[00:06:59] tschüss [04:15:45] New patchset: Mattflaschen; "Fix spelling of rm command, to allow local clones." [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/50146 [04:16:22] New patchset: Mattflaschen; "Fix spelling of rm command, to allow local clones." [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/50146 [05:02:18] New patchset: Mattflaschen; "Recursive clone/update for exts, so their submodules are cloned" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/50148 [06:39:37] Reedy: hi [06:47:57] being on openid-wiki, I cannot wget to openid-wiki2: "Connecting to openid-wiki2.instance-proxy.wmflabs.org (openid-wiki2.instance-proxy.wmflabs.org)|208.80.153.147|:80... failed: Connection time" [06:48:06] can any one help here ? [06:49:55] solved [09:26:09] New patchset: Hashar; "regroup scm macros" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50150 [09:26:24] New review: Hashar; "Patch Set 1: Verified+2 Code-Review+2" [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50150 [09:26:25] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50150 [10:50:05] hashar: what's wrong with gerrit? [10:50:15] DanielK_WMDE_: nothing? :-] [10:50:18] i'm getting timeouts from the web interface [10:50:21] DanielK_WMDE_: what is the trouble? [10:50:22] ahh [10:50:54] in such a case I look at the ganglia graph for the manganese host http://ganglia.wikimedia.org/latest/?r=4hr&cs=&ce=&c=Miscellaneous+eqiad&h=manganese.wikimedia.org&tab=m&vn=&mc=2&z=medium&metric_group=ALLGROUPS [10:51:00] which shows a huge CPU spike [10:51:15] kill kill! [10:51:18] :-D [10:51:55] Are the logs flooded again with "Dispatched Failed!"? [10:52:14] I have no idea what could cause that, last time it resumed after 30 - 40 minutes [10:52:23] qchris: I don't have access to the server :( [10:52:36] Oh. Who has besides ^demon? [10:52:43] ops :-] [10:53:08] lets ask in -operations [10:53:10] k [11:07:09] DanielK_WMDE_: gerrit should be up again, thanks to apergos [11:12:55] qchris: thanks [11:13:31] hashar, qchris: gerrit is backed by a servlet, right? [11:13:44] java GC runs can have an effect like this. [11:13:58] might be useful to trigger GC more often, and/or choose a different GC strategy [11:13:59] Unfortunately, it's not just gc. [11:14:27] what is it, then? [11:14:29] It's a problem of jetty that gerrit hits on some java versions. [11:14:38] hrm [11:14:46] don't use jetty, then? [11:14:53] Maybe... [11:15:06] We'll have to discuss with ^demon, as jetty comes with gerrit [11:15:19] I wish I could help but the whole java world is unknown to me [11:15:25] Switching to something else like tomcat isn't our first choice. [11:15:45] maybe Jetty issue is solved in a later version or the Jetty devs could help in [11:15:54] Most reports say "FIxed in recent Java 7" but that's somewhat of an option I guess. [11:17:17] Tomcat or Enhydra, i suppose [11:18:01] Yes, there are some other engines that do not seem to hit this bug. [11:18:21] But I can see ^demon's point in trying to stay with the container that is shipped with gerrit. [12:13:46] Crap, implementation navigation is exhausting. [12:14:07] It's all built-in in mediawiki, but when you don't have that. [12:14:14] Especially backwards navigation in a reverse chronological list [12:14:25] Got it done though, finally. [12:18:17] New review: Demon; "Patch Set 2: Verified+2 Code-Review+2" [mediawiki/tools/release] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50146 [12:18:18] Change merged: Demon; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/50146 [14:14:09] guillom: so I'll take a look at https://www.mediawiki.org/wiki/Wikimedia_engineering_report/How_to [14:14:23] I see you've updated it this month :) [14:14:26] sumanah: I got delayed on updating it; I'm actually doing it now [14:14:30] I haven't yet [14:14:39] wait no that was me [14:14:47] I mean, I changed a few things earlier, but tehre's still cleanup to do [14:14:48] * sumanah is not the fastest today! [14:15:20] If you want to use the how-to as basis for this discussion, I'll need 15 min to finish updating it [14:16:07] Yeah, that seems like the best way to proceed, guillom -- I want to think about how to split up tasks and who can do what [14:16:14] ok [14:16:24] I'll be back in 15 min then [14:16:29] sounds good [14:18:16] heads-up to people listening: I'd love assistance working on https://www.mediawiki.org/wiki/Wikimedia_engineering_report/2013/February -- the best thing you can do is to use the https://www.mediawiki.org/wiki/Wikimedia_Engineering/Project_documentation_howto#How_to_add_a_status_update_for_the_monthly_engineering_report instructions to update mediawiki.org's activity pages to explain what's new [14:20:09] sumanah: as a quick note, it may not be necessary for you or Tilman to get involved, if I get updates early enough I can rush to publish the report on March 1st. [14:20:41] true, guillom but I would prefer for you to be able to use your time next week on stuff other than heads-down work on this report -- chatting with colleagues, for instance [14:21:08] sumanah: I understand; I'm just saying it's an option; the choice is yours. [14:21:17] OK. [14:21:22] (back in about 10) [14:35:21] I'm back guillom [14:35:41] sumanah: just saved [14:35:46] :-) will reload! [14:36:40] There may be a couple things I would only remember if I updated that page while actually doing the steps, but it should be fairly accurate now. [14:36:56] sure. [14:38:21] OK. So, I'm reviewing it and I think I'd like to try to get a volunteer to help with some of these, and I have an idea of whom I could get [14:38:51] I'll ask Amir whether I could grab some of Priyanka's time, and I can probably get Erika or a similar person to proofread [14:40:11] and you might have seen in my email that I'll be asking Tilman to do the post-publication steps [14:40:18] yup [14:40:38] https://meta.wikimedia.org/wiki/Wikimedia_Blog/Converting_wiki_pages_to_blog_posts -- as far as you know, are all the known issues with that script already listed there? [14:41:01] What do you mean by "issues"? [14:41:25] " Known issues" - bugs, I presume [14:41:34] there are 4 bullet points listed [14:41:59] oh, I didn't remember those were there; let me look [14:42:48] yes, that seems accurate [14:43:10] Also, could you clarify whether we just have the plain English summary or also a separate "highlights"? right now I just see https://www.mediawiki.org/wiki/Wikimedia_engineering_report/How_to#Plain_English_summary . [14:43:32] I feel like I remember when these were two separate things. [14:43:37] They're the same [14:43:52] (the "highlights" that went into the WMF-wide monthly report and the "summary for less technical readers") [14:44:07] It was called "highlights" only the first time, then was renamed to "summary", but they're the same thing [14:44:24] The summary for less technical readers is what goes into the WMF report :) [14:44:54] OK! I wonder what I am remembering, then. [14:45:14] Anyway, I can write that, and I can ask Quim when he comes in whether he could take on the task of writing it [14:45:51] You can probably ask Nemo_bis for help to make the page translatable once it's finished and copied into the etherpad [14:46:59] yep [14:47:52] guillom: have you already done the "Second-to-last week of the month" item "Update the list of projects based on projects listed as "Current" on the team hubs"? [14:48:10] sumanah: I haven't. I've only added new hires [14:48:19] okay. So, here's my plan: [14:48:53] You go ahead and do the rest of the "Second-to-last week of the month" work. I will find a way to take care of the "Last week of the month" tasks, with me, volunteers, and potentially Quim. [14:49:26] I may rope you in for some portion of that next week but I will try not to. [14:49:40] ok [14:50:27] And then the week of Mar 4-8 I'll find a way to take care of "Immediately before publication" and I'll write or get written the "Plain English summary" and I will ask Communications to help me with Publication, After publication, & Set up the next report. [14:55:27] If this gets untenable then I will end up delaying the report, which is not the end of the world [14:55:38] sound reasonable guillom? [14:55:49] yes [14:56:06] I'm updating the list of projects now to send the e-mail [14:57:37] ok! thank you. [14:58:32] sure [15:00:04] Emailing a copy-and-paste of this decision now. :) [15:04:07] I'm getting a bit fed up of my outgoing reviews list: https://gerrit.wikimedia.org/r/#/dashboard/98 [15:04:31] I've had nothing merged for two days now, despite 11 of them being LQT patches [15:08:26] Krenair: what do you think is a reasonable frequency of getting things merged? [15:08:39] (that is, what *would be* a reasonable frequency) [15:13:45] sumanah: list updated and e-mail sent [15:14:06] Thanks guillom - I am about to head off to the coworking space. [15:14:11] ok [15:14:12] See you on-line [15:26:42] New patchset: Jeroen De Dauw; "Wikibase now has a standard entry point via which the tests should run fine" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50164 [15:32:31] New review: Hashar; "Patch Set 1:" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50164 [15:34:12] New review: Hashar; "Patch Set 1: Verified+2 Code-Review+2" [integration/zuul-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50164 [15:34:12] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50164 [15:44:32] New patchset: Jeroen De Dauw; "Actually run the unit tests and report the result" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50169 [15:44:42] New patchset: Jeroen De Dauw; "Actually run the Wikibase unit tests and report the result" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50169 [15:50:04] JeroenDeDauw: do tests pass ? [15:50:23] hashar: they should with https://gerrit.wikimedia.org/r/#/c/50168/ yes [15:52:01] New patchset: Hashar; "trigger test pipeline for wikibase (non voting unit tests)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50169 [15:52:04] JeroenDeDauw: I slightly tweaked it [15:52:09] to make sure it is not going to block [15:52:22] New review: Hashar; "Patch Set 3: Verified+2 Code-Review+2" [integration/zuul-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50169 [15:52:23] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50169 [15:52:42] hashar: but we want to have voting for the tests, that's the whole point :p [15:52:57] JeroenDeDauw: yeah :-] [15:53:02] JeroenDeDauw: once they pass properly. [15:53:20] JeroenDeDauw: at least that will report the status, can make them voting later tonight. For now I need to go grab my daugher [15:53:28] I have deployed that previous change [15:53:35] hashar: ok thnx [15:54:23] hashar: the reasoning behind having it report right away is that we can fix failures in the team without being blocked on external factors (you) [15:54:38] No real issue with jenkins doing stupid -2 for a few minutes [15:54:56] well if jenkins start doing -2 that will block you isn't it ? [15:57:31] hashar: not really, we can just ignore it, and presumably we can fix it fast anyway [15:57:50] hashar: though the issue it is running into now might not fall into that category... [15:58:32] i am off [15:58:35] be back later tonight [15:58:42] happy trails [15:58:49] rats :-D [16:04:10] New patchset: Jeroen De Dauw; "Ask is now a dependency for Wikibase" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50176 [16:05:24] arrrrrrrrr [16:05:28] he ran off :p [16:05:52] If someone could review the one line that is https://gerrit.wikimedia.org/r/#/c/50176/ ... [16:09:27] siebrand: Reedy ^^ [16:09:54] JeroenDeDauw: Review that ps, you mean? [16:10:02] Yeah [16:10:17] I'm eager to finnaly have Jenkins work for Wikibase :) [16:10:26] JeroenDeDauw: I don't know about that part of Jenkins yet. [16:10:44] JeroenDeDauw: Don't feel comfortable approving that. [16:10:55] siebrand: this line lists dependencies of the extension under test [16:11:01] As you can see, I just added one [16:11:08] JeroenDeDauw: Is the order of any importance? [16:11:41] siebrand: Not really [16:12:20] New review: Siebrand; "Patch Set 1: Code-Review+1" [integration/jenkins-job-builder-config] (master) C: 1; - https://gerrit.wikimedia.org/r/50176 [16:12:32] JeroenDeDauw: I +1-ed with a comment. [16:13:21] siebrand: thnx [16:14:37] New patchset: Jeroen De Dauw; "Do vote for Wikibase" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50178 [16:14:57] New review: Jeroen De Dauw; "Patch Set 1:" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50178 [16:16:31] New patchset: Jeroen De Dauw; "Do vote for SMW based on unit tests" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50179 [16:27:16] New review: Mwjames; "Patch Set 1: Code-Review+1" [integration/zuul-config] (master) C: 1; - https://gerrit.wikimedia.org/r/50179 [16:33:09] New review: Reedy; "Patch Set 1:" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50176 [18:02:16] qgil_: ping re meeting [18:51:16] !g 45216 [18:51:16] https://gerrit.wikimedia.org/r/#q,45216,n,z [18:51:20] !g bug:45216 [18:51:20] https://gerrit.wikimedia.org/r/#q,bug:45216,n,z [18:58:19] ^demon: Looks like bug:- index isn't being updated anymore [18:58:42] ^demon: Was that something we configured locally or did Gerrit index any key/value pair on a separate line with a colon? [18:58:57] <^demon> We configured it. [18:59:04] Whatever it was, it isn't being updated anymore. Though the keyword is still auto-completed and the old index is there. [18:59:11] <^demon> I rebuilt that index yesterday after we fixed it.... [18:59:23] <^demon> Since we weren't capturing all the formats. [18:59:49] <^demon> Give me a minute, I'm in a meeting. [19:03:12] Reedy ? [19:03:15] Reedy: ping [19:07:56] <^demon> Krinkle: This is the config var: https://gerrit.wikimedia.org/r/Documentation/config-gerrit.html#_a_id_trackingid_a_section_trackingid [19:08:05] <^demon> It's set in gerrit.config.erb like the others [19:09:00] ^demon: maybe the upgrade broke it? [19:09:12] New patchset: Stefan.petrea; "Add replace_space function" [analytics/webstatscollector] (time_travel) - https://gerrit.wikimedia.org/r/50195 [19:09:13] <^demon> We did some config tweaks to it yesterday. [19:09:19] <^demon> Tried to widen the regex for what it expects. [19:09:31] I've added it to a few commits via the commit message editor but it's not showing up in the search. It's been over 24h [19:10:28] ^demon: https://gerrit.wikimedia.org/r/#/c/50137/ [19:10:35] !g bug:45216 [19:10:35] https://gerrit.wikimedia.org/r/#q,bug:45216,n,z [19:10:55] But this one does work: https://gerrit.wikimedia.org/r/#q,bug:45216,n,z [19:10:57] <^demon> I wonder if it doesn't reindex on a commit msg change. That'd be stupid. [19:11:06] Hm.. maybe [19:11:22] https://gerrit.wikimedia.org/r/#/q/bug:45191,n,z [19:12:00] I need to make a test commit for a documentation example, I'll try it [19:12:11] push bug: 1234 from cli and then edit it to 2345 [19:20:25] New patchset: Stefan.petrea; "Squashed commit of the following:" [analytics/webstatscollector] (time_travel) - https://gerrit.wikimedia.org/r/50197 [19:39:22] Change abandoned: Stefan.petrea; "abandoned. New change will be here https://gerrit.wikimedia.org/r/#/c/50195/" [analytics/webstatscollector] (time_travel) - https://gerrit.wikimedia.org/r/47827 [20:02:43] Reedy: do you want a bug to get Validator deployed on wmf branches ? :-D [20:02:52] since wikidata since to depends on it https://gerrit.wikimedia.org/r/#/c/50176/ [20:03:54] New patchset: Hashar; "Ask is now a dependency for Wikibase" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50176 [20:04:10] New review: Hashar; "Patch Set 2: Verified+2 Code-Review+2" [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50176 [20:04:35] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50176 [20:05:47] New review: Hashar; "Patch Set 2:" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/50176 [20:12:11] hashar: do you have contact to Reedy ? [20:12:32] I want him to give his "ok" for the E:RSS push [20:13:30] hashar: what do you think, can I submit ? [20:16:00] no idea honestly [20:16:18] unless I re review it myself :] [20:17:08] I mean, will he get angry or so ? [20:20:16] ok so here is the deal [20:20:21] whenever you +2 / merge code [20:20:25] and that it lands in production [20:20:37] here what might happen: some dumb code start overloading a piece of the cluster [20:20:46] other parts depending on that piece start overloading [20:20:52] editors complains about slowness [20:20:58] and then can't edit anything :-] [20:21:08] in the case of the RSS extension I don't think it is going to cause such issue :-] [20:21:16] but then there are the security issues and so on [20:21:36] Wikinaut: I would ping whoever reviewed the extension lastly and ask them to +2 / merge / deploy it [20:22:00] Wikinaut: OK, I got it. [20:23:25] it#s okay. [20:23:30] security has been checked [20:23:55] and I do use this code since March 2012 in all my wikis in production ( 1..8000 users) [20:24:02] 10 9 8 .. [20:24:39] for your infoNew review: Wikinaut; "Patch Set 14:" [mediawiki/extensions/RSS] (master) - https://gerrit.wikimedia.org/r/3925 [20:24:41] gerrit-wm Change merged: Wikinaut; [mediawiki/extensions/RSS] (master) - https://gerrit.wikimedia.org/r/3925 [20:24:47] murphy's law says that since we all reassured each other about how it's probably fine on IRC it's going to make the sun explode [20:25:00] Wikinaut: congrats :-] [20:25:23] ok I am out, I need to buy bottled water and build a bunker in my garden [20:25:33] hashar: heh :P [20:25:52] Make LOVE, not war [20:28:55] New review: Hashar; "Patch Set 1: Verified+2 Code-Review+2" [integration/zuul-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/50179 [20:28:55] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/50179 [22:10:47] ^demon: do you know what the use of https://gerrit.wikimedia.org/r/#/c/48258/3 is ? [22:12:15] <^demon> I hadn't looked. [22:12:51] ^demon: want to look at https://gerrit.wikimedia.org/r/#/c/48940/7 ? :) [22:57:14] New review: CSteipp; "Patch Set 2:" [mediawiki/tools/code-utils] (master) - https://gerrit.wikimedia.org/r/50055 [23:24:02] brion: it seems like I'll need to duplicate a ton of code in hasViewableContent if I want to do the checks there [23:24:12] Reedy: so if I want to make a branch of an extension to go with a core branch in ED, I just use the same branch name as the core? [23:24:38] Ryan_Lane: meh, then put it in the easy place. just a thought :) [23:24:49] * Ryan_Lane nods [23:44:13] Hi, before filing a bug, a question [23:44:43] I installed a fresh core and run the installer with an existing database, but [23:44:53] error "Run me from the command line please." [23:45:16] on installation screen "Eine vorhandene Installation aktualisieren" [23:45:30] (update an existing installation) [23:45:33] WHy ? [23:45:53] In former timers, this showed a dump of text lines in an textarea terminal-like window [23:45:59] not today [23:46:04] regression ? [23:48:51] *ping*