»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
AlexDaniel . 00:26
yoleaux 7 Apr 2018 17:29Z <smls> AlexDaniel: I've uploaded my codegolf tests here: github.com/smls/coding-challenges/...r/codegolf
AlexDaniel notable6: release try these tests: github.com/smls/coding-challenges/...r/codegolf 00:27
notable6 AlexDaniel, Noted!
AlexDaniel squashable6: status 00:55
squashable6 AlexDaniel, 🍕🍕 SQUASHathon is in progress! The end of the event in ≈11 hours. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel, Log and stats: gist.github.com/45cb166710239b3142...79a8afaad0
squashable6 🍕🍕🍕 First contribution by MasterDuke17++! ♥ 01:05
Geth roast: b477bee606 | MasterDuke17++ (committed using GitHub Web editor) | CONTRIBUTING.md
Add a couple of suggestions

Wording stolen nearly verbatim from zoffix++.
Geth rakudo: ronaldxs++ created pull request #1706:
Update 05-messages/03-errors.t test for RT #132238		 01:11
squashable6 🍕 ronaldxs++ opened pull request “Update 05-messages/03-errors.t test for RT #132238”: github.com/rakudo/rakudo/pull/1706
synopsebot RT#132238 [new]: rt.perl.org/Ticket/Display.html?id=132238 [LTA] $<> and “Corresponding … was at line”
squashable6 🍕 zoffixznet++ merged pull request “Update 05-messages/03-errors.t test for RT #132238”: github.com/rakudo/rakudo/pull/1706 01:13
Geth rakudo: 27e77c8d03 | (Ronald Schmidt)++ | t/05-messages/03-errors.t
Update 05-messages/03-errors.t test for RT #132238 clearer error message for unclosed <> hash quote index operator
squashable6 🍕 zoffixznet++ wrote a comment on “Update 05-messages/03-errors.t test for RT #132238”: github.com/rakudo/rakudo/pull/1706...-379510886
Geth rakudo: 9ff655645b | (Zoffix Znet)++ (committed using GitHub Web editor) | t/05-messages/03-errors.t
Merge pull request #1706 from ronaldxs/rakudo-t-RT-132238-match-hash-2

Update 05-messages/03-errors.t test for RT #132238
titsuki_ bisectable: say (1 +< -10) 02:04
bisectable6 titsuki_, On both starting points (old=2015.12 new=9ff6556) the exit code is 0 and the output is identical as well 02:04
titsuki_, Output on both points: «0␤»
titsuki_ bisectable: say (1 +> -10) 02:05
bisectable6 titsuki_, On both starting points (old=2015.12 new=9ff6556) the exit code is 0 and the output is identical as well 02:06
titsuki_, Output on both points: «1024␤»
titsuki_ bisectable: say (1 +< -100000000000) 02:10
bisectable6 titsuki_, Bisecting by exit code (old=2015.12 new=9ff6556). Old exit code: 1
titsuki_, bisect log: gist.github.com/779c0cfd73ed4af784...bcc8108d67
titsuki_, (2017-05-10) github.com/rakudo/rakudo/commit/64...68243f0115
titsuki_ bisectable: say (1 +> 100000000000) 02:16
bisectable6 titsuki_, Bisecting by exit code (old=2015.12 new=9ff6556). Old exit code: 1
titsuki_, bisect log: gist.github.com/2c95a78636f5c783d0...cd50eb628a
titsuki_, (2017-05-10) github.com/rakudo/rakudo/commit/ef...459c888f68
squashable6 🍕 titsuki++ opened issue “The bitshift operation o…”: github.com/rakudo/rakudo/issues/1707 02:49
🍕🍕🍕 First contribution by titsuki++! ♥
squashable6 🍕 AlexDaniel++ opened issue “… xx ∞ should return a Seq”: github.com/rakudo/rakudo/issues/1708 03:39
Geth roast: b320464868 | (Aleks-Daniel Jakimenko-Aleksejev)++ | S03-operators/repeat.t
Test xx not dying with large repetition count

Closes RT#126014.		 03:43
synopsebot RT#126014 [open]: rt.perl.org/Ticket/Display.html?id=126014 Too many repetitions with xx operator causes out of memory; should it work lazily?
Geth roast: 19344b3c7f | (Aleks-Daniel Jakimenko-Aleksejev)++ | S03-operators/repeat.t
Oops. Fix RT number in the last commit
 03:45
squashable6 🍕 AlexDaniel++ wrote a comment on commit b320464868d3: github.com/perl6/roast/commit/b320...t-28472097 03:46
squashable6 🍕 zoffixznet++ wrote a comment on “… xx ∞ should return a Seq”: github.com/rakudo/rakudo/issues/17...-379519020 04:19
Geth roast: a7af87465e | (Aleks-Daniel Jakimenko-Aleksejev)++ | S32-io/io-handle.t
Test that nl-in has \n and \r\n

Closes RT#131858, but I still don't understand why it gives an array instead of a Junction. There's probably a good reason, but speccing it this way is probably wrong. Please adjust.		 04:22
synopsebot RT#131858 [open]: rt.perl.org/Ticket/Display.html?id=131858 [REGRESSION] default $.nl-in on IO::Handle does not correctly work in subclasses
squashable6 🍕 AlexDaniel++ wrote a comment on “… xx ∞ should return a Seq”: github.com/rakudo/rakudo/issues/17...-379519219 04:24
AlexDaniel Zoffix: can you take a look at github.com/perl6/roast/commit/a7af...7f1bbff0f1 ? 04:25
AlexDaniel goes for another nap 04:28
araraloren good noon :) 04:51
squashable6 🍕 zoffixznet++ self-assigned issue “… xx ∞ should return a Seq”: github.com/rakudo/rakudo/issues/1708 04:57
Zoffix AlexDaniel: `is` is not an appropriate test here. It'll stringify the return value so even if it returns ((((((((“\n”,)))))), “\r\n”), the test would still pass. As for Array vs. Junction. Junctions aren't introspectable, so I wouldn't say that's an appropriate type, so you're left with array/list and since it's a mutable attribute, array seems a fine choice. 04:58
AlexDaniel Zoffix: I see 04:59
Zoffix and I'll fix the … xx ∞ thing shortly, to behave like the 42 xx 2**99999 case
AlexDaniel Zoffix: so what's the appropriate test sub there? is-deeply will complain about list vs array difference, and I don't feel like this is right 05:00
Zoffix AlexDaniel: what's right? 05:01
AlexDaniel I'd say some sort of positional with the required elements 05:02
Zoffix It's a mutable attribute. If it's not an array, it'll crash
m: class { has $.nl-in = (1, 2, 3) }.new.nl-in = 1, 2, 3 05:04
camelia Cannot modify an immutable Int (1)
in block <unit> at <tmp> line 1
AlexDaniel m: my $z = class Z is IO::Handle { }.new; dd $z.nl-in; $z.nl-in = ("\n", "\r\n"); dd $z.nl-in; $z.nl-in = <a b c>; dd $z.nl-in
camelia Z is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in method new at src/RESTRICTED.setting line 32
in block <unit> at <tmp> line 1
AlexDaniel what the f
e: my $z = class Z is IO::Handle { }.new; dd $z.nl-in; $z.nl-in = ("\n", "\r\n"); dd $z.nl-in; $z.nl-in = <a b c>; dd $z.nl-in
evalable6 $["\n", "\r\n"]
$("\n", "\r\n")
$("a", "b", "c")
AlexDaniel I guess I'm confused about something 05:06
Zoffix m: class Foo { has $.nl-in = (1, 2, 3) }; my Array $x = Foo.new.nl-in
camelia Type check failed in assignment to $x; expected Array but got List ($(1, 2, 3))
in block <unit> at <tmp> line 1
Geth roast: ec2b4c3a7c | (Aleks-Daniel Jakimenko-Aleksejev)++ | S32-io/io-handle.t
Change nl-in test to is-deeply
 05:07
Geth roast: 22001cb924 | (Aleks-Daniel Jakimenko-Aleksejev)++ | S04-statements/for.t
Use is-deeply in hole tests
 05:12
Geth rakudo: 5f256baec2 | (Zoffix Znet)++ | src/core/Rakudo/Iterator.pm6
Implement optimized Iterator methods for `42 xx ...`

Fixes R#1708 github.com/rakudo/rakudo/issues/1708		 05:25
synopsebot R#1708 [open]: github.com/rakudo/rakudo/issues/1708 … xx ∞ should return a Seq
Geth roast: b4b0d584bc | (Zoffix Znet)++ | S03-operators/repeat.t
Test opt methods/sinkage of `xx` with simple value

Closes github.com/rakudo/rakudo/issues/1708 R#1708 Rakudo fix: github.com/rakudo/rakudo/commit/5f256baec2		 05:29
squashable6 🍕 zoffixznet++ closed issue “… xx ∞ should return a Seq”: github.com/rakudo/rakudo/issues/1708
Geth roast: 5fe5978a18 | (Zoffix Znet)++ (committed using GitHub Web editor) | CONTRIBUTING.md
Use implementation-neutral language
 05:51
Geth roast: 03ac03f16c | usev6++ | MISC/bug-coverage-stress.t
[JVM] Skip tests using 'atomicint'
 06:21
squashable6 🍕🍕🍕 First contribution by usev6++! ♥
Geth doc/master: 4 commits pushed by CC++, cc++, (Juan Julián Merelo Guervós)++ 06:36
Geth doc: bd81a89833 | (JJ Merelo)++ | 2 files
Checks links in grammars

Closes #1901 since it was actually fixed. Maybe it was not in the published site yet. Also changes link to what the link points to and rephrases some sentences Atom does not like.		 07:01
Geth rakudo: 4402839adb | (Zoffix Znet)++ | src/Perl6/Actions.nqp
Fix crash in double-sink for for ... { Package::foo }

Fixes R#1693 github.com/rakudo/rakudo/issues/1693
After adding a sink call to the last statement, we UNWANTED, the guts and because of the setup, we ended up adding another sink call into the QAST inside UNWANTED, causing the error in the Issue. ... (5 more lines)		 07:29
synopsebot R#1693 [open]: github.com/rakudo/rakudo/issues/1693 [regression][⚠ blocker ⚠] Cannot find method 'Nil'
roast: 06d2abbc6c | (Zoffix Znet)++ | S04-statements/sink.t
Test sinkage of package call in for

Closes github.com/rakudo/rakudo/issues/1693 R#1693 Rakudo fix: github.com/rakudo/rakudo/commit/4402839adb		 07:30
squashable6 🍕 zoffixznet++ closed issue “Cannot find method 'Nil'”: github.com/rakudo/rakudo/issues/1693
Geth doc: 55eaffe2f1 | (JJ Merelo)++ | doc/Language/concurrency.pod6
Adds headings for react and whenever

While keeping the index. It's level 3 indexing, although it shows up at the same level in the headings. Contributes to improve a bit the tutorial, thus refs #1906, and at the same time fixes #1552		 07:45
synopsebot Link: doc.perl6.org/language/concurrency
AlexDaniel squashable6: status 07:50
squashable6 AlexDaniel, 🍕🍕 SQUASHathon is in progress! The end of the event in ≈4 hours. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel, Log and stats: gist.github.com/d511d8fc3a95d0c3cb...907f4306a1
Geth rakudo: 9bcc8e523e | (Zoffix Znet)++ | src/core/CompUnit/PrecompilationUnit.pm6
Fix thinko in CURR conditional

Part of fixing R#1662 github.com/rakudo/rakudo/issues/1662		 07:52
synopsebot R#1662 [open]: github.com/rakudo/rakudo/issues/1662 [⚠ blocker ⚠] Windows Issues with "Replace error-prone file timestamp check by checksum when loading deps"
jmerelo Can I propose a topic for the next SQUASHathon? I'd like to focus on github.com/perl6/doc/issues/561
buggable New CPAN upload: App-FindSource-0.0.2.tar.gz by ALOREN cpan.metacpan.org/authors/id/A/AL/...0.2.tar.gz 07:54
AlexDaniel “Documentation has changed since you last contributed 9 days ago. Take a look before submitting an issue: Contributing guidelines (Last updated 23 hours ago)” 08:09
heh, that's a very interesting kind of a notification
jmerelo AlexDaniel: nothing too important. Mostly related to HTML layout. 08:10
AlexDaniel jmerelo: yeah you can propose, but there are things to consider. For example, the topic should be approachable for newcomers
it's a doc squashathon next month for sure, feel free to narrow it down if you think that'd be useful 08:11
but maybe just say during the squashathon that you want more attention to that issue, and we'll do our best to help there
jmerelo AlexDaniel: OK, I'll do that. The thing is that I keep coming back to that issue whenever I try to fix things somewhere else.
jmerelo AlexDaniel: I'll try to label some issues as easy to newcomers also, to help people get incorporated. But come this point, there are not many of those... Most issues are quite gnarly. 08:13
AlexDaniel so we are 12+6 tickets down from the start of the squashathon, I think we should push a bit more right now :) 08:13
jmerelo AlexDaniel: I'm quite the newcomer here, but if you feel there's something easy I could do, please feel free. 08:14
AlexDaniel jmerelo: yes, I know. All lhfs were squashathoned away very early I think
jmerelo: well at this point I think you should just scroll through fail.rakudo.party/t/TESTNEEDED and see if there's any issue that can be easily tested 08:15
and if so, just write the test and submit a PR for roast or rakudo
probably skip those that are EXOTICTEST or PERFTEST
jmerelo AlexDaniel: OK. I'll try to do that. Thanks! 08:16
squashable6 🍕 zoffixznet++ labeled issue “A race in HyperRace causing rare crash/hang”: github.com/rakudo/rakudo/issues/1709 08:18
🍕 zoffixznet++ opened issue “A race in HyperRace causing rare crash/hang”: github.com/rakudo/rakudo/issues/1709
🍕 zoffixznet++ labeled issue “A race in HyperRace causing rare crash/hang”: github.com/rakudo/rakudo/issues/1709
squashable6 🍕 zoffixznet++ opened issue “New stresstest floppers”: github.com/rakudo/rakudo/issues/1710 08:26
squashable6 🍕 AlexDaniel++ labeled issue “New stresstest floppers”: github.com/rakudo/rakudo/issues/1710 08:38
jmerelo AlexDaniel: I'm checking out this one rt.perl.org/Public/Bug/Display.htm...et-history but having lots of trouble logging in. If I use GitHub, I can't display the ticket, if I use anything else, it simply does not work. 08:44
Anyway, the only problem I see there is (possibly) the declaration of the sub as "our". I can write a test for that, but it can also be closed safely, I guess. 08:45
commit: releases module foo { our sub bar () { return "baz" }; } say foo::bar; 08:50
committable6 jmerelo, gist.github.com/50a38176af49a2b417...e5c4271b21
jmerelo commit: releases module foo { our sub bar () { return "baz" }; }; say foo::bar; 08:51
committable6 jmerelo, ¦releases (28 commits): «baz␤»
AlexDaniel jmerelo: oh well… I think we'll have to reproduce the whole thing in whateverable to figure out what's going on there 08:53
but that's not too hard, let's see
jmerelo AlexDaniel: I think rt.perl.org/Public/Bug/Display.htm...et-history can be closed. It was probably some problem with the local installation. 08:53
AlexDaniel we'll have to try it to know for sure 08:54
jmerelo AlexDaniel: see above. I think the error was probably local due to some old version of the library not including parse-string in the path 08:55
AlexDaniel e: mkdir ‘sandbox/rt125384/lib/Pod/Perl5’ 08:57
evalable6
synopsebot RT#125384 [open]: rt.perl.org/Ticket/Display.html?id=125384 Could not find symbol '&parse-string' in installed module, local version works fine. 08:58
AlexDaniel e: spurt ‘sandbox/rt125384/lib/Pod/Perl5.pm’, ｢use Pod::Perl5::Grammar; use Pod::Perl5::ToHTML; module Pod::Perl5 { our sub parse-string (Str:D $pod) { Pod::Perl5::Grammar.parse($pod); } }｣
evalable6
squashable6 🍕 zoffixznet++ wrote a comment on “A race in HyperRace causing rare crash/hang”: github.com/rakudo/rakudo/issues/17...-379533340
AlexDaniel e: spurt ‘sandbox/rt125384/lib/Pod/Perl5/ToHTML.pm’, ｢class Pod::Perl5::ToHTML { }｣ 08:59
evalable6
AlexDaniel e: spurt ‘sandbox/rt125384/lib/Pod/Perl5/Grammar.pm’, ｢grammar Pod::Perl5::Grammar { token TOP { .* } }｣
evalable6
AlexDaniel now what… 09:00
e: shell(:cwd(‘sandbox/rt125384’), ｢PERL6LIB=lib perl6 -MPod::Perl5 -e 'say Pod::Perl5::parse-string("=pod\n\n")'｣) 09:01
evalable6 ｢=pod

｣
AlexDaniel oh, that works?
nooow… 09:02
c: 2015.05 shell(:cwd(‘sandbox/rt125384’), ｢PERL6LIB=lib perl6 -MPod::Perl5 -e 'say Pod::Perl5::parse-string("=pod\n\n")'｣)
committable6 AlexDaniel, ¦2015.05: «Unexpected named parameter 'cwd' passed␤ in block <unit> at /tmp/9x3_m2xkii:1␤ in any <unit-outer> at /tmp/9x3_m2xkii:1␤␤ «exit code = 1»»
AlexDaniel haha
c: 2015.05 shell(｢cd sandbox/rt125384; PERL6LIB=lib perl6 -MPod::Perl5 -e 'say Pod::Perl5::parse-string("=pod\n\n")'｣) 09:03
committable6 AlexDaniel, ¦2015.05: «｢=pod␤␤｣␤»
AlexDaniel c: 2015.06 shell(｢cd sandbox/rt125384; PERL6LIB=lib perl6 -MPod::Perl5 -e 'say Pod::Perl5::parse-string("=pod\n\n")'｣)
committable6 AlexDaniel, ¦2015.06: «｢=pod␤␤｣␤»
AlexDaniel c: all shell(｢cd sandbox/rt125384; PERL6LIB=lib perl6 -MPod::Perl5 -e 'say Pod::Perl5::parse-string("=pod\n\n")'｣)
committable6 AlexDaniel, gist.github.com/3b421ae9ffb7fadd93...4b226eff03 09:04
AlexDaniel jmerelo: well, I guess you're right
jmerelo: oh wait, it needs ｢panda install ./｣ ? 09:05
squashable6 🍕 lizmat++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379534089 09:08
lizmat decommute from GPW& 09:09
squashable6 🍕 zoffixznet++ wrote a comment on “New stresstest floppers”: github.com/rakudo/rakudo/issues/17...-379534521 09:15
🍕 ronaldxs++ wrote a comment on “Update S02-types/hash_ref.t test for RT #132238”: github.com/perl6/roast/pull/413#is...-379534530
🍕 ronaldxs++ closed pull request “Update S02-types/hash_ref.t test for RT #132238”: github.com/perl6/roast/pull/413
synopsebot RT#132238 [resolved]: rt.perl.org/Ticket/Display.html?id=132238 [LTA] $<> and “Corresponding … was at line”
AlexDaniel jmerelo: OK! -1 ticket then: rt.perl.org/Ticket/Display.html?id...xn-1545278 09:18
jmerelo: next?
jmerelo AlexDaniel: :-)
AlexDaniel: Sure. 09:19
squashable6 🍕 zoffixznet++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379534906 09:20
squashable6 🍕 AlexDaniel++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379535029 09:22
squashable6 🍕 AlexDaniel++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379535207 09:24
ufobat_ is there any work in progress on some i18n modules? 09:40
AlexDaniel ufobat_: I don't think so: github.com/perl6/perl6-most-wanted...8n-and-nlp 09:41
ufobat_ okay, ty 09:42
Geth roast: JJ++ created pull request #414:
Adds test for RT#126394		 10:02
squashable6 🍕 JJ++ opened pull request “Adds test for RT#126394”: github.com/perl6/roast/pull/414
🍕🍕🍕 First contribution by JJ++! ♥
synopsebot RT#126394 [open]: rt.perl.org/Ticket/Display.html?id=126394 [BUG] Assigning Failure to typed variable swallows Failure information 10:03
squashable6 🍕 zoffixznet++ merged pull request “Adds test for RT#126394”: github.com/perl6/roast/pull/414
🍕 zoffixznet++ wrote a comment on “Adds test for RT#126394”: github.com/perl6/roast/pull/414#is...-379537760
Geth roast: ce173d4c66 | (JJ Merelo)++ | S04-exceptions/fail.t
Adds test for RT#126394

  Checking that it fails and effectively includes the important message.
roast: 295d101a09 | (Zoffix Znet)++ (committed using GitHub Web editor) | S04-exceptions/fail.t
Merge pull request #414 from JJ/master

Adds test for RT#126394		 10:04
roast: 39b4d2a83e | (Zoffix Znet)++ (committed using GitHub Web editor) | S04-exceptions/fail.t
Remove needless ticket title

  (along with possibly-conflicting with fudger `#?` portion)
jmerelo You can't see me now, but I'm smiling with a smile that's 4 inches wider than my face 10:05
OK, I see that. Still smiling. 10:06
Geth roast: d3c04b47ca | (Zoffix Znet)++ (committed using GitHub Web editor) | S04-exceptions/fail.t
Add test description

  (and toss unneeded `say` code)		 10:07
jmerelo Right :-) 10:08
AlexDaniel jmerelo: should I close the ticket for you? 10:12
jmerelo AlexDaniel: Thanks! 10:18
timotimo m: m: (say "hi") xx Inf 10:32
camelia (timeout)hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi
hi…
squashable6 🍕 robertlemmen++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379539551
🍕🍕🍕 First contribution by robertlemmen++! ♥
timotimo OK, so that doesn't use UnendingValue 10:33
ufobat_ is there a way to access the local file pod data? i am looking for a workaround for the NYI pod variable $= 10:37
I've got a bunch of data that's documented in the pod and i need to get the same data into a local datastructe, without repeating myself 10:39
timotimo we do have $= partially implemented 10:41
Zoffix timotimo: yeah, and that one can't be cheated to not sink either, 'cause sinking executes the code 10:42
ufobat_: $=pod should work
timotimo m: =begin pod␤Hello␤=end pod␤say $=POD.perl
camelia 5===SORRY!5=== Error while compiling <tmp>
Pod variable $=POD not yet implemented. Sorry.
at <tmp>:4
------> 3say $=POD7⏏5.perl
timotimo m: =begin pod␤Hello␤=end pod␤say $=pod.perl
camelia [Pod::Block::Named.new(name => "pod", config => {}, contents => [Pod::Block::Para.new(config => {}, contents => ["Hello"])])]
Geth roast: 1f171a9d2f | (Aleks-Daniel Jakimenko-Aleksejev)++ | S07-hyperrace/basics.t
Add tests for hyperized s///

Should also catch other reliability issues of .hyper, including segfaults. Closes RT#126752.		 10:44
synopsebot RT#126752 [open]: rt.perl.org/Ticket/Display.html?id=126752 [BUG] $/ vs .hyper gets confused
ufobat_ it seems not to be documented, or i coudn't find it 10:46
is just $=pod working? 10:47
ufobat_ yeah i guess it will work :) 10:51
AlexDaniel why am I using “RT#…” when everything else in the file has “RT #…” with a space
am I blind or what
jmerelo .tell moritz I still need the docs.per6.org website logs, or the public ones updated, to search for 404s and, if possible, referrers to them. 10:55
yoleaux jmerelo: I'll pass your message to moritz.
ufobat_ jmerelo, i would find it good if the pod section of the docs would mention how to access pod via $= 10:57
jmerelo ufobat_: OK
squashable6 🍕 robertlemmen++ wrote a comment on “rakudo 2018.02 test fail…”: github.com/rakudo/rakudo/issues/16...-379541036
🍕 robertlemmen++ opened issue “intermittent "Bytecode v…”: github.com/rakudo/rakudo/issues/1711 11:00
squashable6 🍕 robertlemmen++ wrote a comment on “Rakudo 2017.10 fails to …”: github.com/rakudo/rakudo/issues/12...-379541231 11:01
AlexDaniel squashable6: status
squashable6 AlexDaniel, 🍕🍕 SQUASHathon is in progress! The end of the event is in 58 minutes. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel, Log and stats: gist.github.com/a6e0160962a391743e...055578c69d
AlexDaniel oh 58 minutes more!
squashable6 🍕 AlexDaniel++ labeled issue “intermittent "Bytecode v…”: github.com/rakudo/rakudo/issues/1711 11:05
dogbert17 [Tux]: I believe that you got lucky today, the SEGV you found is still present 11:20
releasable: next 11:22
releasable6 dogbert17, Next release in ≈13 days and ≈7 hours. 8 blockers. 0 out of 189 commits logged (⚠ 9 warnings)
dogbert17, Details: gist.github.com/724ebb60723cc411e7...e91f78cd34
dogbert17 eight blockers, could be a record 11:23
AlexDaniel dogbert17: we had more, Zoffix++ closed a bunch recently I think 11:30
squashable6 🍕 dogbert17++ wrote a comment on “New stresstest floppers”: github.com/rakudo/rakudo/issues/17...-379543072 11:32
dogbert17 damn, you can't do anything without a bot picking it up :) 11:34
squashable6 🍕 robertlemmen++ edited issue “intermittent "Bytecode v…”: github.com/rakudo/rakudo/issues/1711 11:35
dogbert17 wonders if github.com/rakudo/rakudo/issues/1654 is still present
AlexDaniel dogbert17: you can – focus on testneeded tickets :P 11:36
squashable6 🍕 AlexDaniel++ wrote a comment on “New stresstest floppers”: github.com/rakudo/rakudo/issues/17...-379543361 11:37
AlexDaniel dogbert17: also, you can do something like 11:39
squashable6: shut up
squashable6 AlexDaniel, .oO( Mmmm… pizza! )
AlexDaniel squashable6: be loud again
squashable6 AlexDaniel, ALRIGHT, LET'S DO IT!!!
jmerelo :-) 11:40
squashable6 🍕 dogbert17++ wrote a comment on “New stresstest floppers”: github.com/rakudo/rakudo/issues/17...-379543561 11:40
squashable6 🍕 dogbert17++ wrote a comment on “Fresh-clone builds fail …”: github.com/rakudo/rakudo/issues/16...-379544462 11:55
Geth roast: dogbert17++ created pull request #415:
Add test for RT #127974		 12:15
dogbert17 tada
synopsebot RT#127974 [open]: rt.perl.org/Ticket/Display.html?id=127974 [CONC] sprintf() not threadsafe/reentrant if the format tokens use explicit indices
dogbert17 was probably too late, i.e. no pizza 12:16
jmerelo squashable: time 12:21
squashable6 jmerelo, I cannot recognize this command. See wiki for some examples: github.com/perl6/whateverable/wiki/Squashable
jmerelo squashable6: status
squashable6 jmerelo, Next SQUASHathon in 25 days and ≈21 hours (2018-05-05 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
jmerelo dogbert17: you can still squash as many bugs as you lile :-) 12:22
dogbert17 jmerelo: you're right, and besides, too much pizza isn't good for you :-) 12:23
jmerelo dogbert17: but I _love_ pizza! 12:24
AlexDaniel dogbert17: that test does not seem to be good enough I think 12:29
dogbert17: see gist.github.com/Whateverable/9fce9...048e1ee073
dogbert17: it says “ok” on 2016.01.1
.tell [Coke] can you write a description for ｢new｣ label on github.com/perl6/doc/labels ? Because I think it's currently being misused often 12:32
yoleaux AlexDaniel: I'll pass your message to [Coke].
jmerelo AlexDaniel: I see there are some going back to 2017 12:33
AlexDaniel: I add it to new issues, and try to delete them one month later or so
I'll proceed to delete a few 12:34
AlexDaniel jmerelo: yeah that's not what it was supposed to be I think. The idea was that ｢new｣ is for tickets which require entirely new text to be written, and ｢update｣ is when some text needs a tweak
jmerelo AlexDaniel: Then you're right. A description is needed 12:35
AlexDaniel interestingly, there was a time when we had *a lot* of labels, then we cleaned it up and we had just 4 or 5
jmerelo AlexDaniel: actually, there's a description here github.com/perl6/doc/blob/master/C...IBUTING.md
AlexDaniel now we have a lot of labels again :) 12:36
dogbert17 AlexDaniel: so lives-ok gave that a pass despite the messages?
AlexDaniel jmerelo: IMO put these into github label descriptions, and throw away that section
dogbert17: yea 12:37
jmerelo AlexDaniel: issue that
dogbert17 hmm, thought that would be a runtime error
AlexDaniel … did someone rename ｢meta｣ to ｢style｣?
jmerelo AlexDaniel: where? 12:38
AlexDaniel jmerelo: okay: github.com/perl6/doc/issues/1909 12:40
jmerelo: doc repo, the black ｢style｣ label was previously ｢meta｣ I think
jmerelo AlexDaniel: thanks! 12:42
AlexDaniel .tell [Coke] github.com/perl6/doc/issues/1909
yoleaux AlexDaniel: I'll pass your message to [Coke].
jmerelo AlexDaniel: I don't know if there's a log for activity in the repo to find out who did it, and maybe the reason why. 12:45
AlexDaniel not a big deal 12:48
Geth rakudo: ronaldxs++ created pull request #1712:
Update 05-messages/03-errors.t simpler and clearer tests for RT#132238		 13:03
synopsebot RT#132238 [resolved]: rt.perl.org/Ticket/Display.html?id=132238 [LTA] $<> and “Corresponding … was at line”
AlexDaniel squashable6: log 2018-04-07 13:29
squashable6 AlexDaniel, Log and stats: gist.github.com/454e1080b59a19e162...ecc16907b1
AlexDaniel so we closed 21 this time compared to the previous testneeded squashathon when we closed 18 13:31
not bad at all in my opinion :)
also, 20 people touched roast and rakudo repos during these 50 hours 13:32
that's a lot
not that different from other squashathons though 13:33
AlexDaniel gist.github.com/7d7c5c796d5d48527f...49e03121e9 13:36
35 tickets closed overall
AlexDaniel hmmm www.reddit.com/r/ProgrammerHumor/c..._problems/ 14:28
AlexDaniel the second line in any of these can be changed to “I won't use Perl 6” :) 14:31
RandalSchwartz rakudobrew build moar has been failing for a few days. is this expected? 15:48
AlexDaniel Ran
RandalSchwartz fails at "Configuration FAIL"
AlexDaniel RandalSchwartz: by any chance can it be this? alerts.perl6.org/alert/7
RandalSchwartz lemme loook 15:49
might be... lemme try the fix
APic May the Schwartz be with You! 15:50
RandalSchwartz Nope, still failed. 15:51
Failed running /opt/local/bin/perl5.26 Configure.pl --backends=moar --gen-moar --git-reference="/Users/merlyn/MIRROR/rakudobrew-GIT/bin/../git_reference" --make-install at /Users/merlyn/MIRROR/rakudobrew-GIT/bin/rakudobrew line 57.
not really anything else other than "Command failed" 15:52
anything else I can try? 15:53
maybe git clean in the repo?
timotimo buggable: sauce
huggable: sauce
huggable timotimo, Install latest version of Rakudo from source: github.com/zoffixznet/r#table-of-contents
timotimo RandalSchwartz: ^- better than rakudobrew unless you're a core developer who wants to switch rakudo multiple times per hour 15:54
RandalSchwartz so many years ago, I was building from the rakudo git, then I was told to use rakudobrew, and now it's something different again? :) 15:55
timotimo yes, one of the things problematic about rakuodbrew is you just get whatever's up on git at that moment 15:56
jmerelo RandalSchwartz: Hi! You can use rakudobrew. I use rakudobrew. It's OK for developers. The rationale behind this discouragement has more to do with first time users than with everyone
timotimo also, i can't help but repeat this funny anecdote all the time
at one point, rakudobrew switch used rm -rf ~/.rakudobrew/bin/* (or similar) to clean everything before it creates the shell scripts and links again 15:57
and someone decided to symlink that to their /usr/bin or some other system-wide bin folder
jmerelo I like it, anyway. I have HEAD if I need it and monthlies if I want that. But Rakudo Star is what we're encouraging now.
timotimo: wow
timotimo that was at least two years ago 15:58
RandalSchwartz and I build rakudo star from the zoffixnet github? 15:59
timotimo it automatically gives you the latest released tag
so you'll get newer than rakudo star
but not bleading edge, which isn't guaranteed to not crash
jmerelo RandalSchwartz: you can also download Debian/Ubuntu packages and a docker containers. We have the whole nine yards. 16:00
timotimo huggable: packages
huggable timotimo, CentOS, Debian, Fedora and Ubuntu Rakudo packages: github.com/nxadm/rakudo-pkg/releases
timotimo this also has alpine linux apks
jmerelo timotimo: right. Thanks!
timotimo there's even something there for windows subsystem for linux 16:01
jmerelo BTW, there's also rakudup now. rakudup.github.io/ 16:02
RandalSchwartz well, since I don't run linux anywhere..
timotimo if you're on mac, we do have rakudo star .dmg installers, don't we?
and for windows, see above :)
jmerelo Baseline is: you can't go wrong with rakudo star, rakudobrew is probably OK for most users (including me), there are binary versions for most OSs if you want that. 16:03
RandalSchwartz except that rakudobrew is failing
for the past few days
jmerelo RandalSchwartz: is it?
RandalSchwartz yes... see above
timotimo yeah, that's what he came to tell us
jmerelo OK, I'll go to the log. I came in the middle of the conversation. 16:04
timotimo what you pasted was really all output there was?
RandalSchwartz yeah, failed
no details
I would recognize details :)
I am a bit familiar with programming. :)
timotimo you should be able to cd into the folder it would have used to build stuff in
i did not mean to imply, sorry
RandalSchwartz hehe
perfectly ok
it's just funny when support people start asking me questions 16:05
and I say, well, I've already done a traceroute, and I suspect it's your router
timotimo sir, have you tried removing the plug and rotating it 180 degrees?
RandalSchwartz for 3-prong plugs, that's hard. :)
jmerelo I built HEAD a couple of days ago...I think I have perl 5.22... could that be a Perl error? 16:06
timotimo european engineering at its finest
anyway, if you cd into the folder rakudobrew was trying to use to build your stuff, you can try invoking "perl Configure.pl --gen-moar" manually and seeing if that also fails
that should allow for more fine-grained failure diagnostics 16:07
jmerelo RandalSchwartz: BTW, big admirer here. I bought one of the tees after the 1995 incident, with the names on it.
timotimo if rakudobrew outputs the commands it issues maybe you'll also see the --prefix it was trying to use
RandalSchwartz ok lemme try that 16:08
this is what's failing: Command failed (status 256): /opt/local/bin/perl5.26 Configure.pl --optimize --prefix=/Users/merlyn/MIRROR/rakudobrew-GIT/moar-master/install --make-install 16:10
what dir am I in?
jmerelo It's failing also for me now
RandalSchwartz heh
not just me then
jmerelo perl 5.20 in my case. 16:11
RandalSchwartz 5.26 for me 16:12
jmerelo But it does say something about the libatomic thingie, so it might have something to do with the alert Alex Daniel mentioned before
timotimo which Configure.pl is this, rakudo's? can you tell if it's one of the sub-project's configures that it should be calling?
jmerelo This alerts.perl6.org/alert/7
RandalSchwartz I tried deleting that
at least one of them. there were multiple ones 16:13
jmerelo No way, right?
RandalSchwartz what's the git clean that resets everything back to HEAD?
timotimo git reset --hard
but you might have to manually delete the folder anyway
RandalSchwartz no, that doesn't remove build products
timotimo oh, true 16:14
git clean -xdf or something?
RandalSchwartz too early in the morning to remember the incantation
yeah, that might be it
jmerelo It's working now...
timotimo i *think* this one won't be fixed by a clean, though
jmerelo rm -rf /home/jmerelo/.rakudobrew/moar-HEAD/nqp/MoarVM/3rdparty/ was the thing
timotimo or only if git clean descends into submodules maybe?
RandalSchwartz yeah, that didn't help
timotimo we had code in a folder for a while, then changed the folder to be a submodule instead 16:15
jmerelo I deleted _the whole_ 3rd party directory, not just the libatomic one
It's working now... 16:16
And it took shorter than usual: "Rakudo has been built and installed.
perl6 -v now returns "This is Rakudo version 2018.03-189-g9bcc8e523 built on MoarVM version 2018.03-59-gfc1bb3126" 16:17
... which is HEAD right now... github.com/rakudo/rakudo/commit/9b...116d33ac2f 16:18
RandalSchwartz Ahh... rm -rf **/*3rd*
and now it seems to be building
jmerelo RandalSchwartz: :-)
RandalSchwartz Done, moar-master built 16:21
This is Rakudo version 2018.03-189-g9bcc8e523 built on MoarVM version 2018.03-75-g156a7a6ff
Mine is newer!
RandalSchwartz carefully pets his shiny new perl6 16:22
jmerelo :-) MoarVM is newer, that's right. I'll have to update that...
RandalSchwartz OK, so I'll have to add that to my daily build script
the rm -rf :) 16:23
jmerelo RandalSchwartz: that always helps keep everything clean and tidy :-)
timotimo you'll only need to do that every few years
RandalSchwartz well, the git fetch will restore those tomorrow
timotimo the 3rdparty stuff is fetched by the Configure.pl that moarvm runs 16:25
RandalSchwartz ahh, ok 16:26
timotimo without them your moarvm build is not likely to succeed
jmerelo How can I update the MoarVM version? 16:27
timotimo rakudobrew will by default only take the version that we require 16:28
rakudobrew triple master master master should give you whatever's absolutely latest
jmerelo timotimo: it's deleting 3rdparty/ all by itself... 16:29
timotimo++
Well, that failed... 16:31
jmerelo This is Rakudo version 2018.03-189-g9bcc8e523 built on MoarVM version 2018.03-96-gfa167464a 16:35
RandalSchwartz: well, who's got the newest Perl 6 now? :-) 16:36
Now, I'll benchmark the evolutionary algorithms again :-)
lizmat jmerelo: if you have a --profile output, I'm willing to have a look at that :-) 16:38
jmerelo lizmat: I will give you one in a moment. 16:39
lizmat okido
jmerelo lizmat: crossover now has improved a lot. Not too much for small sizes, but it's 60% faster now for reasonable sizes. 16:43
timotimo how did that happen :) :) 16:44
jmerelo timotimo: it happened by Perl 6 having the most amazing team of developers there is out there :-)
timotimo there's always some way to improve 16:45
for example, i need to figure out how to see unanswered pull requests on all my own github repositories 16:46
timotimo i bet there's been a pull request or two open for months and i just didn't see it out of the >2k github notifications that have piled up again >:( 16:47
jmerelo lizmat: here's the profile jj.github.io/perl6eo/profile-152320...86118.html
timotimo now that's interesting. how come push-all and push aren't getting jitted there?
lizmat will looks after she's moved to the terrace 16:48
jmerelo timotimo: you can have them sent to a Slack channel so that now there's double chance you don't see it among the >2k notifications from Slack channels 16:49
timotimo it should be trivial to just put a search query in github ... 16:50
jmerelo timotimo: BTW, thanks a lot with that. With everything, in general :-)
jmerelo timotimo: maybe have them go to their very own tab in GMail... 16:51
timotimo i have a folder in my thunderbird that has all github messages in it
timotimo it currently has 5.3k unread 16:51
though surely i could filter out the names of my own repos 16:52
jmerelo lizmat: Sorry to say I was comparing it to the string version. It's still faster, though, and scales much better with size. For chromosomes size 4096 it's 30% faster 16:53
jmerelo timotimo: wow. It's probably better to delete them all and start with a blank slate :-) 16:54
lizmat: still the slowest of the bunch, but now only 4x slower than C++, while before it was 5x slower. 16:56
jmerelo timotimo: any advice from the profiler on the implementation you can give? I can't really make head or tails of it. To try and avoid push-until-lazy? 16:58
timotimo the exclusive time is usually more interesting 16:58
timotimo List.pm6:1100 is not being jitted and it has push from Array.pm6 inlined into it, which therefore also doesn't get jitted 16:59
that could be interesting
could you run the code with the env variable MVM_JIT_LOG set to some filename, and put the file up somewhere?
jmerelo timotimo: sure
jmerelo timotimo: here you go raw.githubusercontent.com/JJ/perl6...er/jit.log 17:07
lizmat
.oO( flaky wifi on the terrace, 4G much more reliable out here )
 17:11
timotimo looks like floor_n might cause this to be jitted. why does it try to floor_n in the first place, i wonder 17:12
timotimo ah, it's the iterator of roll 17:14
jmerelo bad, bad roll iterator 17:15
lizmat jmerelo: using floor_n here is probably the reason it got so much faster
timotimo yo jmerelo how's your x86_64 assembly? :) 17:16
jmerelo timotimo: between zero and nonexistent
timotimo: both my daughters have it in class, but I think they mostly hate it. 17:17
timotimo x86 assembly is a very natural thing to despise 17:18
jmerelo timotimo: :-) it really is. Sorry can't be of help in that department
lizmat I used to be fluent in 8086 :-(
mst the only assembly I've ever got on with was arm26
timotimo nah, it's fine
i wonder, is there a notable difference between coerce_ni and floor_n? 17:19
lizmat but that's at least 25 years agi
timotimo coerce_ni uses cvttsd2si
lizmat looks up coerce_ni
timotimo oh, neat, konsole lets me select something and the right click menu has a "search for $x with" -> "google" 17:20
m: use nqp; say nqp::coerce_ni(0.1e2)
camelia ===SORRY!===
No registered operation handler for 'coerce_ni'
timotimo right, we don't expose this one
lizmat right 17:21
timotimo how exactly does the truncation cvttsd2si does match what floor_n does?
i think it might actually be what we need 17:22
lizmat if it doesn't round, I think you may be right :-) 17:23
timotimo well, there's still the question of "floor towards -inf" and "floor towards 0" 17:24
El_Che lizmat: you need one of these: www.ubnt.com/unifi/unifi-ap-ac-pro/
timotimo m: say floor(-19.9e0)
camelia -20
timotimo m: say floor(19.9e0)
camelia 19
lizmat El_Che: probably, but for our new house 17:25
El_Che lizmat: I remember now
timotimo um, huh.
lizmat timotimo ?
timotimo for some reason i get NaN there 17:26
timotimo oh, wait, is floor_n supposed to give you a num back? 17:26
it is.
timotimo what we need is just roundsd 17:27
lizmat que? 17:28
rounded would not produce a uniform distribution *and* out of bounds values ?
timotimo how do you mean? 17:32
lizmat if we're making random ints between 0..N and we're rounding the nqp::rand_n value 17:34
you will underrepresent values < 0.5 and also give N as a possible value (which you shouldn't)_ 17:35
timotimo anyway, we now jit floor_n :)
lizmat whee!
timotimo m: my @all-values = ^100 .roll(1_000_000); say @all-values.grep(0).elems; say @all-values.grep(10).elems; 17:37
camelia 10031
10036		 17:38
timotimo m: my @all-values = ^1000 .roll(10_000_000); say @all-values.grep(0).elems; say @all-values.grep(10).elems;
camelia (timeout) 17:38
timotimo a bit too much, eh? 17:38
m: my @all-values = ^1000 .roll(3_000_000); say @all-values.grep(0).elems; say @all-values.grep(10).elems;
camelia 3075
2977		 17:39
MasterDuke m: say ^100 .roll(1_000_000).Bag.minmax(*.value) 17:42
camelia 7 => 9782..2 => 10280
b2gills RandalSchwartz: Have you been watching Perl 6 development in recent years? For example react/whenever was added somewhat recently. 17:45
timotimo turns out it takes a little while to get 100_000_000 numbers bagged up 17:48
though i think since the memory keeps growing it might not actually be bagging it very good
lizmat could be it's vivifying a complete list of 1M elems first 17:49
lizmat hmmm...not, should not 17:50
timotimo my $bag = bag ^1000 .roll(100_000_000); say $bag{0, 1, 2, 3} 17:51
the best way would have been to keep a running tally and output it regularly 17:52
lizmat one of the reasons jmerelo is allocating so much, is because it's creating Scalars for the values in the arrays 18:02
lizmat however, the values in the array never change 18:02
so technically, they wouldn't need a Scalar container
timotimo yeah, in that case don't assign into an array ;) 18:03
Geth doc: 6e1318831e | (JJ Merelo)++ | README.md
Minor edits to README.md
 18:05
doc: ed5cda5579 | (JJ Merelo)++ | highlights/package.json
Address warnings about the package.json file

Which probably does not fix #1910, but it helps anyway.
doc: 2ae9ffc45d | (JJ Merelo)++ | README.md
Eliminates wishlist stuff closes #1908
jmerelo lizmat: how can I address that? 18:07
timotimo probably assign into a $ container and use .List at the end so you don't just take the Seq
MasterDuke is there a way to clean up/close/garbage collect the threads started because of a .race? 18:08
Geth rakudo: a53811cd4d | (Elizabeth Mattijsen)++ | tools/build/NQP_REVISION
Bump NQP to get the nqp::floor_n jitting
 18:09
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....2-gf79f0ce
d9869e53b5 | (Elizabeth Mattijsen)++ | src/core/Array.pm6

Hopefully that will mean they optimize better in the future
lizmat MasterDuke: nqp::force_gc ? 18:10
jmerelo timotimo: can't use splice then... 18:11
lizmat timotimo: that's not to simple, you can't use splice on a List
*so
MasterDuke lizmat: doesn't seem to help, actually seems to make it worse 18:12
lizmat jmerelo: are you really married to Bools ? Could you work with 0 and 1 instead ?
jmerelo lizmat: I can get rid of that, but I seem to remember that sped up things a list...
lizmat: (thinking of a joke to answer to that)
lizmat: (nope, nothing comes up)
lizmat well, looking at that now
jmerelo: fwiw, with floor_n jitted now, your program became about 5% faster for me 18:13
jmerelo lizmat: random generation in Bools was much faster than for ints... Why should, in this case, changing to Ints help? I can see it in the case of the count-ones benchmarks, but not really here.
lizmat: that's great. Everything helps. Thanks!
lizmat well, I feel mative ints *should* be faster 18:14
jmerelo lizmat: checking now
lizmat jmerelo: they're not 18:15
figuring out why they're not :-)
jmerelo lizmat: Right, they are faster. But just slightly so. I'm using (0,1).roll instead of Bool.roll. 18:16
lizmat: it's going to make count-ones really faster.
MasterDuke (re jmerelo's code, not my .race question) i think i also tried using a Buf, but that wasn't any faster 18:17
jmerelo MasterDuke++
MasterDuke (back to my .race question) `/usr/bin/time perl6-m -e 'for ^10_000 { (^100).race(batch=>1).map({ $_ }).List }'` uses 645140maxresident 18:19
but `/usr/bin/time perl6-m -e 'use nqp; for ^10_000 { (^100).race(batch=>1).map({ $_ }).List; nqp::force_gc() if $_ %% 1_000 }'` uses 923220maxresident 18:20
jmerelo c: Bool.roll(10_000) ||| Bool.roll xx 10_0000 ||| (0,1).roll(10_000) ||| (0,1).roll xx 10_000
committable6 jmerelo, ¦Bool.roll(10_000): «Cannot find this revision (did you mean “0111f10”?)»
jmerelo c: HEAD Bool.roll(10000) ||| Bool.roll xx 10000 ||| (0,1).roll(10000) ||| (0,1).roll xx 10000 18:21
committable6 jmerelo, ¦HEAD(a53811c): «»
MasterDuke jmerelo: you probably mean, bench:
jmerelo bench: HEAD Bool.roll(10000) ||| Bool.roll xx 10000 ||| (0,1).roll(10000) ||| (0,1).roll xx 10000
benchable6 jmerelo, starting to benchmark the 1 given commit
jmerelo, ¦HEAD: «0.1444»
jmerelo MasterDuke++
MasterDuke and 'compare'
jmerelo bench: HEAD compare Bool.roll(10000) ||| Bool.roll xx 10000 ||| (0,1).roll(10000) ||| (0,1).roll xx 10000 18:22
benchable6 jmerelo, starting to benchmark the 1 given commit
jmerelo, ¦HEAD: ««run failed, exit code = 1, exit signal = 0»»
jmerelo bench: compare HEAD Bool.roll(10000) ||| Bool.roll xx 10000 ||| (0,1).roll(10000) ||| (0,1).roll xx 10000 18:24
benchable6 jmerelo, starting to benchmark the 1 given commit
jmerelo, gist.github.com/06c3413d371079365f...90ef5237f1
Cleverson Hello, is there a place to request new modules or vote for wanted modules? I've located only a text file listing the most wanted modules. 18:31
jmerelo Cleverson: do you mean this? github.com/perl6/perl6-most-wanted 18:32
moritz well, you could state your wish(es) here 18:33
yoleaux 10:55Z <jmerelo> moritz: I still need the docs.per6.org website logs, or the public ones updated, to search for 404s and, if possible, referrers to them.
moritz jmerelo: how would you like to get them? 153MB as a compressed tarball 18:35
jmerelo moritz: do you have some way to process them in finer grain, like monthly or so, using the program you already use? The problem is that current statistics are from forever 18:36
moritz: I mean these www.p6c.org/stats/doc.perl6.org#Pag...er%20visit
Cleverson jmerelo: Hi, I had found the most wanted listing, thanks for informing the URL of the entire repo. I'd like basically two functionalities which seems still lacking... 18:37
jmerelo moritz: for the time being I'm mainly interested in having this www.p6c.org/stats/doc.perl6.org#404%20Errors by month. Or for the last few months.
jmerelo Cleverson: just shoot. This place is as good as any other... 18:38
Cleverson: besides, I'm not sure people is really checking that repo on a daily, or even monthly, basis.
Cleverson First, a way to control the entire keyboard on a terminal, so that I could for example get every key press on the keyboard, and take an action depending on the key or key sequence.
The other I'd like is a simpler audio module to play and stop sound files, simpler than the already existing portAudio module. 18:39
jmerelo Cleverson: best way to do that is to start yourself... :-) WRT the second one, you'll probably get more mileage if you post issues on the existing module. 18:41
moritz jmerelo: www.p6c.org/stats/doc.perl6.org.recent
jmerelo moritz++
jmerelo off to dinner o/ See you tomorrow! 18:42
Zoffix .tell jmerelo FWIW, the discouragement against rakudobrew isn't just first time users. It introduces a bug with USAGE. It is known to introduce precomp bugs. It builds random dev commit instead of a release. It requires running a command after installing any binary. And it requires reinstalling modules after each update, which isn't needed anymore for rakudo. 18:46
yoleaux Zoffix: I'll pass your message to jmerelo.
robertle Zoffix: which reminds me of a question I had: is there a way to tell whether precompiled stuff needs re-compiling after upgrading moar+nqp+rakudo? 18:47
Zoffix robertle: after moarvm it doesn't. The rest, I don't know. Ask nine++ 18:48
robertle so you are saying that no moarvm update will require a recompilation? ever?
Zoffix robertle: dunno, but I don't have to recompile rakudo/nqp when I make changes to moarvm, so I guess stuff doesn't need recompilation 18:49
robertle and will ask nine++ of course! the root of the question is how to package perl6 modules foer debian. there seem to be two ideas so far: either only ship the raw sources and compile on installation, plus recompile on (certain) upgrades of moar/nqp/rakudo 18:50
or ship the precopmpiled files with the sourc
es and rebuild the whole ecosystem when (certain) moar/nqp/rakudo upgrades do happen 18:51
lizmat robertle: how about just shipping the precomped files ?
robertle neither sound great
lizmat: right, the problem is that you need to rebuild the whole ecosystem, and guarantee that they get installed together. so very tight coupling of dependencies. essentially you turn the whole ecosystem into one big package
robertle recompiling on the target system when moar/nqp/rakudo get updated avoids that, but of course wastes cycles and has some intereting race conditions as well 18:52
Zoffix robertle: FWIW, I don't need to recompile any modules **manually** when I upgrade rakudo. It figures out they need recompilation and recompiles them automatically.
robertle plus, in both cases, it is unclear how long-running programs would be handled
Zoffix: ah! the problem with that is that it then needs to put the precompiled stuff in you homedir, because you can't write to /usr/lib/perl6 or whatever 18:53
Zoffix probably
robertle so we would duplicate the precompiled stuff for each user...
a can of worms, but apparently it's the same for other languages that do on-the-fly precompilation. (python, haskell) 18:54
robertle there would be some ways to make this easier for distributions: if there was a way to know that a moar/nqp/rakudo combination is incompatible with another one, that 18:56
would be a great start. then we would only need to do the recompilation in the rarer cases (helps in both scenarios) 18:57
robertle is the whole precompilation a rakudo, an nqp or a moar thing anyway? I somehow assumed it would be a moar thing, but thinking about it now I am not so sure anymore... 18:59
geekosaur rakudo, but it generates something appropriate for the backend 19:00
robertle k, thanks!
robertle do you think there is a way to tell whether the 19:01
robertle way rakudo stores these has changed? there must be, after all rakudo can tell when loading them. is there a version or so embedded? 19:01
Zoffix Fairly sure each rakudo version change requires recompilation. 19:03
geekosaur it'll be somewhere in the CompUnitRepo stuff but I don't think that's exposed as such
robertle ok, I'll do some digging! 19:04
timotimo (100123 99647 99620 100327) 19:56
MasterDuke: it would be possible to signal to threads to shut themselves down 19:57
MasterDuke but that's NYI? 19:58
timotimo yup
MasterDuke timotimo: would you expect the code i pasted to just keep creating threads or reuse them? 20:03
timotimo where was that? 20:04
MasterDuke 'for ^10_000 { (^100).race(batch=>1).map({ $_ }).List }' 20:05
timotimo shouldn't just keep creating threads. you can set RAKUDO_SCHEDULER_DEBUG and see what it's thinking
MasterDuke just "created initial...", "supervisor started", "sup thinks there are 4 cores" 20:07
lizmat_: how do i use RAKUDO_REPORT_COLUMNS? `RAKUDO_REPORT_COLUMNS=tcd perl6 -Msnapper -e '<...>'` gives `Some exceptions were thrown in END blocks: X::AdHoc: Cannot unbox a type object (Any) to a str.` 20:08
lizmat_ checks 20:09
lizmat MasterDuke: any particular code I should run? 20:10
timotimo MasterDuke: if that's the output, it shouldn't be creating threads 20:11
MasterDuke lizmat: what i pasted for timotimo immediately above
lizmat ah, ok, checking 20:12
MasterDuke timotimo: heaptrack seemed to suggest that the leak (or maybe just the large amount of mem allocated) was from thread cloning/pthread creations/something like that
lizmat MasterDuke: looks like there's no tcd data and it's not protected against that 20:14
could you file an issue pls?
MasterDuke k 20:15
timotimo um, huh. 20:16
timotimo make: *** No rule to make target '3rdparty/uthash.h', needed by 'src/main.o'. Stop. 20:16
how did i mess this up?
timotimo oh, haha 20:17
i rm -rf'd 3rdparty, it did not like that
MasterDuke lizmat: github.com/rakudo/rakudo/issues/1714 20:20
lizmat MasterDuke++ # will look at it soon 20:21
robertle lizmat: do you have some benchmarks that you use to drive your perf improvements? or some other method to find worthwhile targets? 20:23
lizmat well, these last ones were inspired by jmerelo's benchmarks 20:23
I thought that using native int arrays should be a lot faster 20:24
but alas, it was way slower
so I am delving in that area at the moment
with relatively low hanging fruit for me
buggable New CPAN upload: ANTLR4-Grammar-0.5.2.tar.gz by JGOFF cpan.metacpan.org/authors/id/J/JG/...5.2.tar.gz
lizmat so targeting is really by what people notice as being slow
robertle understood, thanks! 20:25
timotimo MasterDuke: i've begun implementing a "kill a worker" thing 20:44
MasterDuke oooh, nice 20:45
moritz the unions might not like that
timotimo unions are bad for my plans to become a billionaire by benefiting from other's work and barely reimbursing them sufficiently for feeding themselves and their family 20:47
not entirely sure how exactly which part should be responsible for what 20:52
El_Che Perl 6 will not have threading problems but a workers' revolt 20:53
lichtkind timotimo, was great chatting with you unfortunately was not as much time 20:54
lizmat lichtkind o/
timotimo true
your exam is tomorrow?
lichtkind common people no communist propagande here this killed already enough worker 20:55
timotimo, you memorized correctly
timotimo i'll be hoping for your success
lichtkind thank you
yes even today i mostly coded in perl 6 :) 20:56
and even got some interesting possibilty propaganda wise but wil tell if something really happens
lizmat, cheers
AlexDaniel Zoffix: fwiw “it requires reinstalling modules after each update, which isn't needed anymore for rakudo” I don't think that's true, at least I don't do that 21:12
precomp takes some time after updating though 21:13
Zoffix AlexDaniel: that's cause you're building random, potentially buggy, dev commits and it just sticks it to `master` cell. If you tell it to build releases, you'll have to reinstall modules for each release, because it keeps all the things separate. 22:10
AlexDaniel Zoffix: aw yeah that's right

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!