[04:02:44] 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. [04:02:45] T161577: OOM problems with snapshot1007 due to wikidata dumpRdf jobs taking all 60GB+ of main memory - https://phabricator.wikimedia.org/T161577 [04:02:45] T162245: Enable GC for HHVM CLI (at least for dump runners) - https://phabricator.wikimedia.org/T162245 [04:03:04] nope, moritzm said "We're now using 3.18 in production (except snapshot* and video scalers hosts, which are still on trusty)" [04:03:23] which is a very relevant exemption since the bug only affects snapshot* hosts [04:03:33] he said in the previous comment "Note that also requires a migration of the snapshot hosts to jessie" [16:48:02] Anyone want to merge a release notes fix? https://gerrit.wikimedia.org/r/#/c/391583/ [17:23:12] anomie: {{done}} [17:25:52] Thanks [17:27:53] anomie: if you have a minute... https://gerrit.wikimedia.org/r/#/c/387516/ could use a +2. [17:28:25] I can use it unmerged too if it seems goofy to you [17:31:05] * anomie looks [17:58:13] bd808: Found a few issues. [17:58:33] anomie: awesome. I'll take a look in a bit. [17:58:39] thanks for reviewing [19:44:17] Hmm. https://integration.wikimedia.org/ci/job/mediawiki-core-php70-phan-docker/1271/checkstyleResult/ is claiming no warnings or errors, but looking at the console output there actually were some. [19:51:26] * anomie files T180630 [20:00:00] RFC discussion on imported user names at #wikimedia-office in 1 minute! [20:01:00] TimStarling: you awake? [20:01:31] yes [20:02:00] excellent! [20:46:35] legoktm: If you want to cry tears of blood for when you get back, https://gerrit.wikimedia.org/r/#/c/391645/ is going to be fun. Lots of random config maps inside Translate that I'll need to handle. Yay. [22:07:22] RoanKattouw: it turns out to be surprisingly difficult to find meeting times that work for both legoktm and anomie [22:08:33] there is a slot around 6pm PST that we're now putting a weekly team meeting in, but that's too late for europe [22:09:28] Yeah that's 3am [22:12:40] 3 / 4 am :) [22:13:00] legoktm has no school on fridays but that is saturday for me [22:13:47] which would be OK on occasion but I think Angela would probably kill me if it was weekly ;) [22:14:38] hello :) [22:15:37] James_F: I'll take a look, I haven't looked at converting Translate in quite a while [22:15:47] hi legoktm, we were talking about RFC meeting scheduling [22:18:32] I have no real preference for the current time, it's just what I've planned my schedule around [22:18:52] our class picker has a nice feature that lets you block off specific times and then generates schedules that meet that (if possible) [22:23:00] legoktm: No rush. :-) [22:23:48] sounds very flexible [22:25:01] in my undergraduate degree we had very few choices [22:25:47] on semester I ended up with a conflict between two classes, which was sufficiently rare that nobody was interested in fixing [22:25:54] luckily one of the two was "introduction to computing" [22:26:12] I missed every single introduction to computing lecture, just got the lecture notes from someone, still got 96% [22:26:18] hah :D [22:26:54] well, I had a choice between Computing 1A and Introduction to Computing [22:27:32] Computing 1A was the proper software engineering subject, everyone agreed it was hard [22:27:37] it was mostly haskell [22:28:26] I chose introduction to computing because I'm lazy and because I knew I would get better marks in it :) [22:28:45] I didn't know about the scheduling conflict until it was too late though [22:30:16] the other side of the scheduling conflict was quantum physics, which is definitely not the sort of subject you can skip and still pass [22:59:31] :D [23:03:09] haskell in undergrad first year CS? Ours was LISP :) [23:03:24] (then Java, like most schools at the time) [23:05:30] Mine was Java (plus B, awk, and Erlang, I think). [23:06:12] But certainly, the idea that I could schedule uni around my life was not an option. :-) [23:06:53] I did MW stuff during lectures... [23:15:05] yeah haskell in first year, just checked, they've switched to C now, which seems more practical [23:22:10] Reedy: Phase 3 wasn't written yet for me. ;-P