[09:16:11] FYI: we're spreading ZIM (HTML) dumps of it.wiki at a bandwidth that is more than twice what download.wikimedia.org takes. [09:17:00] If you have a good upload bandwidth (10+ Mb/s ?) and can share a 16 GiB file on bittorrent that would be appreciated. http://download.kiwix.org/portable/wikipedia_it_all.zip.torrent [09:30:35] ah [09:30:48] I wish we had support torrent feeding :-] [09:35:44] hashar: archive.org does it ;) [09:35:54] Nemo_bis: great! [09:36:06] so they have millions of torrents [09:36:08] Nemo_bis: so they receive our dumps and then feed them ? [09:36:15] hashar: yes [09:36:21] awesome [09:36:35] but the upload to them is not automatic [09:36:57] and it's not possible to use the same info_hash because their torrent includes also their metadata files so it's slightly differnt [09:37:32] the nice thing is that if one finds a torrent worth preserving they can just upload it to an item and its content gets downloaded by them :) [09:37:38] upload_by_torrent :p [10:31:23] I cannot access Wikipedia from India. Is that an issue? [10:34:45] vibhav: are you at the language summit? maybe siebrand knows :) [10:35:03] Nope [10:35:15] Unfortunately :( [10:36:09] quite often it's an ISP problem so traceroute and so on are useful infos for those who can help you (note me) [10:39:40] Nemo_bis: There is a ??? in the middle [12:37:50] Can someone tell me whether it is possible to have [[Special:LonelyPages]] configured to ignore certain namespace? Even to know where it is configured would be helpful [12:38:46] <- trying to comprehend inconsistent output [12:39:17] 'page_namespace' => MWNamespace::getContentNamespaces(), [12:39:27] so it's content namespaces [12:39:39] not configurable (currently) without changing those [12:40:44] okay, at enWS, it has selective list of say 100 pages from the Index: ns, some from Page: ns + others [12:41:13] do you want to remove some namespaces or to be able to easily see the namespaces you're interested in? [12:41:26] was thinking that they would not show [12:41:40] firstly was trying to comprehend what was showing ad why [12:48:11] do you now? [12:48:15] we have https://bugzilla.wikimedia.org/show_bug.cgi?id=8516 btw [12:48:52] * sDrewth apologises for thinking that knowledge might help [12:49:04] * sDrewth face palms [12:50:56] ...? [12:51:06] it's not clear to me if it's all clear to you now :) [16:12:14] AaronSchulz: did something change in job queue? https://bugzilla.wikimedia.org/show_bug.cgi?id=42715 [16:12:25] This, if confirmed, is a very urgent bug. ^^^ [16:12:52] robla: ^ [16:15:10] please ping Aaron about this when he comes online [16:15:30] I was just about to sign off here, and I'm going to be at an offsite meeting for the next three days [16:16:57] will it involve a spa and massages? [16:21:37] Reedy: ^demon sorry i missed rob but who can decide if/when we try wikibase again for enwiki? [16:21:53] well, not if, but when [16:22:14] As long as we have ops around, I see no reason we can't try again in the wmf9 window [16:22:19] ok [16:22:41] the memcached thing that happened yesterday is okay now, right? [16:22:55] * aude not sure how it was resolved, if a server died or what [16:23:58] * aude probably wants to eat before coming back to do deployment, though [16:24:15] more fooood for the kittens [16:24:19] heh [16:24:27] meta's job queue is also suspicious [16:24:29] * aude survived on cakes on monday [16:26:03] !log Job queue weirdness and TranslationNotifications duplicates again noticed on Meta (bug 42715) [16:26:05] Logged the message, Master [16:26:13] more than this I don't know what to do :) [16:26:30] job queue probably requires aaron to look at it [16:33:57] hi [16:34:16] is it possible to enable transwiki import at uk.wikinews? [16:34:30] and in uk.wikiquote too will be nice [16:39:07] Base-w, you need to add the request to our issue tracker at http://bugzilla.wikimedia.org (with a link to community consensus) [16:40:49] ok if only consensus requred it is ok :) [16:42:02] You can ask again here if you need further help :) [16:56:39] will it be possible to anable moving to transwikiimportes group by sysops? [16:57:17] or some policy allow to it only from meta? [17:07:27] Base-w: hm? all sysops have import right [17:55:31] Nikerabbit: can you fix TranslationNotificationJob::run ? [17:56:34] Nemo_bis: yes i want to have opportunity to give it to nonsysops [18:01:32] Base-w: rarely useful, indeed ask stewards; otherwise you'd need some local policy [18:02:29] ok i will ask community about creating policy [18:11:39] Base-w: why don't you just elect them sysops, poor souls [18:11:53] we have one sysop [18:11:54] me [18:12:10] and i have it temporary [18:12:26] AaronSchulz: in gerrit [18:15:06] Base-w: ah, but then you'll have to ask stewards anyway if you don't have bureaucrats, what's the point? [18:15:22] or maybe sysops could add the flag, but what a mess :) [18:16:37] now i could grant a flag to trusted users (if i will have a possibily) [18:17:20] Nikerabbit: heh I was about to comment before I saw ps2 [18:19:17] Nikerabbit: can the extension be updated to master in wmf9? [18:21:47] AaronSchulz: yes please [18:21:57] there hasn't been any other changes except l10n as far as I know [18:27:48] Hey preilly, was your -1 last week on https://gerrit.wikimedia.org/r/#/c/48071 just about the release notes? Or is there another reason to not merge that patch? [18:28:04] csteipp: yes it was just about the release notes [18:28:11] Cool. Thanks! [18:28:15] csteipp: np [18:28:59] Nikerabbit: done, thanks [18:30:50] AaronSchulz: let's hope this fixes it [18:31:56] well assuming there are no duplicate jobs left it should be fine [18:50:57] Nikerabbit: I also noticed https://gerrit.wikimedia.org/r/#/c/48852/ [18:51:16] is that urgent too? [18:51:58] Nikerabbit: is it ok to merge and update that ext to master too? [18:54:12] AaronSchulz: it's not urgent imho [18:54:41] AaronSchulz: perhaps better to not update it master right now, there's small instability there [18:55:05] ok [18:58:46] Hey Reedy, are you doing the deployment today? [19:01:38] ^demon: https://gerrit.wikimedia.org/r/#/c/48860/ [19:15:43] Reedy ^demon AaronSchulz: who's handling the deployment today? [19:16:55] AaronSchulz: is the change job notification returning null critical for us to fix and backport? [19:17:48] aude: I'm merging the sanity work-around in core anyway [19:17:51] so it won't matter soon [19:18:00] AaronSchulz: ok [19:18:15] we can fix it for the next regular deployment (wmf10) [19:18:43] gerrit acting up doesn't help though [19:18:50] :( [19:20:13] merging into wmf is much slower and clunkier now [19:22:08] ugh [19:23:33] Reedy: around? [19:29:53] AaronSchulz: I haven't heard from him in a while.. are you ok to do the last deployment when you get those fixes merged? [19:30:36] I am now [19:30:41] Reedy, https://bugzilla.wikimedia.org/show_bug.cgi?id=40759#c12 - is that (point #3) a change which needs the namespaceDupes maint script run? [19:30:45] I've just had to drive 50 odd miles in the snow :) [19:32:55] Reedy: too slow :) [19:33:21] Took less than an hour [19:33:21] Fairly impressed [19:33:49] how much snow? [19:34:21] at least in the US, some east coast people couldn't open their doors a few days ago, heh [19:34:28] just over an inch here [19:34:33] but britain falls apart [19:35:05] I recall James_F saying something to that affect [19:35:12] I would kinda be amused to see SF freak out with an inch of snow... [19:35:28] AaronSchulz: Yeah, London deals better. [19:35:47] csteipp: http://foundsf.org/images/4/47/Mission$snow-in-mission-1887.jpg [19:36:05] I was going to say [19:36:12] I thought it'd happened before [19:36:33] http://imgs.sfgate.com/blogs/images/sfgate/parenting/2011/02/16/snow_one495x285.jpg [19:36:35] another good one [19:36:37] Haha, I hand't seen that one.. [19:36:42] hard to imagine [19:37:48] I'd love snow here. [19:38:04] Right up to the point where I'd have to walk up Nob Hill in the icy conditions. :-) [19:39:32] who is able to change fields directly on db? there is a situation with voy, are you aware? [19:40:12] "a situation" [19:40:15] Can you be any more vague? [19:40:49] i'm still looking to a discusson in english [19:41:58] basicly, with the import, some aspects of the license weren't accomplish, so seems that it would be need to change the names that appear on the pages historic [19:42:11] Reedy: so you are updating wikibase? [19:42:18] for userX to userx (wikitravel) or similar [19:42:21] Alchimista: That might be me. Can you point me to what you're seeing? [19:42:42] Krenair: Looks to be, see my comment [19:43:09] csteipp: found. here's one discussion, wich can ilustrate the problem:http://en.wikivoyage.org/wiki/Wikivoyage:Travellers%27_pub#Attribution_at_the_bottom_of_the_page [19:44:19] AaronSchulz: possibly, yeah [19:44:21] http://en.wikivoyage.org/wiki/Wikivoyage:Travellers%27_pub#Attribution_at_the_bottom_of_the_page (clickable link) [19:44:35] Reedy: ok [19:45:24] Krinkle: re: "(FR #615) Skrill/Moneybookers integration" trailing whitespace -- I'm not seeing much (only in skrill_resultswitcher.body.php)... so I'm a bit confused where that came from. [19:45:33] csteipp: in pt.voy is beend discussed how to change the usernames, i could do it with a bot, but it would be needed to create the accounts and rename them, if there would be a simpler and less agressive solution.. [19:45:46] !log test for timestamp. [19:45:48] Logged the message, Master [19:45:52] Krinkle: or rather -- I dont think I know what you're complaining about [19:46:01] That's dead on. [19:46:27] Alchimista: yeah, it looks like Philippe is looking into the legal side (and he's the one who would clear this). I'll ping him about it directly too. [19:46:30] mwalker: https://gerrit.wikimedia.org/r/#/c/23160/8/skrill_gateway/skrill_resultswitcher.body.php [19:46:51] mwalker: Looks like some of it got cleaned up, latest version (#8) only in this file ^ [19:48:09] mwalker: I was more commenting as a general habit. Just because it was fixed, doesn't solve the problem. I have no idea whether this is the first, but assuming it wasn't my comment was to the purpose of "Be aware of this in your editor and/or pre-commit in the future" [19:48:32] i.e. "git diff" always highlights these errors in bold red [19:48:33] Alchimista: Since this is pt and es specific, do you have a link to those discussions? We have a couple people who can get by in both languages, who may be able to help decide what needs to happen. [19:48:40] and most editors have a way to warn and/or auto-fix it as well. [19:48:49] Krinkle: sure! the extra space at the end of the comment lines I'll agree with -- do you know why the double tab blank lines inside the execute function are marked as whitespace errors -- seems to me that you would actually want this behaviour [19:49:00] csteipp: yap, here: http://pt.wikivoyage.org/wiki/Wikivoyage:Bar_dos_viajantes#C.C3.B3pias_da_Wikitravel_sem_atribui.C3.A7.C3.A3o [19:49:13] mwalker: not really, why would you? [19:49:47] mwalker: afaik if you start typing there in most editors, it sets up indentation properly. One could differentiate between trailing space on empty and non-empty lines but we don't do that. [19:49:58] Krinkle: whitespacing is in some form a context keeper -- so; if the context is currently at 2 tabs, why would you break it? [19:50:21] Right now conventions recommend no trailing space, regardless of what is or isn't on the line. Seems hard to verify how many spaces should be on the empty line for the line is empty. [19:50:54] mwalker: I understand what you mean, you can click on the "empty" line and start adding something [19:52:46] but it's not really a form of preference, it is for consistency across our code base. So in this case though both have their ups and downs (the whole thing is a minor thing obviously), as far as I'm concerned we should do it consistently regardless of what may have a slight advantage. [19:54:07] fair point -- commit hook time? :p [19:54:45] mwalker: I don't know which editor you have, maybe ask some people in the ops/platform team what they use to prevent this. I mean, don't think that I'm manually un-indenting empty lines :-), that'd be crazy. [19:55:05] there's probably a setting somewhere. I know there is in Sublime, and probably for your editor as well [19:55:16] which would automatically make sure empty lines are empty [19:55:35] in this case, it's not my editor; it's adam's -- but if your point is consistency across the code base, lets just have it as a global commit hook into gerrit and stop caring about it [19:55:50] mwalker: right, sorry, I meant his. [19:56:34] mwalker: anyway, nice work. Just noticed in in passing, don't get hung up on it. Something to check for later maybe. At some point we want this to be asserted in Jenkins, so then you'd have to have no trailing whitespace or it won't merge. [19:56:56] But for now, let it be and think about it later. no big deal ) [19:57:02] check [19:57:18] (but but but... bikeshed!) [19:57:49] mwalker: Hm.. interesting, you mean a hook to alter committed code? [19:59:35] sure; why not -- I can't think of an instance where it would matter - even python doesn't care about blank lines [20:24:02] Reedy: when/if it looks like wikidata is stable, can I get help deploying https://gerrit.wikimedia.org/r/#/c/48070/ ? [20:25:30] wikidata looks good [20:27:30] csteipp: sure [20:27:35] Wikidata is looking pretty much fine [20:27:57] If you can +2, then I think I can do the synch [20:28:53] :D [20:37:49] doh, Reedy, I actually messed up on that patch [20:38:16] @seek high speed machining [20:38:46] ? [20:41:34] Done. Forgot to add meta, so that we can actually add rules once we give out permissions. [20:45:47] csteipp, so hang on... where are we with global filters now? [20:46:10] oh you're still waiting for deployment? ok [20:46:15] Code has been deployed for... 6 months or so. Just not enabled anywhere [20:46:31] I meant config deployment [20:47:10] :D [20:47:35] Oh, Reedy, want to +2? [20:50:19] csteipp: I'd probably enable it on test2wiki aswell [20:50:51] I can do that.... updating the notes right now [20:51:07] I'll approve it then [20:58:30] Actaully, Reedy, to Krinkles point, I probably should have a bug so other wikis can track this.. I'll do that and deploy this late. [20:58:56] Reedy: Actaully, I'll just go it :) Thanks! [20:59:54] csteipp: I don't mean to interrupt, it's just easier to have a bug ticket saying like "Review stuff" and "Deploy stuff" so that there's a number we can refer to and a place where we eg. can see that comment from Tim you talked about saying he reviewed it [21:00:11] especially given that comments in Gerrit generally are unreachable and get lost. [21:01:00] Krinkle: I totally understand, and there's no rush on this. Would you prefer I put that all in before I deploy? [21:01:03] enabling with just test/test2 (where one is central and the other the client) is probably a good idea, then create a filter and test it. Then go for the real thing. [21:02:17] csteipp: Assuming this is in fact scheduled/approved, I assume it is okay to go ahead. But sure, for transparancy, it is probably easier to have this in before hand, since you can't amend the commit after it has been merged. [21:05:32] James_F|Away: Are you ok with merging Global Filters? [21:05:41] csteipp: Yes, go for it. :-) [23:45:19] how's the job queue? [23:45:31] flooded? [23:45:55] which? [23:46:36] pt.wiktionary's [23:46:51] 0 [23:46:52] categorylinks don't seem to update [23:47:11] actually templatelinks [23:48:12] Seems to be very few entries [23:49:15] in that case the problem must be somewhere else [23:49:42] i did this edit: https://pt.wiktionary.org/w/index.php?diff=1651057&oldid=1646185 [23:49:56] removed link to Wikcionário:Tarefas [23:50:21] it still shows as linking that page: https://pt.wiktionary.org/w/index.php?title=Especial%3AP%C3%A1ginas+afluentes&limit=50&target=Wikcion%C3%A1rio%3ATarefas&namespace= [23:50:25] (first hit) [23:50:40] 270629 / 10013309 [23:51:20] what does that mean? [23:51:31] job queue is 10M in size? [23:52:07] the above edit was made 38 mins ago [23:54:41] null editing doesn't solve [23:55:41] for todays log file, that many many are ptwikt / total job lines