#raku-dev on 21 March 2020 - Raku Programming Language Log

AlexDaniel

Kaiepi: hello

00:01

Kaiepi: maybe you want to join the Raku org on github… :)

00:02

I've sent the invitation again

00:03

Kaiepi

joined

thanks

00:11 hungrydonkey left

Geth

roast: d8a70faf14 | (Ben Davies)++ | S14-roles/mixin.t
Add test for typechecking of mixins of routines against Callable

00:14

MasterDuke

jnthn: lizmat++ found a nice fix for the shaped native array issue, but the new github.com/rakudo/rakudo/issues/3560 bisects to a big moarvm merge you did back in 2017

00:36

00:52 patrickz joined 00:56 patrickb left 01:05 patrickz left 01:20 sena_kun joined 01:22 Altai-man_ left 01:37 tinmarino left 03:19 Altai-man_ joined 03:21 sena_kun left 04:21 greppable6 left, notable6 left, squashable6 left, committable6 left, sourceable6 left, statisfiable6 left, shareable6 left, linkable6 left, tellable6 left, bloatable6 left, quotable6 left, benchable6 left, unicodable6 left, nativecallable6 left, coverable6 left, releasable6 left, evalable6 left, reportable6 left, bisectable6 left 04:22 evalable6 joined, reportable6 joined, linkable6 joined, coverable6 joined, squashable6 joined, benchable6 joined, tellable6 joined 04:23 greppable6 joined, bloatable6 joined, nativecallable6 joined, notable6 joined, sourceable6 joined, quotable6 joined, unicodable6 joined 04:24 statisfiable6 joined, bisectable6 joined, releasable6 joined, committable6 joined 04:25 shareable6 joined 05:20 sena_kun joined 05:22 Altai-man_ left 07:19 Altai-man_ joined 07:22 sena_kun left 09:20 sena_kun joined 09:21 Altai-man_ left 09:28 pmurias left

lizmat

Files=1305, Tests=111218, 205 wallclock secs (28.26 usr 8.37 sys + 2880.88 cusr 272.63 csys = 3190.14 CPU)

09:28

09:36 AlexDaniel left

Geth

rakudo: 0fba27d746 | (Elizabeth Mattijsen)++ | src/core.c/Rakudo/SEQUENCE.pm6
Streamlining Rakudo::SEQUENCE.iterator (1/N)

  - use binding where possible
  - use nqp::eqaddr instead of =:= or ===
  - use nqp::istype instead of ~~
  - check for Failure, instead of letting .self do the throwing
  - use topic inside worker subs

09:45

rakudo: d68fe5eb02 | (Elizabeth Mattijsen)++ | src/core.c/IterationBuffer.pm6
Add IterationBuffer.Slip

Turns out this will become very handy in streamlining Rakudo::SEQUENCE

09:59

lizmat

tyil: looks like the module announcer is awol again

10:11

tyil

ack

10:12

tellable6

2020-03-16T11:10:51Z #raku <chloekek> tyil: I heared you’re on Mastodon. Can I follow you?

2020-03-18T09:17:27Z #raku <patrickb> tyil Do you need help with the rakudo star docker image? jjatria is willing to help!

tyil

heh

.tell chloekek of course, I'm @tyil@soc.fglt.nl

10:13

tellable6

tyil, I'll pass your message to chloekek

tyil

.tell patrickb sure thing,

tellable6

tyil, I'll pass your message to patrickb

tyil

wops

.tell patrickb you can refer jjatria to [email@hidden.address] if he wants to contact me, or reach out on irc :>

10:14

tellable6

tyil, I'll pass your message to patrickb

tyil

I was wondering how I missed messages, since its not like me to miss highlights on irc, but I see the .tell is shared across the raku channels

10:18

v convenient tbh

10:48 AlexDaniel joined, AlexDaniel left, AlexDaniel joined 11:19 Altai-man_ joined 11:22 sena_kun left

MasterDuke

nqp: my @a := nqp::list_i(0,1,2,3,4,5,6,7,8,9); my $b; my int $c := 0; my num $s := nqp::time_n(); while $c++ < 100_000_000 { $b := $c if @a }; say(nqp::sub_n(nqp::time_n(), $s)); say($b)

11:24

camelia

2.0757856369018555
100000000

MasterDuke

nqp: my @a := nqp::list_i(0,1,2,3,4,5,6,7,8,9); my $b; my int $c := 0; my num $s := nqp::time_n(); while $c++ < 100_000_000 { $b := $c if nqp::elems(@a) }; say(nqp::sub_n(nqp::time_n(), $s)); say($b)

camelia

1.2959961891174316
100000000

MasterDuke

nqp: my @a := nqp::list_i(); my $b; my int $c := 0; my num $s := nqp::time_n(); while $c++ < 100_000_000 { $b := $c if @a }; say(nqp::sub_n(nqp::time_n(), $s)); say($b)

camelia

1.2075092792510986

MasterDuke

nqp: my @a := nqp::list_i(); my $b; my int $c := 0; my num $s := nqp::time_n(); while $c++ < 100_000_000 { $b := $c if nqp::elems(@a) }; say(nqp::sub_n(nqp::time_n(), $s)); say($b)

camelia

0.39176368713378906

lizmat

nqp: my @a := nqp::list_i(); my $b; my int $c := -1; my num $s := nqp::time_n(); while ++$c < 100_000_000 { $b := $c if nqp::elems(@a) }; say(nqp::sub_n(nqp::time_n(), $s)); say($b)

11:25

camelia

0.36582326889038086

lizmat

prefix increment faster still :)

11:26

MasterDuke

true, but i was really comparing `if @a` vs `if nqp::elems(@a)`

lizmat

oki :-)

11:35

looks like you've found another winner :-)

fwiw, spectest was down significantly, which may well have been caused by faster parsing

MasterDuke

these are really pico-optimizations though, less than 1s difference for 100m calls

11:37

lizmat

they're not so pico if I see results in make spectest

11:39

11:55 AlexDaniel left

Geth

rakudo: 1d57d17da0 | (Elizabeth Mattijsen)++ | src/Perl6/Metamodel/MethodContainer.nqp
Hide methods that are an implementation detail

  - add a :implementation-detail parameter to ^methods
  - if set, will *also* show methods that are implementation details
  - by default, will not produce such methods

13:04

13:06 travis-ci joined

travis-ci

Rakudo build errored. Elizabeth Mattijsen 'Streamlining Rakudo::SEQUENCE.iterator (1/N)

13:06

travis-ci.org/rakudo/rakudo/builds/665138020 github.com/rakudo/rakudo/compare/a...ba27d74685

13:06 travis-ci left

lizmat

hmmm.. 3 jobs failed to run the build script

13:07

MasterDuke

yeah, happened on a commit or two of mine. travis seems to be acting up more than usual

13:08

lizmat

maybe overload... so many home workers :-)

13:13 travis-ci joined

travis-ci

Rakudo build errored. Elizabeth Mattijsen 'Streamlining Rakudo::SEQUENCE.iterator (1/N)

13:13

travis-ci.org/rakudo/rakudo/builds/665138020 github.com/rakudo/rakudo/compare/a...ba27d74685

13:13 travis-ci left 13:20 sena_kun joined 13:22 Altai-man_ left 13:37 travis-ci joined

travis-ci

Rakudo build errored. Elizabeth Mattijsen 'Add IterationBuffer.Slip

13:37

travis-ci.org/rakudo/rakudo/builds/665139571 github.com/rakudo/rakudo/compare/0...8fe5eb02cb

13:37 travis-ci left

lizmat

same 3 jobs: An error occurred while generating the build script.

13:37

seems the MacOS backend is borked

13:39

13:54 travis-ci joined

travis-ci

Rakudo build errored. Elizabeth Mattijsen 'Hide methods that are an implementation detail

13:54

travis-ci.org/rakudo/rakudo/builds/665173679 github.com/rakudo/rakudo/compare/d...57d17da09c

13:54 travis-ci left

lizmat

again, all MacOS builds

13:55

being on MacOS myself, and seeing that a re-configure works fine, going to assume that Travis is borked

m: m: use v6.c; say "foo" # looks like this is borked also

14:13

camelia

5===SORRY!5=== Error while compiling <tmp>
Too late to switch language version. Must be used as the very first statement.
at <tmp>:1
------> 3m: use v6.c7⏏5; say "foo" # looks like this is borke

lizmat

m: use v6.c; say "foo" # looks like this is borked also

camelia

Don't know setting RESTRICTED
at gen/moar/ModuleLoader.nqp:257 (/home/camelia/rakudo-m-inst-2/share/perl6/lib/Perl6/ModuleLoader.moarvm:previous_setting_name)
from gen/moar/ModuleLoader.nqp:261 (/home/camelia/rakudo-m-inst-2/share/perl6/lib/…

lizmat

bisectable6: use v6.c # probably ok

14:26

bisectable6

lizmat, On both starting points (old=2015.12 new=1d57d17) the exit code is 0 and the output is identical as well

lizmat, Output on both points: «»

sena_kun

hmm, I just did all spec tests (master, 6.c, 6.d) and it was fine?

14:27

s/just/a couple of hours ago/

lizmat

sena_kun: it's related to the RESTRICTED setting

14:28

Geth

rakudo: 0c36feb07d | (Elizabeth Mattijsen)++ | src/core.c/PseudoStash.pm6
PseudoStash introspection omits implementation detail

- applies to .keys, .values, .kv, .pairs and .elems
- specify :implementation-detail to include *all* entries

14:36

rakudo: 8fe30b9c9f | (Elizabeth Mattijsen)++ | src/core.c/Rakudo/SEQUENCE.pm6
Streamlining Rakudo::SEQUENCE.iterator (2/N)

  - use &producer instead of $code
  - to improve readability
  - to improve pre-inlining performance
  - also remove some unneeded .defined tests

15:11

lizmat

afk for a few hours&

nine

releasable6: status

releasable6

nine, Next release in ≈7 days and ≈3 hours. 2 blockers. Changelog for this release was not started yet

nine, Details: gist.github.com/ca47dfe8b3da48e47c...d59810488d

sena_kun

Release is next week. So fat it looks pretty ok, though we as usual have a number of regressed modules. The previous ticket was not resolved, so I'll update the blin data soon (hmm, why not start it right now instead of waiting evening). Resolving regressed modules and giving moarvm is in shape for a release, it's a wrap time then.

15:15

15:16 grayrider joined

nine

Why is github.com/rakudo/rakudo/issues/3559 a blocker for rakudo when it's apparently an issue with rakudo star's build system?

15:17

sena_kun

s/fat/far/

15:19 Altai-man_ joined 15:20 travis-ci joined

travis-ci

Rakudo build errored. Elizabeth Mattijsen 'PseudoStash introspection omits implementation detail

15:20

travis-ci.org/rakudo/rakudo/builds/665200174 github.com/rakudo/rakudo/compare/1...36feb07d8a

15:20 travis-ci left

Altai-man_

re blocker: We should move the ticket to star repo, I think. But given our "once in a three month" release is broken for some system, I'd say we want to make a hotfix next month (this one) instead of leaving BSD with a break for another months.

15:21

Of course, if that's possible to do such a "hotfix" for star.

15:22 sena_kun left

Altai-man_

As in, if there are volunteers with enough time on hands.

15:22

15:22 grayrider left 15:23 grayrider joined

vrurg

Altai-man_: theoretically, it is possible to stick to make on *BSD. It was rather problematic at some point due to OpenBSD make erh... specific.