🦋 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: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
00:51 buildable6 left 00:54 buildable6 joined
Geth rakudo/ugexe-patch-1: bdd77290d6 | (Nick Logan)++ (committed using GitHub Web editor) | src/core.c/CompUnit/Repository/Installation.rakumod
Remove largely unknown flags from installed raku script shims

These were missed in github.com/rakudo/rakudo/commit/73...9628617885
rakudo: ugexe++ created pull request #5434:
Remove largely unknown flags from installed raku script shims
01:54 buildable6 left 01:57 buildable6 joined
nemokosch ugexe: do you recognize this Github CI error? github.com/2colours/WebDriver2/act...7918561303 02:07
ugexe is it reproducible if you re-run it? at first look it seems like the extracting failed (even though it extracted fine for previous distributions) in a way that isn't letting zef report the error at all (which seems to happen a lot on windows with github actions) 02:14
to be clear, what i see happen a lot on github actions with windows are those random exit code 1s. even with one liners that don't use any non-core modules 02:16
02:19 nativecallable6 left, nativecallable6 joined
nemokosch okay, rerunning the two failed tasks with "debug logs" 02:22
ugexe if it is reproducible then it is probably a bug in the most recent release of zef, although those changes should only be exercised after extraction has succeeded
nemokosch github.com/2colours/WebDriver2/act...7918857009 seems quite non-deterministic 02:23
something failed but not exactly the same
02:24 nativecallable6 left, nativecallable6 joined 02:57 buildable6 left 02:59 buildable6 joined
ugexe two hail marys would be setting RAKUDO_MAX_THREADS to something lowish (to potentially prevent help avoid any weird file access issues that happen in the background like precompilation), and/or setting ZEF_EXTRACT_TIMEOUT=0 (which would avoid a Promise based timer in the extractor, but i dont have much reason to believe that shouldnt work correctly on windows) 03:06
03:59 buildable6 left 04:00 buildable6 joined 05:00 buildable6 left 05:01 buildable6 joined 06:01 buildable6 left 06:02 buildable6 joined
Geth nqp/main: f9e114e93d | (Christian Bartolomäus)++ | src/vm/jvm/runtime/org/raku/nqp/runtime/Ops.java
[JVM] Look at type check mode of checked value

The old code could fail with values that had a different type check mode from the type that they are checked against. A simple example is the code from github.com/rakudo/rakudo/issues/2714:
   class C { };
... (11 more lines)
roast: 8ca18e1577 | (Christian Bartolomäus)++ | S03-metaops/hyper.t
Add test for issue 4700 in old issue tracker

roast: 91997f55b3 | (Christian Bartolomäus)++ | S12-methods/delegation.t
Add test for issue 3721 in old issue tracker
bartolin bisectable6: my %h{Str(Cool)} = a => 42 ## github.com/Raku/old-issue-tracker/issues/6176 06:51
bisectable6 bartolin, Will bisect the whole range automagically because no endpoints were provided, hang tight
bartolin, Output on all releases: gist.github.com/38c732905eff0e7066...140a356729
bartolin, More than 3 changes to bisect, please try a narrower range like old=2020.10 new=HEAD
bartolin bisectable6: old=2020.10 new=2020.11 my %h{Str(Cool)} = a => 42 06:52
bisectable6 bartolin, Bisecting by exit code (old=2020.10 new=2020.11). Old exit code: 1
bartolin, bisect log: gist.github.com/ee754254616caf5045...96c65a512a 06:53
bartolin, (2020-11-15) github.com/rakudo/rakudo/commit/f2...e8f9076d6e
Geth roast: 1f824745f2 | (Christian Bartolomäus)++ | S12-coercion/coercion-types.t
Add test for issue 6176 in old issue tracker

07:02 buildable6 left
Geth roast: Dev7g++ created pull request #847:
README: Added Our Contributors
07:03 buildable6 joined 08:01 MasterDuke left 08:03 buildable6 left 08:05 buildable6 joined
Geth roast: 8f7ef334ed | (Christian Bartolomäus)++ | README.md
Use tags for new releases, not branches

That's how we did it for "6.d": We have a tag for the state of roast when the relaease was cut and a corresponding branch "6.d-errata" for errata.
roast: fc44f14d99 | (Christian Bartolomäus)++ | README.md
Tweak grammar/style
09:05 buildable6 left
Geth rakudo/main: b5c9c9b31f | (Elizabeth Mattijsen)++ | 3 files
RakuAST: add French localization

Courtesy of Lucien Grondin
09:07 buildable6 joined 09:44 sena_kun joined 10:04 Geth__ joined 10:05 lizmat_ joined, Geth left 10:06 RakuIRCLogger left, Geth__ left, Geth joined 10:07 buildable6 left, lizmat_ left, lizmat joined 10:09 RakuIRCLogger__ left 10:10 buildable6 joined, Geth left 10:11 Geth joined
Geth rakudo: usev6++ created pull request #5435:
Check if requested backtrace is null
11:05 Geth left, Geth joined 11:10 buildable6 left 11:11 buildable6 joined
Geth roast: b0c9b23cf8 | (Christian Bartolomäus)++ | S32-io/lock.t
[JVM] Unfudge now passing tests for file locking

Fixed with github.com/Raku/nqp/commit/5af47c154f.
rakudo/main: 76bfebf590 | (Christian Bartolomäus)++ (committed by Elizabeth Mattijsen) | src/core.c/CallFrame.rakumod
Check if requested backtrace is null

  ... before trying to extract a value. This will avoid an internal error
on the JVM backend when exploring call frames that don't exist. For details see github.com/rakudo/rakudo/issues/3607.
The old code works on MoarVM, because calling on a object doesn't explode there. But adding the extra check shouldn't hurt, either.
rakudo/main: fe8123d045 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: make quote-lang markers rx m s localizable

  - add tokens for quote-lang markers
  - separate m// and s/// into separate tokens
  - add actions for these new tokens
  - add helper functions in actions to simplify these tokens
rakudo/main: 92ed414c90 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
Don't use underscores in modern identifiers
12:11 buildable6 left 12:12 buildable6 joined
Geth rakudo/lizmat-quotelangextenders: c1d959b88e | (Elizabeth Mattijsen)++ | src/Perl6/Grammar.nqp
Remove undocumented quotelang extenders

The primary quotelang markers q, qq, and Q allow for a number of extenders according to the documentation: w, ww, and x. So instead of qw, one could also say q:w, and instead of qqww one could also say qq:ww.
The grammar however **ALSO** allows for: to, s, a, h, f, c, b, o. Allowing ... (13 more lines)
rakudo: lizmat++ created pull request #5436:
Remove undocumented quotelang extenders
rakudo/main: f712bdf8a2 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: make quote-lang markers q Q qq localizable

  - add tokens for quote-lang markers
13:12 buildable6 left 13:13 buildable6 joined
vrurg m: say "=head1 test\n=item foo\n".AST.DEPARSE 13:28
camelia =head1 test

=item foo

vrurg It doesn't look good that DEPARSE inserts extra newlines. 13:29
14:13 buildable6 left 14:16 buildable6 joined 14:42 nativecallable6 left, nativecallable6 joined 14:46 nativecallable6 left, nativecallable6 joined 15:01 codesections1 joined
Geth roast: 17b56c62b2 | (Christian Bartolomäus)++ | S06-advanced/callframe.t
[JVM] Unfudge test for exploring call frames

Fixed with github.com/rakudo/rakudo/commit/76bfebf590.
15:16 buildable6 left
Geth roast: 18298f92cf | (Christian Bartolomäus)++ | S05-metasyntax/regex.t
[JVM] Unfudge test for capturing back references

Workaround added with github.com/Raku/nqp/commit/1c4951ff57.
15:18 buildable6 joined 15:45 codesections1 left 16:18 buildable6 left 16:20 buildable6 joined 16:44 notna joined
Geth rakudo/main: c0eb6b6f21 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: streamline some grammar methods

  - cheat-heredoc
  - quibble
  - babble
17:20 buildable6 left 17:22 buildable6 joined 17:47 notna left 17:59 MasterDuke joined 18:22 buildable6 left 18:25 buildable6 joined
Geth rakudo/main: bdd77290d6 | (Nick Logan)++ (committed using GitHub Web editor) | src/core.c/CompUnit/Repository/Installation.rakumod
Remove largely unknown flags from installed raku script shims

These were missed in github.com/rakudo/rakudo/commit/73...9628617885
rakudo/main: 7efb6219c1 | (Nick Logan)++ (committed using GitHub Web editor) | src/core.c/CompUnit/Repository/Installation.rakumod
Merge pull request #5434 from rakudo/ugexe-patch-1

Remove largely unknown flags from installed raku script shims
18:32 buildable6 left, buildable6 joined
Geth rakudo/main: 45b1a26b30 | (Elizabeth Mattijsen)++ | src/Raku/Grammar.nqp
RakuAST: refactor quote-language handling

Before, the "quote-lang" method would take 2 optional arrays: one with the base tweaks ('q' in the case of q/foo/, ['q','w'] in the case of qw/foo/. And one with additional tweaks, coming from any adverbs ( so [['f', 1],] in the case of q:f/bar/, and [['f', '0'],] in the case of q:!f/bar/. ... (15 more lines)
ugexe github.com/rakudo/rakudo/blob/45b1...od#L33-L36 - i wonder if we even need the `sub MAIN` here 18:36
lizmat probably not
ugexe actually we don't even need the wrapper at all anymore i dont think 18:38
i guess if two distributions provide the same bin/foo then one would get overridden 18:39
lizmat that would even be better :-) 18:43
MasterDuke evalable6: say "LICENSE".IO.modifed 18:44
evalable6 (exit code 1) No such method 'modifed' for invocant of type 'IO::Path'. Did you mean
in block <unit> at /tmp/B9SQ3iJQ7k line 1
MasterDuke evalable6: say "LICENSE".IO.modified
evalable6 Instant:1696867989
MasterDuke bisectable6: say "LICENSE".IO.modified 18:45
bisectable6 MasterDuke, Will bisect the whole range automagically because no endpoints were provided, hang tight
MasterDuke, Output on all releases: gist.github.com/f3dc268199a1c9c37a...9bf3bf4cd7
MasterDuke, Bisecting by output (old=2016.06 new=2016.07.1) because on both starting points the exit code is 0
MasterDuke, bisect log: gist.github.com/9f074b66d080cad530...21180e86fb 18:46
MasterDuke, (2016-07-08) github.com/rakudo/rakudo/commit/5b...9e7b07223c
MasterDuke, Output on all releases and bisected commits: gist.github.com/ad9e9505ac88e16eea...39b4a1f40d
MasterDuke huh. does it really have no fractional part?
evalable6: say "IN-MEMORIUM".IO.modified 18:47
evalable6 (exit code 1) Failed to find '/srv/IN-MEMORIUM' while trying to do '.modified'
in block <unit> at /tmp/2IYJ4ZJ8Mq line 1
MasterDuke evalable6: say "IN-MEMORIUM.md".IO.modified 18:48
evalable6 (exit code 1) Failed to find '/srv/IN-MEMORIUM.md' while trying to do '.modified'
in block <unit> at /tmp/yNe9JwUvjW line 1
MasterDuke evalable6: say "IN-MEMORIAM.md".IO.modified
evalable6 (exit code 1) Failed to find '/srv/IN-MEMORIAM.md' while trying to do '.modified'
in block <unit> at /tmp/B0sDyGDQyF line 1
MasterDuke arg
evalable6: say "README.md".IO.modified
evalable6 Instant:1696867989
lizmat m: say "README.md".IO.modified.DateTime 18:54
camelia Failed to find '/home/camelia/README.md' while trying to do '.modified'
in block <unit> at <tmp> line 1
lizmat say "README.md".IO.modified.DateTime
evalable6 2023-10-09T16:12:32Z
lizmat not an issue with Instant.gist
19:05 vrurg left, vrurg joined 19:24 Nemokosch joined 19:25 buildable6 left 19:26 buildable6 joined 19:48 codesections1 joined 19:52 codesections1 left 20:26 buildable6 left 20:28 buildable6 joined
Geth rakudo: 2colours++ created pull request #5437:
Hungarian localisation of core
21:28 buildable6 left 21:31 buildable6 joined 22:04 Nemokosch left
nemokosch two notes about the Hungarian translation 22:29
1. one thing I always cry about with translation is the preposition-oriented phrasing of (eg.) English-first stuff 22:30
it's really hard to not "sound like an idiot" if your templates assume that you have prepositions - but in fact you have postpositions
same thing would probably arise with Turkish 22:31
22:31 buildable6 left
2. I kinda hope that the translations show how untenable the "negation lifting" is 22:31
lizmat well, that's why there are separate entries for block-if and modifier-if 22:32
nemokosch in a lot of languages (I'm inclined to say, in most languages) there isn't a word like English "any"
so any(1, 2) != 2 turning into none(1, 2) == 2 will really look like a nonsense 22:33
lizmat well.. I see this as a first attempt at localization
22:33 buildable6 joined
lizmat it won't be ideal in some languages, and maybe even impossible in other 22:34
nemokosch and I'm deliberately trying to keep the "negativity" down about the mere idea
trying to focus on the positives, and in my eyes, ditching "negation lifting" could be a beneficial side effect of this effort 22:35
having to face how untenable it is once somebody escapes English
lizmat well, one of the side-effects has already been internal :-)
nemokosch nighties 22:36
actually, is / ist is pretty painful in a language that has zero copula for X is Y type of sentences 22:39
I guess Russian is even worse in this regard
22:52 sena_kun left 23:23 vrurg_ joined 23:25 vrurg left 23:31 vrurg_ left, vrurg joined 23:33 buildable6 left 23:35 buildable6 joined