🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | log inspection situation still under development | For MoarVM see #moarvm
Set by lizmat on 22 May 2021.
Geth rakudo/rakuast: b5a116fa20 | (Stefan Seifert)++ | 2 files
Rename RakuAST::Call::Safe to ::MaybeMethod for clarity

Makes it clearer that this is about calling methods and that such a call may or may not actually happen.		 07:19
rakudo/rakuast: 9184b3672a | (Stefan Seifert)++ | 3 files
Implement RakuAST::Call::VarMethod, i.e. .& operator

Call a callable from a variable with methodish syntax		 07:28
timo do we have a clue why liz' test-t came to an all-time low but tux' test-t was not faster than usual? 08:51
i haven't even checked if there was a bump tho
lizmat well, my test-t has always been lower than tux's: different hardware / OS 08:54
timo i was thinking your change for leave-ish phasers would be, like, a percentual improvement thtat would translate to other hardware easily 08:56
lizmat well... the leavish fixes I didn't expect much to do with test-t, tbh... 08:58
Text::CSV only has one LEAVE phaser, and it is on the public method to ensure closing any file opened for reading 08:59
I'd expect the std handles to be more of an improvement, in reduced startup time
and there is one LEAVE phaser for every rakudo run, as part of a Lock.protect making sure all END phasers are run only once, instead of by multiple threads 09:00
so I guess my leavish fixes meant 4 hash lookups less for test-t 09:01
timo mhh okay i guess i kinda mixed stuff up a little bit :) 09:02
|Tux| Rakudo v2022.04-40-g3051eefa8 (v6.d) on MoarVM 2022.04-2-g997302
csv-ip5xs0.839 - 0.885
csv-ip5xs-205.209 - 5.210
csv-parser3.500 - 3.565
csv-test-xs-200.395 - 0.580
test6.618 - 7.181
test-t1.495 - 1.547
test-t --race0.929 - 0.941
test-t-2020.615 - 20.826
test-t-20 --race6.942 - 7.843
 10:35
Geth rakudo/rakuast: 346ed9a398 | (Stefan Seifert)++ | 3 files
Support traits on variables
 13:31
Geth rakudo/rakuast: 8245f4f042 | (Stefan Seifert)++ | src/Raku/ast/variable-declaration.rakumod
Fix missing native type on lexicalrefs

Need to set the variable lookup's return type, so the QAST compiler will use the right register kinds and ops.		 13:54
rakudo/rakuast: 4141d562cc | (Stefan Seifert)++ | src/Raku/Grammar.nqp
Support coercion types
 14:13
lizmat notable6: weekly 17:56
notable6 lizmat, 4 notes: gist.github.com/68b93365567e9bc281...68036083ba
lizmat notable6: weekly reset 17:59
notable6 lizmat, Moved existing notes to “weekly_2022-05-09T17:59:28Z”
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2022/05/09/2022-19-docublast/ 18:42
Geth rakudo/rakuast: f474a026f3 | (Stefan Seifert)++ | 2 files
Set precompilation-mode on QAST::CompUnit if appropriate
 19:51
rakudo/rakuast: cd58014824 | (Stefan Seifert)++ | src/Raku/ast/compunit.rakumod
Fix segfault due to missing description on CompUnit

If getscdesc is called on a comp unit without description, it puts a NULL into a string register. This may lead to segfaults or hash key lookup failures. While this is clearly a bug in the VM, it also shows that setting the description was missing from the RakuAST implementation.		 20:00
Geth rakudo/rakuast: 4a755a4ceb | (Stefan Seifert)++ | src/Raku/ast/name.rakumod
Fix superfluous generated nqp::who

IMPL-QAST-PACKAGE-LOOKUP will already add the nqp::who op, so all that's left to do is return the original stash for empty name parts.		 20:23
rakudo/rakuast: 3313e8f9c8 | (Stefan Seifert)++ | src/Raku/ast/compunit.rakumod
Add a repo_conflict_resolver to generated QAST::CompUnits
 20:29

Powered by Cro and the Raku® Programming Language.

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

Thank you!