Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-release, 2016-02-19

| Channels | #perl6-release index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:27 stmuk_ joined #perl6-release
02:48 ilbot3 joined #perl6-release
02:48 Topic for #perl6-release is now »r̈« - Discussions about Perl 6 and Rakudo release strategies - http://irclog.perlgeek.de/perl6-release/today
07:17 MadcapJake joined #perl6-release
07:22 FROGGS joined #perl6-release
09:31 lizmat joined #perl6-release
13:25 [Coke] jnthn, lizmat - how do you feel about a 2016.02 this weekend?
13:26 lizmat pretty good, just about commit my final Str.trans opts
13:26 [Coke] Ok. If we don't have any 6.d stuff yet, I can definitely cut the release.
13:27 nine Do we pass all 6.c tests again?
13:27 [Coke] we will before a release is cut, for sure.
13:27 [Coke] (testing that now on os x)
13:27 jnthn Yeah, "we pass the 6.c tests" is The Thing to ensure.
13:27 jnthn I'm good with a release that does that
13:28 [Coke] if NQP/moar have things to bump, let's bump them now.
13:28 lizmat will test 6.c tests when I get back from some cycling&
13:28 jnthn Yeah, Moar has moar leak fixes
13:28 jnthn I can bump it this evening if nobody beats me to it
13:30 [Coke] danke.
13:52 nine I get failures in: t/spec/S02-types/WHICH.t t/spec/S32-io/dir.rakudo.moar t/spec/S32-temporal/DateTime.t t/spec/S32-list/reduce.t t/spec/integration/weird-errors.rakudo.moar and a hang in t/spec/integration/advent2013-day14.t
13:55 jnthn Is last one a reliable hang or heisenbug?
14:03 nine A heisenbug of course. Everything else would have been too easy :)
14:04 jnthn Well, it's easy as in "I doubt that's a real regression"
14:04 jnthn the others warrant attention.
14:12 [Coke] I also have failures on: t/spec/S11-modules/require.t t/spec/S17-supply/unique.t
14:13 [Coke] be sure to run stresstest, not just spectest
14:13 nine require.t is just a passing TODO
14:18 [Coke] jnthn: you had said you want 6.c to be a tag, not a branch in roast, yes? and that we have an errata branch that we actually make changes to?
14:21 jnthn [Coke]: Yeah, that's my thinking
14:21 jnthn 6.c is a tag, 6.c-errata is a branch
14:24 perlpilot joined #perl6-release
14:54 [Coke] ok, someone needs to fix that, then.
15:20 [Coke] working on it.
15:23 [Coke] ok, we have a 6.c tag.
15:23 [Coke] Oh. Um. so when I made the 6.c branch, I updated the README to reflect what it was.
15:24 [Coke] going forward, We'll have to make sure that happens pre tagging (since tagging will be happening off master, I assume)
15:26 [Coke] I cannot delete the 6.c branch.
15:27 [Coke] there is now a 6.c errata branch.
15:27 [Coke] there is now a 6.c-errata branch.
15:28 [Coke] figured out how to delete the branch.
15:59 jnthn :)
15:59 jnthn Fun Git fact: you can have a branch and a tag with the same name.
15:59 jnthn A few years ago when I was teaching Git I could have told you the order of precedence... :)
16:07 [Coke] yup. well, we did for about 5 minutes. :)
16:08 [Coke] had to delete with :refs/heads/6.c instead of just :6.c
17:37 cognominal_ joined #perl6-release
17:59 Juerd joined #perl6-release
18:18 cognominal joined #perl6-release
18:18 b2gills joined #perl6-release
18:25 rjbs joined #perl6-release
18:27 masak joined #perl6-release
18:27 llfourn joined #perl6-release
19:52 * lizmat apparently lacks the git fu to change 6.c roast
19:53 lizmat jnthn: re bumping Moar / nqp: shouldn't those be releases really ?
19:54 jnthn lizmat: For sure, I thought the idea was to bump them now for testing
19:55 lizmat ok, *I* can do that
19:55 jnthn And then releases this weekend
19:55 jnthn [Coke]: When in the weekend were you pondering the release? I'm no quite up for cutting a MoarVM release tonight :)
19:55 [Coke] jnthn: "scheduled" for tomorrow, we're in no rush
19:55 [Coke] lizmat: we shouldn't wait until just before the release to bump
19:55 lizmat bumping now
19:56 [Coke] we should bump in advance, even if we later tag and update to point to the tag.
19:56 [Coke] gets more testing in.
19:57 jnthn [Coke]: OK, I can do a Moar release tomorrow
19:57 jnthn I've time to do it now but...I'm tired and fear I'll mess something up :)
19:58 [Coke] yup, that's fine.
20:12 lizmat nom is now on latest/greatest MoarVM/NQP
20:23 jnthn \o/
20:27 lizmat jnthn: seems the bump is causing quite significant slowdowns :-(
20:30 jnthn The Moar one or NQP one?
20:31 jnthn It seems quite unlikely to be the Moar commits, given they're all memory leak patches.
20:31 * jnthn just look through them all, and many only apply with --full-cleanup
20:32 lizmat then it must be nqp :-(
20:33 jnthn The only patch that I might suspect is 5ac92837
20:33 lizmat in nqp ?
20:33 jnthn Any chance you could git revert that one and try a build?
20:33 jnthn Yes
20:33 lizmat will try in a mo
20:35 lizmat building now...
20:36 [Coke] agreed, that's the likely culprit
20:48 lizmat hmmm... *or* I'm not building correctly, *or* it is not the culprit
20:53 [Coke] I haven't verified the original slowness claim;

| Channels | #perl6-release index | Today | | Search | Google Search | Plain-Text | summary