01:06 kylese left 01:07 kylese joined 01:08 sibl joined 01:23 [Coke] left 01:52 sftp left 02:15 kylese left, kylese joined
ugexe hmm github.com/rakudo/rakudo/commit/840e3eaf5b made it so $?FILE.IO.absolute is non-deterministic, i.e. based on runtime cwd 02:55
so e.g. debuggers may show stack traces from the wrong place 02:58
03:00 sibl left 03:02 sibl joined
ugexe i noticed this because `t/spec/S24-testing/line-numbers.t` fails with a RAKUDO_RAKUAST=1 built rakudo when run with `RAKUDO_RAKUAST=1 make spectest`, but not when run with `RAKUDO_RAKUAST=1 ./install/bin/raku -Ilib t/spec/S24-testing/line-numbers.t` 03:06
03:20 [Coke] joined
disbot2 <cokebot9000> Blin run results (short version): gist.github.com/coke/72c5436fbaa23...f954bae46a 03:21
03:29 sibl left 03:30 guifa joined 03:32 sibl joined
ugexe I’m inclined to think the net telnet one is a fluke 03:44
The commit it blames is rakuast only 03:46
[Coke] Yup, the bisect is typically suspect, unfortunately. 03:48
working rechecking those x by hand.
sorry, "worth", not "working"
We need to fix Testo for the unit/class/role issue you cleaned up. Also, it's very weird that we have a module called Testo that we are maintaining that is advertising itself as the best way to test stuff with Raku 03:50
Testo - Raku Testing Done Right
zzz 03:51
03:56 guifa left 04:06 guifa joined 04:11 guifa left 04:13 Aedil left 04:15 annamalai joined 04:18 vasko4535586 joined 04:28 Aedil joined
patrickb re $?FILE, the debugger UI I'm working on does not rely on the path of $?FILE. It only looks at the basename. If that matches it compares the checksum to determine if it's the same file. 05:53
Reason being that when debugging it's not necessarily the case that the local sources are located in the same folder as the sources that the VM is working with. So the path is not a good thing to look at to determine files being the same. 05:55
So I don't care much whether FILE is absolute or relative. 05:56
06:29 abraxxa joined 06:46 Sgeo left 07:56 sibl left 08:19 sibl joined 09:17 sibl left 11:56 oodani left 11:58 oodani joined
ugexe i just mean a debugger in general 13:42
that being said i dont have a suggestion for how to fix the reproducibility issue the $?FILE shape change was meant to fix 13:45
a different suggestion rather
i'm inclined to just try to change the failing test, but i wanted to point out the potential issue in case someone recognized it to be particularly problematic 13:50
there are only a couple of `RAKUDO_RAKUAST=1 make spectest` failures left. we'll need to start running a CI spectest for rakuast soon to keep that from regressing 13:57
github.com/Raku/roast/blob/master/...TTING-6e.t - i feel like this isn't an appropriate test for the roast 14:02
is the number of scopes really part of the spec or an implementation detail? i would argue it is the later given rakuast only installs one level of scope per language revision and legacy installs 2, but that difference makes that test impossible to write in a way that makes the test useful at all 14:03
arguably it could be a test in the core, where at least it ok to detect the difference between which backend is being used and adjust test expectations accordingly 14:04
disbot2 <melezhik.> o/ 14:14
<melezhik.> I would like use perl5 regexp with comb operator
<melezhik.> github.com/melezhik/Sparrow6/blob/...kumod#L223 14:15
<melezhik.> Is it possible ?
<melezhik.> m: / "ok" ~~ m:Perl5/[a-z] 14:17
evalable6 (exit code 1) ===SORRY!=== Error while compiling /tmp/0uoVLGNjGr
Unrecognized regex modifier :Perl5
at /tmp/0uoVLGNjGr:1
------> / "ok" ~~ m:<HERE>Perl5/[a-z]
disbot2 <melezhik.> m: / "ok" ~~ m:Perl5/[a-z]/
evalable6 (exit code 1) ===SORRY!=== Error while compiling /tmp/MpWLY_PnXw
Unrecognized regex modifier :Perl5
at /tmp/MpWLY_PnXw:1
------> / "ok" ~~ m:<HERE>Perl5/[a-z]/
disbot2 <melezhik.> m: “ok" ~~ m:P5/[a-z]/ 14:19
evalable6 (exit code 1) ===SORRY!=== Error while compilin…
melezhik., Full output: gist.github.com/0a64efdf5b3e422987...3a1f4be914
timo m: say "hello".comb(rx:P5/[a-z]/) 14:27
evalable6 (h e l l o)
disbot2 <melezhik.> timo thanks
<melezhik.> m: my $data = "ok"; my $pattern = "ok"; my $matched = $data.comb(/<mymatch=$pattern>/,:match)>>.<mymatch>; 14:30
evalable6
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "ok"; my $matched = $data.comb(/<mymatch=rx;P5$pattern>/,:match)>>.<mymatch>; 14:31
evalable6 (exit code 1) ===SORRY!=== Error while compilin…
melezhik., Full output: gist.github.com/e95e8a9cae0bf25c26...a7740d42c4
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "ok"; my $matched = $data.comb(/<mymatch=rx:P5$pattern>/,:match)>>.<mymatch>;
evalable6 (exit code 1) ===SORRY!=== Error while compilin…
melezhik., Full output: gist.github.com/dfd1c129b433ef0b61...963a5b741b
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "ok"; my $matched = $data.comb(rx:P5/$pattern>/); 14:33
evalable6
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "ok"; my $matched = $data.comb(rx:P5/$pattern>/); say $matched; 14:34
evalable6 ()
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "[a-z]"; my $matched = $data.comb(rx:P5/$pattern>/); say $matched; 14:35
evalable6 ()
[Coke] You can do this privately as well. :)
disbot2 <melezhik.> m: my $data = "ok"; my $pattern = "([a-z])"; my $matched = $data.comb(rx:P5/$pattern>/); say $matched; 14:36
evalable6 ()
15:05 [Coke] left 15:15 human-blip left 15:17 human-blip joined 15:44 tgt joined 15:50 tgt left 16:27 abraxxa left
disbot2 <melezhik.> Sudo passwordless check - chat.deepseek.com/share/is2ey6vgbhvdbyqtoa 17:45
18:49 human-blip left 18:51 human-blip joined 20:26 vrurg joined 20:27 annamalai left 20:29 vrurg_ left 20:30 vrurg_ joined 20:33 vrurg left
gfldex Can I get evalable6 to run RakuAST? 20:34
20:42 ShimmerFairy left 20:43 ShimmerFairy joined 20:55 itaipu left 21:10 itaipu joined 22:07 Sgeo joined 23:28 itaipu left 23:48 itaipu joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!