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.
timotimo d'oh, the scope registered with .symbol on a QAST::Block doesn't distinguish between lexical and lexicalref, for instance 00:24
vrurg timotimo: what's the difference? (/me is collecting knowledge) 00:32
timotimo having parameters declared as lexicalref is related to "is rw" 00:34
vrurg thanks! 00:35
timotimo changing the scope from blah to blahref or backwards in a QAST::Var means the difference between being able to operate on a container or its value, at the very least for native refs, i.e. "int $a is rw" 00:38
vrurg Wish there be more docs around for the internals than github.com/edumentab/rakudo-and-nq...als-course provides. 00:41
Anyway... The more I think about CoreClass:ver<6.x> notation the less I like it. It's not applicable to module versions but people would expected it to be. Perhaps having CORE.x:: namespaces would be better solution. Correspondingly, CORE.e::PseudoStash would provide the correct class. 00:45
timotimo uuughhhh the definitions stmts has so many different possible kinds of entry 00:53
just gotta build a fully recursive solution, aren't i 00:57
AlexDaniel shareable6: HEAD 04:50
shareable6 AlexDaniel, whateverable.6lang.org/HEAD
AlexDaniel kawaii: soo, two things to try 06:01
AlexDaniel kawaii: first, git pull in blin, I've added a thing that will print *some* output 06:03
kawaii: and then, I wonder if maybe you should delete the `data/` directory and let it reinitialize some stuff
kawaii: but let's see the error first 06:04
AlexDaniel` though that doesn't exactly explain anything 06:17
or at least, I have no idea how starting from scratch would make things better 06:19
AlexDaniel greppable6: \/.*\<\{ 06:30
greppable6 AlexDaniel, No! It wasn't me! It was the one-armed man! Backtrace: gist.github.com/22838f8653317c2274...97e2a60d0c
AlexDaniel greppable6: foo
greppable6 AlexDaniel, 10731 lines, 457 modules: gist.github.com/06592cc635e5b17b76...d13874ce36 06:32
AlexDaniel greppable6: \/.*\{ 06:35
greppable6 AlexDaniel, No! It wasn't me! It was the one-armed man! Backtrace: gist.github.com/23113929079284badf...5934333de9 06:38
AlexDaniel whatever :D
moritz Hi all 12:07
I got an email
Geth nqp: 50e249647b | (Kent Fredric)++ | README.pod
Fix documented minimum JDK to JDK9

Closes: github.com/perl6/nqp/issues/571
moritz from github: kentfredric has requested to join Perl 6’s nqp team.
ah, seems that somebody already acted on it :D
Geth rakudo/master: 4 commits pushed by (Vadim Belman)++ 13:27
kawaii AlexDaniel`: found the issue my version of libc6 was a little too old 14:43
not sure how though since I'd run Blin on this same box before
oh well, I pulled the latest from source and now Blin is happy 14:44
tbrowder hi all 15:05
recent pod activity has highlighted an issue i think needs some clarification: should pod config keys be allowed to have empty values? from reading S26 i don't find an explicit answer, but my gut says the user shoud get an exeption in all such cases. i'm going to ask damian via e-mail. 15:09
tbrowder i just sent email to damian, but i would like to hear other opinions from this group. 15:23
ugexe i would think the difference between empty string, null, and non-existence would be an important distinction. it often is in apis 15:29
tbrowder ugexe: concur. i was not being precise, but i would reject: empty strings, empty hashes or arrays. zero integers would be ok as values. 15:55
and i would reject empty strings as keys. 15:56
S26 calls the config entries pairs and notes they must be constants, so i don't think rejecting empty values should be unexpected. 15:59
ugexe why reject them? does allowing them cause confusion? 16:06
i can do e.g. `FOO="" perl6 -e 'say %*ENV<FOO>:exists'` for instance, and get what i expect even if someone might interpt that the same as `FOO=` 16:07
then again you are talking about *keys*, and my example is values 16:08
tbrowder config accepts keys as True or False, hence no need for an empty key. 16:09
tobs yeah, I think there is no way to write a colonpair with an empty key. This issue is about empty values to non-empty keys, as far as I'm aware. 16:09
tbrowder *key with empty value
AlexDani` for those who saw the syntax error puzzle I had today, I figured it out! 17:26
AlexDaniel m: sub ss() {}; given ‘abc’ -> $_ is copy { say s/b/d/ } 17:26
camelia ｢b｣
AlexDaniel m: sub s() {}; given ‘abc’ -> $_ is copy { say s/b/d/ }
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing required term after infix
at <tmp>:1
------> 3 given ‘abc’ -> $_ is copy { say s/b/d/ 7⏏5}
expecting any of:
prefix
term
AlexDaniel so having a sub named `s` is bad
especially so if it's exported from another file, then it's a bit more confusing xD 17:27
Kaiepi *poke* github.com/rakudo/rakudo/pull/3008 github.com/perl6/nqp/pull/558 github.com/MoarVM/MoarVM/pull/1126 17:47
AlexDaniel Kaiepi: hmm, may I ask what exactly this poke is trying to achieve? :) 17:51
Kaiepi looking for a review, it's been up for a while 17:52
i need this merged before i can start on the next stage of my grant work, implementing support for synchronous udp/unix/raw sockets 17:53
AlexDaniel Kaiepi: a review from? I looked through the changes and it makes sense 17:54
Kaiepi i was thinking jnthn maybe, it's the moarvm changes in particular that might need some tweaking if i chose the wrong type to represent socket families 17:56
AlexDaniel given that it touches moarvm stuff the list of people is kinda small
yeah
well, jnthn is in front of me… but I don't think a review is going to come any time soon :) 17:57
Kaiepi damn
AlexDaniel Kaiepi: so what exactly do you need for the grant? Having them merged? 17:58
I can merge
Kaiepi yeah
AlexDaniel what's up with lib/MAST/Ops.nqp file? Is it supposed to be changing that much? 17:59
Kaiepi ?
AlexDaniel Kaiepi: github.com/MoarVM/MoarVM/pull/1126...22cb9c3e68 18:00
ahhh 18:01
offsets
of course, right :)
AlexDaniel Kaiepi: let's merge then, we'll see if anybody complains :) 18:05
good amount of time before the release too
Kaiepi thanks
Geth nqp: da80b36e2b | (Ben Davies)++ | 2 files
Make socket family handling portable

MoarVM requires that the socket family be a separate parameter for nqp::bind_sk and nqp::connect_sk from the port; update the JVM accordingly and use the same platform-independent constants here as in MoarVM.
Related to github.com/rakudo/rakudo/issues/3007		 18:06
nqp: c90e0304c2 | (Ben Davies)++ | 3 files
Use nqp::const for socket family constants
nqp: a83caea486 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | 4 files
Merge pull request #558 from Kaiepi/portable-socket-families

  [IP6NS Grant] Make socket family handling portable		 18:07
rakudo/master: 4 commits pushed by (Ben Davies)++, (Aleks-Daniel Jakimenko-Aleksejev)++
nqp: 8fb861b167 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/templates/MOAR_REVISION
[MoarVM Bump] Brings 2 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...g352ae27e4 352ae27e4 Merge pull request #1126 from Kaiepi/portable-socket-families d189b29bc Make socket family handling portable		 18:08
rakudo: 90ffa349cc | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/templates/NQP_REVISION
[NQP Bump] Brings 15 commits

NQP bump brought: github.com/perl6/nqp/compare/2019....g8fb861b16 8fb861b16 [MoarVM Bump] Brings 2 commits a83caea48 Merge pull request #558 from Kaiepi/portable-socket-families 50e249647 Fix documented minimum JDK to JDK9 2add6edf3 bump moar revision for some stability fixes ... (11 more lines)
travis-ci Rakudo build failed. Aleks-Daniel Jakimenko-Aleksejev 'Merge pull request #3008 from Kaiepi/portable-socket-families 18:53
travis-ci.org/rakudo/rakudo/builds/568991778 github.com/rakudo/rakudo/compare/a...8372e04a68
Kaiepi it was just the builds that weren't using nqp on master that failed 18:55
and the jvm... but i don't think that null pointer exception happened at all when i tested it 18:57
vrurg Oh, 'use v6.e.PREVIEW; say CORE::C::PseudoStash.^ver' works as I'd expect it to work... ;) 19:11
Geth rakudo: vrurg++ created pull request #3112:
Implement CORE::<lang-rev> pseudo-packages		 19:21
kawaii AlexDaniel`: I have some Blin results for you friend 22:14
give me a moment to package them up 22:15
AlexDaniel kawaii: can't wait :) 22:16
I'll only manage to review them properly tomorrow, maybe
but I do want to see it right now :) 22:17
kawaii usercontent.irccloud-cdn.com/file/...-07.tar.gz 22:18
AlexDaniel: you even get the SVG you always wanted, aren't you lucky? ;)
AlexDaniel wow that looks very clean 22:19
Unicode::GCB is probably something about the unicode update 22:21
kawaii likely, but otherwise I think it's a good run :)
AlexDaniel some stuff “Could not locate compile-time value for symbol Block”
oh and the graph is actually useful 22:22
as in, maybe there's some issue in Distribution::Common_0.0.2
vrurg: nothing to see here I think 22:23
vrurg: you can check the results of course, but I was expecting a bunch of modules being broken 22:24
and it's just a few, much less than we normally see
oh yeah, unicode one bisected exactly to github.com/rakudo/rakudo/commit/f5...007736835c 22:25
vrurg Is the above archive link is the link to the log?
Ah, I see, – yes, it is! 22:26
Thank, I'll have a look.
AlexDaniel vrurg: ok, now that I checked the bisection results, there's actually some stuff for you :P 22:27
vrurg: github.com/rakudo/rakudo/issues/3113
ok, now I'll get some sleep, finally :) 22:28
tobs oops, my OEIS broke? 22:30
ah, the Maps and containers thing. I guess that makes sense. 22:32
kawaii AlexDaniel: we could do a Blin run once a week if you like 22:35
every Sunday?

Powered by Cro and the Raku® Programming Language.

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

Thank you!