[05:38:18] <TimStarling>	 legoktm: Installer::dirIsExecutable()
[07:09:22] <Krinkle>	 TimStarling: https://gerrit.wikimedia.org/r/#/c/391162/
[08:17:07] <TimStarling>	 wmf prefix?
[08:17:52] <TimStarling>	 I know it is pretty common but I'm not a fan, sorry to nitpick
[08:18:29] <TimStarling>	 originally wf prefix = wiki function, wg prefix = wiki global
[08:18:45] <TimStarling>	 then we introduced wmf = wikimedia function, wmg = wikimedia global
[08:19:07] <TimStarling>	 but I guess some people read wmf as wikimedia foundation and start applying that to other things
[16:12:11] <DanielK_WMDE_>	 anomie: will you be available for the RFC discussion about imported user names tomorrow?
[16:12:35] <DanielK_WMDE_>	 we moved the time to be 2 hours earlier
[16:15:58] <anomie>	 DanielK_WMDE_: Most likely. There's a very slight chance I'll have to disappear for about 10 minutes after the first 10 minutes.
[16:24:30] <DanielK_WMDE_>	 anomie: ok, thanks!
[17:04:10] <anomie>	 "Fun" fact for the day: in JavaScript, '💩'.length === 2.
[17:04:29] <Reedy>	 that's a big poo
[17:21:30] <DanielK_WMDE_>	 anomie: oh shit!
[18:11:12] <legoktm>	 anomie: are you developing MW on postgres these days? :)
[18:14:27] <anomie>	 legoktm: No, but I have PG set up so I can test update.php relatively easily.
[18:14:58] <legoktm>	 :) it's much appreciated
[18:15:19] <legoktm>	 I was gonna ask you to review the ip_changes postgres patch and when I found the link I saw you had already reviewed it :D
[20:35:39] <TimStarling>	 anomie: apparently the two things standing in the way of your idea of a minimum requirement of "PHP 7 or HHVM" are T117534 and T161577
[20:35:39] <stashbot>	 T161577: OOM problems with snapshot1007 due to wikidata dumpRdf jobs taking all 60GB+ of main memory - https://phabricator.wikimedia.org/T161577
[20:35:39] <stashbot>	 T117534: DCAT-AP: XML output no longer valid - https://phabricator.wikimedia.org/T117534
[20:49:30] <bd808>	 TimStarling: We still have to move Wikitech to HHVM as well, but we are working towards that. Blocker right now is time + a database migration from silver to s3
[20:50:08] <bd808>	 Once wikidata is split out to the new s8 I think the DBAs will have time to help us move the db
[21:01:02] <TimStarling>	 the question is: is it worth doing before June?
[21:09:44] <bd808>	 Oh... we are totally planning to move Wikitech by like the end of January or so at the latest
[21:10:49] <bd808>	 dumps might be a good first move candidate for stretch + php7
[21:11:12] <bd808>	 it doesn't have the web interface things to worry about. Just a ton of cron jobs
[21:12:23] <bd808>	 it would be worth asking Ariel about. They are in the middle of refactoring for Cloud Services taking over the data serving side of the dumps process so probably would be Jan/Feb as well
[21:18:24] <Niharika>	 Is there a fixed call precedence for gadget scripts vs user scripts? 
[21:38:01] <anomie>	 TimStarling: It looks like T161577 was tracked down to T162245, and the last comment on T162245 seems like it might be fixed now that we're on 3.18.
[21:38:05] <stashbot>	 T161577: OOM problems with snapshot1007 due to wikidata dumpRdf jobs taking all 60GB+ of main memory - https://phabricator.wikimedia.org/T161577
[21:38:06] <stashbot>	 T162245: Enable GC for HHVM CLI (at least for dump runners) - https://phabricator.wikimedia.org/T162245
[21:51:36] <anomie>	 TimStarling: As for T117534, it looks like it could be worked around by changing the calls that pass null for $url to ->writeAttributeNS( $prefix, $name, $url, $value ) to call ->writeAttribute( "$prefix:$name", $value ) instead, until someone gets around to rewriting it to use Purtle or whatever.
[21:51:36] <stashbot>	 T117534: DCAT-AP: XML output no longer valid - https://phabricator.wikimedia.org/T117534
[21:52:23] <anomie>	 Or, potentially, it could pass the actual URLs to ->writeAttributeNS() instead of null. I don't know if that'll make it repeat the xmlns properties on every element though.