Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
AlexDaniel lizmat: related discussion: github.com/perl6/doc/issues/1425 05:14
[Tux] Rakudo version 2018.09-417-g2c43d2bf5 - MoarVM version 2018.09-124-g588873f7a
csv-ip5xs0.895 - 0.903
csv-ip5xs-207.559 - 7.663
csv-parser22.008 - 24.150
csv-test-xs-200.436 - 0.440
test8.266 - 8.698
test-t1.774 - 1.805
test-t --race0.833 - 0.870
test-t-2031.133 - 32.639
test-t-20 --race11.233 - 11.672
07:48
lizmat Files=1256, Tests=76252, 337 wallclock secs (15.92 usr 5.47 sys + 2368.35 cusr 233.54 csys = 2623.28 CPU) 10:56
Geth rakudo: 266fcc7818 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
add info
11:53
Geth rakudo: bea099889e | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
add info
11:59
rakudo: 5b84611e72 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
add to list
12:03
Geth rakudo: 9908c7ee8d | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
add note
12:08
Geth rakudo: c0bb8c7ccc | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
order NYI items by my priority
12:12
Geth rakudo: 9408080c96 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
Update rakudo-nqp-notes.md
12:25
rakudo: 7668d8c760 | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
add info
12:31
rakudo: 64a777404e | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
update
12:32
travis-ci Rakudo build passed. Tom Browder 'update' 13:49
travis-ci.org/rakudo/rakudo/builds/440988239 github.com/rakudo/rakudo/compare/7...a777404e66
tbrowder___ hm, do we need a travis rebuild for doc changes? if not, can the tooling be changed to prevent it? 13:58
s/doc/all doc/ 13:59
Geth rakudo: 8e4d17a79b | (Tom Browder)++ (committed using GitHub Web editor) | docs/rakudo-nqp-notes.md
Update rakudo-nqp-notes.md
14:10
tbrowder___ here’s a question fo grammar experts: in Rakudo Grammar.nqp there are several instances of “<pod_content=.pod_textcontent>**0..1” and TimToady said the **0..1 forces the object to be an array. Both “pod_content” and “pod_textcontent” are tokens and both have methods in Actions.nqp. I think the “.token” construct means the match isn’t to be saved but does it still trigger the action method? 14:24
tbrowder___ my guess is no, but i have no proof yet 14:27
geekosaur mm, iirc it means it's not saved as both pod_content and pod_textcontent. I would thnk action methods still fire, it just doesn't get saved twice in the final Match object 14:30
Geth roast: 2b3f279afb | (Zoffix Znet)++ | S32-io/open.t
[v6.d REVIEW] Don't create a monster filename in bug cover

975-char path still reproes the issue
Orig: github.com/perl6/roast/commit/11fb4415f
15:18
dogbert11 .seen lizmat 15:28
hmm
timotimo about 5 hours ago in here 15:33
4 hours ago in #perl6
dogbert11 timotimo: do we have a dead bot? 15:35
.seen timotimo 15:36
timotimo dunno
yoleaux isn't here, maybe has never been here?
pmurias Zoffix: nothing obvious comes to my bind regarding the blocker 15:38
dogbert11 timotimo: colabti.org/irclogger/irclogger_log...10-03#l166 15:39
dogbert11 yay 15:39
<someone>++
timotimo cool 15:40
dogbert11 indeed
pmurias Zoffix: I fixed :U :D in 4b60df8f84fdf2ed8f5eab1426247fbc649d062c but I think that was put in master before the merge 15:42
yoleaux 12 Oct 2018 14:56Z <Zoffix> pmurias: would you have any guesses on this release blocker github.com/rakudo/rakudo/issues/23...-427702563 ? It bisects to js branch merge, yet is fine on both parents. Seems to be something with precomp.
Geth roast: cf93c5a351 | (Zoffix Znet)++ | S32-io/io-handle.t
[v6.d REVIEW] Fix up fh-close-on-exit tests

The first write to a handle is unbuffered, so the original test likely doesn't cover the bug.
Orig: github.com/perl6/roast/commit/08ac054de
17:03
roast: 75d4e63be9 | (Zoffix Znet)++ | S32-io/out-buffering.t
[v6.d REVIEW] Remove dupe close-on-exit test

First write to a handle is unbuffered, so the test doesn't cover the bug. Instead of fixing it, remove it, as there's already another test covering the same ticket on S32-io/io-handle.t:214
Orig: github.com/perl6/roast/commit/40ed...d7b1482a8a
17:04
AlexDaniel oh! It's possible to have multiple issue templates on github 17:14
Geth roast: 635d660465 | (Zoffix Znet)++ | S17-promise/basic.t
Remove trailing whitespace
17:43
roast: 60344f5892 | (Zoffix Znet)++ | S17-promise/basic.t
[v6.d REVIEW] Test ex. payload instead of .message

Orig: github.com/perl6/roast/commit/6eb5a91c9
Geth roast: 8d35f3f525 | (Zoffix Znet)++ | S13-overloading/operators.t
[v6.d REVIEW] Reword test desc to avoid impl details

Orig: github.com/perl6/roast/commit/730b5c823
18:03
roast: 19723fa1e0 | (Zoffix Znet)++ | 3 files
[v6.d REVIEW] Remove Bag/Set/Mix.new !=:= tests

In earlier discussions[^1] we agreed that it's sensible to allow implementations to intern objects that have the same .WHICH and these tests would block such interning.
Orig: github.com/perl6/roast/commit/cbfc93971
  [1] github.com/rakudo/rakudo/issues/19...-400833085
18:08
Zoffix .tell timotimo IIRC you worked on async socks in the past. Would you be able to review and merge github.com/perl6/nqp/pull/505 ? 18:22
yoleaux Zoffix: I'll pass your message to timotimo.
Geth 6.d-prep: d92a18d4cc | (Zoffix Znet)++ (committed using GitHub Web editor) | TODO/README.md
Add Pod to non-in-depth review status of Zofreviews
18:30
Kaiepi .tell timotimo don't merge the asyncsocket op pullreqs yet, i need to tweak it a bit so it'll actually be possible to get the fd of sockets in the jvm later on
yoleaux Kaiepi: I'll pass your message to timotimo.
Zoffix .tell samcv would you double-check this commit? github.com/perl6/roast/commit/6c5aaa7e7 `man sprintf` suggests it's right and output should have a plus, but Perl 5 `perl -e 'printf("%+ 10b\n", 9)'` does not produce a `+`. Wondering if P5 just gets it wrong or if there's something missing 18:57
yoleaux Zoffix: I'll pass your message to samcv.
Geth roast: bf3d4c484a | (Zoffix Znet)++ | S02-types/native.t
[v6.d REVIEW] Fix incorrect skip count

And considering dogbert never noticed this, suggests $?BITS isn't actually 32 on 32-bit platforms
19:00
roast: c5f03163d1 | (Zoffix Znet)++ | 2 files
[v6.d REVIEW] Appendicize questionable `int` limits test

It might be fine, but I'm unsure whether it's accurate to say
  `int` can be 32 bits on 32-bit platforms (and it isn't on MoarVM),
which the text of the test suggests.
Shove it into appendices, until our natives are more solidly implemented.
Orig: github.com/perl6/roast/commit/f0dbe1c97
19:03
dogbert11 Zoffix: you're right. 'say $?BITS' returns 64 on 32 bit systems
Zoffix m: dd Kernel.bits 19:04
camelia Cannot look up attributes in a Kernel type object
in block <unit> at <tmp> line 1
Zoffix m: dd $*KERNEL.bits 19:05
camelia 64
Zoffix dogbert11: what about that one ^
dogbert11 Zoffix: 32
Zoffix Thanks. 19:06
R#2367 19:08
synopsebot R#2367 [open]: github.com/rakudo/rakudo/issues/2367 [6.d review][@LARRY] What's the deal with $?BITS ?
samcv .tell Zoffix I think that is right. all our other tests are that way where we print the + 19:27
yoleaux 18:57Z <Zoffix> samcv: would you double-check this commit? github.com/perl6/roast/commit/6c5aaa7e7 `man sprintf` suggests it's right and output should have a plus, but Perl 5 `perl -e 'printf("%+ 10b\n", 9)'` does not produce a `+`. Wondering if P5 just gets it wrong or if there's something missing
samcv: I'll pass your message to Zoffix.
samcv so i'd think we'd have to change the spec otherwise
.tell Zoffix ok this is why: sprintf('%b', -1) #> 1111111111111111111111111111111111111111111111111111111111111111 19:32
yoleaux samcv: I'll pass your message to Zoffix.
samcv that's on Perl 5. so it prints it out the bits stored for a 64 bit integer 19:33
so basically our %b is 100% broken because it doesn't print things out as the exact 64 bit bits the computer stores them as. or our %b implementation works like our other integer format prints and is just a plain number and not two's complement notation 19:35
timotimo is sprintf able to differentiate between an int and an Int being passed? 19:36
yoleaux 18:22Z <Zoffix> timotimo: IIRC you worked on async socks in the past. Would you be able to review and merge github.com/perl6/nqp/pull/505 ?
18:30Z <Kaiepi> timotimo: don't merge the asyncsocket op pullreqs yet, i need to tweak it a bit so it'll actually be possible to get the fd of sockets in the jvm later on
AlexDaniel hah
Kaiepi timotimo: should be good for review now 19:45
Geth roast: c820202a23 | (Zoffix Znet)++ | S32-str/utf8-c8.t
[v6.d REVIEW] Check we actually match right stuff

Orig: github.com/perl6/roast/commit/1ba17759b
20:17
Geth roast: 0e059d4078 | (Zoffix Znet)++ | S12-enums/basic.t
[v6.d REVIEW] Test actual type of enum.enums

Orig: github.com/perl6/roast/commit/1fbabe9b5
21:06
Geth roast: 4a740af6f1 | (Zoffix Znet)++ | S02-types/range.t
[v6.d REVIEW] Remove pointless comment
21:39
Geth 6.d-prep: 45532479ca | (Zoffix Znet)++ (committed using GitHub Web editor) | TODO/README.md
Update zofcommit in roast review
23:19
Geth roast: 6279673fa1 | (Zoffix Znet)++ | S32-num/is-prime.t
[v6.d REVIEW] Toss .is-prime tests on unreduced rats

It's now a requirement for Rationals to be reduced on creation, so these tests don't make sense and .is-prime is well-covered already
Orig: github.com/perl6/roast/commit/a71839ea2
23:41