[00:15:22] !ask [00:15:23] Please feel free to ask your question: if anybody who knows the answer is around, they will surely reply. Don't ask for help or for attention before actually asking your question, that's just a waste of time – both yours and everybody else's. :) [00:16:39] but seriously though, anybody know why Special:Statistics says 684,361,502 edits on Wikipedia, but count(*) on revisions is 546,626,420 [00:16:44] that's a pretty big discrepancy [00:20:25] mahmoudhashemi: !shrug [00:20:39] hahaha [00:21:01] mahmoudhashemi: Because stats lie. [00:21:01] <^d> mahmoudhashemi: Those statistics are nothing more than a rough guess. [00:21:08] ^d: No. [00:21:14] <^d> Yes [00:21:16] No. [00:21:16] mahmoudhashemi: you can answer your own question by rephrasing it a little [00:21:19] !didUknow that wikipedia grows at a rate of appx 0.12% per week? [00:21:25] <^d> Yes [00:21:34] ^d: What's a rough guess? [00:21:35] over 100x the rate of the average american savings account [00:21:37] "Why do we not run count(*) on revisions on Special:Statistics"? [00:21:44] mahmoudhashemi: although latest revid I saw for a new edit on enwiki was 592.252.096 [00:21:45] because that is very slow [00:21:57] 1 row in set (5 min 53.45 sec) [00:21:57] Special:Statistics uses stored values. [00:22:07] And over time they've drifted. [00:22:11] <^d> The metrics have changed over time too. [00:22:18] Not really. [00:22:23] <^d> Yes they have. [00:22:25] Not really. [00:22:27] <^d> Wanna bet? [00:22:29] Maybe one metric has changed. [00:22:35] Can you name more than one? [00:23:14] mahmoudhashemi: It's difficult to count edits. The revision table includes non-edits, per se. For example, page protections and page moves. [00:23:54] yeahhh, i hear ya [00:24:20] i was mostly curious if maybe filters or something were getting mixed in to that counter [00:24:22] We use a stored statistic that gets incremented on page edit. [00:24:32] But why is then the Special:Statistics value HIGHER instead of lower compared to count(*) [00:24:40] Edits are deleted. [00:24:44] The revision table contains live edits. [00:24:51] The archive table contains deleted edits. [00:24:59] Count the rows in both tables. :-) [00:25:43] seems legit, thought about only 2min ago^^ [00:26:15] 140million deleted edits? [00:27:35] mahmoudhashemi: Maybe! [00:27:44] we shall soon see [00:27:45] ^d: What are we betting? :-) [00:27:59] only 420k [00:28:12] mahmoudhashemi: What? [00:28:13] haha wait [00:28:23] 42 mill [00:28:25] <^d> Gloria: I lost interest :p [00:28:26] I've personally deleted more than 420K pages. ;-) [00:28:48] <^d> I deleted > 1mil jobs before ;-) [00:29:15] so there are still 100 million edit counter increments unaccounted for [00:29:38] do null edits count? [00:30:23] No [00:32:21] <^d> Gloria: mid-air collision! [00:32:27] <^d> I was in the middle of explaining how it was setup ;-) [00:33:05] Heh. [00:33:09] I collided twice. [00:33:27] ^d: How do I see the members of the ldap/wmf group? [00:33:42] Wouldn't that be all wmf employees? [00:33:50] <^d> Nope. [00:33:55] That's confusing! [00:34:11] <^d> (more confusing, OIT has their own ldap too ;-)) [00:34:30] it's a hidden group [00:34:39] <^d> Anyway, ldap groups aren't terribly public. [00:34:41] <^d> Which sucks. [00:34:45] ^d: People like Steven and Terry Chay have +2. [00:34:52] And they're not full-time engineers. [00:34:58] So I'm confused. [00:35:49] Eh? [00:35:50] https://gerrit.wikimedia.org/r/#/admin/groups/11,members [00:36:02] oh, duh [00:36:03] ldap/wmf [00:36:05] <^d> ;-) [00:36:08] * Gloria pets Reedy. [00:36:14] Y U NO CLEAR [00:36:18] <^d> This data's not really private. I can pastebin it somewhere. [00:38:36] <^d> Gloria: https://gerrit.wikimedia.org/wmf-20140124.txt [00:39:02] ^d: Thanks. :-) [00:39:06] <^d> yw [00:39:55] bharris, swalling, tychay, aklapper, [00:40:03] sup dawg [00:40:16] StevenW: Looking at the +2 list. [00:40:19] y u no code review? [00:40:29] StevenW: Sorry for the pings. :P [00:40:53] Guillaume? [00:40:56] Bizarre. [00:41:00] dsc, heh. [00:41:12] <^d> Yeah, this needs cleaning. [00:41:14] <^d> I shall do that now. [00:41:23] <^d> (Consider any former wmf-ers now purged) [00:41:25] Mmm [00:41:59] * Gloria installs wget. [00:42:05] gj [00:42:28] Is it actually impossible to find commits via gerrit search that someone merged, rather than merely were a reviewer on which were merged by someone [00:42:30] ? [00:43:02] StevenW: yes [00:43:08] incredible, i know [00:43:36] you can find them with some git log grep hackery if you really really want to [01:03:10] James_F|Away: Were you able to purge that page? [01:25:55] hai [01:30:30] Hey there can anyone help me with AWB database dumps? [01:32:25] AWB database dumps? [01:33:12] Yeah I'd like the January Database dump of Wikipedia processed for the WP:AWB/Typo list [01:33:14] if that's ok [06:50:55] hello [06:52:28] I seem to recall a bug filed about changing the behavior of prefix: in Special:Search, can anyone find that for me please? [07:04:38] !bug summary:search summary:prefix [07:04:38] https://bugzilla.wikimedia.org/buglist.cgi?quicksearch=summary%3asearch+summary%3aprefix [07:05:00] huh, is the bot down? https://bugzilla.wikimedia.org/buglist.cgi?quicksearch=summary%3Asearch%20summary%3Aprefix [07:05:05] any of these? [07:07:23] Thanks, i found it at https://bugzilla.wikimedia.org/show_bug.cgi?id=24775 [07:07:46] hm, according to that though it's fixed by the new CirrusSearch [07:07:56] i've just had a problem with it [10:31:41] mahmoudhashemi: no vhost? that's ghetto. there's a dress code, you know. [10:39:43] * Nemo_bis joins ori in glaring disapprovingly [10:48:05] ori: good morning [10:48:09] or evening [10:48:33] hello [10:53:55] ori: https://test.wikipedia.org/wiki/User:Gryllida/common.js feedback please? [11:00:10] gry: i'll try to look at a moment but it is very, very late [11:00:29] ori: I tried to reach you all week. if you name a time, I'll do everything to be here then [11:07:14] gry: well, because i'm always reviewing this when i'm very tired, it's always something random that catches my eye, so i'm not sure if it's useful to bring it up [11:07:34] ori: would you like to look at it later and ping me with feedback? [11:07:40] I idle here. [11:07:44] but you should use a tool like jshint; it'll help you write in a consistent style and it will also flag a lot of small code quality issues [11:07:52] http://www.jshint.com/ [11:08:23] Ok, thanks, I will look into it and what it says. [18:40:56] apergos: would mirroring the 25 TB of Commons tarballs on a WMF server cost more or less than 2300 $/month? http://calculator.s3.amazonaws.com/calc5.html#r=DUB&s=S3&key=calc-3CED90FC-C841-421A-A0E7-A2BC66EBBC47 [18:41:43] I couldn't find anyone providing this service for free yet, who knows why :P [18:41:54] I don't know but I have been told that's a waste of our space and that we should provide tarballs on demand [18:42:32] sounds unrealistic [18:44:02] for about 1000 € I could probably do that from home ;) [18:44:33] (just the cost of disks) [19:12:18] sorry, 1200 € [19:12:59] (recycling a PSU and case I have already and my home fiber connection, or "stealing" an ethernet port at university) [19:55:41] oh well https://meta.wikimedia.org/wiki/Grants:IdeaLab/Commons_tarballs_seedbox