🦋 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 00:05 reportable6 joined 04:39 nebuchadnezzar joined 06:02 reportable6 left 06:05 reportable6 joined 06:19 linkable6 left 06:20 linkable6 joined 07:30 patrickb joined
Geth rakudo: 168da39ae2 | (Elizabeth Mattijsen)++ | src/core.c/Iterable.pm6
Simplify Iterable eqv Iterable

By removing the specific Positional handling. This every so lightly improves performance on comparing arrays, while not affecting any other performance.
This also removes the use of AT-POS in eqv handling, making it easier to later introduce proper eqv handling of sparse arrays.
07:53 patrickb left 08:27 nebuchadnezzar left 09:31 dogbert11 left
Geth nqp: MasterDuke17++ created pull request #728:
Simplify MoarVM backend runner's Makefile template
rakudo/sparse-array-raku: 484baa6a57 | (Elizabeth Mattijsen)++ | 4 files
Allow Nil to represent holes in sparse arrays

This commit affects the way arrays are rendered with .raku, and it affects the way Nil is interpreted as a value when initializing an array.
This is achieved by the following conceptual changes: ... (31 more lines)
rakudo: lizmat++ created pull request #4404:
Allow Nil to represent holes in sparse arrays
10:38 linkable6 left, evalable6 left 10:39 evalable6 joined 10:40 linkable6 joined 11:02 morayj joined
lizmat we currently have no way to force a control exception to happen on another thread, right ? 11:56
12:03 reportable6 left, reportable6 joined
Geth rakudo/repl-control-c: 313f0cc1f4 | (Elizabeth Mattijsen)++ | src/core.c/REPL.pm6
Allow CTRL-c to stop entry in REPL

And a second CTRL-c to exit the REPL if nothing else was entered.
This in response to #4402.
This does *not* achieve the suggested behaviour of being able to stop a runaway piece of code, as Raku's signal handler runs on a different thread and we currently do not have a way to force an exception on anything else but the current thread. It *does* allow one to restart entry without having to remove all characters.
rakudo: lizmat++ created pull request #4405:
Allow CTRL-c to stop entry in REPL
rakudo/repl-control-c: c52e11dbec | (Elizabeth Mattijsen)++ | src/core.c/REPL.pm6
Don't need to check if signal tap is installed

As the REPL loop is actually inside the repl-loop method.
12:51 morayj left 13:03 stoned75 joined
stoned75 hi 13:04
I'm looking at rakudo's Num.Str implementation and I must admit I do not understand how it works :-( 13:06
reading github.com/rakudo/rakudo/blob/mast...m.pm6#L120 13:07
I understand that a "numeric" is unboxed and directly feeded to p6box_s() which does... a string boxing. 13:08
but where the stringification occurs ?
in unbox_n() or in p6box_s() ?
If I believe github.com/rakudo/rakudo/blob/mast...wn#p6box_s it can't be p6box_s(), am I wrong ? 13:18
MasterDuke aiui, nqp sees a num being fed into a register that expects a str, and so calls coerce_ns 13:22
which ends up here github.com/MoarVM/MoarVM/blob/mast...#L257-L278 13:24
stoned75 oooh
13:25 djinni` left, morayj joined 13:26 djinni` joined
stoned75 I guess MASTCompilerInstance.coerce()/coercion() does this coercion dispatch ? 13:30
MasterDuke yeah github.com/Raku/nqp/blob/master/sr...#L412-L649 13:34
stoned75 Ok I guess I more or less got it. thanks ! 13:45
MasterDuke np 13:46
13:53 morayj left
Geth nqp: MasterDuke17++ created pull request #729:
Remove no longer needed JVM workaround
releasable6 Next release in ≈4 days and ≈3 hours. 1 blocker. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 15:00
15:06 vrurg_ joined 15:07 vrurg left
Geth nqp: e305efcb31 | (Daniel Green)++ | src/HLL/SysConfig.nqp
Remove no longer needed JVM workaround

The JVM has been re-bootstrapped and `nqp::execname` now works with it.
nqp: 0ac7daada6 | MasterDuke17++ (committed using GitHub Web editor) | src/HLL/SysConfig.nqp
Merge pull request #729 from MasterDuke17/remove_no_longer_needed_jvm_workaround_in_HLL__SysConfig

Remove no longer needed JVM workaround
15:35 morayj joined
Geth rakudo/rakuast: 893bb80f03 | (Jonathan Worthington)++ | 3 files
RakuAST node for s/// and friends

There's rather a lot of semantics in here, although the implementation in the current compiler frontend seems to make things a little more complex or less efficient than they could be. Of note:
  * It seems we can emit code to truth-test the `.match` result, rather
... (6 more lines)
[TuxCM] Rakudo v2021.05-17-g168da39ae (v6.d) on MoarVM 2021.05-14-g0a62f0c6a
csv-ip5xs0.836 - 0.859
csv-ip5xs-208.603 - 8.918
csv-parser26.232 - 26.591
csv-test-xs-200.370 - 0.373
test7.393 - 8.511
test-t1.983 - 2.089
test-t --race0.882 - 0.943
test-t-2034.969 - 35.304
test-t-20 --race9.631 - 10.305
17:40 morayj left 18:02 reportable6 left 18:05 reportable6 joined 18:15 stoned75 left 20:23 rypervenche left 20:40 rypervenche joined 21:54 Voldenet joined