Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
Kaiepi greppable6, PROTO_TCP 00:16
greppable6 Kaiepi, 4 lines, 3 modules: gist.github.com/b6e482a7177ea9bc26...87e5cc3533
Kaiepi greppable6, PROTO_UDP 00:19
greppable6 Kaiepi, Found nothing!
Kaiepi greppable6, SOCK_PACKET 00:24
greppable6 Kaiepi, 2 lines, 2 modules: gist.github.com/c73cef66b9281379a7...6fb9068468
00:27 Xliff left
Kaiepi greppable6, ProtocolType 00:31
greppable6 Kaiepi, Found nothing!
00:31 lucasb left
vrurg greppable6: Perl6::BOOTSTRAP 00:50
greppable6 vrurg, Found nothing!
Geth rakudo: vrurg++ created pull request #3138:
Improve multi-revision support		 02:17
02:31 sortiz left 03:05 Xliff joined
Xliff \o 03:05
my @a = ( 03:31
a
*snarl*
m: my @a = ('a' ... 'q')».race.say 03:32
camelia No such method 'race' for invocant of type 'Str'. Did you mean any of these?
cache
rand
take

in block <unit> at <tmp> line 1
03:52 Xliff left 04:14 pamplemousse joined 06:10 pamplemousse left 06:21 Voldenet left 06:26 Voldenet joined, Voldenet left, Voldenet joined 07:22 Voldenet left 07:27 Voldenet joined, Voldenet left, Voldenet joined 07:52 cygx joined
cygx So, I can now build standalone NQP applications 07:53
Cuts hello world runtime by 1/3
I'll see if I can beat this thing into shape over the next week 07:54
(by standalone I mean single-file executable packaging all bytecode) 07:58
Geth rakudo: dumarchie++ created pull request #3139:
Throw if keep/break is invoked on a vow more than once		 07:59
[Tux] Rakudo version 2019.07.1-193-g71646ead9 - MoarVM version 2019.07.1-118-g3e715b424
csv-ip5xs0.700 - 0.758
csv-ip5xs-205.213 - 5.283
csv-parser22.554 - 22.644
csv-test-xs-200.430 - 0.437
test7.208 - 8.029
test-t1.662 - 1.740
test-t --race0.877 - 0.996
test-t-2029.701 - 31.287
test-t-20 --race9.168 - 9.182
 08:32
AlexDaniel cygx: amazing 08:34
cygx Well, there are still some moving pieces, but in contrast to my last attempt a couple of years ago, I have a better idea what needs to happen 08:52
Geth roast: 8b0db37e6c | (Elizabeth Mattijsen)++ | spectest.data
Oops, merged tests too soon.

Please revert when gb18030 encoding / decoding support has trickled through MoarVM / NQP / Rakudo		 09:09
09:09 Ven`` joined
Geth problem-solving/path-to-raku: 6af3f53042 | (Elizabeth Mattijsen)++ | solutions/language/PATH-TO-RAKU.md
Remove Wikipedia / Rosettacode reference
 09:15
09:28 Ven`` left 09:55 cygx left 10:48 lucasb joined 11:09 squashable6 left 11:14 squashable6 joined 12:51 evalable6 left 12:52 evalable6 joined 13:22 Ven`` joined 14:20 chloekek joined 14:42 chloekek left 14:56 Ven`` left 14:59 Ven`` joined 15:06 Ven`` left
Geth nqp: bdf722467f | usev6++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java
[JVM] Decontainerize in param unboxing

For positional parameters this has been done with d242880fe6.
Makes code like the following (from S12-attributes/native.t) work (again?):
   class C { has int $.foo = 42 }		 15:50
16:04 pmurias joined 16:45 pmurias left
bartolin .tell lizmat Do you maybe know/remember, why we don't use 'my int $int' here: github.com/rakudo/rakudo/blob/7164...u.pm6#L218 (same on line 405). I stumbled upon this accidentally while looking for a JVM bug. It looks like it was introduced with github.com/rakudo/rakudo/commit/b706b84312 -- and maybe it was an oversight? 17:00
tellable6 bartolin, I'll pass your message to lizmat
Geth roast: 0b586b99ad | usev6++ | S02-types/native.t
Adjust wrong test for RT #124084

It's correct that trying to assign a Rat to a native variable of type num dies with X::Syntax::Number::LiteralType. We have tests for that behaviour in S02-types/type.t.		 17:29
synopsebot RT#124084 [resolved]: rt.perl.org/Ticket/Display.html?id=124084 Must initialize num/num32/num64 ?
18:57 pmurias joined 19:38 Xliff joined
Xliff \o 19:38
19:45 chloekek joined
Xliff is bored/. 19:47
Xliff also has fat fingers. ;P
19:47 pmurias left
Xliff
.oO( Seriously! They are getting as large as hotdogs. /o\ )
 19:47
19:48 pmurias joined
Xliff pmurias: o/ 19:48
Kaiepi so i'm making it so on moarvm, it tries to use all records received when resolving a hostname whenever it needs to, not just the first 20:01
the problem is this breaks a couple tests in roast that expect localhost to only ever resolve to one address
is this actually breaking or is it an issue with the test?
timotimo that's totally a configurable thing 20:02
so i'd say it's an issue with the test
Kaiepi aight
20:22 pamplemousse joined
pmurias Xliff: \o 20:25
Geth nqp: a70b135dc0 | (Paweł Murias)++ | src/vm/js/nqp-runtime/io.js
[js] Noop stderr/stdout instead of failing
 20:27
nqp: 5fb8113c24 | (Paweł Murias)++ | src/vm/js/nqp-runtime/nqp-str.js
[js] Make nqp::decont_s work on nqp str
pmurias bartolin: seems like it's accidently omitted 20:30
bartolin: adding it in and seeing if the spectest passes seems like a good way to find out
MasterDuke pmurias, bartolin: passes a spectest 20:45
20:52 Kaiepi left 20:55 Kaiepi joined 20:59 pmurias left
Geth rakudo: c0b6cafc22 | (Daniel Green)++ | src/core/Mu.pm6
Explicitly make a variable a native int again...

since it looks like the type was inadvertently dropped in b706b84312.
  bartolin++ for the spot.		 21:00
21:01 pmurias joined 21:18 lucasb left 21:24 pmurias left 21:34 pamplemousse left 21:43 pamplemousse joined
vrurg Oh, my, I finally got it! 'use v6.e.PREVIEW; grammar Foo { token TOP { \d+ } }; say Foo.parse("aaa").WHAT' -> (Failure) 21:56
AlexDaniel yay 21:59
22:06 travis-ci joined
travis-ci Rakudo build failed. Daniel Green 'Explicitly make a variable a native int again... 22:06
travis-ci.org/rakudo/rakudo/builds/576292820 github.com/rakudo/rakudo/compare/7...b6cafc228c
22:06 travis-ci left
vrurg AlexDaniel: depends on R#3138 and perl6/problem-solving#94. So, depends on jnthn approval. 22:07
synopsebot R#3138 [open]: github.com/rakudo/rakudo/pull/3138 [6.e][CORE.setting][WIP] Improve multi-revision support
22:11 pamplemousse left
vrurg .tell jnthn I know you're overloaded, but I'd like to get your approval for problem-solving#94 and its implementation R#3138. It changes directory tree and therefore pretty prone for possible merge conflicts if not pulled in soon. 22:17
synopsebot R#3138 [open]: github.com/rakudo/rakudo/pull/3138 [6.e][CORE.setting][WIP] Improve multi-revision support
tellable6 vrurg, I'll pass your message to jnthn
22:23 chloekek left 22:49 pamplemousse joined
Geth rakudo: vrurg self-assigned [WIP] Flapping test t/spec/S02-types/subset-6e.t github.com/rakudo/rakudo/issues/3141
tbrowder++ created pull request #3142: add better err msgs for invalid open brackets for inline and leading …
 22:55
jnthn vrurg: I'm probably more sick (nothing serious) than overloaded; hopefully taking it easy today will have helped. I'll try to take a look tomorrow; I had some reservations based on what was described at the issue, but seeing what's implemented might reduce them to nothing. :) 23:10
vrurg jnthn: thanks! And get well soon. 23:11
jnthn: Just a thing to mention: I've got 6.e version of Grammar working well with the PR. 23:12
jnthn Cool :) 23:13
I think my general comment would be that bootstrap/metamodel I prefer not to extend too much when we can do things at the Perl 6 level.
But pragmatically, it can be easier to do some things in NQP 23:14
vrurg jnthn: Nah, I don't like metamodel being version-dependent. But bootstrap is different thing. To my view it even belongs more to core than to metamodel. I was tempring moving it into core.<rev> :) 23:15
*tempting 23:16
Anyway, you should relax today. That's what helps with cold the most, to my experience.
jnthn Your feeling on bootstrap is pretty much spot on 23:17
It's really doing the initial wiring together of stuff so we can write the rest in Perl 6
I've no objection to moving it to communicate that better 23:18
I'm curious (but will wait for tomorrow to find out) how versioning it helps over doing All The Things in the CORE.<not c>.setting helps (since ideally we'd be bootstrapped enough in the nested settings to not need the bootstrap stuff) 23:19
vrurg jnthn: First thing, Grammar needs HLL::Compiler and it's an NQP thing which it can't get with `use NQPHLL:from<NQP>` because it breaks serialization. Attaching own BOOTSTRAP fixes this issue. 23:21
jnthn: Second thing, new revisions might introduce new things which would need bootstrapping. 23:22
jnthn Well, there's indeed a long tradition of avoiding fixing serialization. :) 23:23
vrurg Otherwise, 6.d bootstrap/exporthow are empty and are there only to support building. 6.e does 'use Perl6::Metmodel; use QRegex` and nothing else.
jnthn OK, that sounds happily minimal
vrurg growl on serialization...
jnthn I think it's everyone's least favorite thing to have to debug/work on... 23:24
vrurg A lot of things could be skipped, run-away closures are hurting a lot of people. Would I know moar inside-out, I'd get this fixed asap.
*but run-away... 23:25
jnthn Urgh, yes. 23:37
Well, going to sleep. 'night o/
vrurg jnthn: g'night! 23:38
23:47 pamplemousse left 23:53 Xliff left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!