🦋 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: ... | log inspection situation still under development | For MoarVM see #moarvm
Set by lizmat on 22 May 2021.
00:02 reportable6 left 00:03 reportable6 joined 02:04 dogbert11 joined 02:08 dogbert17 left 02:44 frost joined 02:51 dogbert11 left 02:52 dogbert11 joined 03:44 dogbert11 left, dogbert11 joined 03:53 dogbert11 left 03:54 dogbert11 joined 05:43 coverable6 left, notable6 left, linkable6 left, committable6 left, nativecallable6 left, greppable6 left, statisfiable6 left, bisectable6 left, evalable6 left, unicodable6 left, squashable6 left, benchable6 left, sourceable6 left, releasable6 left, reportable6 left, tellable6 left, shareable6 left, bloatable6 left, quotable6 left, sourceable6 joined, unicodable6 joined, linkable6 joined 05:44 notable6 joined, tellable6 joined, benchable6 joined, nativecallable6 joined, shareable6 joined 05:45 bisectable6 joined, committable6 joined, bloatable6 joined, evalable6 joined, statisfiable6 joined, squashable6 joined
nine [Coke]: I'm leaning towards a no there for the reasons zoffix already gave 06:36
bartolin nine: would you be ok with me merging github.com/Raku/nqp/pull/722 (the second part of the workaround for precompilation of native arrays for the JVM backend)? 06:46
06:53 coverable6 joined 08:03 quotable6 joined 08:04 reportable6 joined 08:11 greppable6 joined 08:56 releasable6 joined 10:29 sena_kun left 10:30 sena_kun joined 10:31 Altai-man joined 10:37 nuno joined 10:53 nuno left
nine bartolin: oh, sure! 11:03
Geth nqp: 163db9a942 | (Christian Bartolomäus)++ | 4 files
[JVM] Keep correct REPR.name for native arrays

It turned out that f3dc977 doesn't really fix the usage of native arrays in the core setting. For some reason during the compilation of blib/CORE.d.setting.jar native array are deserialized twice.
The first time worked as expected, but since we did reset REPR.name ... (6 more lines)
nqp: db3af058a9 | (Christian Bartolomäus)++ | 3 files
[JVM] Use more generic name for new REPR field

This allows to use the new field for other REPRs than VMArrays as well, if the need arises.
bartolin thanks for looking 11:19
Geth rakudo/faster-regex-grep: bf82d828a0 | (Elizabeth Mattijsen)++ | 2 files
Make .grep(Regex) and .first(Regex) about 40% faster

By not needing to create relative expensive .Match objects when only a Bool can do.
For historic reasons, Regex.ACCEPTS is *not* expected to return a Bool (contrary to all other .ACCEPTS methods) but a Match object. ... (7 more lines)
11:20 Kaiepi left
Geth rakudo: lizmat++ created pull request #4383:
Make .grep(Regex) and .first(Regex) about 40% faster
11:21 Kaiepi joined 11:31 Kaiepi left 12:02 reportable6 left
Geth rakudo/workers-for-long-queues: 39e5834dc4 | (Jonathan Worthington)++ | src/core.c/ThreadPoolScheduler.pm6
Tweak worker addition for many small tasks

Today we add additional workers to the thread pool based upon lack of progress. This works well when work items are relatively large, and so hold on to a worker for a while. For applications that schedule lots of small pieces of work that are well below the threshold at which the supervisor runs, however, some work is always completed, and so the ... (24 more lines)
12:54 frost left 14:44 evalable6 left, bloatable6 left, releasable6 left, sourceable6 left, shareable6 left, committable6 left, bisectable6 left, statisfiable6 left, tellable6 left, notable6 left, benchable6 left, unicodable6 left, greppable6 left, squashable6 left, nativecallable6 left, coverable6 left, linkable6 left, quotable6 left, releasable6 joined 14:45 shareable6 joined, statisfiable6 joined, reportable6 joined, unicodable6 joined, notable6 joined, linkable6 joined, nativecallable6 joined, tellable6 joined 14:46 squashable6 joined, committable6 joined, evalable6 joined, quotable6 joined, coverable6 joined, bisectable6 joined, benchable6 joined 14:47 sourceable6 joined, bloatable6 joined, greppable6 joined 14:57 b2gills left 15:38 ilogger2 joined, kevin1 joined, SmokeMachine joined 15:39 smarton joined 15:44 sena_kun joined
Geth rakudo/faster-regex-grep: cfe58e5fdf | (Elizabeth Mattijsen)++ | 2 files
Do the cursoring inside .grep / .first

Get rid of the implementation-detail method, do the raw cursoring in the .grep / .first logic directly for performance.
This now makes Joelle Maslak's[1] example about 25% faster without any changes to the code.
  [1] twitter.com/jmaslak/status/1397706847658778627
17:41 Kaiepi joined 18:04 reportable6 joined 18:13 Kaiepi left
Geth roast: df696b48da | (Christian Bartolomäus)++ | S09-typed-arrays/native-shape1-str.t
[JVM] Fudge (todo) recently added test
roast: 1233ff95ea | (Christian Bartolomäus)++ | S11-modules/require.t
[JVM] Unfudge now passing test

Fixed with github.com/Raku/nqp/commit/00408eea93.
20:50 rba joined 21:26 rypervenche joined 22:31 evalable6 joined 22:32 linkable6 joined 23:32 evalable6 left, reportable6 left, linkable6 left, unicodable6 joined, linkable6 joined, coverable6 joined 23:33 committable6 joined, quotable6 joined, evalable6 joined, releasable6 joined, notable6 joined, bisectable6 joined, nativecallable6 joined 23:34 tellable6 joined, shareable6 joined, sourceable6 joined, reportable6 joined 23:35 greppable6 joined, squashable6 joined, statisfiable6 joined, bloatable6 joined