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.
buggable 🎺🎺🎺 It's time for the monthly Accidental /win Lottery 😍😍😍 We have 4 ballots submitted by 4 users! DRUM ROLL PLEASE!... 00:00
And the winning number is 42! Congratulations to lizmat! You win a can of WD40!
Geth nqp: a911a64b26 | (Elizabeth Mattijsen)++ | tools/build/MOAR_REVISION
Bump MoarVM to get Unicode 11.0
08:24
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...6-g80397b2
brrt ohai 08:37
t/05-messages/moar/01-errors.t relies on an implementation detail of MoarVM
I changed that implementation detail and now it fails
|Tux| 31839 Segmentation fault (core dumped) perl6 -Ilib csv-ip5xs.pl < /tmp/hello20.csv 08:49
Rakudo version 2018.09-66-g479348f31 - MoarVM version 2018.09-116-g80397b20f
csv-ip5xs0.925 - 0.926
csv-ip5xs-207.311 - 7.311
csv-parser20.909 - 21.058
csv-test-xs-200.447 - 0.464
test7.947 - 8.185
test-t1.674 - 1.712
test-t --race0.810 - 0.811
test-t-2028.730 - 28.864
test-t-20 --race10.595 - 11.240
lizmat close, but no cigar
|Tux| size matters. lizmat can you investigate? 08:51
$ cat /tmp/hello20.csv | perl6 -Ilib csv-ip5xs.pl
1000000
$ cat /tmp/hello20.csv{,} | perl6 -Ilib csv-ip5xs.pl 08:52
Segmentation fault (core dumped)
brrt |Tux| - MVM_JIT_DISABLE=1 ? 08:54
|Tux| That is a *lot* slower! but it PASSs 08:55
brrt hmmpf 08:56
|Tux| have I just caused work for you?
brrt yes
:-)
how do I reproduce
lizmat [Tux]: I'm afraid I can't reproduce :-( 08:57
|Tux| gist.github.com/Tux/b688fb0c152a0b...7eb18a806e
|Tux| This is perl 5, version 28, subversion 0 (v5.28.0) built for x86_64-linux-thread-multi-ld 08:58
This is Rakudo version 2018.09-66-g479348f31 built on MoarVM version 2018.09-116-g80397b20f
implementing Perl 6.c.
200000 lines of that CSV went OK, 400000 did not 08:59
Geth rakudo: acd8325cbc | (Elizabeth Mattijsen)++ | tools/build/NQP_REVISION
Bump NQP to get unicode 11

This breaks a number of Grapheme tests at the moment, and it makes a number of TODO tests pass. Since the breakage seems limited to Unicode, I'm bumping anyway, in the hopes that the breakage will soon be fixed.
09:00
¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....1-ga911a64
|Tux| It did not core under gdb :( 09:08
5000000
[Thread 0x7ffff62f6700 (LWP 12305) exited]
[Inferior 1 (process 12281) exited normally]
timotimo it's not as useful as having a "still almost alive" process, but if it dumps core, you can add gdb after the fact 09:09
|Tux| there is no core dump :(
timotimo perhaps check /proc/sys/kernel/core_pattern 09:11
brrt oh, hmm, i don't have Inline::Perl5 installed
|Tux| (free) time ran out $work, sorry 09:12
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Bump NQP to get unicode 11 10:00
travis-ci.org/rakudo/rakudo/builds/435499797 github.com/rakudo/rakudo/compare/4...d8325cbcc3
buggable [travis build above] ☠ All failures are due to: failed make test (6 failures). Across all jobs, 4 unique test files failed. 10:01
samcv lizmat: i will bump roast now 10:12
Geth roast/Unicode-11.0: 4 commits pushed by (Samantha McVey)++ 10:13
roast: 94e7daa351 | (Samantha McVey)++ | 6 files
Update Emoji and Grapheme Break Tests to version Unicode version 11.0
10:14
roast: e9c44baedd | (Samantha McVey)++ | 6 files
Update Unicode Collation tests for Unicode 11.0
roast: e23f1d88dc | (Samantha McVey)++ | docs/unicode-generated-tests.asciidoc
Make sure to mention the updated GraphemeBreakTest.t in docs
Geth rakudo: samcv self-assigned Need to Regenerate Property mappings with tools/build/makeUNIPROP.p6 github.com/rakudo/rakudo/issues/2337
5c272ee418 | (Samantha McVey)++ | 7 files

Now that MoarVM has Unicode 11 support we now use the 11 files instead. This should fix some test failures due to the recent Moar bump.
10:24
samcv tracking unicode stuff here github.com/rakudo/rakudo/issues/2337 10:33
lizmat samcv++ 10:38
spectest clean again for me 10:45
samcv lizmat: and don't unfudge passing unicode tests atm. i will get to it tomorrow and then look more closely :) 10:50
lizmat samcv: okidoki! 10:50
sleep well!
lizmat notable6: weekly 11:14
notable6 lizmat, 13 notes: gist.github.com/a73b8c05159fc9d3e6...9a6206b08e
lizmat notable6: weekly reset 11:22
notable6 lizmat, Moved existing notes to “weekly_2018-10-01T11:22:36Z”
AlexDaniel lizmat: reportable weekly report: gist.github.com/dc2705c33363655590...4ee7eadb2d
AlexDaniel lizmat: reportable monthly report: gist.github.com/d2e8948e82d6af63ef...50f8b12938 11:23
lizmat AlexDaniel++
travis-ci Rakudo build failed. Samantha McVey 'Update the Unicode tests to use 11.0 datafiles 11:28
travis-ci.org/rakudo/rakudo/builds/435533168 github.com/rakudo/rakudo/compare/a...272ee418b9
AlexDaniel lizmat: also we probably need a mention of the upcoming squashathon 11:28
there's no guide for it yet, unfortunately 11:29
buggable [travis build above] ☠ All failures are due to: failed make test (6 failures). Across all jobs, 2 unique test files failed.
lizmat :-(
AlexDaniel lizmat: but I'm on it right now and it will be here: github.com/perl6/ecosystem/wiki/SQUASHathon 11:30
lizmat AlexDaniel: cool
AlexDaniel: s/Octobre/October/ 11:33
AlexDaniel yeah it's not the only thing, I'm rewriting :)
lizmat oki! 11:34
AlexDaniel lizmat: updated, but I'm still working on it 11:45
lizmat: btw is there a picture of *one* camelia? :)
lizmat ok... will publish in 15 minutes or so
AlexDaniel: www.facebook.com/photo.php?fbid=20...&ifg=1 ? 11:46
AlexDaniel lizmat: omg they are so beautiful 11:47
lizmat And yet another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2018/10/01/...-than-two/ 12:07
Geth rakudo: 64137bd21c | (Elizabeth Mattijsen)++ | src/core/Mu.pm6
Fix problem with custom bless() found by Xliff
12:20
robertle weekly: yet another benchmark github.com/robertlemmen/p6bench 12:22
notable6 robertle, Noted!
lizmat robertle: sorry, too late, but you're indirectly mentioned in a comment on Wendy's blog post, right?
robertle sure, not worried about it, just thought I'll mention it 12:24
Geth roast: c9efd5c98f | (Elizabeth Mattijsen)++ | S12-construction/construction.t
Add test for github.com/rakudo/rakudo/commit/64137bd21c
12:29
brrt .tell Zoffix I broke t/05-messages/moar/01-errors.t, but we really need to talk about that test 17:08
yoleaux brrt: I'll pass your message to Zoffix.
Geth rakudo: ab83b9d82b | (Elizabeth Mattijsen)++ | src/Perl6/World.nqp
Use same QAST sequence for native attributes

Before, both opaque and native attributes used a 2-step approach to initializing attributes: first check with nqp::existskey if the argument has been specified, then if it exists, do an nqp::atkey. This was changed recently with 2fd8ffe56ad18324b for opaque attributes. This commit does the same for native attributes.
This makes initialization of native attributes only very marginally faster.
19:55
lizmat .tell jnthn I wonder whether our approach in BUILDALL could be improved by also generating a custom new method with a dedicated signature 20:09
yoleaux lizmat: I'll pass your message to jnthn.
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Use same QAST sequence for native attributes 20:54
travis-ci.org/rakudo/rakudo/builds/435806899 github.com/rakudo/rakudo/compare/6...83b9d82b23
buggable [travis build above] ☠ All failures are due to: failed make test (6 failures). Across all jobs, only t/05-messages/moar/01-errors.t test file failed. 20:55
Geth roast: 99a225b2e8 | (Samantha McVey)++ | 6 files
Update mass-roundtrip-*.t for Unicode 11.0

Tests haven't changed, but I altered the test generator so it adds the version the test file was generated from at the top to make it easier to tell which version it was generated from.
21:00