🦋 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.
lizmat M1 Files=1353, Tests=116312, 194 wallclock secs (16.21 usr 4.05 sys + 1118.56 cusr 59.49 csys = 1198.31 CPU) 08:37
Xliff m: ($*TZ / 3600).fmt('%02d'); 14:03
camelia ( no output )
Xliff m: ($*TZ / 3600).fmt('%02d').say
camelia 02
Xliff m: (-14400 / 3600).fmt('%02d').say 14:04
camelia -4
Xliff m: (-14400 / 3600).fmt('%03d').say
camelia -04
Xliff Bug?
Sign should not be accounted for in width spec, right?
lizmat: ^^
lizmat Xliff: not sure.. would have to look that up 17:42
SmokeMachine is a point release going to be done? 2022.06 is breaking Red, right? 17:56
people using Red should not update, right?
lizmat SmokeMachine: yeah, looks like 18:08
could you verify that it is not broken on HEAD atm ? 18:09
SmokeMachine I tried that after your fix and that was working... I'll try that again now
SmokeMachine and is there any idea of why Blin haven't found the problem? 18:18
lizmat: it seems to be working: usercontent.irccloud-cdn.com/file/.../image.png 18:21
lizmat ok, good to know
I guess we need to ping jdv
SmokeMachine I'm going to release Red and I don't know if I should add any kind of warning on GH/doc page about 2022.06... 18:29
lizmat perhaps you could code it into Red, to say you cannot run on that ? 18:33
SmokeMachine lizmat: that's a good idea
lizmat: would you agree with this message? usercontent.irccloud-cdn.com/file/.../image.png 18:59
lizmat s/instanciation/instantiation/
SmokeMachine thanks 19:00
usercontent.irccloud-cdn.com/file/.../image.png 19:01
does that make sense? 19:06
lizmat SmokeMachine: sorry, was distracted 19:12
SmokeMachine np :)
lizmat SmokeMachine: perhaps make it more lines of ~70 chars each ? 19:13
SmokeMachine I was thinking on breaking it in multiple lines...
patrickb Any dev around that's able to provide pointers on how to hunt a precomp error? 19:24
lizmat what's the error ?
patrickb I think the app I'm working on fails to recompile a changed source file and happily runs with old precomp files.
SmokeMachine usercontent.irccloud-cdn.com/file/.../image.png 19:25
patrickb I could just nuke the precomp dirs, but maybe the constellation is of value to someone more knowledeable.
lizmat patrickb: been in a similar situation 19:34
you could try raku.land/zef:lizmat/Repository::P...p::Cleanup 19:35
and if that fails, well, nuke install :-) 19:36
SmokeMachine github.com/FCO/Red/commit/bdd5bbdf...b9a726fb2a
lizmat SmokeMachine++ 19:36
afk& 19:38
Geth rakudo: ffdeecd1b9 | (Elizabeth Mattijsen)++ | src/Perl6/bootstrap.c/BOOTSTRAP.nqp
Fix copy-pasto in error message for native unsigned ints
20:19
rakudo: 378a89e67b | (Elizabeth Mattijsen)++ | src/Perl6/bootstrap.c/BOOTSTRAP.nqp
Simplify $has_varname setting

It's not a lot, but it happens for *every* parameter binding
20:28
patrickb Are there any docs about the precomp machinery? 21:04
found docs/module-management.md 21:06
patrickb Argh! Read through large parts of the above docs, gained some understanding to debug this further, only to have the misbehavior disappear, because stupid me modified the repo chain. 22:06
So now I'm waiting for the behavior to reappear ASAP so I won't have forgotten half of what I know now again. 22:07
Meh.
lizmat sounds very familiar :-(
Geth rakudo: 750487c6b1 | (Elizabeth Mattijsen)++ | src/Perl6/bootstrap.c/BOOTSTRAP.nqp
Simplify native value / container check in binder

All tests check for $got_native, so extract that first. And then use a ternary to set up the type of native that didn't match expectations. And then return with error if such a type is set.
This is mainly about having to do fewer checks and reducing the bytecode size. Sadly nowhere near inlining levels yet.
22:13