00:02 Kaiepi left 00:03 Kaiepi joined 01:00 lucasb left
vrurg .ask pmurias When jvm backend serialize closures does it serialize their respective core setting too? 01:11
tellable6 vrurg, I'll pass your message to pmurias
02:02 benjif left, benjif joined 02:58 benjif left
Geth nqp/configure-rebuild-all: fcf5bf5591 | (Vadim Belman)++ | 3 files
Added --force-rebuild command line option

Forces rebuild of MoarVM if `--gen-moar` was used too. Mostly good if bypassed from upstream (Rakudo) which might need total rebuild under certain circumstances.
nqp: vrurg++ created pull request #586:
Added --force-rebuild command line option
nqp/configure-rebuild-all: e304774121 | (Vadim Belman)++ | 3rdparty/nqp-configure
nqp-configure submodule update
rakudo/configure-rebuild-all: 532476f5ca | (Vadim Belman)++ | 3 files
Add --force-rebuild command line option

Forces rebuild of all required components like NQP and, possibly, MoarVM. --force-rebuild tries to pretend we're building in an empty or almost empty environment. I.e.:
perl Configure.pl --backends=moar --gen-nqp --force-rebuild ... (6 more lines)
rakudo: vrurg++ created pull request #3248:
Add --force-rebuild command line option
03:10 benjif joined 06:08 [TuxCM] joined 06:37 [TuxCM] left
Geth roast: Altai-man++ created pull request #588:
Test fake-infix adverbs on a construct
06:55 benjif left 06:56 benjif joined 07:05 benjif left
Geth rakudo: Altai-man++ created pull request #3249:
Test return constraints for role implementors
nine More good news: looks like my changes to NativeCall also fixed issues with running optimized (JIT compiled) native subs at BEGIN time, so we can now lift that long time restriction :) 07:20
That restriction meant that a native sub called at BEGIN time has never gotten the optimized body and instead always went through the old generic code, even later at runtime. 07:21
Will make that change after the release
07:37 pmurias joined 08:34 [TuxCM] joined 08:42 pmurias left 09:14 pmurias joined
AlexDaniel colabti.org/irclogger/irclogger_lo...-10-19#l84 09:42
now I remember
I need to split text by \n but I can't
09:42 pmurias left
AlexDaniel m: say “foo\r\nbar”.split(“\n”)[0] 09:43
camelia fo
AlexDaniel camelia wtf was that
e: say “foo\r\nbar”.split(“\n”)[0]
evalable6 foo␍
AlexDaniel and previously I had this github.com/perl6/whateverable/comm...9ed5779a8f 09:46
so .lines is wrong for example because it splits on single \r
and .split(“\n”) is wrong because it ignores \r\n
I remember raising a ticket at some point, but IIRC the idea there was basically “tough luck, split it yourself” 09:51
which is kinda horrible for something as simple as processing output of unix tools :S 09:52
maybe .lines on IO::Path is different? But I don't have IO::Path 09:54
the output comes from Proc::Async, which I then decode as utf8-c8. I shouldn't? 09:55
e: dd “foo\rbar\r\nbaz\nend”.lines 09:57
evalable6 ("foo", "bar", "baz", "end").Seq
AlexDaniel e: dd “foo\rbar\r\nbaz\nend”.split(“\n”)
evalable6 ("foo\rbar\r\nbaz", "end").Seq
AlexDaniel e: dd “foo\rbar\r\nbaz\nend”.split(/\n/)
evalable6 ("foo", "bar", "baz", "end").Seq
AlexDaniel e: dd “foo\rbar\r\nbaz\nend”.split(/“\n”/)
evalable6 ("foo\rbar\r\nbaz", "end").Seq
AlexDaniel e: dd “foo\rbar\r\nbaz\nend”.split(/“\n”|“\r\n”/) 09:58
evalable6 ("foo\rbar", "baz", "end").Seq
AlexDaniel I mean, that's going to eat \r but I can close my eyes on that for now… 09:59
e: dd “foo\rbar\r\nbaz\nend”.split(“\n”|“\r\n”)
evalable6 any(("foo\rbar\r\nbaz", "end").Seq, ("foo\rbar", "baz\nend").Seq)
AlexDaniel lizmat: ↑ I don't think it's supposed to work this way? :) 10:00
the junction thing
alright, whatever: github.com/perl6/whateverable/commit/9ac71dca89 10:11
10:12 greppable6 left 10:14 greppable6 joined
lizmat m: dd “foo\rbar\r\nbaz\nend”.split(“\n”,“\r\n”) 10:15
camelia ().Seq
lizmat hmmm
11:50 sena_kun joined
[TuxCM] Rakudo version 2019.07.1-416-ge45bb3418 - MoarVM version 2019.07.1-283-ge594fbcbd
csv-ip5xs0.758 - 0.762
csv-ip5xs-206.252 - 6.317
csv-parser20.650 - 21.560
csv-test-xs-200.422 - 0.426
test6.961 - 7.548
test-t1.746 - 1.754
test-t --race0.790 - 0.821
test-t-2029.391 - 29.715
test-t-20 --race9.005 - 9.186
12:24 lucasb joined 12:40 pmurias joined 12:48 pmurias left 12:50 pmurias joined 13:45 [TuxCM] left 14:05 [TuxCM] joined
Geth roast: 8b29706459 | Altai-man++ | S12-class/inheritance.t
Test Parameter subclassing

Spec test for github.com/rakudo/rakudo/issues/2829 as suggested by jnthn.
roast: 0aeb74eef5 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | S12-class/inheritance.t
Merge pull request #585 from Altai-man/test-param-subclass

Test Parameter subclassing
roast: e84dbeca19 | Altai-man++ | S12-meta/classhow.t
Test runtime-created classes being attributes

Closes github.com/rakudo/rakudo/issues/2607
roast: 8fddffc486 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | S12-meta/classhow.t
Merge pull request #586 from Altai-man/test-2607

Test runtime-created classes being attributes
roast: 535cea30d2 | Altai-man++ | S02-types/pair.t
Test fake-infix adverbs on a construct

Just code provided by the ticket, closes
roast: 93cc2e4274 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | S02-types/pair.t
Merge pull request #588 from Altai-man/test-1500

Test fake-infix adverbs on a construct
15:31 sena_kun left 15:36 sena_kun joined 15:54 [TuxCM] left
Geth rakudo: 7dc34148eb | (Stefan Seifert)++ | t/04-nativecall/00-misc.t
Fix new NativeCall test failing on Windows

On Windows calloc isn't as easily available, so use a function of the test library instead as NativeCall target.
16:38 greppable6 left 16:40 greppable6 joined 17:01 pmurias left 17:13 sena_kun left 17:14 sena_kun joined 17:57 haxmeister joined
haxmeister are there any detailed documents describing the minimum requirements for installing modules/dist in terms of what rakudo specifically needs as opposed to what zef needs? 18:00
19:04 pmurias joined
timotimo are you talking about individual modules, like C bindings for something need the C library itself, or something else? 19:07
haxmeister I'm talking about information that can help me with system administration of Raku and the installation of packages or modules on a unix style system... in this application zef doesn't appear to be designed well for it, so I want to clarify what is actually needed by rakudo 19:12
I"m talking about multi-user systems
timotimo ah 19:22
sorry for being AFK unexpectedly
our very own nine has done some work for packaging stuff for the suse distro, there's a thing called CompUnitRepo::Staging or so that does precomp magic so the precompiled files could be distributed in packages i think? 19:23
haxmeister I am using OpenSuse Tumbleweed currently on this laptop, and Funtoo and Ubuntu on other machines. 19:30
but my inquiry was more generic.. looking for the type if information nine would have used to do such work 19:31
would need info to help with distribution on source based linux distros as well 19:32
timotimo well, nine is a core developer who developed that staging thing (which as i understand it is also useful to build stuff in one place and copy it over to another, like you would in a sandboxed build environment) 19:45
so i can't really point you at the questions that were necessary for that work
haxmeister: you could have a look at rakudo's tools/install-dist.p6, but i can't guarantee that that's the right way to go for what you want to do 20:07
haxmeister thank you timotimo 20:09
20:09 pmurias left
timotimo good luck! 20:11
20:11 pmurias joined 20:54 [TuxCM] joined 21:05 [TuxCM] left
nine haxmeister: tools/install-dist.p6 is specifically meant to be used for creating distribution packages. The existing packages show how it's used: build.opensuse.org/package/view_fi...c?expand=1 21:16
There are also scripts for automating the packaging and for putting them onto the Open Build Service: github.com/niner/meta2rpm 21:17
timotimo we could perhaps have a spot for people in the same situation as haxmeister
on the website or where-ever
AlexDaniel nine: thank you for looking into the blockers 👍 21:25
Kaiepi i'm running into the same error as #3045 in my own code 21:29
however you get the bot to link rakudo issues
timotimo that's the right way, i guess it just isn't here 21:30
Kaiepi ah
AlexDaniel yeah because we have to chase down every single bot and tweak it
hmm do I feel like it or not… 21:33
I can quickly just turn it into something able6
21:34 ggoebel joined
AlexDaniel actually, let me do it 21:35
timotimo synopsable
japhb A bot that chases down other bots and moves them to new channels?
shepherdable 21:36
AlexDaniel rather recruits them
timotimo: linkable?
japhb piedpiper 21:37
timotimo it does a little more than just linking, i think?
AlexDaniel like
AlexDaniel deletes code, deletes code, deletes code… 21:39
timotimo i don't remember :) :) 21:40
nine AlexDaniel: the epoll issue must be one of the weirdest ever 21:54
Nothing about it makes real sense: github.com/rakudo/rakudo/issues/32...-544200300
AlexDaniel nine: let me guess, minor changes to the source code causing issues?
nine yep
Almost any change really 21:55
AlexDaniel I remember some issue where changing some whitespace in the source code made it reliably fail or pass
that was fixed many years ago, but still :) 21:56
was fun
22:31 patrickb joined
Geth rakudo: Altai-man++ created pull request #3250:
Test indexing with whatever for CArray
22:33 lucasb left
patrickb .tell nine wrt the Inline::Perl5 with `make test`. I think I have an idea what's going wrong and what needs to be done to fix it. I'll report back in the next days. 22:35
tellable6 patrickb, I'll pass your message to nine
Geth rakudo: bc8ca4e6e7 | Altai-man++ | t/04-nativecall/05-arrays.t
Test indexing with whatever for CArray

Closes github.com/rakudo/rakudo/issues/2682
rakudo: df657a3552 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | t/04-nativecall/05-arrays.t
Merge pull request #3250 from Altai-man/test-carray-star-index

Test indexing with whatever for CArray
22:36 patrickb left 22:57 pmurias left 23:11 sena_kun left 23:14 ggoebel left 23:28 ggoebel joined
Geth rakudo: 9b181a700e | (Ben Davies)++ | src/Perl6/Metamodel/Naming.nqp
Fix another MOP-related StrLexRef serialization bug

Perl6::Metamodel::Naming's name and shortname attributes are expected to be literal strings, but if, say, a Raku module uses any given new_type method on a Rakudo HOW with :name bound to a Scalar, this doesn't end up being the case. Decont names and shortnames given to Perl6::Metamodel::Naming's methods similarly to f13ac2c03630ad0b56658f29c722fd8f2a45409c.
Kaiepi nine++, i wouldn't have figured this out without that commit