Perl 6 language and compiler development | Logs at | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by AlexDaniel on 12 June 2018.
00:39 samcv left 00:40 samcv joined 00:44 lizmat left 01:01 fake_space_whale joined 02:25 dalek left 02:50 synopsebot left, Undercover left, dalek joined, ChanServ sets mode: +v dalek, SourceBaby left 02:51 p6lert left, Geth joined, ChanServ sets mode: +v Geth, p6lert_ joined, synopsebot joined, ChanServ sets mode: +v synopsebot, SourceBaby joined, ChanServ sets mode: +v SourceBaby, Undercover joined, ChanServ sets mode: +v Undercover 02:58 p6lert_ left, p6lert joined 03:23 fake_space_whale left 04:31 skids joined 04:42 bartolin_ left, bartolin joined 04:51 benjikun left 04:57 benjikun joined 05:03 skids left 05:07 bartolin left 05:15 bartolin joined 05:17 ufobat_ joined 05:22 llfourn left 05:25 llfourn joined 05:44 dct joined 06:01 dct left 06:05 benjikun left 06:07 lizmat joined 06:33 robertle joined
[Tux] Rakudo version 2018.06-248-g434ede491 - MoarVM version 2018.06-318-gb8f9f85f2
csv-ip5xs1.121 - 1.239
csv-ip5xs-208.373 - 8.846
csv-parser25.263 - 26.260
csv-test-xs-200.465 - 0.489
test10.118 - 10.453
test-t2.235 - 2.313
test-t --race0.975 - 0.987
test-t-2040.652 - 41.320
test-t-20 --race13.365 - 14.126
08:46 literal left 08:47 literal joined 08:49 Ven`` joined 08:53 pmurias joined
lizmat did we just lose the #perl6 channel ? 09:06
Files=1246, Tests=76415, 383 wallclock secs (16.23 usr 6.06 sys + 2692.28 cusr 263.18 csys = 2977.75 CPU)
pmurias lizmat: it worked for me once I identified 09:13
lizmat yeah, I realize that's what's going on now, judging from the log
Geth rakudo: stmuk++ created pull request #2136:
Since merge of rescalar branch Java build needs 3G see 80dca6b7
09:38 woolfy joined 09:44 pmurias left, pmurias joined
Geth rakudo: 8feb13cd08 | (Steve Mynott)++ |
Since merge of rescalar branch Java build needs 3G see 80dca6b7
rakudo: ee44d4b42a | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) |
Merge pull request #2136 from stmuk/master

Since merge of rescalar branch Java build needs 3G see 80dca6b7
09:52 woolfy left 10:11 Ven`` left
Geth rakudo: 1188f801d4 | (Elizabeth Mattijsen)++ | lib/experimental.pm6
Remove hack from "is cached"

Any onlystar sub will be marked so that it won't be optimized away.
rakudo: d60be5e4a1 | (Elizabeth Mattijsen)++ | src/Perl6/World.nqp
Make error less LTA with e.g. sub infix:{"foo"}

Spotted by dakkar++
lizmat looked at the remaining blockers and realized she's out of her league there 12:01
12:10 x[LGWs4x4i]uG2N0 left 12:13 x[LGWs4x4i]uG2N0 joined
Geth roast: 714be9c2ff | (Elizabeth Mattijsen)++ | S02-types/set.t
Add test for R#2068
synopsebot R#2068 [open]: Set.Hash returns badly typed Hash
12:37 j3nnn1 joined 13:16 Ven`` joined 13:20 ufobat_ left
Geth rakudo: 3158ae7e0c | (Elizabeth Mattijsen)++ | src/core/Attribute.pm6
Add "does role" handling for Attributes
lizmat Oops, forgot to mention that was for R#2130 13:22
synopsebot R#2130 [open]: [LTA] Can't `does` an Attribute even tho you can `does` a variable
13:22 yoleaux left, yoleaux joined 13:24 Ven`` left 13:35 brrt joined
brrt \o 13:35
13:38 Ven`` joined 14:00 skids joined 14:06 fake_space_whale joined
Geth rakudo: AlexDaniel self-assigned tools/ is a perl 6 script but has Prolog file extension
9a45b2b8f9 | (Elizabeth Mattijsen)++ | src/core/Bool.pm6

Judging from the error reports, the most common pitfall is !%h<a>:exists. This patch only takes care of *that* particular case. Perhaps other cases could be handled this same way, like for infix:<+> and infix:<-> and prefix:<->.
14:37 Ven`` left, brrt left 14:43 lizmat left 14:45 lizmat joined 14:48 brrt joined 14:53 Ven`` joined 15:00 Ven`` left
synopsebot R#2098 [open]: [LTA] !%hash<a>:exists doesn't suggest to use %hash<a>:!exists
R#1814 [closed]: have X::OutOfRange gist the faulty value too, fixes #1813
15:17 robertle left
lizmat . 15:26
15:26 fake_space_whale left
Geth rakudo/post-release-2018.07: 21 commits pushed by 7 authors
roast/post-release-2018.07: 5 commits pushed by (Elizabeth Mattijsen)++, (Timo Paulssen)++, (Zoffix Znet)++
synopsebot R#2111 [closed]: [regression] `please 'use experimental :pack'` even if code does exactly that
R#2068 [closed]: Set.Hash returns badly typed Hash
Geth rakudo: 2c7692fc7d | (Elizabeth Mattijsen)++ | src/Perl6/Grammar.nqp
Make sure we show the op if it is too fiddly

Fixes R#2093
synopsebot R#2093 [open]: [LTA] Too-fiddly error missing op name
15:55 robertle joined 16:13 brrt left 16:15 ufobat_ joined 16:36 brrt joined
Geth roast: bdc9fd57b3 | (Elizabeth Mattijsen)++ | S12-class/attributes-required.t
Add tests for R#2083
synopsebot R#2083 [open]: [LTA][testneeded] `is required` change removed useful feature
16:46 diakopter joined 16:52 literal left 16:53 literal joined
Geth roast: 8995cd75eb | (Elizabeth Mattijsen)++ | S02-names/pseudo.t
Add tests for R#2058
synopsebot R#2058 [open]: [regression][testneeded] MoarVM inlining regression found in P5quotemeta module (goes away with MVM_SPESH_INLINE_DISABLE=1)
16:58 brrt left
timotimo Blol i logged: 16:59
Geth roast: ceb0380ba0 | (Elizabeth Mattijsen)++ | S05-match/make.t
Add test for R#2057
synopsebot R#2057 [open]: [regression][testneeded] Yapsi module is failing its tests
17:03 travis-ci joined
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Make sure we show the op if it is too fiddly 17:03
17:03 travis-ci left
buggable [travis build above] ☠ Did not recognize some failures. Check results manually. 17:03
17:06 stmuk_ joined 17:08 stmuk left
Geth roast: 8e9c588f8b | (Elizabeth Mattijsen)++ | S03-binding/attributes.t
Add tests for R#2130
synopsebot R#2130 [open]: [LTA][testneeded] Can't `does` an Attribute even tho you can `does` a variable
17:27 AlexDani` joined 17:30 AlexDani` left, AlexDani` joined, AlexDaniel left, AlexDani` is now known as AlexDaniel 18:04 ufobat_ left
Geth nqp: 3fc2378248 | MasterDuke17++ (committed using GitHub Web editor) | docs/ops.markdown
Fix the backends for nqp::<eqatic eqatim eqaticim>
18:22 Ven`` joined 18:41 Ven`` left 19:01 Kaiepi left 19:02 Kaiepi joined 19:07 Kaiepi left
Geth rakudo: 08b449e1ab | (Elizabeth Mattijsen)++ | 4 files
Make parameterization errors show where they occur

Fixes R#2032. If a ^parameterize sub returns an (instantiated) string, it will be taken as the payload of an X::AdHoc exception to be properly thrown with all the bells and whistles of a normal compilation error.
synopsebot R#2032 [open]: [LTA] Error from `Array[42]` lacks location
19:37 Kaiepi joined
Geth nqp/truffle: eed41d4f4b | (Daniel Green)++ | 8 files
[truffle] Implement nqp::<cmp_ins eqat(ic) not_i>
20:11 brrt joined
brrt \o 20:11
yoleaux 17:14Z <MasterDuke> brrt: the windows problem is just with the expression jit, correct? if so, could just it be disabled?
brrt .tell MasterDuke I believe it is in the JIT, but not in the expression JIT 20:12
yoleaux brrt: I'll pass your message to MasterDuke.
MasterDuke ah
yoleaux 20:12Z <brrt> MasterDuke: I believe it is in the JIT, but not in the expression JIT
brrt .tell Zoffix - repsectfully, and given the heat with the intent of not letting annoyance grow bigger, that is not what I said
yoleaux brrt: I'll pass your message to Zoffix.
brrt .tell Zoffix what I said was, since there is an intent to release a version, and there is a blocking bug that is hard to fix, we may circumvent it for now, and do a point release later 20:13
yoleaux brrt: I'll pass your message to Zoffix.
MasterDuke brrt, timotimo: while you're here, any idea why a --profile that bench code i was running on moar/jvm/truffle says it's not being jitted for moar? 20:18
brrt no
MasterDuke npq: sub foo() { my $s := nqp::time_n(); my $a; my $i := 1; while $i++ < 1_000_000 { $a := ~(($i - 1) / $i) }; say(~$a); say(nqp::time_n() - $s) }; foo() 20:19
nqp: sub foo() { my $s := nqp::time_n(); my $a; my $i := 1; while $i++ < 1_000_000 { $a := ~(($i - 1) / $i) }; say(~$a); say(nqp::time_n() - $s) }; foo()
camelia 0.999999
MasterDuke except i was doing 50_000_000 iterations
well, the routines tab showed 'foo' as green. but the overview tab said 92% interpreted (only 1 out of the 13 frame) 20:21
20:23 robertle left 20:25 travis-ci joined
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make parameterization errors show where they occur 20:25
20:25 travis-ci left 20:34 Ven`` joined 20:47 Ven`` left 20:49 brrt left
Geth roast: 0139961320 | (Elizabeth Mattijsen)++ | S32-basics/warn.t
Add tests for R#1833
synopsebot R#1833 [open]: [testneeded] Spurious warning when smartmatching Slip ~~ Empty
AlexDaniel c: 2018.06 21:08
committable6 AlexDaniel,
AlexDaniel c: 46ef1b5^,46ef1b5,HEAD
c: 46ef1b5^,46ef1b5,2018.06,HEAD 21:09
committable6 AlexDaniel,
21:12 skids left
Geth roast: 30c042c102 | (Elizabeth Mattijsen)++ | S06-advanced/callframe.t
Add tests for R#1871
synopsebot R#1871 [open]: Cannot stringify this: 1.::(1)
Geth roast: 86d825be58 | (Julien Simonet)++ | S03-smartmatch/hash-hash.t
Add test for smartmatch hash with Mu values.
roast: 8afb8fb36b | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | S03-smartmatch/hash-hash.t
Merge pull request #423 from jsimonet/master

Add test for smartmatch hash with Mu values.
21:18 brrt joined
Geth ¦ rakudo: lizmat self-assigned Baggy cross operators 21:21
timotimo MasterDuke: the percentage in the overview tab is for "number of entries" rather than "time spent in one or the other kind" 21:23
so you can get 99% interpreted, but still be spending 99% of your time in jitted code
to get that to be more accurate, i'll have to track times for the three states separately
MasterDuke timotimo: heh, i kind of suspected that after explaining what i saw in the different tabs 21:24
but good to have confirmation
Geth ¦ rakudo: lizmat self-assigned BUILDPLAN sharing optimization appears to have broken ancestor attribute initialization
lizmat stops looking at issues for today 21:25
brrt the one thing i can tell about the windows jit bug, is that it seems to be related to Test::Helpers is-run, somehow 21:28
Geth nqp: MasterDuke17++ created pull request #491:
Sort QAST::VM annotations when dumping them
21:29 brrt left 21:48 pmurias left
Geth nqp: 70194018a6 | MasterDuke17++ (committed using GitHub Web editor) | src/QAST/VM.nqp
Sort QAST::VM annotations when dumping them (#491)

Fixes #439. I looked at the other node type's dump methods and they all looked fine.
synopsebot NQP#439 [closed]: [LHF] QAST dump methods need to sort annotation dumps
22:11 x[LGWs4x4i]uG2N0 left 22:15 x[LGWs4x4i]uG2N0 joined
AlexDaniel lizmat++ reviewing older tickets 22:21
lizmat: fwiw I think there's no need to remove labels from closed tickets
22:36 Kaiepi left, Kaiepi joined
japhb buggable: speed :4 22:38
buggable japhb, █ ↑ ▇ ↑ ▃ dates: 2018-07-06–2018-07-26 22:39
japhb, ▇▂▄▃▂▂ ▇█▇▅█ ▃ ▃█▂▅ █▂▁▂█ ▂▅▂ ▃ range: 12.801s–18.257s
japhb, ██████▅███████▂▃█ ████▅█████▃ ▄▄▄███▂▇ █▁ ▃▄ ▅▄ speed: 6% faster (widths: 9/3)
japhb, █████████████████████████████▆███████████▆▃▁██▆██▅
22:42 skids joined 22:57 lucasb joined 23:09 lucasb left 23:21 j3nnn1 left