🦋 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.
Geth rakudo/new-disp: 82e4d57578 | (Jonathan Worthington)++ | src/vm/moar/dispatchers.nqp
Correct a comment

  nine++ for noticing
09:33
rakudo/new-disp: 4 commits pushed by (Elizabeth Mattijsen)++ 09:52
moon-child m: say Version.new(5) 09:54
camelia v5
moon-child m: say Version.new(5, 6)
camelia Default constructor for 'Version' only takes named arguments
in block <unit> at <tmp> line 1
lizmat moon-child: what did you expect? 09:55
moon-child clearer error message
lizmat perhaps make an issue for that... Version is kinda special in that way 10:00
gfldex m: say Version.new(5|6); 10:03
camelia any(v5, v6)
gfldex .oO( Raku, a language for the irresolute. )
[Coke] Glad to see we're removing nqp invokes to speed things up! 11:45
That seems like an important point in our optimization strategy history. 11:46
Geth rakudo/new-disp: 11 commits pushed by (Elizabeth Mattijsen)++
review: github.com/rakudo/rakudo/compare/6...f26ae4af5e
Geth nqp/new-disp: fda44c5974 | (Timo Paulssen)++ | 2 files
use dispatcher-drop-n-args syscall in nqp dispatchers
12:07
rakudo/new-disp: 9c587d92d0 | (Timo Paulssen)++ (committed by Jonathan Worthington) | src/vm/moar/dispatchers.nqp
Use new dispatcher-drop-n-args syscall

Can reduce allocations of temporary MVMCapture and MVMCallsite objects.
12:08
DBIish: ancientwizard++ created pull request #211:
All Oracle Issues 210, 209, 207, 204, 203
12:22
[Tux] Rakudo v2021.09-4-gba15173fd (v6.d) on MoarVM 2021.09-1-gb516702c4
csv-ip5xs0.882 - 0.899
csv-ip5xs-209.217 - 9.409
csv-parser28.957 - 29.701
csv-test-xs-200.374 - 0.380
test7.860 - 8.790
test-t2.076 - 2.177
test-t --race0.928 - 1.004
test-t-2034.559 - 34.740
test-t-20 --race10.123 - 10.395
12:44
Geth rakudo/new-disp: 5 commits pushed by (Elizabeth Mattijsen)++ 13:02
Geth nqp/new-disp: c99c5bc93b | (Jonathan Worthington)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp
Add mapping for ctxnt op

And also map `curlexpad` to it rather than `ctx`.
15:07
Xliff \o 17:12
lizmat: I see you are making a lot of \ -> $ optimizations on new-disp. I would have thought \a was faster than $a. So the opposite is true? 17:13
lizmat Xliff: it will be in new-disp
as long as the parameter can not be an Iterabl;e
Xliff OIC. Thanks for the info! 17:14
lizmat at least, that's my current understanding :-)
Xliff I take it new-disp will start migrating to master over the next few weeks? 17:15
lizmat yes, that's the idea :-)
Xliff: if you want to be up-to-date on new-disp, #moarvm is the channel to be atm 17:16
Xliff Ah, yes. I keep forgetting about that. :> 17:25
lizmat: Did you finish your RakuAST implementation of sprintf? 17:29
lizmat yes, that's sitting in the RakuAST branch 17:31
for a while now already :-) 17:32
Xliff lizmat: Excellent! I think I'm going to play with it for a bit. :) 17:44
Geth URI: fe4031e9ad | (David Warring)++ | 3 files
add directory method
18:17
Xliff Anybody wanna play around on a 64-core Google Engine instance? 18:24
Looks like Raku may have a parallelism bug ... or maybe my script does.
Working script conks out when parallel compiling my p6-ICal project with no reason.
I have no idea what to look for. 18:25
If no takers, I'll pack it up and shut it down. It's costing me $3/hour 18:27
I'll check again in a couple of hours. 18:28
lizmat I think some people would be interested, but they're all busy doing something else, or having dinner :-)
also, I think you should also advertise / plan this on #moarvm :-)
[Coke] Also, where is the script that might be having trouble? Will get more eyes if it's in a gist or repo somewehre. 18:35
Xliff [Coke]: Same place it's always been: github.com/Xliff/p6-GtkPlus/blob/m...-build.pl6 18:38
[Coke]: That script has been used without issue (concurrently level 20) for well over a year now.
ugexe fwiw i dont think the trick of ordering provides works anymore since hash randomization was added years ago 18:39
timo we as a species need to think about ordered hashes in serialization and deserialization in all our modules like json, messagepack, corbs, rompf, arson, etc 19:06
Geth URI: b1cf9b72b2 | (David Warring)++ | lib/URI.rakumod
add !make-path() method

to avoid side affects in directory() method
19:10
rakudo/new-disp: 18 commits pushed by (Elizabeth Mattijsen)++
review: github.com/rakudo/rakudo/compare/4...2bea44920f
19:25