged: Reedy; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/70213 [17:35:34] New review: Krinkle; "Release note "action=watch may now return errors" is not very useful, please clarify." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68462 [17:44:21] New patchset: Siebrand; "Localisation updates from http://translatewiki.net." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70176 [17:44:43] New review: Siebrand; "Replied. CR2 will contain an update of MessagesAst.php." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70176 [17:45:22] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69950 [17:46:36] New patchset: Krinkle; "Correct release notes for I9dd9f0fd499c37f2" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [17:46:46] New review: Krinkle; "Fixed in I9dd9f0fd499c37f2." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68462 [17:48:51] New patchset: Krinkle; "Correct release notes for I9dd9f0fd499c37f2" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [17:54:26] ori-l, YuviPanda, i would have loved to reuse existing json config classes, but we needed lots of extras on top of it. I'm still all for joining it and creating one big json configuration system [18:00:55] New patchset: Subramanya Sastry; "Added new + updated old tests for preventing regressions in Parsoid" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70209 [18:01:13] Krenair: I just sent an email to you. [18:04:09] New patchset: Aaron Schulz; "Proposed SQL schema." [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/69958 [18:06:01] New review: Cmcmahon; "This is great, thanks Tomislav" [qa/browsertests] (master) C: 2; - https://gerrit.wikimedia.org/r/70045 [18:06:01] Change merged: Cmcmahon; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/70045 [18:06:44] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70209 [18:14:26] New patchset: Nemo bis; "(bug 50103) "Do not invalidate translations" checkbox should be checked by default" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/70219 [18:14:28] Tpt_: "Fatal error: Call to a member function exists() on a non-object in /usr/local/apache/common-local/php-1.22wmf8/extensions/ProofreadPage/ProofreadPage.body.php on line 160" [18:14:57] hey Reedy [18:14:58] tell me [18:15:16] how bad would it be if i used a bot to null edit some 100k articles on a wikipedia? [18:15:37] (i want https://pl.wikipedia.org/wiki/Specjalna:UnconnectedPages refreshed) [18:15:48] New review: Siebrand; "This is not the way this feature should be implemented. This encourages under-fuzzing, which is real..." [mediawiki/extensions/Translate] (master) C: -2; - https://gerrit.wikimedia.org/r/70219 [18:15:49] Firstly, why [18:15:51] Ah [18:16:14] it's generated in real-time [18:16:23] but the data in page_props is not there for all of those arciles [18:16:44] The question is how quickly you're wanting to do all those null dits [18:16:47] * bawolff wonders if that implies something is broken in job queue? [18:17:12] Maaybe [18:17:12] reedy@tin:/a/common$ mwscript showJobs.php plwiki [18:17:13] 3875 [18:17:40] There was a complaint on commons about a week ago that adding a category to a template didn't update the pages using the template [18:18:21] yurik: +1 [18:18:25] Sounds suspicious [18:19:10] Change abandoned: Nemo bis; "(no reason)" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/70219 [18:19:14] https://commons.wikimedia.org/wiki/Category:ACC_needing_decomposition was the category in question, still only 2 files when it should have hundreds [18:19:17] yurik: +1 [18:19:19] no, i think that wikibase simply doesn't update those on wikidata edits [18:19:23] only on local edit [18:19:25] bawolff: ^ [18:19:27] Reedy: ^ [18:19:32] * bawolff was going to file a bug about it if it didn't populate itself in about a week [18:19:57] Reedy: how quickly would it be okay to do those edits? :P [18:20:15] MatmaRex: I would still call that a bug, just a different one from my bug. [18:20:51] bawolff: sure, but i see why they don't do that [18:21:03] -> Bugzilla [18:21:14] bawolff: there's a lot of edits happening on wikidata, and these was even more at the beginning of the project, and this could flood the queues [18:21:27] i think there was a discussion about WD purging mechanisms somewhere [18:23:05] 15 null edits per minute would take 4-5 days [18:26:35] * bawolff wonders how long it would take to go through 13858 jobs... [18:27:15] or actually if the job listing at siprop=statistics is even accurate now that we use redis [18:27:31] qgil: are you around? [18:27:40] hi, yes [18:27:58] I'm about t send an emai with more details about GSoC AllHands [18:28:06] hi :) I sent you a confirmation email, but I forgot to mention something you may enjoy too -- I wrote a blog post about my adventures with rtl/ltr in VE [18:28:16] cool! [18:28:39] http://moriel.smarterthanthat.com/tips/the-issues-of-coordinate-system/ <-- hope you like it :) [18:29:45] * bawolff does his standard spiel about trying to get people to add their blogs to en.planet.wikimedia.org :P [18:30:25] * mooeypoo didn't hear the spiel before [18:30:35] what's en.planet.wikimedia.org ? :P [18:30:41] * mooeypoo goes to check [18:30:55] New patchset: Reedy; "ProofreadPage Fatal error: Call to a member function exists() on a non-object" [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/70226 [18:31:41] ooooh [18:32:04] https://meta.wikimedia.org/wiki/Planet [18:32:04] Anyone want to review 70226 (easy one line change) so I can deploy it? [18:32:50] bawolff: not all of my posts are wiki-related, though. [18:32:55] most are now that I do GSoC [18:33:07] but not all.. mmm i can have the category in there, though. [18:33:10] New patchset: Umherirrender; "Add list=allfileusages" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61817 [18:33:30] mooeypoo: GSoC is considered wiki-related [18:33:43] Yippie, build fixed! [18:33:43] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6 build #70: FIXED in 12 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6/70/ [18:33:48] But most people use some soft of category to filter out the non-wiki posts [18:34:35] bawolff: yah, I think I can do that, it's Wordpress.. basically fool-proof system [18:34:46] yay! [18:35:19] you may want to verify I'm good enough first, though :D hee [18:35:40] I'm pretty sure you are [18:35:51] hey bawolff, https://bugzilla.wikimedia.org/show_bug.cgi?id=50119 might interest you [18:36:05] (i'm making a patch for it already) [18:36:11] And we don't exactly have standards for planet in any case... [18:36:24] MatmaRex: Good :) [18:37:49] New patchset: IAlex; "Fix fatal error when $wgValidateAllHtml is enabled and apache_request_headers() does not exists" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63631 [18:38:44] New patchset: IAlex; "Fix fatal error when $wgValidateAllHtml is enabled and apache_request_headers() does not exists" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63631 [18:39:13] New patchset: Matmarex; "Add broken-file-category for file links inside " [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70227 [18:40:08] bawolff: wanna review? :P ^ [18:40:39] Not right now, but sometime today I will [18:40:51] hm silly question, but just to make it clear, I should add the blog address "below" -- meaning to the "Not yet added" list in that page bawolff ? [18:41:05] or is there a specific request-to-be-added list i'm missing? [18:41:45] Yeah, in the not yet added section https://meta.wikimedia.org/wiki/Planet#Not_yet_added [18:42:03] Change abandoned: Ori.livneh; "Thanks for the reviews. You are both correct, of course. I was careless when investigating this issu..." [mediawiki/extensions/CoreEvents] (master) - https://gerrit.wikimedia.org/r/69966 [18:43:21] New patchset: IAlex; "Rebuild MessagesEn.php" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70228 [18:47:29] bawolff: https://meta.wikimedia.org/wiki/Planet_Wikimedia#Not_yet_added <-- added! :) thanks [18:49:20] New review: IAlex; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [18:54:07] Change merged: Tpt; [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/70226 [18:59:36] Reedy: I've merged your change. Thanks a lot. [19:00:48] Needs merging to production next [19:03:41] New review: Anomie; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/70215 [19:04:03] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70228 [19:07:01] New review: Ori.livneh; "(1 comment)" [mediawiki/extensions/Campaigns] (master) C: -1; - https://gerrit.wikimedia.org/r/69841 [19:11:35] hey Platonides [19:12:08] re: https://gerrit.wikimedia.org/r/#/c/70058/ -- what am I looking at, here? I don't remember asking for that feature -- I wasn't even aware of this tool prior to this code-review request. [19:12:33] what does this do that chown doesn't? [19:13:05] (or petan) [19:13:25] hi [19:13:27] * YuviPanda makes up a song about ori-l giving out CR like candy [19:13:45] ori-l it was Oren, not you [19:13:55] ori-l: http://lists.wikimedia.org/pipermail/labs-l/2013-June/001318.html + thread. [19:15:00] wow, this seems like a really bad idea [19:15:12] the existance of the tool? [19:15:16] or the rewrite? [19:15:20] or the patch? [19:15:31] the existence of the tool. i'm still not sure what it does that chown doesn't. [19:15:34] hmm I just heard about this a few mins ago [19:15:39] guess I should read the thread [19:16:10] ori-l it's for people who don't have root to overtake files in their folder [19:16:18] it is neccesary for it to exist [19:16:25] given the restrictions of apache server [19:16:37] I'm not particularly sure why, but I do know that chown doesn't work but take does [19:16:39] (on tool-labs) [19:16:48] because chown requires root [19:17:49] it does? [19:18:52] Reedy: Is there a way to backport the change without updating to the current master? We have merged a change today that should be tested before being launched to production. [19:19:28] Yeah, if we make a deployment branch or similar to we can not include that revision [19:19:35] yes it does (require root) [19:21:10] well, if so, then it's presumably for a good reason? [19:21:36] you don't want peple arbitrarily taking ownership of files, bviously [19:21:48] (and then being able to write them) [19:22:44] I wonder what files in what directory people need to own [19:22:51] that this tool is necessary [19:23:48] apergos it only allow you to overtake files in folder which you own [19:24:02] for example, you ssh using your regular account and upload files to tool/public_html [19:24:15] then you switch to tool account and you need to transfer ownership [19:24:34] well if you own the folder you can already just make a copy of said file (presuming you have read access) and own that, tossing the original [19:24:42] a bit kludgy but works perfectly fine [19:24:48] apergos that is kludgy [19:25:00] we want people in tools project to have simple way to do things [19:25:05] what is creating files in people's home dir anyways? [19:25:09] James_F: RoanKattouw: https://bugzilla.wikimedia.org/show_bug.cgi?id=50123 [19:25:13] huh? [19:25:20] well you said in a folder they own [19:25:24] yes [19:25:45] what service is creating files (owned by something else) in directories owned by someone else? [19:25:50] if you believe there is a security hole, you are welcome to prove it in toolsbeta project [19:26:04] sftp? [19:26:16] you can't upload files using different account [19:26:26] you need to use personal account to upload stuff to tool's folder [19:26:39] which make its owner not being tool [19:27:15] so... (keep going) [19:27:37] apergos there are many things which suck on unix permissions system which is... just outdated [19:27:48] IMHO even windows have more advanced file system permissions [19:27:58] role-based with many options [19:28:19] unix is using model invented long time ago [19:28:56] what's the actual problem? [19:29:33] like, it's very hard to create a file on unix which can be read by group blah, written by group moo, permission of it can be changed by group foo and its files can be executed by user bar only [19:30:03] yes but in the specific instance, what needs to happen and why? [19:30:13] which instance [19:30:16] someone uploads a script (I guess) to... [19:30:31] somewhere, for use by whom? via sftp [19:30:31] someone uploads a script directly to public_html of tool [19:30:45] it is owned by user who uploaded it [19:30:49] not by a tool [19:31:01] you need it to be owned by a tool [19:31:22] there are some tricky ways to do this, but you should have seen these labs-l threads about people crying how hard it is [19:31:33] some of them even claimed it to be "impossible to use" this way [19:31:37] ori-l: nobody other than petan thinks there is an actual problem, btw :) This thing has been going on for 3 days now :) Rest of us are fine with things as they are [19:31:38] that's why take exist [19:31:44] well a script mght make those tricky ways simple [19:31:52] YuviPanda no, you don't get the point [19:32:01] YuviPanda we aren't talking about -g we are talking about take as it is [19:32:05] ah, okay [19:32:20] if the file isn't owned by a tool, what happens? [19:32:30] apache server give you 500 [19:32:37] because the permissions are wrong [19:32:45] it matters for php scripts only [19:32:49] AFAIK [19:33:09] Coren set a lot of restrictions which I don't think that are necessary [19:33:47] anyway, given to poor possibilities of linux FS permissions system (let's call it FSPS cuz I am really lazy) [19:33:48] well it seems that talking with him about whether some sort of defaut group setting for sftp uploads into certain directories is feasible, might be ne approach [19:33:51] take is a nice extension [19:34:00] *default [19:34:02] *one [19:34:12] New patchset: Tpt; "ProofreadPage Fatal error: Call to a member function exists() on a non-object" [mediawiki/extensions/ProofreadPage] (1.22wmf8) - https://gerrit.wikimedia.org/r/70240 [19:34:15] we have default group [19:34:29] I gues he is going to have to be in the loop if he made changes to the apache configs for php scripts [19:34:30] but the apache requires OWNER to be corrent, it doesn't care so much about group [19:34:46] New review: Tpt; "Backport" [mediawiki/extensions/ProofreadPage] (1.22wmf8); V: 2 C: 2; - https://gerrit.wikimedia.org/r/70240 [19:34:47] Change merged: Tpt; [mediawiki/extensions/ProofreadPage] (1.22wmf8) - https://gerrit.wikimedia.org/r/70240 [19:35:31] I don't really see anything wrong on this approach, FSPS suck and this extension improves them at some point :> [19:35:39] it's very secure so it can't really cause any troubles [19:35:50] it basically only allows you to do what you could do anyway, just hard way [19:35:59] it let you do that simply [19:36:25] but, at some point it allows you to do things which you couldn't do [19:36:41] if you had a folder in your folder, which would have same group as you but permissions 700 [19:36:47] you coulds till overtake it using take [19:37:38] there are a dozen ways of approaching this problem that don't require writing a low-level tool in C that mucks with file ownership [19:37:40] take itself was invented by Coren he might give a better explanation why we need it, but I see a lot of use cases for it [19:37:50] ori-l such as? [19:37:53] it would be best to avoid using special scripts like this that we write ourselves [19:38:04] ori-l: keep in mind it needs to be "linux n00b" friendly [19:38:28] it's c++ not c :D [19:38:34] why everybody think it's c [19:38:43] :P [19:39:01] because C sucks less :-P [19:39:05] I have weird passion for low level tools, so I kind of like it but maybe you are correct [19:39:18] I don't share this opinion :P [19:39:23] c++ <3 [19:39:28] petan, fix apache configs: https://httpd.apache.org/docs/2.2/suexec.html [19:39:33] petan: you are trying to rewrite memcached in C#. I think that sortof goes behind 'weird passion' [19:39:33] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70176 [19:39:33] * AaronSchulz likes having a String type [19:39:36] :P [19:39:42] ori-l: they are "purposefuly fucked" [19:39:50] of course you half to use .c_str() to use a lot of libraries... [19:39:56] petan: well, we can purposely un-fuck them [19:39:58] *have [19:40:12] petan: cron job that runs every minute and chowns public_html/* to tool user [19:40:21] * AaronSchulz goes back to staring at a sql file [19:40:23] that is creepy and resource expensive [19:40:29] and even more dangerous than take utility [19:40:47] but I also had that idea in past [19:40:52] not exactly this but something similar [19:41:08] users could trigger a job which chown -R their home or something like that [19:41:11] you and coren should have a chat about what can be done within the bounds of the apache config and regular unix permissions and utils to work around this [19:41:28] propose something and mail the list [19:41:35] yes we should have a lot of chat :) [19:41:43] so it can be security reviewed [19:41:56] security really needs a lot more eyes to be done right [19:41:56] well, I don't know why should I propose something, I actually like doing it this way :P [19:42:04] maybe people who don't like it should propose something better [19:42:18] New patchset: Tpt; "(bug 50117) Backport for ProofreadPage" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70245 [19:42:28] this is pretty secure, fast and resource inexpensive [19:42:33] and it's newbie friendly [19:43:23] marc volunteered to do a cmplete security review; I expect chris steipp would be willing to look at it too [19:43:26] * petan suggest apergos propose a better ping which doesn't need to run as root [19:43:39] reedy: https://gerrit.wikimedia.org/r/#/c/70245/ that use the branch https://git.wikimedia.org/log/mediawiki%2Fextensions%2FProofreadPage.git/refs%2Fheads%2F1.22wmf8 [19:43:51] ok if you want to do that, here is source code: https://github.com/benapetr/take [19:43:52] :P [19:43:57] I suspect they will both prefer that we do not use a homegrown program though [19:44:14] no, this is something you need to do, as the person making the patches [19:44:42] wait, who you call "marc" now [19:44:42] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [19:44:42] Marc-Andre or Mark? :P [19:44:47] Marc A. Pelletier [19:45:03] but... he MADE THE TOOL why should he a) review it b) oppose it [19:45:04] who volunteered to do a full security review [19:45:13] of your patches [19:45:15] I just made a new version of it [19:45:19] ah that, yes [19:45:30] New patchset: Krinkle; "Correct release notes for I9dd9f0fd499c37f2" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [19:45:36] but I don't understand why he should be againsted the philosophy of it, when he invented it [19:45:37] so take him up on it, and ask chris, who is our security dev, to do so as well [19:45:46] and follow their recommendations [19:45:54] I am trying to take him up on it since saturday :) [19:45:57] which may be 'don't write it yourself, do X insead' [19:46:00] it's not easy [19:46:00] *instead [19:46:04] New review: Krinkle; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [19:46:15] petan: to be fair that was saturday, and today is his day off :) [19:46:25] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [19:46:30] New patchset: Reedy; "Update ProofreadPage to track 1.22wmf8 branch" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70246 [19:46:39] New review: Krinkle; "Fixed message and added release note for the changing of the public method for watching, it is not u..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [19:46:43] saturday, sunday = I had time, monday = neither of us have time, tuesday... friday = I am busy... :P [19:47:14] petan: also four other people pointed out security flaws to you and you are still refusing to acknowledge them [19:47:21] YuviPanda what [19:47:25] ^ see [19:47:47] YuviPanda I am not "refusing to acknowledge them" I just don't see where they are? maybe they fail to point them out? [19:47:55] ^ see again :) [19:48:04] lot of people just said "it is insecure" but nobody told me why or how [19:48:10] petan: I gave you four links [19:48:11] IMHO sudo is insecure [19:48:18] YuviPanda where they are [19:48:32] if you mean that link about race conditions, you realize that is covered in source code? [19:48:46] it is race condition-proof [19:48:48] it isn't, which is what the 4 of us are trying to tell you [19:48:57] elaborate [19:48:58] and since you aren't really listening, nevermind. [19:49:04] we did, for multiple hours yesterday [19:49:06] I"m gonna say it again. get your code security reviewed by marc and chris. follow their recommendations *before* deploying it. [19:49:18] New review: Spage; "(1 comment)" [mediawiki/extensions/Campaigns] (master) - https://gerrit.wikimedia.org/r/69841 [19:49:20] i'm not going to do that again. look over wikimedia-labs logs if you want [19:49:23] and +infinity to apergos [19:49:23] you were telling me yesterday multiple hours that it is insecure and nobody told me HOW it IS [19:49:44] apergos: I am trying to get it reviewed since Saturday [19:49:54] New patchset: Spage; "Add campaign to login form's signup link." [mediawiki/extensions/Campaigns] (master) - https://gerrit.wikimedia.org/r/69841 [19:49:57] it's not that I don't want to have it reviewed, I can't get anyone to do that [19:50:13] Platonides reviewed it, ran it, used strace to point out the race condition [19:50:21] people have weekends. if you need to wait a little, then you wait a little. [19:50:21] which you are still refusing to acknowledge. [19:50:25] YuviPanda I can have a look on logs, but I won't see the answers where they are not... [19:50:54] * YuviPanda goes back elsewhere to doing productive work. [19:51:10] apergos I know, I am not saying that anybody did something bad, not reviewing it. I am just trying to tell you that I know I should have it reviewed and I am trying to. I definitely agree it should be reviewed [19:51:12] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70246 [19:51:13] @notify TrevorParscal [19:51:13] I'll let you know when I see TrevorParscal around here [19:51:22] OK, so we all agree [19:51:30] not in all stuff :P [19:51:43] I still can't see the vulnerabilities which YuviPanda and others talk about [19:52:04] of course they were telling me for hours it is vulnerable yesterday, but nobody told me how and where [19:52:13] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [19:52:23] it's like saying "your code suck" without any details [19:52:47] petan: just give it a bit of time. I can understand the process can be frustrating, especially if you've worked hard to write careful code. [19:53:10] I think you should give it a bit of a rest at the moment because the conversation is getting needlessly heated. [19:56:00] I'm going to give my eyes a bit of a rest here soon.. for about 8 hours [19:56:05] feel very shorted on sleep [19:59:24] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [20:00:31] Reedy: Thanks a lot :-) [20:02:28] New review: Anomie; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [20:12:31] New patchset: Jeroen De Dauw; "Do not allow PHP 5.5 failures anymore" [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/70311 [20:12:59] Change merged: Jeroen De Dauw; [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/70311 [20:13:07] Project browsertests-en.wikipedia.org-linux-chrome build #144: FAILURE in 4 min 10 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-linux-chrome/144/ [20:15:33] Change abandoned: CSteipp; "Starting over using the reference library directly" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/68001 [20:33:23] New review: Spage; "Works for me in normal efLogServerSideEvent() use, but the exception message has a problem and the t..." [mediawiki/extensions/EventLogging] (master) C: -1; - https://gerrit.wikimedia.org/r/70062 [20:35:40] hey ho :) [20:35:41] does anyone know how to dump the resource loader cache content ? [20:36:27] hashar: "dump" as in "delete"? TRUNCATE objectcache; ;) [20:36:54] the ext.vector.collapsibleTabs module has been moved from the Vector extension into MediaWiki core and it seems to be the root cause of a problem on beta [20:36:59] or `delete from objectcache where key like "%resourceloader%", or something along these lines [20:36:59] ah objectcache [20:37:09] I want to find out what is cached [20:37:12] but i think wmf is using memcached for this [20:37:15] and why the cache hasn't invalidated itself [20:37:39] you probably want to poke Krinkle|detached or RoanKattouw_away [20:37:41] I am not even sure where the load.php URL are cached [20:37:54] (aka is it on text squids or on bits ?:D ) [20:38:20] sounds it is on bit cache [20:39:26] aharghg [20:39:40] beta serves load urls from the text cache when it should be bits [20:40:13] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [20:43:01] hashar: argh, i just learned this is happening for some users on pl.wikisource [20:43:15] so if you happen to know how to unbreak it [20:43:17] it'd be lovely [20:43:29] Reedy: have you deployed a new wmf branch today ? [20:43:39] New patchset: SuchABot; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [20:43:45] nearly 3 hours ago [20:43:48] Well, "new" [20:43:56] it was on mediawikiwiki, testwiki and test2wiki last thursday [20:44:09] New patchset: CSteipp; "Add OAuth reference library" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70314 [20:44:09] New patchset: CSteipp; "Style Cleanup of library" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70315 [20:44:09] New patchset: CSteipp; "Basic MediaWiki integration, library cleanup" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70316 [20:44:18] Reedy: PURGE ALL THE CACHES [20:44:21] Reedy: we might have a weird RL cache issue [20:44:32] Reedy: https://gerrit.wikimedia.org/r/#/c/55524/ is breaking stuff. [20:44:38] on beta http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page the collapsible menus in the sidebar are not collapsible [20:44:55] though if you add ?debug=true , that by pass the load.php cache and you get them [20:45:00] The RL cache was updated on thursday post deploy [20:45:09] goood [20:45:33] isn't it purging by itself when the js / css files are changed ? [20:45:37] * MatmaRex is working with a pl.wikisource user with similar issue [20:45:39] or does that need a manual purging? [20:46:27] oh, that's the RL message cache [20:46:37] We cache too much [20:46:45] [21:44:18] Reedy: PURGE ALL THE CACHES [20:46:46] ^ that [20:48:45] Reedy: /* cache key: plwikisource:resourceloader:filter:minify-js:7:4f71a68f01c144d82798d7c8c82ba20b */ [20:48:48] Reedy: can you purge this for me? [20:50:40] > delete plwikisource:resourceloader:filter:minify-js:7:4f71a68f01c144d82798d7c8c82ba20b [20:50:40] MemCached error [20:50:42] Helpful [20:51:35] Uhhhhhhh [20:51:43] Why is mcc.php still using MemcachedClient? [20:52:14] $mcc = new MWMemcached( array( [20:52:15] heh [20:52:22] why do we have to purge them manually ? :( [20:52:37] Reedy: ping me if you manage to do it [20:52:57] I seem to recall there being a bug for that one [20:53:36] MatmaRex: [20:53:36] reedy@tin:/a/common$ mwscript eval.php plwikisource [20:53:36] > $wgMemc->delete( 'plwikisource:resourceloader:filter:minify-js:7:4f71a68f01c144d82798d7c8c82ba20b' ); [20:53:42] > [20:53:54] * Reedy considers this a success [20:54:58] Maybe I should've var dumped [20:55:21] so this is potentially affecting anyway [20:55:45] or is that just plwikisource did not get purged? [20:56:16] maybe RL is not purging properly on dependency changes [20:56:17] https://gerrit.wikimedia.org/r/#/c/55524/13/resources/Resources.php [20:56:41] stuff's throwing "TypeError: $(...).delayedBind is not a function @" [20:57:00] and the module is missing from RL URLs [20:57:06] so we would be shipping the correct file [20:57:12] but not the dependency ? [20:57:12] Reedy: that seems to have fixed it for the affected user at pl.wikisource [20:57:15] yes [20:57:26] but the longer i tink about it, the longer i think it shouldn't have [20:57:31] the more i think* [20:57:38] the longer / the more* [20:57:39] dammit. [20:58:46] * hashar tries to find out the key of the skins.vector.js module :D [20:59:32] the "broken" URL was https://bits.wikimedia.org/pl.wikisource.org/load.php?debug=false&lang=pl&modules=ext.centralNotice.bannerController%7Cjquery.client%2Ccookie%2CmwExtension%7Cmediawiki.legacy.ajax%2Cwikibits%7Cmediawiki.notify%2Cutil%7Cmediawiki.page.startup%7Cskins.vector.js&skin=vector&version=20130624T022110Z&*:3 [20:59:46] sorry, no: https://bits.wikimedia.org/pl.wikisource.org/load.php?debug=false&lang=pl&modules=ext.centralNotice.bannerController%7Cjquery.client%2Ccookie%2CmwExtension%7Cmediawiki.legacy.ajax%2Cwikibits%7Cmediawiki.notify%2Cutil%7Cmediawiki.page.startup%7Cskins.vector.js&skin=vector&version=20130624T022110Z&*:33 [21:00:08] but it still seems "broken" [21:00:27] (obviously not loading delayedBind as it's missing in the modules= list) [21:00:49] but the user's reporting it works now [21:00:52] so i amcompletely at loss [21:01:04] well at least it happens on http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [21:01:09] though it does not when adding ?debug=true [21:01:23] (I guess that by pass the caches entirely) [21:02:02] yes, it does [21:02:05] (sort of) [21:02:21] acutally, no, scratch the"sort of", it does afaik [21:03:09] the two types errors I got are with: http://en.wikipedia.beta.wmflabs.org/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki%2CSpinner%7Cjquery.triggerQueueCallback%2CloadingSpinner%2CmwEmbedUtil%7Cmw.MwEmbedSupport&only=scripts&skin=vector&version=20130621T002832Z [21:03:09] and http://en.wikipedia.beta.wmflabs.org/w/load.php?debug=false&lang=en&modules=site&only=scripts&skin=vector&* [21:03:33] hashar: hey, this actually looks like a different error [21:03:35] aint that lovely [21:03:41] yup :-) [21:03:52] and the cache is 3 days old [21:04:09] Exception thrown by skins.vector.js: Object [object Object] has no method 'collapsibleTabs' [21:04:10] ah [21:04:18] RD: {{poke}} [21:04:19] which points to http://en.wikipedia.beta.wmflabs.org/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki%2CSpinner%7Cjquery.triggerQueueCallback%2CloadingSpinner%2CmwEmbedUtil%7Cmw.MwEmbedSupport&only=scripts&skin=vector&version=20130621T002832Z [21:04:56] one of those modules might have the wrong dependency [21:05:31] hashar: is the problem in the minimization code only, or is there a reason you are uding debug=false? [21:05:42] that is debug=false by default [21:05:53] oh, I thought you manually had added it [21:06:19] hashar: What's up? [21:06:21] hashar: Make sure both core and extensions are up to date [21:06:23] various extensions depend on latest master [21:06:35] incl. Vector who recently did a cross-repo change [21:06:40] yeah [21:06:43] Project browsertests-test2.wikipedia.org-windows-internet_explorer_9 build #375: ABORTED in 9 min 5 sec: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_9/375/ [21:06:53] I think we are hit by that [21:06:59] so, did you update both core and extensions at the same time?. [21:07:10] Where is this [21:07:11] it is done automatically more or less [21:07:16] verify it [21:07:20] It wouldn't be the first time [21:07:31] Krinkle: there was a problem like the one described at the last comment to https://gerrit.wikimedia.org/r/#/c/55524/13 on pl.wikisource as well [21:07:37] (a different one that hashar is having) [21:07:40] yeah I need to verify it but it happened on production too [21:07:48] bug report is https://bugzilla.wikimedia.org/show_bug.cgi?id=49911 [21:07:50] Reedy purged a cache key and the user having issues claims it's now working [21:07:50] (for beta) [21:08:03] way to reproduce: anonymous access to http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [21:08:12] the sidebar shows no collapsible menus [21:08:15] and there is a few errors [21:09:36] New patchset: Ori.livneh; "Validate adherence to 'enum' constraint in server-side logging" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70062 [21:10:37] Krinkle: Vector and core are both up to date on core [21:10:39] New review: Matmarex; "@Hoo, if this is still happening, can you run update.php and recheck? This might be a caching issue…" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/55524 [21:10:40] on beta [21:10:51] hoo: ^^ [21:11:26] MatmaRex: I think that solved itself after some time [21:11:42] did a lot of JS development this weekend and didn't see this error anymore [21:12:19] hoo: it just happened for a user on pl.wikisource [21:12:26] so this does happen, apparently [21:12:33] :/ [21:12:39] and purging some RL caches apparently fixed it [21:18:21] so I have no idea [21:18:32] restart memcached? ;) [21:18:39] na [21:18:48] ideally I would want to know why it did not get purged properly [21:19:24] the only "special" thing about that patch is changing module's dependencies [21:19:36] which mean something is wrong in RL [21:20:35] "There are only two hard things in Computer Science: cache invalidation and naming things." [21:20:51] And off-by-one errors. [21:21:03] three things then. :P [21:21:08] No no, two things. [21:21:11] :D [21:21:17] And shiny red uniforms. [21:21:22] don't you mean four things, brion_ [21:21:23] ? [21:21:34] there are... FOUR... LIGHTS! [21:24:54] got some $wgLoadScript tweaking for beta at https://gerrit.wikimedia.org/r/70322 :) [21:25:03] New patchset: Brion VIBBER; "Bug 48995: Add acceptable use 'tutorial' to app" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [21:27:44] New review: Krinkle; "(1 comment)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/69669 [21:27:49] New review: Yuvipanda; "w00t. " [apps/android/commons] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/69606 [21:27:52] ResourceLoader.php: $key = wfMemcKey( 'resourceloader', 'filter', $filter, self::$filterCacheVersion, md5( $data ) ); [21:27:52] Change merged: Yuvipanda; [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/69606 [21:27:52] Does not seem it uses the dependencies to generate the cache key [21:29:34] New patchset: Krinkle; "Call error on unknown dependencies in mw.loader.using()" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60346 [21:29:51] Change abandoned: Krinkle; "(no reason)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60346 [21:30:54] hashar: i think that the URL itself should chance when dependencies are changed [21:31:02] change* [21:31:19] (i think based on behavior observation, not reading code) [21:32:25] New patchset: Krinkle; "Convert underscores to spaces in {{SUBJECTSPACE}}" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61936 [21:33:05] New review: Krinkle; "Using str_replace on the text to replace the underscores seems like a wrong implementation. Is there..." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/61936 [21:34:23] MatmaRex: so the output content is passed through a minify filter [21:34:32] that takes as argument the module content (aka some javascript) [21:34:40] and generate a key according to that contet [21:37:23] not sure why changing the dependencies does not change the content though [21:37:52] hashar: my wild guessis that the problem is somewhere around makeResourceLoaderLink() and its callers [21:37:55] New review: Spage; "wfm, phpunit OK. Perhaps uiMessage() should return wfMessage(....)->text(), see comment." [mediawiki/extensions/EventLogging] (master) C: 2; - https://gerrit.wikimedia.org/r/70062 [21:38:05] hashar: but i'm still trying to figure out when dependencies are handled [21:38:13] no ideas [21:38:19] Change merged: jenkins-bot; [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70062 [21:38:23] during RL link generation, or during generating the content for that link [21:38:28] hashar: Why would dependencies be related to that in any way? [21:38:34] What are you guys trying to solve. [21:38:40] https://gerrit.wikimedia.org/r/#/c/55524/13/resources/Resources.php [21:38:51] New patchset: Hiong3-eng5; "SpecialPage:OWDownloads" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/70325 [21:38:52] Krinkle: TypeError: $(...).delayedBind is not a function @ [21:39:09] If dependencies change, the package needs to be invalidated, the minified string of the individual file is fine and needs to reminifying, it won't be any different [21:39:15] skin.vector.js had some dependencies changes and its minified cached version has not been purged / invalidated / assigned a new key [21:39:15] Krinkle: as noted on https://gerrit.wikimedia.org/r/#/c/55524/13 , and as it happened today at pl.wikisource [21:39:24] needs *no* reminification. [21:39:43] cause the content hasn't changed right? [21:39:48] Krinkle: and apparently miraculously fixed by purging cache entry for the minified code, although it might have been a coincidence [21:40:01] mostly a coincidence but not exactly [21:40:16] Either way, purging that key won't do anything because the key includes the md5 hash of the contents [21:40:29] next update (every 5 minutes globally) it will use a different cache key [21:40:47] by design RL cache keys must not and need not be invalidated every, doing so will do nothing but regenrate the exact same cache again [21:41:06] ok I got that point [21:41:19] if the contents changed, it'll regenerate in < 5 minutes with a different key. If the contents didn't change it'll end up the same. [21:41:26] now why isn't the module invalidated when its parameter changes (such as position or dependencies ? ) [21:41:35] New patchset: Mattflaschen; "Don't log when schemaAction is falsy." [mediawiki/extensions/GettingStarted] (master) - https://gerrit.wikimedia.org/r/70327 [21:41:53] hashar: maybe it's squid caching [21:42:01] When dependencies change, the package is invalidated. If it doesn't that's a bug. Unfortunately it's hard to debug at this point since it seems there have been already things been messed with and resolved. [21:42:16] I haven't touched anything on beta :) [21:42:16] position won't and shouldn't invalidate the package [21:42:26] hashar: keeping the old URLs in HTML [21:42:29] it'll change the position of the entry in the load queue on the html output for wiki pages. [21:42:31] but dependencies should isn't it ? [21:42:47] the request the loader triggers is still the same and shouldn't be invalidated. [21:43:14] changing position of a module will cascade slowly (~ 30 days) because it is data that is in the html itself. [21:43:34] either load() from head or load() from body respectively [21:43:42] on a per page basis, too. [21:45:49] New patchset: Ori.livneh; "Allow explicit & implicit null returns from hook handlers" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70328 [21:48:41] New patchset: Aaron Schulz; "Proposed SQL schema for OAuth tables" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/69958 [21:49:43] select md_module from module_deps where md_deps LIKE '%collapsibleTabs%'; [21:49:44] Empty set (0.00 sec) [21:49:44] :( [21:49:52] I am such a noob [21:50:12] but I have ext.vector.collapsibleNav :) [21:52:03] New patchset: Hiong3-eng5; "SpecialPage:OWDownloads" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/70325 [21:52:33] New review: Ori.livneh; "Thanks" [mediawiki/extensions/GettingStarted] (master) C: 2; - https://gerrit.wikimedia.org/r/70327 [21:52:41] Change merged: jenkins-bot; [mediawiki/extensions/GettingStarted] (master) - https://gerrit.wikimedia.org/r/70327 [21:52:57] ^ greg-g do you think I could deploy that change (70327, above) during LD? [21:55:25] New review: Mattflaschen; "recheck" [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/69817 [22:07:03] qgil: You around? [22:07:24] yes abut about to start a meeting bawolff [22:07:49] with greg-g :) [22:08:00] ok, its not super important. I just noticed last issue of "tech news" had an inaccuracy, and I'm not sure the best way of "correcting" it [22:08:29] bawolff, guillom coordinates, if that helps [22:08:39] ok, I'll ask him [22:08:52] its a wiki, right? [22:08:55] sorry [22:09:04] :P [22:09:09] the wiki part yes, but the part that was sent to people... [22:09:19] also technically a wiki but more complex to fix [22:09:34] Well for starters I guess I should edit the wiki part :) [22:09:40] :) [22:09:43] New patchset: Ori.livneh; "Fix QUnit test failure" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:09:47] wow, translate markup... [22:09:51] oh right, it does copy/paste it to the usertalk pages [22:10:09] Is this what people feel like when they complain about edit box filled with weird templates scaring newbies? [22:10:45] :) no you know [22:19:02] New patchset: Nemo bis; "Sort groups alphabetically in Special:Translate export tab" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/70332 [22:20:00] New patchset: Ori.livneh; "Fix QUnit test failure" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:24:57] New patchset: CryoKey; "Updated the login flow to attempt to flush credentials" [mediawiki/extensions/CryoKey] (master) - https://gerrit.wikimedia.org/r/70334 [22:26:11] New patchset: Ori.livneh; "Fix QUnit test failure" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:26:45] Change merged: CryoKey; [mediawiki/extensions/CryoKey] (master) - https://gerrit.wikimedia.org/r/70334 [22:27:32] New patchset: Ori.livneh; "Fix QUnit test failure" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:28:24] Change merged: jenkins-bot; [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:29:52] New review: Ori.livneh; "recheck" [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/69817 [22:33:59] Change merged: Aaron Schulz; [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70314 [22:36:46] Change merged: Aaron Schulz; [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70315 [22:39:40] New review: Krinkle; "I'm making a note here: HUGE SUCCESS." [mediawiki/extensions/GuidedTour] (master) C: 1; - https://gerrit.wikimedia.org/r/69817 [22:39:59] New patchset: Ori.livneh; "Make QUnit tests for EventLogging voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70335 [22:40:32] New patchset: Ori.livneh; "Make QUnit tests for EventLogging voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70335 [22:40:55] ori-l: Shall I deploy that? [22:41:11] Krinkle: sure, that'd be awesome [22:42:32] Change merged: Krinkle; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70335 [22:44:04] New patchset: Krinkle; "Support inter-page JSON references" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/68941 [22:44:19] Verifying that it works [22:44:35] New patchset: GWicke; "Disable Parsoid fetch-parsertests.js call" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70337 [22:44:37] Yep [22:46:35] New review: Mattflaschen; "(1 comment)" [mediawiki/extensions/Campaigns] (master) - https://gerrit.wikimedia.org/r/69841 [22:46:50] New patchset: Krinkle; "Clean up: -qunit is voting by default" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70338 [22:47:01] Change merged: Krinkle; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70338 [22:48:12] New review: Ori.livneh; "I still don't think this is a great pattern -- I'd prefer something like:" [mediawiki/extensions/GuidedTour] (master) C: 2; - https://gerrit.wikimedia.org/r/69817 [22:48:34] Change merged: jenkins-bot; [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/69817 [22:48:59] New patchset: Bsitu; "Add HTML email support to Echo notification" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/70112 [22:52:43] New review: Mattflaschen; "(1 comment)" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70330 [22:52:52] New patchset: Yaron Koren; "Fixed handling of date filter inputs in SQL" [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70339 [22:54:36] Change abandoned: Yaron Koren; "Contains error" [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70339 [22:54:46] New patchset: Ori.livneh; "Update GettingStarted to master" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70340 [22:54:57] New patchset: Ori.livneh; "Update GettingStarted to master" [mediawiki/core] (wmf/1.22wmf7) - https://gerrit.wikimedia.org/r/70341 [22:55:30] New patchset: Yaron Koren; "Fixed handling of date filter inputs in SQL" [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70342 [22:56:52] Change restored: Yaron Koren; "I don't really understand Gerrit." [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70339 [22:57:05] Change merged: Yaron Koren; [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70339 [22:59:18] Change merged: Yaron Koren; [mediawiki/extensions/SemanticDrilldown] (master) - https://gerrit.wikimedia.org/r/70342 [23:00:01] going ahead with the GettingStarted sync [23:00:16] Change merged: Ori.livneh; [mediawiki/core] (wmf/1.22wmf7) - https://gerrit.wikimedia.org/r/70341 [23:00:22] Change merged: Ori.livneh; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70340 [23:02:44] Yippie, build fixed! [23:02:45] Project MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox build #32: FIXED in 1 min 7 sec: https://wmf.ci.cloudbees.com/job/MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox/32/ [23:05:31] done [23:09:46] New review: Ori.livneh; "recheck" [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/67034 [23:10:28] New review: Mattflaschen; "Yes, this was just about getting the tests to pass. I am still planning on doing a broad refactorin..." [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/69817 [23:16:06] Change merged: jenkins-bot; [mediawiki/extensions/GuidedTour] (master) - https://gerrit.wikimedia.org/r/67034 [23:22:34] Change merged: Krinkle; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/70337 [23:22:47] New review: Ori.livneh; "(1 comment)" [mediawiki/extensions/Campaigns] (master) - https://gerrit.wikimedia.org/r/69841 [23:35:34] New review: Krinkle; "$success = WatchAction::doWatch();" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70215 [23:37:37] New review: Krinkle; "Be sure to bind to document ready when using an element in the DOM. Don't use the load position to r..." [mediawiki/extensions/Campaigns] (master) C: -1; - https://gerrit.wikimedia.org/r/69841 [23:38:22] New patchset: Inez; "Adjust parserTests.txt for bug 43015" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70348