00:12:34  * jgijoined
00:13:37  <jbergstroem>seems good
00:30:01  * jgiquit (Quit: jgi)
00:31:17  * jgijoined
00:37:28  <jbergstroem>i'm moving a linter from DO to joyent since we don't hae tu run tests on it
00:43:27  * jgiquit (Quit: jgi)
01:09:12  * jgijoined
01:16:48  <jbergstroem>ok, almost done refactoring/renaming linters and freebsd bots!
01:16:54  <jbergstroem>i will retire our voxer stuff.
01:25:34  <jbergstroem>if someone wants to chip in at https://github.com/nodejs/build/issues/254 it'd be great
01:25:47  <jbergstroem>i'd like to start walking through that
01:29:08  <jbergstroem>rvagg: any opinion on retiring our freebsd voxer stuff?
02:36:09  <rvagg>jbergstroem: make it so
02:39:34  <jbergstroem>rvagg: thanks
03:16:19  <rvagg>have I touched too many languages in these v0.10 and v0.12 PR to get reviews and signoff from collaborators or do you think everyone is just too busy? https://github.com/nodejs/node/pull/3642 https://github.com/nodejs/node/pull/3965
03:33:16  * jgiquit (Quit: jgi)
03:51:59  * bengljoined
04:32:19  <jbergstroem>did a quick round
04:46:22  <orangemocha>I have been quite busy, sorry, will try to take a look tomorrow
04:46:30  <orangemocha>ARM cross-compile is failing consistently: https://ci.nodejs.org/job/node-cross-compile/588/nodes=cross-compiler-pi1p/console
04:47:05  <orangemocha>on master, it looks legit
05:08:51  * jgijoined
05:27:54  * rmgquit (Remote host closed the connection)
05:28:28  * rmgjoined
05:32:55  * rmgquit (Ping timeout: 250 seconds)
07:01:49  * jgiquit (Quit: jgi)
11:32:08  <jbergstroem>rvagg: was planning on digging into https://github.com/nodejs/node/pull/3965 a bit later; will you update it to reflect some of the changes in 0.12 or are those without merit there?
11:47:24  <rvagg>jbergstroem: it's almost exactly the same as the v0.12 one but without icu
11:47:41  <rvagg>the v0.10 one isn't so urgent, but we really need to get v0.12.8 out
14:31:49  * rmgjoined
14:36:28  * rmgquit (Ping timeout: 244 seconds)
15:44:08  * rmgjoined
16:21:21  * jgijoined
16:37:25  * jgiquit (Quit: jgi)
17:01:34  * jgijoined
17:06:48  * jgiquit (Quit: jgi)
17:08:36  * jgijoined
17:08:36  * jgiquit (Client Quit)
17:38:35  * jgijoined
18:49:06  <thealphanerd>hey rvagg are you still around? I was curious if I could land the PR on citgm, and what the deal is with publishing.
18:49:22  <thealphanerd>(if I can have access / am allowed to publish to npm)
19:26:29  * jgiquit (Quit: jgi)
19:33:59  * jgijoined
19:49:55  * orangemocha_joined
19:51:26  * orangemochaquit (Ping timeout: 240 seconds)
21:30:26  * jgiquit (Quit: jgi)
22:05:12  * jgijoined
22:34:12  <rvagg>thealphanerd: yeah, just land it
22:34:45  <thealphanerd>multiple commits?
22:34:47  <rvagg>thealphanerd: only james has publish permissions so you'll have to wait for him to publish
22:35:05  <rvagg>thealphanerd: have a look at the commit log and follow along, no need to be as strict as we are on core
22:35:17  <thealphanerd>I’ve been for the most part sticking to that
22:35:26  <thealphanerd>I’ll look one more time to make sure it is reasonable
22:36:04  <thealphanerd>I’m mucking about in jenkins right now
22:48:21  <thealphanerd>I just noticed that the jenkins config for node-test-commit is still pointing at nodejs/io.js/
22:48:40  <thealphanerd>is this the kind of thing I should make an issue about on nodejs/build?
22:50:09  <rvagg>thealphanerd: no, you're an admin, just fix it and let us know in here and it should be fine
22:51:40  <thealphanerd>ok. I updated it
22:52:05  <thealphanerd>wasn’t entirely clear if you were giving me admin to experiment or if you geniunely wanted me to do admin’y things. more than happy to pitch in
22:52:28  <thealphanerd>this is the dev ops equivelent to a spelling error in the readme lul
23:11:32  <rvagg>thealphanerd: ideally the smoke test jobs are your domain, but if you see easy things to pitch in on then feel free, just as long as we're in the loop
23:20:46  <thealphanerd>so I’m alrady hitting my first wall, of sorts. the top level job (for PR or commit) spawns a a test-commit job for each platform. Each of those platform specific jobs maintains a number of scripts that are executed. One for git stuff, on for build stuff, as well as scriipts for logging / diagnosis.
23:20:47  <thealphanerd>For the smoker we essentailly need the exact same setup, but want to run citgm rather than the node test suite.
23:20:47  <thealphanerd>What would be the best way to avoid creating two sources of truth that need to be maintained. Is there a way of injecting an execute script from a parent job? Should we include it in each platforms commit job behind a flag?
23:25:02  <rvagg>thealphanerd: careful using test-commit stuff as an example of how to do this best, it's way overcomplicated
23:25:35  <rvagg>look at iojs+release as an example of it being self-contained, all platforms in one job, no git stuff getting in the way, there's some complicated bash stuff in there but you'll see that it's mostly about juggling <=v0.12 vs >v0.12 stuff
23:26:56  <thealphanerd>are the same machines used in iojs+release? (I’m in the process of looking)
23:35:58  <jbergstroem>sorry 'bout all the freebsd attention; i'm pretty much done now
23:36:17  <jbergstroem>moving onto centos6,7 and a debug build (will require a 4g vm though)