🦋 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.
00:00 reportable6 left 00:02 reportable6 joined
Geth rakudo/ugexe/curfs-sha1-meta6: 5ec81cad35 | (Nick Logan)++ | src/core.c/CompUnit/Repository/FileSystem.pm6
Consider META6.json contents when calculating CURFS id

Currently when the contents of META6.json change, precompilation may not reoccur and thus a stale $?DISTRIBUTION may be getting used. This takes the contents of the META6.json into consideration when checksumming for differences, which should avoid such a situation.
rakudo: ugexe++ created pull request #5283:
Consider META6.json contents when calculating CURFS id
01:02 evalable6 left, evalable6 joined 01:13 heartburn left, [Tux] left, timo left, samebchase left, gfldex left 01:15 guifa left, camelia left 01:16 patrickb left, sjn left, japhb left 01:17 summerisle left, nebuchadnezzar left 01:18 RakuIRCLogger left, jgaz left, lizmat_ joined, guifa_ joined 01:20 bisectable6 left, shareable6 left, shareable6 joined, bisectable6 joined, codesections left, timo1 joined, releasable6 left, bloatable6 left, tellable6 left, reportable6 left, unicodable6 left, Geth left, nativecallable6 left, coverable6 left, sourceable6 left, gfldex_ joined, nebuchad` joined, heartburn joined, kurahaupo left, lizmat left, sjn_ joined, [Tux] joined, lizmat joined 01:21 camelia joined, shareable6 left, bisectable6 left 01:22 lizmat_ left, timo1 left 01:23 sjn_ left, nebuchad` left, heartburn left, sjn joined, timo1 joined 01:24 gfldex_ left, heartburn joined 01:25 gfldex joined 01:29 summerisle joined 01:30 patrickb joined 01:33 japhb joined, codesections joined 01:37 jgaz_ joined 02:00 japhb left 02:01 japhb joined 02:04 releasable6 joined, tellable6 joined, reportable6 joined 02:05 unicodable6 joined, bloatable6 joined 02:13 nativecallable6 joined 02:14 coverable6 joined, sourceable6 joined 02:15 bisectable6 joined 02:17 shareable6 joined 02:34 Xliff left 05:10 vrurg_ joined 05:12 vrurg left 06:00 reportable6 left 06:03 reportable6 joined 07:37 Geth joined
Geth rakudo/lizmat-deparse-dutch: 490f13825e | (Elizabeth Mattijsen)++ | 3 files
RakuAST: move ::Deparse::NL -> ::Deparse::L10N::NL

As suggested by vrurg++ and ugexe++
08:19 Geth left 08:20 Geth joined
Geth rakudo/main: e5568f9322 | (Elizabeth Mattijsen)++ | 2 files
RakuAST: move ::Deparse::CORE to ::Deparse::L10::CORE

to make its purpose (localization) more clear
rakudo/main: 34854eb407 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | 2 files
RakuAST: add Dutch in foreign language deparsing
10:17 linkable6 left, evalable6 left 10:20 evalable6 joined, linkable6 joined
lizmat idea for future RakuAST optimization 10:27
annotate any routine with a marker that if it is just returning any of the input parameters unchanged 10:28
so that new-disp will be able to just replace it with that parameter *without* needing to call anything 10:29
think methods returning self (such as Str.Str)
or hook methods such xsyn returning the second positional (aka 3rd arg) by default 10:30
also think allowing --> self as syntax 10:31
and the routine is otherwise empty 10:36
11:24 codesections left
[Tux] Rakudo v2023.05-115-g34854eb40 (v6.d) on MoarVM 2023.05-6-gb039b913b
csv-ip5xs0.857 - 0.902
csv-ip5xs-205.365 - 5.551
csv-parser3.880 - 4.407
csv-test-xs-200.382 - 0.467
test6.569 - 6.668
test-t1.421 - 1.519
test-t --race0.878 - 0.903
test-t-2021.829 - 23.179
test-t-20 --race6.336 - 7.228
12:00 reportable6 left 12:02 reportable6 joined
nine lizmat: how many such identity functions are there? How would we communicate this from the AST to the dispatcher? Anyway, spesh should be doing a pretty good job at inlining such functions and optimizing away all the overhead. 12:06
lizmat I would guess a hint in the QAST ? doesn't the current static optimizer already look at some hints? timo1 ? 12:07
well, if a dispatcher could already do it, why leave it to spesh? 12:08
nine QAST can carry annotations, but the dispatcher only has byte code and meta objects to work with 12:09
lizmat then in the meta object? 12:10
nine Actually I meant MVMCode objects 12:23
Low level is low :)
13:02 evalable6 left, linkable6 left 13:03 evalable6 joined 13:04 linkable6 joined 13:51 uzl[m] left 13:58 uzl[m] joined 14:01 nine left, nine joined, codesections joined 14:04 kurahaupo joined 14:16 kurahaupo left 14:18 kurahaupo joined 14:31 nebuchadnezzar joined 14:48 squashable6 left 14:50 ab5tract joined, squashable6 joined
lizmat . 15:09
hoi 15:10
16:04 evalable6 left, linkable6 left 16:06 evalable6 joined, linkable6 joined 17:48 sena_kun joined 18:00 reportable6 left 18:02 reportable6 joined 19:08 ab5tract left 20:00 codesections left
Geth rakudo: patrickbkr++ created pull request #5285:
Precomp Pipeline: Fix building on CentOS 7
patrickb I have now fixed all issues that prevented the precompiled binaries on rakudo.org (and via `rakubrew download`) to be built. 20:10
AFAIK it's less than two weeks until 2023.06 is scheduled for release. Should we just skip the binaries for 2023.05 or should we do a point release? 20:11
jdv: Is it intentional that the next release date in docs/release_guide.pod is a Monday and not a Saturday? 20:15
lizmat: Bumping moar will get the Windows tests working again. Can you do the honors? 20:25
japhb I'd vote for the point release, if nothing else just to ensure everything works as expected now.
lizmat patrickb: am unsure what you want me to do 20:52
nothing changed with Moar since June 4th ? 20:53
Geth rakudo/main: 0ff1118a8c | (Elizabeth Mattijsen)++ | src/Raku/ast/base.rakumod
RakuAST: .DEPARSE now accepts multiple roles

To make it easier to specify any language and any highlighting roles seperately.
21:13 jgaz_ left 21:14 jgaz joined 21:57 sena_kun left 22:40 kurahaupo left, kurahaupo joined