[01:58:18] New patchset: Rfaulk; "E3 Analysis modules used for ETL of experimental data." [analytics/E3Analysis] (master) - https://gerrit.wikimedia.org/r/24108 [01:58:18] New patchset: Rfaulk; "adding __init__.py to sub-path to enable import for modules under "classes" sub-path" [analytics/E3Analysis] (master) - https://gerrit.wikimedia.org/r/24109 [01:58:50] The analytics repo has announcements in here? [02:02:36] ugh. [06:55:49] New review: Ori.livneh; "Holy fuck, that's a lot of Python. It could use some refactoring, but I don't think Gerrit is the id..." [analytics/E3Analysis] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/24108 [06:56:46] i sometime forget that reviews are echoed over irc [12:39:44] thedj[work]: back around. Looks like we want
to be made
? [12:41:21] hashar: yes [12:41:45] i submitted a patchset [12:41:57] so I will get your patch against 1.20wmf12 and move it to master as well :) [12:47:41] will merge it [12:47:47] and deploy [13:23:29] hashar: it was deployed ? [13:23:40] thedj[work]: yes I did [13:23:46] roughly 30 minutes ago [13:23:50] should have pinged you sorry [13:24:05] rendering test has been made on https://test.wikipedia.org/wiki/Bug40306 [13:24:11] looked fine so I merged it in [13:24:24] hmm, caching must be strong... [13:24:25] https://nl.wikipedia.org/w/index.php?title=Sjabloon:Tabel_tijdlijn_luchtvaartfabrikanten_Verenigd_Koninkrijk&oldid=32272359 [13:24:55] probably the object cache or something similar that needs to expire. [13:25:22] I have no idea where it can be cached :( [13:25:52] maybe that is in the parser cache [13:26:37] http://www.mediawiki.org/wiki/Manual:$wgParserCacheExpireTime [13:26:44] default 24hours [13:27:25] that's the problem with cache :/ [13:28:22] also explains why it took so (relatively at least) long for the first reports to surface... [13:44:08] hashar: i don't get it, i still see the problem.... [13:44:09] https://en.wikipedia.org/wiki/User:TheDJ/sandbox [13:44:58] the div gets margins, but it shouldn't... [13:45:40] bahh [13:46:14] need to put in on wmf11 as well ;:) [13:46:19] Silly gerrit. [13:46:20] https://gerrit-review.googlesource.com/#/c/37920/ [13:46:22] ah duh ! [13:46:31] [13:46:37] Incoming/Outgoing reviews is awful terminology [13:49:34] TomDaley: dashboard have been private? Seriously ? :( [13:49:55] It was never supposed to be public really. [13:50:03] It's why it was exposed via ugly uids. [13:50:12] I liked to see what other people were working on / submitted ;) [13:50:20] You can still search all this. [13:50:26] sure [13:50:28] And add custom dashboard header thingies. [13:50:33] hence I am not sure why it needed to be made private [13:50:51] I guess saved search will fit my needs [13:51:30] Yeah, https://gerrit-dev.wmflabs.org/r/Documentation/user-custom-dashboards.html [13:54:46] ahh [13:54:57] can we save them in the Gerrit GUI ? [13:55:02] kind of a bookmark? [13:55:10] No save within gerrit yet. [13:55:24] will use my browser bookmarks so ;-] [13:55:29] +1 / +1 [13:55:31] But that is probably doable. Cool thing is they're sharable too :) [13:55:34] please merge in production! [13:55:40] It's already coming in 2.5 [13:55:45] I mean [13:55:49] deploy 2.5 NOW [13:55:57] Waiting for 2.5rc1. [13:56:04] rc0 has some regressions I need fixed. [13:57:12] rc0 is on labs already though. [13:57:16] If you wanna play with it. [14:00:04] I should [14:00:25] still thinking about the jenkins/gerrit workflow [14:50:27] "Since dashboards are private now" O_o they won't on our install, will they? [14:55:22] The dasboard/ links are private starting in 2.5 [14:55:37] Actually, that URL construction doesn't work anymore at all. [14:56:58] I think it's supposed to try and guess, but I've not gotten that to work [14:59:33] Actually, I misread that. That behavior changed with custom dashboards. The /dashboard/ query is now for custom dashboards only. [15:02:08] hmpf I guess complaints in the "feature" request were useless [15:05:18] Well, dashboards are sharable links. You can easily construct a dashboard that you want to share with someone. [15:05:49] The concept of "my dashboard" is now private (although it's completely re-createable via custom dashboards) [15:06:32] And user links now take you to owner:Foo queries (which is consistent with other UI elements) rather than the defunct dashboard/ [15:07:39] so I guess we'll have to mass-create shared custom dashboards which reproduce the old dashboard [15:09:14] TomDaley: the dashboards are useful :( [15:09:21] essential, rather [15:09:42] TomDaley: how does the new context affect https://code.google.com/p/gerrit/issues/detail?id=1482 ? [15:10:03] how else can i see my review queue? (i'm sure there's some search query) [15:10:37] aude: It's still accessible from My->Changes like now [15:10:48] * aude sure that denny (my boss) would like to see what the team is doing, etc. [15:10:59] is there a way to do that? [15:11:13] TomDaley: ok :/ [15:11:20] Yes, via custom dashboards. The only thing that's changing is exposing it via dashboard/ [15:11:34] https://gerrit-dev.wmflabs.org/r/Documentation/user-custom-dashboards.html [15:12:03] The site's security certificate is not trusted! (yikes!) ;) [15:12:15] Nemo_bis: Well custom dashboards let you do that yourself. But with the refactoring I think it'll make it easier to fix that for the default self-view. [15:12:22] aude: Sorry, it's self-signed for labs :\ [15:12:26] Haven't fixed that yet. [15:12:54] TomDaley: so it's not superseded/made impossible? :) [15:13:06] No, I don't believe so. [15:13:18] * aude thinks people will get confused [15:13:20] * aude confused [15:13:26] oh, they're used to [15:13:36] * Nemo_bis would like https://code.google.com/p/gerrit/issues/detail?id=873 WONTFIXED [15:13:51] aude: The only reason you find those pages now (unless you bookmark them) is via links to your username in the UI. [15:14:09] Those links now do owner:Foo queries, which is more consistent with other links like that (eg: project, topic, etc) [15:14:26] please aok [15:14:27] ok [15:15:13] what's the timeline for deploying this? [15:17:10] Haven't scheduled a window yet. "Soon" is roughly accurate. [15:17:17] ok [15:17:26] I want the next release candidate out at least before we upgrade. [15:17:41] please don't surprise us though :) [15:18:07] * aude curious what else will change [15:18:18] Lots of cool stuff. Lemme throw the release notes up somewhere for you. [15:18:29] ok [15:19:44] Hrm, can't get in to gerrit-dev :\ [15:21:50] Eh, https://gerrit.googlesource.com/gerrit/+/stable-2.5/ReleaseNotes/ReleaseNotes-2.5.txt is the best I can do at the moment. [15:21:50] TomDaley: some of us would like this: https://code.google.com/p/gerrit/issues/detail?id=1567 [15:22:07] Chad told me to file a bug but I never did so [15:22:25] 1567 won't be possible until lucene stuff is done, at the least. [15:22:42] I can't imagine it going in right now, at least. [15:23:16] why is it so hard? [15:23:26] is lucene also useful for full-text search? [15:24:24] Yes. [15:24:29] Lucene is useful for many many things [15:24:47] which is why it doesn't mean much :p [15:24:58] good to know it will help also on that side here [15:26:07] Lucene allows us to index pretty much everything about a change in a sane way for searching. [15:26:11] TomDaley: where is the info about reviews and comments stored? could the data be replicated elsewhere (like Labs or Toolserver) to be queriable? [15:26:19] good [15:26:43] A combination of being in the database, for now. [15:26:47] Plus some git metadata. [15:28:12] what's in git metadata? [15:28:25] Um, final review values (who +2'd, who submitted, etc) [15:28:29] Comments are in the db. [15:29:07] I'm pretty sure we could replicate the DB. Only table I'd want to exclude is probably account_ssh_keys. [15:29:13] Other than that, there's nothing really private. [15:29:16] so every patchset deletes those git metadata? [15:29:23] oh, nice [15:29:58] TomDaley: thank [15:29:59] s [15:30:18] can I file a bug about replicating the db somewhere then? :) [15:30:19] As far as git metadata, https://gerrit-review.googlesource.com/Documentation/refs-notes-review.html should explain it better. [15:30:28] I don't really know how it works that well tbh. [15:30:31] reading [15:31:02] quite a lot of changes [15:31:06] Nemo_bis: You can toss it in BZ, sure. Might want to throw Asher or Ryan on the CC too. [15:31:13] TomDaley: thanks [15:32:20] You're welcome. And now, I think I'll take an early lunch :) [15:34:34] hmm no more olympics athletes names in the channel [19:05:27] soooo manyyy changes [19:05:30] https://gerrit.wikimedia.org/r/#/q/project:operations/mediawiki-config+is:open,n,z [19:18:34] hi RoanKattouw [20:36:07] hashar: silly questions day, how do I get git diff to show me whitespace changes too (in a readable way if possible)? [20:37:04] doesn't it show whitepsaces already ? [20:37:09] I got errors in red [20:37:33] or do you mean replacing space with dots and tabs with >> ? [20:37:51] hashar: I mean git diff, not gerrit [20:38:00] no idea [20:38:05] hmm ok [20:38:09] thanks [20:38:09] I sometime use 'git diff --check' [20:38:15] will report any WS error [20:38:45] there is that tip : http://stackoverflow.com/questions/5257553/coloring-white-space-in-git-diffs-output [20:40:32] thanks, will try both [23:34:54] Is there a way to get the current server time at page load on the client side? Like mw.config.get( 'wgTimestamp' )? [23:38:47] can't find anything, so I'll just stick it in our ResourceLoaderGetConfigVars