Camelia, the Perl 6 bug

IRC log for #parrot, 2012-11-19

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:16 xenoterracide joined #parrot
00:16 xenoterracide I'm sure people realize this... but 4.9.0 isn't actually on the mirrors
00:28 whiteknight I did not realize that. that was dukeleto's release, right?
01:00 xenoterracide not sure, I've just been debugging arch's latest release and realized that there was a later released mentioned on the home page that is not actually on the ftp mirror
01:01 xenoterracide (also no idea what happened on arch that I can't get rakudo to run on it's build or using it's pkgbuild of parrot)
01:17 Mike-PerlRecruiter_ joined #parrot
01:56 woosley joined #parrot
02:11 cotto xenoterracide, the release is in the right place.  It's just the link from the announcement that's wrong.
02:11 cotto ftp://ftp.parrot.org/pub/pa​rrot/releases/stable/4.9.0/
02:12 cotto .0, .3, .6 and .9 releases are stable
02:12 xenoterracide ah
02:17 cotto looks like I'm getting a similar set of test failures with parrot master
02:21 xenoterracide and yet I built off of 4.9.0 gitcheckout using perl Configure.pl --gen-parrot=<tag> --gen-nqp
02:21 xenoterracide and that works
02:21 xenoterracide doesn't really make sense to me
02:22 cotto the spectest_regression test failures look identical on Rakudo on parrot master and the threads branch
02:22 xenoterracide ah
02:22 cotto Rakudo
02:22 cotto 's config system knows how to grab parrot
02:22 cotto that's what you were telling it to do
02:22 xenoterracide yeah I know
02:23 xenoterracide it just seems to work when I do that instead of point it to arch's parrot
02:23 cotto It could be that you've either got an old version or arch has something configured surprisingly
02:23 xenoterracide I think it's the latter
02:23 xenoterracide idk, something changed between today and yesterday
02:24 xenoterracide they built a new version
02:24 xenoterracide and it changed something
02:24 xenoterracide but I'm not sure what
02:24 xenoterracide they are also using 4.8.0 but that wasn't a problem last week
02:24 cotto I don't think too much changed between those two releases.
02:25 xenoterracide I don't think it has anything to do with 4.8.0 and 4.9.0
02:25 xenoterracide I'm not sure what changed in the arch build
02:25 xenoterracide I don't see much change in the pkg build
02:26 xenoterracide but all of a sudden it's broke
02:27 xenoterracide maybe it's because arch makes it a shared library
02:27 xenoterracide that's my only thought
02:30 cotto libparrot.so should be a shared library
02:31 xenoterracide well idk
02:31 xenoterracide I spent enough time already trying to figure out why it stopped working
02:35 cotto I'll most likely be around if you want to resume figuring out what broke what.
02:35 xenoterracide cotto: probably another day, fighting sleep deprivation
02:35 cotto go sleep
02:36 cotto lack of sleep makes people do stupid things
02:36 xenoterracide yeah, well if I go to sleep now I'll wake at 4am
02:36 xenoterracide or 2am
02:36 xenoterracide so I'm just going to sit up and be stupid and write play code
02:37 cotto your cal
02:37 cotto l
02:49 Reini_ joined #parrot
02:49 Reini_ 3 new thread errors: http://smolder.parrot.org/app/​projects/report_details/32335
02:51 cotto and here I was hoping to merge it soon
02:53 Reini_ it's a -j4 test artifact. single it works.
02:54 Reini_ wonder what resource blocks it.
02:54 Reini_ all 134 t/src/extend_vtable.t fail if parallel, but pass if single...
02:56 cotto I'd still call that a blocker.
02:56 Reini_ we'd need some kind of mutex in Test::More for non-parallel safe tests
02:57 Reini_ interestingly the same 3 tests failed at the 2nd time in parallel
02:57 Reini_ Looks like a new ticket to me
03:02 cotto that sounds like it could be rough to repro
03:06 Reini_ I'm trying now with asan and tsan
03:11 xenoterracide http://privatepaste.com/1130814e00 O.o
03:11 xenoterracide oops wrong chan
03:12 dalek parrot/threads: df25d15 | cotto++ | config/auto/warnings.pm:
03:12 dalek parrot/threads: add -fpermissive so that nqp builds under gcc 4.7.2
03:12 dalek parrot/threads: review: https://github.com/parrot/parrot/commit/df25d15a38
05:19 Reini_ pbc_to_exe.winxed does link -s (strip) which is deprecrated (and wrong)
05:19 Reini_ strip should be done later, in make install
05:22 cotto Is that causing problems or did you just happen to notice it?
05:25 Reini_ i could not debug into tadzik's threads failure. now I can.
05:25 Reini_ https://gist.github.com/3817314#comments
05:25 Reini_ Parrot_NQPLexPad_set_pointer
06:35 zby_home joined #parrot
09:01 bouncy joined #parrot
09:58 brrt joined #parrot
10:30 schm00ster joined #parrot
10:34 brrt joined #parrot
11:34 brrt joined #parrot
11:41 brrt left #parrot
12:14 mtk joined #parrot
12:29 mtk joined #parrot
13:17 Mike-PerlRecruiter_ joined #parrot
14:07 whiteknight joined #parrot
14:10 whiteknight good morning, #parrot
14:11 tadzik hello whiteknight
14:13 PacoAir joined #parrot
14:21 jlaire joined #parrot
14:25 Reini_ joined #parrot
15:23 jlaire left #parrot
16:56 elmex joined #parrot
17:51 contingencyplan joined #parrot
18:00 Reini I am afraid but the make -j4 test problems with threads only appear on darwin
18:01 Reini darwin also only had the shared IO problem of GH #856. maybe it's related
18:01 schm00ster joined #parrot
18:08 Reini uuh, all my bsd vm's only have one cpu. So I cannot confirm that
18:47 dmalcolm joined #parrot
18:50 Reini_ joined #parrot
18:56 davidfetter joined #parrot
19:23 alester joined #parrot
19:56 cotto ~~
20:17 dalek rakudo/qast-sink-1: 6cc0d6c | moritz++ | src/core/Junction.pm:
20:17 dalek rakudo/qast-sink-1: work around nqp::can(Junction, $anything) always returning True
20:17 dalek rakudo/qast-sink-1: review: https://github.com/rakudo/rakudo/commit/6cc0d6c4d9
20:18 dalek rakudo/qast-sink-1: 318e56c | moritz++ | src/ (3 files):
20:18 dalek rakudo/qast-sink-1: Revert "only sink when --sink command line option is provided"
20:18 dalek rakudo/qast-sink-1:
20:18 dalek rakudo/qast-sink-1: This reverts commit 3f12074d13a51b2602039ece5f95454055e8abd0.
20:18 dalek rakudo/qast-sink-1: review: https://github.com/rakudo/rakudo/commit/318e56c9d9
20:46 dalek rakudo/nom: bf472b0 | jonathan++ | src/Perl6/Actions.pm:
20:46 dalek rakudo/nom: Make :($a, $b) := \(1, 2) update $a and $b.
20:46 dalek rakudo/nom:
20:46 dalek rakudo/nom: Previously, the spec had such a construct doing nothing. We'll try it
20:46 dalek rakudo/nom: like this for a bit to see how it works out. Note, shortest patch that
20:46 dalek rakudo/nom: could possibly work; probably needs some tuning.
20:47 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/bf472b05a0
23:23 whiteknight joined #parrot
23:23 whiteknight good evening, #parrot
23:30 cotto hi whiteknight
23:30 whiteknight Hello cotto
23:33 sorear hello whiteknight.
23:33 whiteknight hello sorear
23:36 whiteknight My stress level may be going down significantly in the coming days, and my available hacking time should be going up too
23:37 whiteknight so that's good for me

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced