Camelia, the Perl 6 bug

IRC log for #parrot, 2013-02-28

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:25 kid51 joined #parrot
02:07 kid51 Just ran smolder in sixparrot branch
02:07 kid51 Only one failure: t/pmc/integer.t test 129
02:08 kid51 not ok 129 - i_divide overflow promotion
02:08 kid51 # Have: 0
02:08 kid51 # Want: 2
02:10 dalek parrot/sixparrot: efd27df | jkeenan++ | src/pmc/parrotinterpreter.pmc:
02:10 dalek parrot/sixparrot: [codingstd] Un-cuddle that else.
02:10 dalek parrot/sixparrot: review: https://github.com/parrot/parrot/commit/efd27df7aa
02:34 cotto kid51: that feels like a leftover from removing bignum.
02:41 kid51 cotto: I've got 3 codingstd test failures in master -- but not in sixparrot
02:41 kid51 (I would have expected it to be the other way around ;-) )
02:47 kid51 2 of those 3 were spurious
02:48 kid51 But still failing t/codingstd/pccmethod_deps.t; test seems to be ending prematurely
03:03 kid51 I now realize that even that last test failure was spurious
03:05 kid51 What happened was that I built and smoked in sixparrot branch, then ran make codetest in that branch, then switched back to master and ran codetest *there* -- without first doing 'make test-clean' -- which, when I did it, cured 2 of the 3 "failures" -- and then started from scratch -- which cured the remaining one.
03:41 woosley joined #parrot
04:35 preflex_ joined #parrot
04:38 benabik joined #parrot
05:49 woosley joined #parrot
07:07 Mike-PerlRecruiter_ joined #parrot
07:22 bouncy joined #parrot
09:26 dalek rakudo/vmarray: c54039f | (Arne Skjærholt)++ | / (8 files):
09:26 dalek rakudo/vmarray: Start VMArrayHOW, expose it in the grammar.
09:26 dalek rakudo/vmarray: review: https://github.com/rakudo/rakudo/commit/c54039faa5
09:28 dalek nqp: 3dd264f | (Arne Skjærholt)++ | src/6model/reprs/VMArray. (2 files):
09:28 dalek nqp: First implementation of compose() for VMArray.
09:28 dalek nqp: review: https://github.com/perl6/nqp/commit/3dd264f322
09:53 xcombelle joined #parrot
10:27 woosley left #parrot
10:35 PacoAir joined #parrot
11:04 donaldh joined #parrot
12:13 dalek nqp: 63b55a0 | (Timo Paulssen)++ | src/QRegex/P6Regex/Grammar.nqp:
12:13 dalek nqp: complain about negatives in range quantifiers.
12:13 dalek nqp: review: https://github.com/perl6/nqp/commit/63b55a0aad
12:51 dalek nqp: 82b7e17 | (Timo Paulssen)++ | src/QRegex/P6Regex/Grammar.nqp:
12:51 dalek nqp: fix error report with negative quantifiers.
12:51 dalek nqp: review: https://github.com/perl6/nqp/commit/82b7e17576
13:19 PacoAir joined #parrot
13:30 darbelo joined #parrot
13:36 dalek rakudo/nom: 22b3cf3 | moritz++ | t/spectest.data:
13:36 dalek rakudo/nom: run S02-magicals/vm.t, timotimo++
13:36 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/22b3cf3d74
14:21 Psyche^ joined #parrot
14:44 bluescreen joined #parrot
15:35 benabik joined #parrot
16:52 arnsholt_ Is there some particular reason that Parrot has -Waggregate-return in its CFLAGS?
16:55 darbelo joined #parrot
16:56 cotto arnsholt: alester is the resident gcc flag guy.  I don't know off the top of my head.
16:57 cotto You might look through the history of config/auto/warnings.pm
16:58 arnsholt Right. I'll do that, then
16:59 arnsholt I'm removing warnings from NQP's compile, and that one can't be removed, since one of the sixmodel functions actually returns an aggregate
17:00 cotto Is that a recent change?
17:00 arnsholt Nope. Been there as long as sixmodel I think
17:00 cotto Ah.  It's just a noisy warning that you don't need.
17:01 arnsholt Yeah. It's mostly cosmetic, but I like to keep the warnings to a minimum, since they do come in handy every once in a while
17:02 cotto arnsholt: that's a very good idea.
17:05 arnsholt And of course a single warning in an ops file will cause many warnings during compile
17:05 cotto yeah
17:05 arnsholt Speaking of ops files, what's the problem with empty cases in switch blocks?
17:06 cotto arnsholt: opsc turns it into valid C but the old ops2c ops processor isn't smart enough.
17:07 arnsholt Ah, right
17:07 arnsholt What's a non-empty enough case for it to not care? Is something like "default: ;" enough, or does it need a better statement after the label?
17:08 cotto It depends on what gcc accepts.  The ; seemed to be enough to make it happy.
17:11 PacoAir joined #parrot
17:13 arnsholt Pretty sure gcc wil accept that (the empty statement is valid C, after all)
17:16 cotto arnsholt: It complained about a label before an empty statement, which make some sense.  If there's no code, there's nowhere for the label to go.
17:17 cotto It's a bit odd.
17:17 PacoAir joined #parrot
17:23 davidfetter joined #parrot
18:19 zby_home joined #parrot
18:20 chekkaa joined #parrot
18:27 contingencyplan joined #parrot
18:47 benabik joined #parrot
19:06 Mike-PerlRecruiter_ joined #parrot
19:10 bouncy joined #parrot
20:41 bluescreen joined #parrot
22:57 benabik joined #parrot
23:31 donaldh joined #parrot
23:40 dalek rakudo/nom: 357e603 | jnthn++ | src/core/Cool.pm:
23:40 dalek rakudo/nom: Remove bogus coercion from Cool.substr.
23:40 dalek rakudo/nom:
23:40 dalek rakudo/nom: Found by labster++.
23:41 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/357e603cae
23:54 kid51 joined #parrot

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

Parrot | source cross referenced