[00:06] *** reportable6 left [00:08] *** reportable6 joined [01:16] *** linkable6 left [01:16] *** evalable6 left [01:32] I got 2 implementations for .assuming that are way faster then what we got right now: https://gist.github.com/647eb0925ec65bdba11107381557d974 [03:18] *** linkable6 joined [04:18] *** evalable6 joined [04:37] https://www.reddit.com/r/perl6/comments/apksgq/i_made_a_partial_applicator/ I did this a while ago [04:37] idk performance nor compatibility, but it doesn't use EVAL [05:01] *** Altai-man joined [05:03] *** sena_kun left [06:06] *** reportable6 left [06:47] *** Xliff left [07:07] *** reportable6 joined [07:07] *** Util left [07:20] *** Util joined [07:42] *** TheAthlete joined [08:21] ¦ nqp: c4b121432a | (Daniel Green)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp [08:21] ¦ nqp: Register `decont_u` as an op [08:21] ¦ nqp: [08:21] ¦ nqp: Missed in the recent unsigned work. [08:21] ¦ nqp: review: https://github.com/Raku/nqp/commit/c4b121432a [08:21] ¦ nqp: 1fcf95ef90 | MasterDuke17++ (committed using GitHub Web editor) | src/vm/moar/QAST/QASTOperationsMAST.nqp [08:21] ¦ nqp: Merge pull request #758 from MasterDuke17/register_decont_u_as_an_op [08:21] ¦ nqp: review: https://github.com/Raku/nqp/commit/1fcf95ef90 [08:27] ¦ rakudo: b3a71fbc61 | (Daniel Green)++ | tools/templates/NQP_REVISION [08:27] ¦ rakudo: Bump NQP to pick up `decont_u` registered as an op [08:27] ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/b3a71fbc61 [08:27] ¦ rakudo: e5b67c1d86 | (Daniel Green)++ | src/Perl6/Optimizer.nqp [08:27] ¦ rakudo: Eliminate p6decontrv ops caused by `*_u` ops [08:27] ¦ rakudo: [08:27] ¦ rakudo: Otherwise `my uint $a; sub foo { $a = 0 }` gives [08:27] ¦ rakudo: `===SORRY!=== [08:27] ¦ rakudo: p6decontrv expects a QAST::WVal as its first child`. [08:27] ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/e5b67c1d86 [08:27] ¦ rakudo: 9a17526341 | MasterDuke17++ (committed using GitHub Web editor) | 2 files [08:27] ¦ rakudo: Merge pull request #4736 from MasterDuke17/eliminate_p6decontrv_ops_caused_by__u_ops [08:27] ¦ rakudo: review: https://github.com/rakudo/rakudo/commit/9a17526341 [08:59] *** evalable6 left [08:59] *** linkable6 left [09:19] ¦ rakudo: lizmat self-assigned Various methods in CompUnit::RepositoryRegistry are not thread safe https://github.com/rakudo/rakudo/issues/4737 [10:10] ¦ rakudo/lizmat-fix-custom-lib-race: 2be3bc6bfa | (Elizabeth Mattijsen)++ | src/core.c/CompUnit/RepositoryRegistry.pm6 [10:10] ¦ rakudo/lizmat-fix-custom-lib-race: Fix custom-lib race [10:10] ¦ rakudo/lizmat-fix-custom-lib-race: [10:10] ¦ rakudo/lizmat-fix-custom-lib-race: As described in #4737 [10:10] ¦ rakudo/lizmat-fix-custom-lib-race: review: https://github.com/rakudo/rakudo/commit/2be3bc6bfa [10:10] ¦ rakudo: lizmat++ created pull request #4738: Fix custom-lib race [10:10] ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/4738 [11:00] *** evalable6 joined [11:02] ¦ rakudo/lizmat-chomp-needle: 667ab79cfa | (Elizabeth Mattijsen)++ | 2 files [11:02] ¦ rakudo/lizmat-chomp-needle: Give .chomp the possibility to specify a needle [11:02] ¦ rakudo/lizmat-chomp-needle: [11:02] ¦ rakudo/lizmat-chomp-needle: Basically: [11:02] ¦ rakudo/lizmat-chomp-needle: [11:02] ¦ rakudo/lizmat-chomp-needle: say "foobar".chomp("bar"); # foo [11:02] ¦ rakudo/lizmat-chomp-needle: say "foobar".chomp("baz"); # foobar [11:02] ¦ rakudo/lizmat-chomp-needle: review: https://github.com/rakudo/rakudo/commit/667ab79cfa [11:02] ¦ rakudo: lizmat++ created pull request #4739: Give .chomp the possibility to specify a needle [11:02] ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/4739 [11:34] *** TheAthlete left [12:01] *** linkable6 joined [12:07] *** reportable6 left [12:33] ¦ rakudo/lizmat-fix-custom-lib-race: 8e034779d4 | (Elizabeth Mattijsen)++ | src/core.c/CompUnit/RepositoryRegistry.pm6 [12:33] ¦ rakudo/lizmat-fix-custom-lib-race: s/register-custom-lib-repo/register-custom-lib-repository [12:33] ¦ rakudo/lizmat-fix-custom-lib-race: review: https://github.com/rakudo/rakudo/commit/8e034779d4 [13:08] *** [Coke] left [13:08] *** sena_kun joined [13:53] *** discord-raku-bot left [13:54] *** discord-raku-bot joined [14:09] *** reportable6 joined [14:30] *** RakuIRCLogger joined [14:35] *** RakuIRCLogger left [14:35] *** RakuIRCLogger joined [14:39] ¦ rakudo/master: 5 commits pushed by (Vadim Belman)++ [14:39] ¦ rakudo/master: 2486960b70 | Further improvements to smartmatch dispatch [14:39] ¦ rakudo/master: abfb9f861d | Use smartmatch dispatcher for as chained fallback [14:39] ¦ rakudo/master: 6148b75c7a | Remove unused statement [14:39] ¦ rakudo/master: cd20a2bc34 | Allow optimizations over negated chained smartmatch [14:39] ¦ rakudo/master: 4365d9b6d4 | Merge pull request #4735 from vrurg/new-disp-smartmatch-2 [14:39] ¦ rakudo/master: review: https://github.com/rakudo/rakudo/compare/9a17526341f9...4365d9b6d456 [14:39] ¦ roast/master: 4 commits pushed by (Vadim Belman)++ [14:39] ¦ roast/master: bb95336dce | Fix incorrect tests [14:39] ¦ roast/master: d40dd18445 | Implement randomized order of indirect smartmatch tests [14:39] ¦ roast/master: 3a138500e3 | Add more tests for negated smartmatch [14:39] ¦ roast/master: 82594845a8 | Merge pull request #787 from vrurg/new-disp-smartmatch-2 [14:39] ¦ roast/master: review: https://github.com/Raku/roast/compare/b28a53b19eaf...82594845a83b [15:17] *** sena_kun left [15:17] *** sena_kun joined [15:37] *** TheAthlete joined [17:08] *** RakuIRCLogger left [17:08] *** TempIRCLogger__ joined [17:09] *** TempIRCLogger left [17:10] *** lizmat left [17:11] *** lizmat joined [17:11] *** [Tux] left [17:15] *** lizmat left [17:15] *** lizmat joined [17:18] *** [Coke] joined [17:21] *** Geth left [17:23] *** [Tux] joined [18:07] *** reportable6 left [18:09] *** reportable6 joined [18:24] *** sena_kun left [19:07] *** TheAthlete left [20:07] *** squashable6 left [20:07] *** evalable6 left [20:07] *** quotable6 left [20:07] *** notable6 left [20:07] *** linkable6 left [20:07] *** reportable6 left [20:07] *** unicodable6 left [20:07] *** shareable6 left [20:07] *** sourceable6 left [20:07] *** statisfiable6 left [20:07] *** committable6 left [20:07] *** coverable6 left [20:07] *** releasable6 left [20:07] *** nativecallable6 left [20:07] *** greppable6 left [20:07] *** tellable6 left [20:07] *** bloatable6 left [20:07] *** bisectable6 left [20:07] *** benchable6 left [20:08] *** releasable6 joined [20:08] *** shareable6 joined [20:09] *** evalable6 joined [20:09] *** benchable6 joined [20:09] *** unicodable6 joined [20:09] *** sourceable6 joined [20:10] *** greppable6 joined [20:10] *** notable6 joined [20:10] *** tellable6 joined [20:23] <[Tux]> Rakudo v2021.12-161-g4365d9b6d (v6.d) on MoarVM 2021.12-87-gec1ddee66 [20:23] <[Tux]> csv-test-xs-20 0.398 - 0.410 [20:23] <[Tux]> csv-ip5xs 0.755 - 0.791 [20:23] <[Tux]> test-t --race 0.838 - 0.850 [20:23] <[Tux]> test-t 1.525 - 1.531 [20:23] <[Tux]> csv-parser 3.880 - 3.975 [20:23] <[Tux]> csv-ip5xs-20 4.938 - 5.059 [20:23] <[Tux]> test 6.498 - 6.521 [20:23] <[Tux]> test-t-20 --race 7.554 - 7.599 [20:23] <[Tux]> test-t-20 22.627 - 22.704 [20:24] <[Tux]> 5 2021-11-08 13:21:09 test-t 1.540 [20:24] <[Tux]> 4 2021-10-29 17:47:15 test-t 1.536 [20:24] <[Tux]> 3 2021-10-28 16:30:08 test-t 1.534 [20:24] <[Tux]> 2 2022-01-26 20:04:45 test-t 1.531 [20:24] <[Tux]> 1 2022-01-26 20:06:28 test-t 1.525 [20:24] <[Tux]> 20211108 1.540❙20211029 1.536❙20211028 1.534❙20220126 1.531❙20220126 1.525❙ [20:39] nice [21:04] MasterDuke, nine: Looks like Digest still doesn't work at HEAD. Updated https://gist.github.com/japhb/68dded3c8576a1d8fcf968fee0c2ff79 (-145 is Rakudo before the p6decontrv fix, -161 is after; now it looks like a uint problem.) [21:07] "Cannot unbox negative bigint into native unsigned integer" i wonder if it's doing something wrong that just happened to work before [21:08] *** nativecallable6 joined [21:08] *** bloatable6 joined [21:09] *** squashable6 joined [21:09] *** reportable6 joined [21:13] m: use nqp; my str $s = (100..1_000).pick.Str; my int $i; my $n = now; $i = $s for ^10_000_000; say now - $n; say $i [21:13] rakudo-moar 4365d9b6d: OUTPUT: «0.394074819␤161␤» [21:14] a pretty simple change just about halves ^^^ runtime and gets rid of 10m temporary allocations, and i think it should be not too difficult to get even faster [21:17] and in a related question, wouldn't it be convenient for MVMStrings to both know their size and be null terminated? then, depending on encoding of course, the storage could be passed directly to C functions [21:19] guess it depends on how frequently we actually call C functions on strings to determine if the extra overhead (space, code complexity, etc) is worth it [21:19] but iirc, that's how D does it. or maybe C#? [21:24] oh, that would have made more sense over in #moarvm [22:09] *** linkable6 joined [22:10] *** coverable6 joined [22:10] *** quotable6 joined [22:11] *** committable6 joined [22:47] *** timo1 joined [22:48] *** timo left