[00:13:29] Special:CategoryTree is broken [00:13:30] http://en.wikipedia.org/w/index.php?title=Special%3ACategoryTree&target=Meteorites&mode=all&namespaces= [00:13:40] gives a database error [00:36:53] wolfgang42, filed: https://bugzilla.wikimedia.org/show_bug.cgi?id=43253 [00:37:02] My guess is it's our db setup, not the extension itself. [00:37:08] But it can be reclassified if it is. [00:56:41] gn8 folks [01:00:44] anyone still deploying? [01:03:17] kaldari: who's on the calendar for deploy? [01:03:24] no one right now [01:03:28] right [01:03:31] but who was last? [01:03:36] scap takes ages [01:03:37] lightning [01:03:48] which is conveniently unspecific [01:03:50] may want to ping those people directly [01:03:53] heh [01:04:17] I think mobile was the last official one [01:04:23] I'll check the logs [01:05:37] yeah, looks like scap finished [01:06:02] awjr: you guys are totally done, right? [01:06:22] we didn't have time to finish our deployment earlier, so we aborted [01:06:52] nevermind, I think arthur is at House of Shields by now [01:11:01] kaldari: i may need to touch/sync a file shortly, but otherwise we're clear :) [01:11:37] ok, thanks [01:40:20] kaldari: i need to touch/sync a file - will that be ok? [01:40:38] yes [01:41:27] thanks! [02:47:45] about to run scap [05:17:27] aharoni: how's Priyanka doin'? [05:18:38] argh, have to go. Good luck! [05:52:41] Does anyone know how to do a force push to Gerrit with no common ancestry? [05:53:03] The reason is that I want it to be a fork/branch of an upstream repo on GitHub (Guiders). [07:09:41] * Nemo_bis is still interested in aharoni's answer to sumanah's question :) [13:59:32] hi there - http://ganglia.wikimedia.org/latest/graph_all_periods.php seems to have gone 404 [14:00:01] index script is in there, but this one went missing somehow [14:10:58] Logged https://rt.wikimedia.org/Ticket/Display.html?id=4165 for what Joy mentioned [17:31:23] Anyone know a JavaScript variable that tells you someone just saved an edit (besides wgPostEdit)? [17:44:33] superm401: no [17:44:49] that's why we made the extension set a global var [17:45:15] Yeah, but E3Experiments doesn't depend on it, so it's logically incorrect even though it happens to work. [17:45:18] because we figured the need for detecting post-edit state client-side is one that others were likely to have too [17:45:46] It will be added to the page, but we don't know the order. [17:46:19] ah. that's because the code used to live in E3Experiments before p12n [17:46:39] well, it has position: top [17:46:58] just declare the dependency. [17:48:16] That won't work, since then ext.postEdit will always run. [17:49:24] it will work [17:50:17] because loading the extension isn't the same as showing the message [17:52:50] You're right. It's still kind of blech, though (depending on a whole client-side script to get one server-side variable that lots of scripts could use). [17:52:52] I kind of want to add it to core. What do you think? [17:53:04] ^ ori-l [17:53:39] it's a hack, but i'm not sure there's a better way to do it [17:53:53] you can add a query param to the post-edit URL, but people hated that [17:54:11] so maybe it belongs in core. Krinkle, opinion? [17:54:24] * ori-l has to run for a bit. [18:20:34] AaronSchulz: ^demon|lunch Any chance one of you could do the wmf6 changeovers tonight? [18:20:53] <^demon|lunch> Sure, which batch is it? [18:21:02] tonight? :) [18:22:36] Rest of the 'pedias [18:23:16] Shouldn't be simple [18:24:25] Thanks [18:26:39] should or shouldn't? [18:27:51] ori-l: ? [18:28:30] <^demon|lunch> What time? [18:29:09] Starts in 31 minutes [18:29:18] <^demon> mmk. [19:09:55] ^demon: I guess you're doing the *pedia switch today? [19:10:38] <^demon> Just finished it. Nothing seems to be melting so far. [19:12:39] <^demon> [19-Dec-2012 18:19:40] Fatal error: Invalid host name (wikipedia), can't determine language. [19:12:39] <^demon> at /usr/local/apache/common-local/multiversion/MWMultiVersion.php on line 355 [19:12:45] <^demon> ^ That doesn't sound great though. [19:13:07] <^demon> Or "[19-Dec-2012 19:12:43] Fatal error: wikiversions.cdb has no version entry for `DB connection error: No working slave server: Unknown error (10.0.6.48)`." [19:13:23] <^demon> AaronSchulz: Why would wikiversions be using a db error for a wiki name? [19:13:54] ^demon: maybe something called nextJobDB and it gave back and error and that thing just passed it on to runJobs? [19:14:30] <^demon> It came from tmh1. [19:14:55] it may do something similar [19:15:09] actually those are just transcode job runners [19:41:06] hey [19:41:23] 'mod' in #expr does not truncate numbers anymore, and this is breaking things [19:41:49] how should I deal with it? [19:42:06] example: coords template on https://ru.wikipedia.org/wiki/Аморгос [19:47:45] hi! [19:48:16] was there some software upgrades @ ukwiki & ruwiki recently? [19:48:49] <^demon> Yes, all wikipedias were updated to 1.21wmf6 today. [19:52:27] coordinates tempaltes are broken now in some wikis [19:52:32] i.e. http://ru.wikipedia.org/wiki/%D0%A2%D0%BE%D1%80%D0%BE%D0%BD%D1%82%D0%BE [19:52:50] or here: http://uk.wikipedia.org/wiki/%D0%A2%D0%BE%D1%80%D0%BE%D0%BD%D1%82%D0%BE [19:52:56] Ahonc, I’m on it already [19:53:11] but hey, it’s 'mod' that misbehaves! [19:53:26] <^demon> There was a change to fix a division-by-zero case in this release. [19:53:26] O_O [19:53:37] <^demon> https://bugzilla.wikimedia.org/show_bug.cgi?id=35866 and related in gerrit. [19:53:55] enhydra, do you have a test case on how it misbehaves? [19:54:37] {{#expr:5.5 mod 6}} should give 5, it gives 5.5 [19:55:59] all help pages say that 5 is the expected behavior [19:56:21] enhydra: huh. why should that give 5? [19:56:26] Reedy, see what you’ve done! :) [19:56:51] <^demon> Reedy is not around right now. [19:57:31] valhallasw, per http://meta.wikimedia.org/wiki/Help:Calculation and http://meta.wikimedia.org/wiki/Help:Mod,_round,_floor,_ceil,_trunc [19:58:20] <^demon> On the meta page, what about {{#expr:30.5mod7.9}} 6.8? [19:58:26] enhydra: ah, so the definition was that it should misbehave, and it's behaving now. At least that makes sense :-) [19:59:04] ^demon, it’s rendered by 1.21wmf6, isn’t it? [19:59:16] <^demon> Yes, all wmf wikis are on 1.21wmf6 now. [20:00:17] valhallasw, but many crucial templates rely on this [mis]behavior [20:00:32] should I rush to enverbose them all? [20:01:29] enhydra: heh. yes, that was to be expected - even though I'm not sure what extra uses people get from this crazy definition... [20:02:24] valhallasw, for example, converting decimal degrees to dms form is shorter with it [20:02:46] (and that’s what I used in coord template that is now in many wikis) [20:03:07] ah, of course. that makes sense [20:03:45] (…and is now broken on all of them!) [20:03:45] essentially everywhere where you want int(a mod b) [20:04:15] I couldn’t even think of non-int b [20:10:30] this also broke the main page for us [20:10:43] (automatic calculation of featured article) [20:12:10] ^demon, as it’s breaking so much, could you revert this fix for now? [20:12:10] php % returns an int, so we should do the same in the template [20:14:35] <^demon> I'm seeing what version we have deployed & what version is in master. I think there's some more changes beyond what we branched. [20:14:57] there is the change from float to double [20:15:19] but that's also wrong here [20:15:35] <^demon> Pfft, it was already reverted in the following commit. [20:16:18] readded on I32c9eca6 [20:17:11] <^demon> Yeah I see it now. [20:17:37] <^demon> So, deploy d96a17c maybe? [20:18:33] uh? that's a i18n change [20:20:27] <^demon> Yeah. Last change before the divide-by-zero changes. [20:22:31] yes, please do so until the fix itself is fixed [20:23:30] I support that [20:29:41] <^demon> Done. [20:32:32] excellent, thank you!