[00:15:14] Was trying to review those new Echo changes, but I've got stuck on this error: http://pastebin.com/mz5HGjMf [00:16:19] (this is latest master core and Echo) [00:16:23] I've basically given up with it tonight, will try to work it out tomorrow [00:16:49] now that echo has moved from the planning stage to the development stage, i need to remove the word "echo" from my irc keywords, i think. [00:17:53] heh. gnight [00:32:08] * marktraceur wishes Gerrit had a way to one-click revert a file in a patchset [08:58:55] New review: Ori.livneh; "Chew on this, wm-bot." [analytics/glass] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/30315 [08:58:55] Change merged: Ori.livneh; [analytics/glass] (master) - https://gerrit.wikimedia.org/r/30315 [09:00:37] aude: looks like we need more docs https://translatewiki.net/wiki/Thread:Support/About_MediaWiki:Wikibase-item-summary-wbcreateclaim-value/it,_MediaWiki:Wikibase-item-summary-wbcreateclaim-novalue/it,_MediaWiki:Wikibase-item-summary-wbcreateclaim-somevalue/it [09:12:50] Nikerabbit: красавчег [09:16:13] pastakhov_: haven't heard that before, but cool [10:09:47] Nemo_bis: https://bugzilla.wikimedia.org/show_bug.cgi?id=41680 [10:10:12] put it into bugzilla so it's on our todo and someone looks at translation documentation [11:13:26] hmm... git review wants to be updated, but the update script fails. [11:13:26] "error: could not create '/usr/local/man': File exists" [11:13:28] (well, duh) [11:13:32] anyone else have this problem? [11:38:02] DanielK_WMDE: I didn't even try updating yet... last time I did that everything broke horribly [11:38:13] I guess I need someone to debug our AbuseFilter on zhwiki [11:38:14] I can't get some filters work [12:25:37] updates didn't break anything here apparently (but don't trust me) [16:10:20] Reedy: around ? ;-° [16:10:39] Reedy: anomie has patched up our mediawiki-config files to get rid of $cluster : 30792 [16:10:42] https://gerrit.wikimedia.org/r/#/c/30792/ [16:10:54] I guess we could deploy that monday [16:11:03] (I dislike doing production stuff on friday) [16:22:01] zeljkof: I'm told you can get to the other channel now [16:22:18] apergos: thanks, will try [16:22:39] james f set it up [16:23:38] apergos: no, still does not work [16:23:39] hm [16:24:28] passed on the message [16:24:28] zeljkof: What doesn't work? [16:24:46] James_F: I can not join #wikimedia-staff [16:25:01] zeljkof: As in, does it say you're not on the access list, or that you cannot join without an invite, or something else? [16:25:18] James_F: It says the channel is invite only [16:25:35] James_F: The room "#wikimedia-staff" on "irc.freenode.net" is invite only. [16:25:56] zeljkof: Oh, whoops. I typo'ed. Try again? [16:29:02] James_F: still no luck, the same error message [16:30:08] zeljkof: Now? [16:33:51] James_F: I am in, thanks :) [16:52:51] ugh. It looks like the multiversion subdirectory from operations/mediawiki-multiversion is supposed to be placed into the root directory of operations/mediawiki-config to function properly. Is there any way to have it actually work with both being pulled from git, or does one of the two have to be manually copied into place? [17:07:22] I guess you could actually submodule it... [17:10:04] Reedy: But you'd need to move things around after cloning, because it's a subdirectory of one project being moved into the root directory of another project [17:11:46] fix the repo then [17:11:54] move it up a level [17:12:10] mediawiki-multiversion.git having a root folder of multiversion is daft [17:13:50] "daft" is a very good way to describe it [17:15:15] As long as someone "fixes" production too, it shouldn't be an issue [17:17:15] How does production do it? [18:35:12] spagewmf: Thanks for adding the vim setup to the mw.org/wiki/CC page, it has made the vim transition a little easier for me :) [19:42:26] not sure if this is the place to ask: if I abandon a change in gerrit, it doesn't seem to erase the commit I made for that change in git. now I want to git-review a new commit, but the old abandoned commit is still there, and I don't know how to remove it. [19:44:09] chrismcmahon- I may be wrong, but I think the thing to do is create a whole new changeset, since you abandoned the old one. But let's see if anyone more familiar with how gerrit works replies. [19:59:03] I guess I could git reset to before the abandoned change, but that seems a little heavy [20:00:05] chrismcmahon, just 'git checkout master' away from the branch you made the change in? [20:01:01] Krenair: heh. I made and abandoned the change in master. guess it's one of those "don't do that!" things [20:01:30] I don't commit in master for precisely things like this :p [20:01:32] I used to [20:03:09] yeah, intuitively one would think that abandoning a commit in gerrit would erase that commit back down the line upon say "git pull" but that doesn't seem to be the case [20:03:39] I always seem to be starting changes in master. Fortunately "git checkout -b bug/xxxxx master" doesn't complain about uncommitted changes, and just carries them over to the new branch. [20:04:00] chrismcmahon, why would anyone think that.. [20:05:18] Krenair: you mean, why would *I* think that :-) [20:07:47] you can probably get away with using reset, but what I did was delete my copy and clone again [20:09:33] qgil: I'm abstaining--ohloh hides bits of their code [20:09:46] yep, fixed it with a new clone [20:10:22] marktraceur, hi, I was suspecting the possibility given your email address. All fine. You are an independent and not Wikimedia Foundation employee, right? [20:10:29] :) [20:10:45] qgil: I'm an employee, actually, does that change the landscape? [20:11:19] marktraceur, is the first time you send a commit to MediaWiki core? I'm just trying to verify Ohloh's claims for new contributors [20:12:20] qgil: Sorry, I didn't understand the question--I've committed pretty infrequently to core, but it happens, and has happened as early as June or July of this year [20:12:36] marktraceur, well there is this question of WMF development vs independent - 3rd parties. One way to calculate that is finding out who is actually contributing code as WMF employee [20:13:00] marktraceur, but this is all opt-in of course [20:13:43] marktraceur, maybe you used a different username or email address for previous commits? Ohloh plays safe and treats any change as a potential new contributor [20:14:23] qgil: It's possible, actually, that I committed from my desktop and that it has a different value for my email address. Let me check! [20:14:34] marktraceur, but bottom line: no worries. I just want to make sure contributors are aware of our metrics. If you are aware then you are free to do whatever you think suits best :) [20:16:32] qgil: Fair enough! I think I recently *merged* my first commit into core (as in +2 + submit), so maybe the Gerrit interface has a different email address for me. [20:16:48] Hm, nope. Odd. [20:16:53] (Usually it is good practice to contribute work related code from the email address of your employer. It's not just attribution, leaves also the legal side clearer. But in the WMF there seems to be plenty of exceptions, almost a counter-norm. :) [20:17:01] Oh, but it does have a different *username* [20:17:52] qgil: Yeah, I also tend to steer away from the wm.org email address for much the same reason :) [20:18:08] qgil: Also, I registered on Gerrit when I was a contractor. So there's that as well. [20:19:18] marktraceur, steer away, WMF should be _safer_ contributing from WMF addresses, isn't it. [20:19:32] Anyway, not my call or even my problem :) [20:20:12] qgil: I mean, reasons['I don't use ohloh'] == reasons['I don't use google email'], or at least very similar [20:20:44] But it would be good to know if there is a policy, recommendation in place, or if this topic of commits from WMF addresses was ever discussed. [20:20:56] marktraceur, ah, I get you know. Sorry for the misunderstanding. [20:20:59] Sounds like a good thing to clear up, yes :) [20:21:12] haha, I even applied for shell access from my gmail address:) [20:21:13] I [20:21:18] ll ask at wikitech [20:38:06] * marktraceur is excited to cause a wikitech thread [20:43:12] marktraceur, oh, if you were the first one I saw... :) [20:44:01] average_drifter: around ? :-] [20:44:18] average_drifter: drdee pinged me to get your Jenkins access enabled :-] [20:53:52] hashar: hey man [20:53:54] yeah I'm here [20:54:04] I'm gonna try it again now [20:54:36] hashar: thanks man, it works ! [20:54:42] average_drifter: you are the best : -] [20:54:44] I can now finally employ some serious testing [20:54:56] average_drifter: ho I am Antoine Musso [20:55:10] hashar: I suspected, thanks Antoine [20:55:13] been maintaining our jenkins setup for a year or so [20:55:24] I am in France GMT+1 currently [20:55:33] barely connecting during the evening though :/ [20:55:55] average_drifter: what are you going to work on ? ;-] [20:56:36] hashar: wikistats tests [20:56:45] hashar: and wikistats [20:56:45] hashar: squid scripts to be exact [20:57:00] average_drifter: nice :-) [20:57:16] average_drifter: if you need any information about our setup, let me know [20:57:20] ^demon might help too [21:06:59] hashar: ok thanks :) [21:42:06] kaldari, imo https://bugzilla.wikimedia.org/show_bug.cgi?id=41713 should be considered part of notification deduplication [21:42:45] qgil, now you login [21:43:11] I wanted to send back to you the same question you are asking around :P [21:43:11] hi Platonides [21:43:34] Platonides, ? [21:43:46] With which hat are you acting? (volunteer/working for company X...) [21:44:13] Platonides, :) these days I'm a volunteer [21:44:44] Platonides, https://flors.wordpress.com/2012/08/27/moving-up-in-the-freedom-stack/ [21:44:49] I first thought you were working for Ohloh [21:45:03] in this one case, I think it would be easier just to suppress the event, since we would never need to do anything with the information, i.e. it would never take priority over a simultaneous event, and we wouldn't ever need to combine it with another event. [21:45:15] Platonides, er... not at all [21:46:19] your mention of Ohloh admins was odd in such case :P [21:46:21] Platonides, I just think Ohloh is useful providing some metrics we are kind of missing [21:46:56] Platonides, Ohloh admins, you can go and ask them. They respond. [21:47:25] Platonides, one of them was in a oss-metrics mailing list and there everything sarted a few weeks ago :) [21:47:28] started [21:48:32] kaldari, we should have a list of notifications in a certain order, check it when a new event comes through and only allow a notification for the highest one [21:49:12] Platonides, here: http://lists.wikimedia.org/pipermail/wikitech-l/2012-October/063879.html [21:49:45] Krenair: we're working on a spreadsheet for that in Google Docs to figure out the priorities [21:50:16] Krenair: in certain cases we want to combine the information though [21:50:32] kaldari, is this https://docs.google.com/spreadsheet/ccc?key=0AookVJK45wXddHJ3anhPLXJyNFJRM0N2YUdGREdEdlE#gid=0 ? [21:50:37] like "Your article was reviewed and tagged for clean-up" [21:51:19] yep [21:51:39] take a look at Sheet 7 [21:51:51] it's a bit messy right now though [21:51:59] I notice some of these are very enwiki-centric [21:54:46] yes, the nominate for deletion and tagging are [21:54:51] and reviewing for now [21:55:39] All of the Started page - * ones are except for Linked [21:56:24] Also we missed the qqq documentation for a couple of revert event messages, this has been fixed on translatewiki [21:57:16] oh good [21:58:22] someone else already partially fixed one and that has already gone into the via through i18nbot [21:58:40] gone in via i18nbot*