01:06 sxmx left, gfldex left, |Tux| left, djinni` left, krunen left 01:09 sxmx joined, gfldex joined, |Tux| joined, djinni` joined, krunen joined 01:10 sxmx left, sxmx joined 01:15 leont left 01:20 patrickb left 02:20 coverable6 left, benchable6 left, greppable6 left, sourceable6 left, linkable6 left, evalable6 left, shareable6 left, statisfiable6 left, tellable6 left, releasable6 left, unicodable6 left, bloatable6 left, squashable6 left, notable6 left, quotable6 left, committable6 left, nativecallable6 left, bisectable6 left 02:21 statisfiable6 joined, bloatable6 joined 02:22 sourceable6 joined, tellable6 joined, nativecallable6 joined, releasable6 joined, committable6 joined 02:23 squashable6 joined, evalable6 joined, quotable6 joined, unicodable6 joined, benchable6 joined, bisectable6 joined, shareable6 joined, coverable6 joined 02:24 notable6 joined, linkable6 joined, greppable6 joined 02:25 klapperl left 02:31 klapperl joined 02:35 [Coke]_ joined, [Coke]_ left, [Coke]_ joined, tyilanmenyn joined 02:36 vrurg_ joined, samcv_ joined 02:37 djinni`_ joined, HarmtH_ joined 02:38 gfldex_ joined, japhb_ joined 02:43 eaterof joined, literal_ joined, klapperl left, vrurg left, eater left, japhb left, [Coke] left, samcv left, HarmtH left, tyil left, literal left, gfldex left, |Tux| left, djinni` left, krunen left 02:49 krunen joined, klapperl joined, |Tux| joined 03:56 Xliff left 05:32 evalable6 left, linkable6 left 05:34 linkable6 joined 05:35 evalable6 joined 06:39 evalable6 left, linkable6 left, evalable6 joined 06:40 linkable6 joined 08:33 gfldex_ is now known as gfldex 08:49 tyilanmenyn is now known as tyil
bartolin lizmat: there where some dying spectests with your branch on the JVM backend, because adverbs on array slices didn't work. I hope that's easy-ish to fix -- I've added a comment to the PR 09:33
r: my @array = <A B>; say @array[0,1]:p ## died with 'java.lang.RuntimeException: This type does not support positional operations'
camelia (0 => A 1 => B)
09:46 sena_kun joined
Geth nqp: JJ++ created pull request #692:
Minor corrections
10:04 patrickb joined
patrickb o/ 10:09
bartolin: Can you help me do a rebootstrap of the JVM backend? (Is there any documentation besides nqp/docs/bootstrapping.pod?) 10:11
Here is a short explanation of the problem I face: colabti.org/irclogger/irclogger_lo...01-02#l150 10:12
10:47 frost-lab left 11:30 sena_kun left 11:54 leont joined
Geth rakudo/faster-slice-access: 76d3e33811 | (Elizabeth Mattijsen)++ | src/core.c/Rakudo/Internals/PostcircumfixAdverbs.pm6
Hopefully fix JVM backend issues
[Tux] Rakudo v2020.12-32-g8150a2864 (v6.d) on MoarVM 2020.12-19-gf36eaf0a6
csv-ip5xs0.825 - 0.827
csv-ip5xs-208.122 - 8.185
csv-parser24.698 - 26.312
csv-test-xs-200.381 - 0.383
test7.812 - 7.966
test-t1.884 - 1.893
test-t --race0.893 - 0.903
test-t-2033.014 - 33.347
test-t-20 --race9.350 - 9.520
13:28 lucasb joined 13:50 patrickb left 15:06 sena_kun joined
bartolin patrickb: I'm struggling with the rebootstrap for the JVM backend, too. As far as I remember, one has to add the new implementation first (parallel to the old implementation), generate new stage0 files afterwards and only then use the new implementation in the compiler. At least that's how I did it when I wanted to change the order of arguments for bindlexdyn: 1.) github.com/Raku/nqp/commit/db3e4a043f 2.) github.com/Raku/nqp/commit/0610 15:44
tellable6 bartolin, I'll pass your message to patrickb
bartolin but it seems to be more complicated if the op is used by nqp itself (as in your case). You can't just change it in src/vm/jvm/QAST/Compiler.nqp and src/core/testing.nqp at the same step (or so it seems) 15:45
I'd think the following should work: 1.) Add a new op 'spawnprocasync2' to src/vm/jvm/QAST/Compiler.nqp and new versions of the Java methods AsyncProcessHandle and spawnprocasync -- all with the new calling convention 15:48
2.) make j-bootstrap-files 15:49
3.) change src/core/testing.nqp to use the new op spawnprocasync2 and run make akain 15:50
4.) change the definition of 'spawnprocasync' in src/vm/jvm/QAST/Compiler.nqp (no unused) to used the additional parameter and run make j-bootstrap-files again 15:51
5.) change src/core/testing.nqp back to use the updated version of spanprocasync and run make again 15:52
6.) cleanup (remove spawnprocasync2 and now unused versions of the Java methods) and run make j-bootstrap-files again
7.) commit new stage0 files 15:53
8.) commit remaining code
No guarantee that this works. And maybe there is a much simpler way. 15:54
Oops, step 2.) should be 'make && make j-bootstrap-files'
(Not making commits for the intermediate steps would avoid the temporary versions of stage0 files to be committed. But it's not transparent how the change was done.) 15:56
15:57 Altai-man joined 16:00 sena_kun left
Altai-man releasable6, status 16:58
releasable6 Altai-man, Next release in ≈20 days and ≈2 hours. 1 blocker. Changelog for this release was not started yet
Altai-man, Details: gist.github.com/672b8ee6587b0309f0...cb131c1af6
timotimo i added a "if stage 0" (or something) thing to gen-cat so that you can have the same source code including "new" features 17:02
17:21 JRaspass joined 18:28 tobs left 18:29 tobs joined 18:32 tobs left 18:35 tobs joined 18:38 lucasb left 19:11 [Coke]_ is now known as [Coke] 19:30 Xliff joined 20:55 ilogger2 joined 21:06 b2gills joined 22:02 vrurg joined 22:32 Kaiepi joined
gfldex m: say Routine.^methods(:local); 22:35
camelia (The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside The 'ForeignCode' class is a Rakudo-specific
implementation detail and has no serviceable parts inside The 'ForeignCode' class is a Rakudo-spec…
MasterDuke m: .name.say for Routine.^methods(:local); 22:36
camelia <anon>
gfldex m: say Routine.^methods(:local).grep(?*)».name; 22:38
camelia (<anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> <anon> onlystar candidates cando multi soft wrap unwrap package leave gist raku BUILDALL)
gfldex m: say Routine^.can('dispatcher'); 22:41
camelia one((Routine), ())
gfldex Routine.dispatcher is quite stealthy.