»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:02 araujo left, araujo joined, araujo left, araujo joined 00:05 firstdayonthejob joined 00:07 laz78 left 00:10 firstdayonthejob left 00:12 ZoffixWin joined
ZoffixWin Any idea how to link to the Perl 6 RT queue without getting that stupid "possible cross-site forgery" error? Like this one: rt.perl.org/Public/Search/Results....lled%27%29 00:13
(preferably sorted by most recent first)
Guess not. Even if I copy the search URL and open it in a new tab in the same browser, I get the forgery error 00:14
00:18 Sgeo joined, vinnix left 00:21 go|dfish left 00:25 colomon left 00:27 arlenik joined
dalek c: fbdcc94 | util++ | doc/Language/5to6-perlop.pod:
Fixed missing `C` in `C<->`
 00:28
c: 7196c85 | util++ | / (9 files):
Fix typos
c: dcf8944 | util++ | doc/Language/nativecall.pod:
Fix typos
00:28 alwpex left 00:29 ELBeavers joined 00:34 ELBeavers left
Juerd I find myself wanting 'unit react;' :) 00:36
ZoffixWin Well volunteered! :D 00:38
00:39 BenGoldberg joined 00:41 pierre-vigier joined
ZoffixWin New post: blogs.perl.org/users/zoffix_znet/20...erl-6.html 00:43
00:49 go|dfish joined 00:50 spider-mario left, cdg left
Util ZoffixWin++ # Excellent blog post; it will be a good reference to explain this aspect of Perl 6's coolness. 00:53
ZoffixWin \o/
Juerd ZoffixWin: Maybe change that last link to the actual RT queue 00:58
The home page of rt.perl.org is almost a dead-end. How to navigate to the list of bugs is very non-obvious. 00:59
ZoffixWin Juerd, can't. I get the "This looks like an attempt at cross-site forgery" error -_-
00:59 yeahnoob joined
Juerd Aargh 00:59
Juerd really doesn't like that site. 01:00
ZoffixWin It's terrible indeed.
01:10 espadrine_ left
MadcapJake seems like you can't really position command line arguments with MAIN 01:11
Juerd What do you mean by 'position'?
MadcapJake i'm trying to do «p6dx complete -d <text>» and it won't let me do it that way 01:13
Juerd Subcommands? 01:14
MadcapJake 'multi sub MAIN ('complete', Bool :$m!, $filepath)' says "Cannot put required parameter $filepath after variadic parameters"
01:15 kanishka joined
Juerd $m is -d? 01:15
01:15 skids joined
MadcapJake yeah Juerd, i wrote that wrong, it should be `-m` 01:15
01:15 laz78 joined
MadcapJake but i want to allow more flags after `complete` and before `$filepath` 01:16
Juerd If you write :$filepath and trigger a "Usage:" thing, you'll see why
MadcapJake but i don't want to do `--filepath="<filepath>"`
Juerd I didn't say you should want that :)
MadcapJake it puts complete last> 01:17
o_O
Juerd By the way, why is 'm' required?
MadcapJake my plan was to have a separate multi for each flag
Juerd $ perl6 -e 'multi sub MAIN ("complete", $filepath, Bool :$m!) { say "m=$m, fp=$filepath" }' -m complete foo 01:18
m=True, fp=foo
I don't know why 'complete' can't go before -m though
MadcapJake yeah 01:19
i just realized i can do `("complete -m", $filepath)` :D
kind of cheating but it works 01:20
Juerd That'll separate 'complete -m'?
AlexDaniel ZoffixWin: great post, very motivational
ZoffixWin \o/
Juerd MadcapJake: You'd need to pass complete\ -m or "complete -m" in a shell.
MadcapJake i just want the order preserved, complete is like a subcommand with a series of flags that follow it and then a path or a string
Juerd Subcommands are a weak spot in MAIN signatures 01:21
MadcapJake gah
01:22 llfourn joined
Juerd Maybe Bool :$complete! 01:23
That would require --complete
MadcapJake i think all positional args have to go first so i would just need to make them all named or only have named at the end 01:24
Juerd Or fix the feature :P 01:25
Is there any usecase for literals in MAIN's signature, other than subcommands?
Oh, the error message has nothing to do with MAIN 01:26
I don't understand what's variadic about "complete" or :$m!, though
MadcapJake yeah it's just the rule for all signatures 01:27
Juerd Apparently, :$m! is considered variadic. But why?
MadcapJake Juerd: I definitely agree, they should just be treated as positional
i think because it's named
Juerd Maybe I don't understand the meaning of 'variadic'
MadcapJake the 'complete' one i don't understand
variadic (i think) just means it's either optional or named 01:28
anything with a default value, `?`, or `:`
Juerd I don't understand why named required is variadic.
01:29 addison left
MadcapJake agreed it's probably either an oversight or something's blocking it from being considered as such 01:29
01:29 cpage_ joined
Juerd Could you just put the $m at the end 01:30
MadcapJake because it is named so it can be slurped up by a slurpy named arg
Juerd ?
MadcapJake so even though it's required, it's really not positional so it has to be considered variadic
i don't really know, i'm just guessing xD
Juerd I understand that it's not positional, but I don't understand why that matters for positionals. 01:31
MadcapJake well i do know that most languages have the same requirement that positional then variadic for arguments, so i'm wondering if it's just a challenging PL problem
Juerd sub MAIN ("complete", $fp, Bool :$m?) { say "m=$m fp=$fp" } 01:32
$ perl6 complete.p6 -m complete a
m=True fp=a
1;0 juerd@cxie:~$ perl6 complete.p6 complete -m a
Usage:
complete.p6 [-m] complete <fp>
:(
Put it at the end, and it requires it at the beginning :P
MadcapJake lol
i think what i'll go with is «script.pl6 --complete=<Str> [--meta=<IO::Path>]» 01:34
tony-o_ m: 'die'.die .die .die .die .die .die
camelia rakudo-moar 775271: OUTPUT«Method 'die' not found for invocant of class 'Str'␤ in block <unit> at /tmp/rxeeOk4pKE line 1␤␤»
tony-o_ i win
Juerd tony-o_: It did die.
ZoffixWin m: die .die .die .die .die .die .die 01:35
camelia rakudo-moar 775271: OUTPUT«Method 'die' not found for invocant of class 'Any'␤ in block <unit> at /tmp/nZq75iOEq7 line 1␤␤»
Juerd It dies perfectly. 01:36
tony-o_ m: 'die'.die; 01:38
camelia rakudo-moar 775271: OUTPUT«Method 'die' not found for invocant of class 'Str'␤ in block <unit> at /tmp/RB3UZxljoh line 1␤␤»
tony-o_ m: use MONKEY-TYPING; augment class Str { method die { die $.Str; } }; "die".die.die.die.die.die; 01:41
camelia rakudo-moar 775271: OUTPUT«die␤ in method die at /tmp/smzpRK4wC8 line 1␤ in block <unit> at /tmp/smzpRK4wC8 line 1␤␤»
01:45 addison joined
MadcapJake how do you see what exports a module has? 01:48
llfourn MadcapJake: does the module have unit module blah at the top?
MadcapJake yeah
01:48 Actualeyes joined
tony-o_ you can access that stuff through GLOBAL::<module>:: 01:48
llfourn m: use Test; say Test::EXPORT::DEFAULT::.keys; 01:49
camelia rakudo-moar 775271: OUTPUT«(&isnt &plan &pass &cmp-ok &flunk &does-ok &subtest &unlike &like &use-ok &todo &skip-rest &eval-dies-ok &is-deeply &throws-like &ok &is &diag &done-testing &is-approx &skip &dies-ok &lives-ok &eval-lives-ok &MONKEY-SEE-NO-EVAL &nok &is_approx &isa-ok &can…»
tony-o_ guess you don't need GLOBAL:: anymore
llfourn I didn't know you ever did 01:51
tony-o_ i don't think that works on non-core modules
llfourn it should. As long as they have unit module blah
when you put 'is export' it adds it to EXPORT DEFAULT of each package it's in including UNIT:: 01:52
so the only caveat is that if 'is export' is not inside a package you won't see it like that
tony-o_ does that not work on the module A { } syntax?
llfourn no it should work with that as well 01:53
I made github.com/LLFourn/p6-CompUnit-Uti...ure-import for other situations
MadcapJake that keys method is showing my exported routine, but i'm getting "Could not find symbol &blah" 01:54
llfourn m: use Test; say Test::EXPORT::DEFAULT::<&ok>;
camelia rakudo-moar 775271: OUTPUT«sub ok (;; Mu | is raw) { #`(Sub|83321952) ... }␤»
llfourn that should work
tony-o_ llfourn++ ; having trouble with a local module but that might just be me having worked 14 hours on totally unrelated crap today 01:55
01:55 _Vi left
llfourn tony-o_: you were doing something similar? 01:56
MadcapJake i figured out my problem, i was calling it as if i had used `our` declaratore
declarator*
01:56 _Vi joined
MadcapJake llfourn++ thanks for the help! 01:56
llfourn no worries I learnt that trick from someone on here iirc
tony-o_ llfourn: it's a problem with what i was using to test 01:57
how is koalatee determined on modules.perl6.org? 01:59
MadcapJake a band of koala's bash some keys when looking at your code, how many keys they bash divided by total number of possible key presses, is the result
tony-o_ i'm for it 02:00
MadcapJake xD
i've no idea, iirc it was discussed on here, check the logs: irclog.perlgeek.de/perl6 02:01
02:01 yeahnoob left
sortiz m: module Foo { sub foo() is export {}; }; say UNIT::EXPORT::DEFAULT::.keys # Without unit, the export uses the current one 02:02
camelia rakudo-moar 775271: OUTPUT«(&foo)␤»
02:08 captain-adequate joined
sortiz m: unit module Foo; sub foo() is export {}; say Foo::EXPORT === UNIT::EXPORT # unit makes the declared package the current UNIT 02:09
camelia rakudo-moar 775271: OUTPUT«True␤»
02:11 addison left
llfourn sortiz: interesting I didn't know that.. 02:12
m: module Foo { sub foo() is export {} }; say Foo::EXPORT === UNIT::EXPORT
02:12 captain-adequate left
camelia rakudo-moar 775271: OUTPUT«True␤» 02:12
llfourn hmm maybe we are using === wrongly 02:13
m: module Foo { sub foo() is export {} }; say Foo::EXPORT =:= UNIT::EXPORT
camelia rakudo-moar 775271: OUTPUT«False␤»
llfourn m: unit module Foo; sub foo() is export {}; say Foo::EXPORT =:= UNIT::EXPORT 02:14
camelia rakudo-moar 775271: OUTPUT«False␤»
sortiz llfourn, I was that someone.
llfourn m: unit module Foo; sub foo() is export {}; say UNIT::EXPORT =:= UNIT::EXPORT
camelia rakudo-moar 775271: OUTPUT«True␤»
Juerd MadcapJake: I'm nearing completion on a patch for your command line parsing thing. Would you care to test it?
[Coke] ZoffixWin++
llfourn though to me === should do the trick but I dunno. I don't think unit module Foo atm; actually makes Foo the same as UNIT (though there is an RT saying it should). 02:15
Juerd MadcapJake: It consists of two changes: 1. allow -named after positionals and only stop switch parsing at '--' (instead of at '--' or any positional), 2. assume in "Usage: " that string literals come before named switches.
ZoffixWin tony-o_, github.com/perl6/modules.perl6.org...tee.pm#L21
Juerd MadcapJake: I think that these changes combined make it better suited for subcommands. 02:16
llfourn sortiz: rt.perl.org/Public/Bug/Display.html?id=127305
02:17 ELBeavers joined
llfourn sortiz: the one who taught me about the Test::EXPORT trick? Yes! Now I remember :) 02:17
sortiz m: module Foo { BEGIN { say ::?PACKAGE} }; say ::?PACKAGE # module declares a new package, but alone don't change UNIT 02:20
camelia rakudo-moar 775271: OUTPUT«(Foo)␤(GLOBAL)␤»
llfourn sortiz: what do you mean with regards to UNIT? 02:21
sortiz UNIT is a virtual scope, related to the compunit, i.e, file compiled. 02:22
02:22 ELBeavers left
llfourn ah yes I thought you were saying somethig regarding unit module. I think the situation there is that unit module probably should make the module and UNIT the same but atm it doesn't. 02:23
sortiz And all "is export", operates at that level. In a file can be any numbers of packages (package, module, class, etc.) but only one UNIT.
02:24 n0tjack joined, Averna joined
llfourn yes. But is export also works on the package level but I think you know that because you were the one who showed me :P 02:24
sortiz Not the "same", a real package is a Stash, but UNIT is a pseudo one.
llfourn yep. We need to document and explain this eventually... probably soon. 02:25
sortiz So you can only have a "unit" declaration per file, that "aliases" the named package with the current UNIT. 02:28
02:29 n0tjack left
MadcapJake Juerd: sweet! I'd love too! 02:30
Gotta get my son ready for bed, i'll be back in about an hour 02:31
Juerd MadcapJake: By that time I'll have a patch ready for you, and I'll be asleep :) 02:32
MadcapJake okie dokie! xD
02:33 khw joined
Juerd MadcapJake: You'll still have to put the :$m! at the end, though :) 02:34
02:37 dfcarpenterak left
sortiz llfourn, Would make the attempt to document it, but my English isn't good 02:37
llfourn sortiz: ok well, maybe you review what I write :)
(but it seems good to me!) 02:38
sortiz m: unit module Foo; say Foo::.WHAT, UNIT::.WHAT 02:40
camelia rakudo-moar 775271: OUTPUT«(Stash)(PseudoStash)␤»
llfourn m: unit module Foo; BEGIN note UNIT::.WHAT: # it doesn't work at begin from what I remember 02:41
camelia rakudo-moar 775271: OUTPUT«(PseudoStash)␤»
llfourn or not
m: unit module Foo; BEGIN note UNIT::.keys: # it doesn't work at begin from what I remember
camelia rakudo-moar 775271: OUTPUT«(SIGSEGV WhateverCode Slip &callframe Pair SIGHUP &flat PromiseStatus CurrentThreadScheduler &infix:<∖> &RETURN-LIST utf8 StringyEnumeration &splice Distribution PF_INET6 Backtrace &GATHER &callsame &postcircumfix:<{ }> &print &infix:<lt> &sinh &GENERATE…»
llfourn cool.
sortiz m: say UNIT.WHO # Some more info 02:43
camelia rakudo-moar 775271: OUTPUT«PseudoStash.new(("!UNIT_MARKER" => !UNIT_MARKER,"\$!" => Nil,"\$/" => Nil,"\$=finish" => Mu,"\$=pod" => [],"\$?PACKAGE" => GLOBAL,"\$_" => Any,"\$¢" => Nil,"::?PACKAGE" => GLOBAL,:EXPORT(EXPORT),:GLOBALish(GLOBAL)))␤»
Juerd MadcapJake: github.com/Juerd/rakudo/commits/nom
llfourn :: is the same at .WHO afaik
Juerd MadcapJake: Have fun! :)
02:43 gtodd joined 02:45 n0tjack joined 02:47 kid51 left
gtodd is there anyway to represent a time period as an object? instead of Date.new([some value that means now]) .. Date.new([some value that means date of now + 1 year]) 02:49
02:49 n0tjack left
ZoffixWin Wasn't there a Duration type? 02:50
hu{TAB}... where's huggable :/ 02:51
geekosaur I thought so
ZoffixWin yeah
Juerd m: my $a = now; sleep 2; my $b = now; say ($b - $a).WHAT; 02:52
camelia rakudo-moar 775271: OUTPUT«(Duration)␤»
Juerd Yes, there are durations! :)
geekosaur Temporal mentions Instant but not Duration. I don't see where either is defined
ZoffixWin It is in the type docs, but I don't see any usage docs docs.perl6.org/type/Duration 02:53
02:53 huggable joined
geekosaur m: Duration.^methods.say 02:55
camelia rakudo-moar 775271: OUTPUT«(new Bridge Rat Num narrow sign sqrt sin tan cotan acosech abs conj atan2 cosec base pred asec acotan cosh acos acosec sech ceiling unpolar log10 atanh log exp acosh truncate sinh tanh acotanh round Int Real sec floor isNaN succ rand asin cos cis polymod c…»
geekosaur sigh
02:55 _Vi left
gtodd my $yearsworthofmoments = now + Duration.year() 02:55
:)
geekosaur m: Duration.^methods(:local).say 02:58
camelia rakudo-moar 775271: OUTPUT«(new Bridge Rat Num narrow sign sqrt sin tan cotan acosech abs conj atan2 cosec base pred asec acotan cosh acos acosec sech ceiling unpolar log10 atanh log exp acosh truncate sinh tanh acotanh round Int Real sec floor isNaN succ rand asin cos cis polymod c…»
sortiz llfourn, is the same iff the invocant is a scope ;) 02:59
02:59 _Vi joined 03:00 FreezerburnV joined, Juerd left
FreezerburnV o/ 03:00
llfourn sortiz: i know all too well :) github.com/LLFourn/p6-CompUnit-Util#descend-who
03:01 sno left 03:03 colomon joined, cpage_ left
sortiz llfourn++ # Nice util 03:04
FreezerburnV Is there a recommended license for Perl6 projects? (I'm personally a fan of MIT, but considering I'm wrapping a module without a license, figured it would be a good idea to ask) 03:05
llfourn Artistic v2 I think
ZoffixWin FreezerburnV, I use same as Perl: Artistic 2
FreezerburnV Not heard of that one, actually
03:06 Juerd joined
ZoffixWin FreezerburnV, github.com/zoffixznet/perl6-Color/...er/LICENSE 03:06
Juerd Meh, connection issues :(
I typed a long line to TimToady; did that get through before I disconnected?
ZoffixWin FreezerburnV, is the licenseless module you're wrapping in the ecosystem? Probably should open an Issue on it, suggesting to add a license. 03:07
Juerd, nope
03:07 yeahnoob joined
FreezerburnV ZoffixWin: Yep, SDL2::Raw. I'll go ahead and do that 03:07
Juerd Hurrah for input history then!
TimToady: S06 says "As usual, switches are assumed to be first, and everything after the first non-switch,", but in systems that support --long-names and --, that's not at all "usual". In fact, Perl 6 deviating from the convention of allowing switches after positionals might be problematic because it violates the 'least surprise' thing. Is it too late to change this? Changing it would also help a lot with subcommands.
ZoffixWin: Thanks
MadcapJake: github.com/Juerd/rakudo/commits/nom in case it got lost in my server's network disconnection 03:09
I'm off to bed 03:10
03:10 yqt left
Juerd Good night, all(@nicks) 03:11
ZoffixWin night
sortiz night
FreezerburnV Night!
AlexDaniel ZoffixWin: so what happened with huggable?
ZoffixWin: did he crash or something?
Juerd Good any(@times-of-day Z @nicks)
03:12 laz78 left
ZoffixWin AlexDaniel, no idea. She just dies after a couple of days. I come back and do screen -r and just get "there's no screen to return to". Haven't had the time to investigate that yet. 03:12
AlexDaniel ZoffixWin: dammit… 03:13
ZoffixWin: actually, I have exactly the same issue with Net::IRC::Bot
03:13 Zero_Dogg left
ZoffixWin Weird. 03:13
03:14 Zero_Dogg joined
AlexDaniel ZoffixWin: I was blaming Net::IRC::Bot but now I don't know what to say 03:14
ZoffixWin Do you use DBIish too ?
AlexDaniel not in that bot
ZoffixWin hm
geekosaur at a guess, bothbots don't handle netsplits or server disconnects well? 03:15
AlexDaniel I've also noticed that it tends to leave together with other people… And sometimes it can happen even after a few minutes, but other times it stays up for days
geekosaur: yeah, something like that 03:16
maybe
geekosaur *los* of bots seem to die or get stuck when the server goes away
ZoffixWin Yeah, that's probably it. There's no reconnect mechanism in huggable yet. 03:17
AlexDaniel also, sometimes it crashes but other times it just starts eating the memory…
ZoffixWin: if you figure it out, please leave me a note that it is fixed
ZoffixWin sure
AlexDaniel thanks 03:18
03:21 n0tjack joined
ZoffixWin Man, seems C people aren't fans of stuff we Perl folks put into Acme:: 03:22
Want to play with NativeCall, but having trouble finding some funny C lib to play with
03:26 n0tjack left 03:37 vendethiel joined 03:38 laz78 joined 03:45 noganex joined 03:47 Calibellus left 03:48 noganex_ left
MadcapJake Juerd: multi sub MAIN ('complete', Str $begin, Bool :$m!, Int :$depth) #-> complete -m [--depth=<Int>] <begin> 03:50
Juerd: perfect!
that's exactly the form i want, string at the end, subcommand at the beginning, flags in the middle 03:51
multi sub MAIN ('complete', Str $begin, Bool :$m) #-> complete [-m] <begin> 03:53
ZoffixWin Found a lib to learn NativeCall with (by making a bindings module): liballegro "portable library for cross-platform game and multimedia development" 03:57
I think this is something I played with a decade ago, when learning C :) Gonna finish what I started lol 03:58
04:00 vendethiel left
ZoffixWin or maybe not.. it has a lot of stuff that's not very interesting 04:01
FreezerburnV So I'm pretty darn sure I'm not doing any output during compile time to stdout or stderr, though I am doing "./SDL2.dll".IO.e either at compile or run time, depending on the state of the code. And I'm still getting the precomp-id error 04:10
And after that error, it will give me an error about something not existing in the .precomps directory 04:11
(if I try and run the file again)
I'm on whatever the latest thing rakudobrew installed is (I updated tonight)
04:12 Psyche^ joined 04:17 Psyche^_ left 04:21 arlenik left 04:25 pierre-vigier left 04:26 n0tjack joined 04:31 n0tjack left
ZoffixWin Can you pastebin the actual error and maybe the code too? 04:33
FreezerburnV Error: pastebin.com/B8L3VfA5 04:35
sdl2.pm6: github.com/Freezerburn/SDL2-p6/blo...r/sdl2.pm6 04:36
I have all the code on Github, so you can see all of it
Except the OpenGL module. Need to make that repo...
ZoffixWin: Ok now it's all on github :) 04:40
ZoffixWin FreezerburnV, have you tried commenting out this line? github.com/Freezerburn/SDL2-p6/blo...dl2.pm6#L6 04:43
Damn. You wrote all that stuff since I pointed out the snake demo?
FreezerburnV Still doesn't seem to like it :( 04:44
And no
I wrote all of thise in December
Most of it, at least
ZoffixWin Ah *phew*... I was about to feel unproductive :)
FreezerburnV Haha
I've been mostly busy today. Work + Bible study = not much time for Perl today
I'm just trying to get it to compile again
It was working earlier! I swear 04:45
There was a blue screen being redrawn and everything
ZoffixWin Bible study? God hates everyone. The end. What's there to study? :) 04:46
04:46 Herby_ joined
Herby_ Good evening, everyone! 04:46
yoleaux 21 Jan 2016 06:16Z <MadcapJake> Herby_: I think it's just sudo apt-get install libgumbo1
ZoffixWin Well, no idea why it breaks with that error. I'd start commenting out all of those debug() statements and anything else that can produce output, if that were my code.
FreezerburnV That's a rather simplistic view that seems to ignore the entirety of the New Testament :)
But I feel like if you wanted to argue religion, it would be best done over PM or something instead of spamming the perl6 IRC channel
ZoffixWin Or all of the 1000s of other holy books :P 04:47
Herby_ Still having issues with the Gumbo module 04:48
ZoffixWin :(
Herby_ I installed the Gumbo Parser library
FreezerburnV ZoffixWin: Well feel free to talk to me about that outside of #perl6, I'm always open for a discussion. I'll keep my talk in this channel to getting my scripts to compile for the sake of everyone else :)
Herby_ getting the error "Cannot locate native library 'libgumbo.so.1'" 04:49
not sure where to go from here
ZoffixWin Herby_, what OS? 04:50
Herby_ ubuntu 14.04
AlexDaniel libgumbo-dev ?
Herby_ trying to use the Perl 6 module Gumbo
AlexDaniel ah no
you should not really need the dev package…
ZoffixWin Hm. I don't have "gumbo" anything on my Bodhi 04:51
AlexDaniel libgumbo1 package should actually do it…
ZoffixWin FreezerburnV, is that an invitation to PM right now? :)
Herby_ github.com/Skarsnik/perl6-gumbo
FreezerburnV ZoffixWin: Feel free! I just don't want to turn a programming channel into a large argument about religion for everyone's ake
Herby_ "sudo apt-get install libgumbo1" is returning "unable to locate package libgumbo1" 04:52
FreezerburnV Because it's usually not just like 3-4 sentences on each side :P
AlexDaniel ah no, libgumbo1 package is only in debian testing and unstable.
Herby_: yeah right. :(
ZoffixWin There's no gumbo anything on my Bodhi, which is a fork of Ubuntu 14.04
Herby_, have you installed ANY libs?
Herby_ only thing I've installed so far on my fresh linux install is rakudobrew and Gumbo Parser 04:53
ZoffixWin I mean... for that gumbo thing
Herby_ github.com/google/gumbo-parser
just trying to play around with a little html parsing as a perl6 exercise 04:54
Gumbo looked to be my best (if only) option
but I'm open to other ideas
AlexDaniel ZoffixWin: what's your problem? It's a good book! :) www.youtube.com/watch?v=oC-fsFT7ZKs
MadcapJake Herby what ubuntu are you on? 04:56
Herby_ ubuntu 14.04
MadcapJake ahh libgumbo is only on 15.10 i think
Herby_ well thats a stinker
MadcapJake have you tried installing manually
Herby_ no clue how to do that
but i'm game if there are instructions somewhere
MadcapJake git clone github.com/google/gumbo-parser.git 04:57
cd gumbo-parser
rest of instructions right on git repo: github.com/google/gumbo-parser#installation
Herby_ yep, I've installed that
appeared to install succesfully
MadcapJake you did sudo make install 04:58
Herby_ yeppers
followed the instructions to the letter
well, through sudo make install 04:59
MadcapJake ok try `sudo updatedb && locate libgumbo.so.1`
ZoffixWin Herby_, when you figure it out, I would suggest you make a PR to the P6 Gumbo module so that instructions could be included in the docs :) 05:00
Herby_ MadcapJake, it found it in /home/herby/gumbo-parser/.libs/libgumbo.so.1 05:01
and /usr/local/lib/libgumbo.so.1
05:01 quester joined
Herby_ Zoffix, I'll definitely try! 05:01
MadcapJake ok use the /usr/local one
take that path and put it in a routine at top of your perl6 file `sub lib { '/usr/local/lib/libgumbo.so.1' }` 05:02
quester Good *, perl6!
MadcapJake wait, i'm not really sure how you're to use this, let me look at the perl6 module
Herby_ ok thanks. I'm trying to run the second example in the module docs. The one talking about "fancy" 05:03
05:03 jack_rabbit joined
MadcapJake ok Herby_ i'll give it a go, i'm building from source too, just to make sure it works the same 05:04
Herby_ sweet, thanks. Let me know if your "panda install Gumbo" works with successful testing 05:05
if you're doing that step
MadcapJake yep, it's installing right now 05:06
05:06 cpage_ joined
MadcapJake can't locate library here too 05:07
Herby_ k
05:07 maslan left
Herby_ any luck? 05:12
MadcapJake no :( i've got the same situation as you, it doesn't make sense that native wouldn't look in /usr/local/lib 05:13
05:15 yeahnoob left
Herby_ Well thanks for taking time to investigate it 05:16
any other modules you'd recommend for simple html parsing?
in perl 6
geekosaur ideally it'd check /etc/ld.so.conf on linux / *bsd, and appropriate other files on other OSes (use crle to dump /var/ld/ld.conf on Solaris, for example) 05:17
%PATH% on Windows, etc.
05:19 balazs left
geekosaur (actually to do this right on Solaris you'd want to be even more clever, but that would work for a start. ideally you'd look for a config for the program being run first and fall back to ld.conf) 05:20
MadcapJake you know what 05:21
i think this is the same thing that Linenoise went through. 05:22
05:23 lustlife` joined
Herby_ he ever find a solution? 05:24
05:25 AlexDaniel left
MadcapJake If you want to get a bit technical Herby_, git clone the Gumbo P6 module, change `constant LIB = ('gumbo', v1)` to `constant LIB = sub { '/usr/local/lib/libgumbo.so.1' }, then just `panda install <path-to-custom-gumbo-module>`. 05:26
05:26 lustlife left
MadcapJake Herby_: he did but it involves building the linenoise.so via a Build.pm and the c/h linenoise files 05:27
05:27 BenGoldberg left
Herby_ hm 05:27
hmmmmmm
let me try your tech tip
05:27 Gentoolman joined
MadcapJake i'll try to, in case i have the syntax off, i can help ya 05:28
s/to/too/
Gentoolman How to connect it to apache web server
geekosaur nativecall is not real swift right now. I wanted to poke at it but it was like a week before Christmas so I let it go. still need to go back and see if I can put together a coherent proposal for its proper behavior
although it does seem like some of what I said at the time got into it since we have api versions now 05:29
MadcapJake yeah that's neat! i didn't even know you could do that til today, i think the only thing I'd like to see is supplying custom paths to libs via the use statement or via a perl6/panda commandline flag 05:30
i've started a few wrappers but right now i just have LIB just return an absolute path to my library 05:31
s/just// xD 05:32
Herby_ u oh 05:33
MadcapJake tests pass for me Herby_
Herby_ ha hah! 05:34
you're the man :)
success!
and now the module example script works no problem 05:35
Thanks, MadcapJake 05:36
Now I just have to figure out how to use the module, beyond the example
05:36 yeahnoob joined
Herby_ I'm not too familiar with Git. Should we submit a PR or something with our findings? 05:37
MadcapJake I'm posting an issue, there needs to be a more universal solution
Herby_ great, thanks 05:38
MadcapJake no prob! Glad I could help!
05:40 xpen left 05:41 xpen joined 05:43 Cabanossi left 05:45 trustfundbby96 joined 05:47 Cabanossi joined 05:48 skids left 05:51 xpen left 05:54 ELBeavers joined 05:56 lustlife` left 05:59 ELBeavers left 06:02 Amnez777- joined, Herby_ left 06:03 rickbike joined, laz78 left 06:04 ChoHag_ is now known as ChoHag 06:05 jme` left 06:06 Amnez777 left 06:09 pierre-vigier joined 06:12 n0tjack joined 06:13 FreezerburnV left 06:16 n0tjack left 06:20 Gentoolman left, xpen joined 06:23 pierre-vigier left 06:26 pierre-vigier joined
rickbike Hello perl6ers 06:43
06:47 domidumont joined
rickbike Trying to use a number as a hash key 06:47
sub MAIN (*%h) { %h.keys.say; };
then test.pl --foo=bar --42=bat
returns (foo -42) 06:48
why -42?
06:50 mr-foobar left 06:52 domidumont left 06:53 mr-foobar joined, domidumont joined
gfldex rickbike: please file a bug report. It's either a bug in the options parser or a bug in the docs. 07:00
rickbike OK 07:01
07:02 quester left
rickbike any suggestion on a catchy subject/title of bug report? 07:02
gfldex "fail to handle numbers as option name for MAIN" 07:03
07:04 CIAvash joined, khw left
rickbike thanks gfldex 07:04
ugexe you cant have variable names that start with numbers 07:09
:42(bat) would fail just the same
gfldex he didn't ask for that
ugexe how would you access named variable 42 then? 07:10
gfldex either %h<--42> or %h<42>
07:12 geraud left 07:13 mr-foobar left 07:16 mr-foobar joined
dalek osystem: f54022b | azawawi++ | META.list:
ImageMagick's MagickWand API bindings are here :)

See github.com/azawawi/perl6-magickwand		 07:18
07:22 RabidGravy joined, Zero_Dogg left, Zero_Dogg joined 07:24 _mg_ joined 07:26 vendethiel joined 07:28 nakiro joined 07:30 vendethiel left
sortiz Someone remember where is ::?CLASS documented? 07:30
07:30 mr-foobar left
llfourn sortiz: in design documents? 07:31
07:32 mr-foobar joined
llfourn sortiz: design.perl6.org/S14.html 07:32
07:32 domidumont1 joined
sortiz llfourn, thanks! 07:33
07:34 pierre-vigier left 07:35 domidumont left 07:36 trustfundbby96 left, mr-foobar left 07:38 sjoshi joined 07:40 mr-foobar joined 07:43 firstdayonthejob joined 07:45 FROGGS joined
[Tux] csv-ip5xs 17.704 07:46
test 22.555
test-t 12.409
csv-parser 49.714
RabidGravy is it actually intentional that a regex/token/rule defined in a package can only be used if that package is a grammar? 07:47
i.e. 07:48
m: class Foo { our token valid-db-name { ^<[a .. z]><[a .. z0 .. 9_$()+/-]>*$ } }; say so "ab1-foo" ~~ /<Foo::valid-db-name>/; 07:49
camelia rakudo-moar 775271: OUTPUT«Method '!cursor_init' not found for invocant of class 'Foo'␤ in block <unit> at /tmp/Qf5wMBUigj line 1␤␤»
FROGGS RabidGravy: the problem is the invocant 07:53
m: class Foo { our token valid-db-name (Mu:) { ^<[a .. z]><[a .. z0 .. 9_$()+/-]>*$ } }; say so "ab1-foo" ~~ /<Foo::valid-db-name>/;
camelia rakudo-moar 775271: OUTPUT«Method '!cursor_init' not found for invocant of class 'Foo'␤ in block <unit> at /tmp/IPsUYRxh1Q line 1␤␤»
FROGGS errr
m: class Foo { our token valid-db-name (Mu:) { ^<[a .. z]><[a .. z0 .. 9_$()+/-]>*$ } }; say so "ab1-foo" ~~ /<&Foo::valid-db-name>/; 07:54
camelia rakudo-moar 775271: OUTPUT«True␤»
dalek c: 21754ba | (Wenzel P. P. Peppmeyer)++ | doc/Language/variables.pod:
name ::?CLASS
FROGGS m: class Foo { our token valid-db-name { ^<[a .. z]><[a .. z0 .. 9_$()+/-]>*$ } }; say so "ab1-foo" ~~ /<&Foo::valid-db-name>/;
camelia rakudo-moar 775271: OUTPUT«True␤»
FROGGS so, it is not just about the invocant :o)
RabidGravy that clears something up though 07:55
just seemed overkill to define a grammar for basically a single token 07:56
you may have realised that I try to avoid using regular expressions if at all possible :) 07:57
08:04 trustfundbby96 joined
FROGGS hmmmm, I like regexes... 08:04
masak morning, #perl6 08:06
moritz \o masak, *
FROGGS morning m&m 08:07
moritz feels like an endangered food right now
FROGGS is too tired to be hungry 08:08
masak is too tired to be tired 08:09
FROGGS is soo tired he cannot even say "tired" 08:10
masak .u tired
yoleaux U+1F62B TIRED FACE [So] (😫)
masak 😫
08:12 darutoko joined 08:16 pierre-vigier joined
llfourn that looks more like SOMEONE BROKE LEGOS face than tired 08:16
08:23 firstdayonthejob left
TEttinger .u cat 08:28
yoleaux U+0ECB LAO TONE MAI CATAWA [Mn] (◌໋)
U+A2B6 YI SYLLABLE CAT [Lo] (ꊶ)
U+101EC PHAISTOS DISC SIGN CAT [So] (𐇬)
TEttinger no cat face laughing?
08:29 cc9989 joined, rickbike left 08:30 abaugher left, leont joined 08:31 abaugher joined 08:32 zakharyas joined
cognominal .u butterfly 08:34
yoleaux No characters found
cognominal .u bug
yoleaux U+1A00 BUGINESE LETTER KA [Lo] (ᨀ)
U+1A01 BUGINESE LETTER GA [Lo] (ᨁ)
U+1A02 BUGINESE LETTER NGA [Lo] (ᨂ)
cognominal .u ant 08:35
yoleaux U+1CE9 VEDIC SIGN ANUSVARA ANTARGOMUKHA [Lo] (ᳩ)
U+20D4 COMBINING ANTICLOCKWISE ARROW ABOVE [Mn] (◌⃔)
U+20DA COMBINING ANTICLOCKWISE RING OVERLAY [Mn] (◌⃚)
08:37 darutoko left, ely-se joined, darutoko joined, trustfundbby96 left 08:39 araujo_ joined
ely-se Hello, world! 08:40
jdv79 good morning
08:40 araujo left 08:41 abraxxa joined, araujo_ left, araujo_ joined 08:43 araujo_ left, araujo_ joined 08:44 mr-foobar left, leont left 08:45 mr-foobar joined 08:48 rindolf joined
RabidGravy gosh, I started writing this module in April, better get finishing it then 08:55
08:55 mr-foobar left 08:56 sno joined 08:58 mr-foobar joined 09:00 rindolf left
sortiz m: use Test; plan 2; sub falla() { fail "nope" }; my $f = falla(); ok not $f.defined, "Handling"; ok $f.handled # Known bug? 09:02
camelia rakudo-moar 775271: OUTPUT«1..2␤ok 1 - Handling␤nope␤ in sub falla at /tmp/DQ2VSG5Sls line 1␤ in block <unit> at /tmp/DQ2VSG5Sls line 1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp line 3041␤ in block <unit> at /tmp/DQ2VSG5Sls line 1␤␤# Looks lik…»
RabidGravy sorry not quite sure what the point is 09:04
it will throw the failure when you assign it to the variable
sortiz Its documented that testing for .defined in a failure, make its "handled". 09:05
RabidGravy sure 09:06
sortiz So $f.hanldled should returns True, not throw it 09:07
RabidGravy m: sub f() { fail "boom" }; say "before"; my $f = f(); say "after"; 09:08
camelia rakudo-moar 775271: OUTPUT«before␤after␤»
RabidGravy okay
ely-se need a tool that warns on unused variables 09:09
this is too error-prone
09:09 CIAvash left, rindolf joined
sortiz A failure must throw when sinked or used as an invocant if not handled. 09:09
m: sub f() { fail "boom" }; say "before"; f(); say "after"; 09:10
camelia rakudo-moar 775271: OUTPUT«before␤boom␤ in sub f at /tmp/Yo9UGSpWvj line 1␤ in block <unit> at /tmp/Yo9UGSpWvj line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/Yo9UGSpWvj line 1␤␤»
09:10 leont joined, domidumont1 left 09:20 maxb_ joined 09:21 maxb_ is now known as Guest30572
Guest30572 quick question regarding slurpy behaviour - I have this sub definition: 09:21
sub slurpyPerl5 (*@_) { @_.join('/').say } 09:22
09:22 virtualsue joined
Guest30572 ...which I call like that: 09:22
slurpyPerl5([1 , [1.1,1.2],2,3] , 10 , 20)
and I get this: 1/1.1 1.2/2/3/10/20
09:23 yeahnoob left
Guest30572 I always figured the flattening happens hierarchical so i should be: 1/1.1/1.2/2/3/10/20 09:23
sortiz Just checking the code, not .handled method declared so docu bug or NYI. 09:24
Guest30572 did I get anything wrong,- the documantation says: one asterisk will flatten arguments by dissolving one or more layers of bare Iterables
RabidGravy sortiz, which documentation as a matter of interest? 09:25
masak Guest30572: yes, I think you have a wrong assumption somewhere. it doesn't flatten all the way like that
sortiz doc.perl6.org/type/Failure#method_handled
leont Guest30572: try (**@) 09:26
(**@_)
masak um
I'm not sure that's implemented
leont Ow
masak and I suspect even that one respects the Array boundary in one of the arguments
09:27 mr-foobar left
Guest30572 already tried **@_ - it does even less flattening: 1 1.1 1.2 2 3/10/20 09:27
masak m: say [1, 2, (3, 4), 5].flat
camelia rakudo-moar 775271: OUTPUT«(1 2 (3 4) 5)␤»
masak say [1, 2, [3, 4], 5].flat
m: say [1, 2, [3, 4], 5].flat
camelia rakudo-moar 775271: OUTPUT«(1 2 [3 4] 5)␤»
masak m: say (1, 2, (3, 4), 5).flat 09:28
camelia rakudo-moar 775271: OUTPUT«(1 2 3 4 5)␤»
masak m: say (1, 2, [3, 4], 5).flat
camelia rakudo-moar 775271: OUTPUT«(1 2 3 4 5)␤»
masak m: say (1, 2, [3, 4, [5, 6]], 7).flat
camelia rakudo-moar 775271: OUTPUT«(1 2 3 4 [5 6] 7)␤»
jdv79 is there a new chan or are a lot of people just taking a break lately?
masak jdv79: do we talk too little? :P
moritz jdv79: there's #perl6-toolchain and #perl6-release
masak what's #perl6-release for? 09:29
09:29 pierre-vigier left, pierre-vigier joined
jdv79 moritz: thanks 09:29
Guest30572 say [1, 2, [3, 4], 5].flat gives me (1 2 [3 4] 5) - not hierarchical either :-( 09:30
jdv79 i just skimmed my backlog and i didn't see larry or liz or jonathan. i'm guessing a break then.
09:30 araujo_ left, araujo joined, araujo left, araujo joined, ELBeavers joined
ely-se say [1, 2, |[3, 4], 5].flat 09:30
m: say [1, 2, |[3, 4], 5].flat 09:31
camelia rakudo-moar 775271: OUTPUT«(1 2 3 4 5)␤»
Guest30572 quit 09:32
09:32 Guest30572 left
sortiz Array.flat don't flat common Arrays, see: github.com/rakudo/rakudo/blob/nom/...ay.pm#L366 09:33
09:33 pierre-vigier left
RabidGravy I'm not quite sure why you'd want to flatten to an arbitrary depth 09:33
09:34 pierre-vigier joined, lnrdo joined 09:35 ELBeavers left 09:38 n0tjack joined 09:41 mr-foobar joined 09:43 n0tjack left 09:44 brrt joined
sortiz RabidGravy, for arbitrary depth, I want an adverb :deep in flat :) 09:44
masak a named parameter 09:46
sortiz Yes 09:47
With default 1
masak I'm pretty sure deep flattening by default must have been discussed during the GLR, and rejected 09:48
09:49 cpage_ left
ely-se deep flattening by default is a horrible idea 09:50
there should be defaults only if one behaviour is *far* more common than all the others, and the default should then be that behaviour, and in the case of flattening it's shallow flattening
sortiz Sure, but I been investigating about Array.flat for some time, see RT#127121. My current view is that the current implementation is a leftover from GLR, needed to keep roast passing, there are 9 tests that depend on it. 09:51
masak interesting.
my current stance is that I tend to stay away from most kinds of flattening in my code.
at most, I lean on the one-argument rule
the 007 has two explicit `flat` calls, both of them in .t files 09:52
s/007/007 codebase/
sortiz The real problem is that the people expect that "flat" flatten. 09:54
09:56 mr-foobar left, mr-foobar joined
sortiz During GLR a .List method must be added to Array to compensate the fact that .list broke, I tough that Array.flat should be Array.Seq and the code needed changed. 09:57
09:57 domidumont joined
jast personally I don't expect deep flattening 09:58
brrt i thought we had the slip thingy for that
sortiz On the other side Iterator.flat just DWIM
09:59 mr-foobar left, mr-fooba_ joined
ely-se the problem with DWIM is that it doesn't know what you mean 09:59
09:59 Amnez777- left, Amnez777- joined
sortiz "Flat" and "Flatten" are a well defined words. 10:01
gfldex the main problem is that operators return LoLs even if you don't want them to _and_ there is no in place list destructuring beside flat, so you have to use a points or sub to get the destructuring
s/points/pointy/ 10:02
sortiz BTW, I don't expect deep flattening, neither
gfldex i would expect deep flattening if i ask for it and prefer slippy versions of metas 10:03
10:03 espadrine_ joined
gfldex m: sub infix:<|Z>(List:D \l, List:D \r){ gather for l Z r -> [\re, \le] { take re; take le; } }; dd (1,2,3) |Z <a b c d>; 10:03
camelia rakudo-moar 775271: OUTPUT«(1, "a", 2, "b", 3, "c").Seq␤»
gfldex m: dd (1,2,3) Z <a b c d>; # <-- this can be hard to chain with other metas 10:04
camelia rakudo-moar 775271: OUTPUT«((1, "a"), (2, "b"), (3, "c")).Seq␤»
sortiz Yes, right. I seen your comments about it yesterday.
gfldex m: dd ((1,2,3) Z <a b c d>).join 10:05
camelia rakudo-moar 775271: OUTPUT«"1 a2 b3 c"␤»
gfldex and it's hard to chain with methods aswell
10:05 pfortuny joined
pfortuny uffff. I am running a multithreaded code and having issues with opening files for writing... 10:05
Failed to open file [...]: no such file or directory 10:06
ely-se Sounds like a race condition.
pfortuny BUT it is open "$filename", :w
and the filename is unique
avalenn_ is there any emacs mode for perl6 ? 10:07
10:07 brrt left
pfortuny but what I do not understand is why "no such file" 10:07
10:07 leont left
pfortuny avalenn_: yep, in one of the packages 10:07
10:07 eyck left
pfortuny avalenn_: github.com/hinrik/perl6-mode 10:08
10:08 avalenn_ is now known as avalenn
avalenn pfortuny: thank you 10:09
pfortuny avalenn: not at all.
sortiz The other problem that I see, is that people are accustomed to use "@foo" in signatures, temps, etc., and they expect the pre GLR behavior. 10:11
10:11 jaffa4 joined
jaffa4 hi all 10:11
I have a code the worked previously
Type check failed in binding @rec; expected 'Positional' but got 'Hash' the codei is : for @.tokens -> @rec ..... 10:12
10:15 pierre-vigier left
RabidGravy jaffa4, @.tokens is an array of Hashes? 10:17
sortiz Yes, to bind to anything with a @ sigil, you need something that implements that role, and for the % sigil an Associative.
jaffa4 yes
ely-se m: my %h := { a => 1 }; 10:19
camelia ( no output )
RabidGravy not sure how it would have worked before but changing that to %rec should do
ely-se m: my @h := { a => 1 };
camelia rakudo-moar 775271: OUTPUT«Type check failed in binding; expected Positional but got Hash␤ in block <unit> at /tmp/KgRxg8oNOc line 1␤␤»
10:19 virtualsue left
sortiz Or use a simple $ (Scalar), that works too. 10:20
10:23 cpage_ joined
sortiz Unless you need the "special" semantic of @ or %, a simple scalar is good container of any type. 10:24
10:25 Averna left 10:26 zakharyas left
pfortuny ely-se: surprising (?) if I protect the open() with a lock, the issue seems to go away 10:26
fascinating because the names are unique always
is there any explanation for this?
RabidGravy without seeing the code no
pfortuny yep, but the code is a mess... 10:27
I've tried to simplify it but the issue goes away
10:27 zakharyas joined
RabidGravy are you doing this in a .tap (or whenever block) on a supply? 10:28
pfortuny in a tap
RabidGravy if .tap then change it to .act
pfortuny oh!
nope. 10:29
RabidGravy ..tap will run them as fast as they come in .act will only run one at a time
pfortuny it also fails
konobi ooohhh... fun with backpressure
pfortuny which means?
RabidGravy thus demonstrating how hard it is to speculate on code that you haven't seen
pfortuny :D 10:30
:DDDDD
10:30 TEttinger left
pfortuny thanks anyway, the lock thing fixed it and it serves because it is only for the "open" 10:30
ely-se if the code is a mess then there's no hope fixing bugs in it. clean it up first
pfortuny no, it is a mess because of all the things I need to do
inside the tap
(and because the async process I am running IS a mess: maxima) 10:31
but I am tied up to it
RabidGravy consider using a whenever instead
pfortuny what would that do? (sorry novice here) 10:32
konobi pfortuny: reading faster than you can deal with stuff coming in
RabidGravy potentially distributing work via other supplies
pfortuny ??
RabidGravy react { whenever $supply { } }
pfortuny ah, I see. I'll try that.
konobi: mind explaining a bit?
RabidGravy in certain applications it might make the code neater at least 10:33
pfortuny thanks a lot, RabidGravy !
konobi pfortuny: mmm... tricky to explain... lemme see if I can find some docs 10:34
pfortuny what I am doing is just trying to open a file to save the output of a process in it.
each process has a file and I save the output there
but for whatever reason, sometimes, after some iterations, the open($fname, :w) fails 10:35
with a "No such file or directory" (mind this message!)
konobi ah, sorry... yeah, i was just recalling an issue that was had with this async behaviour before
pfortuny the funny thing is the file DOES NOT exist and I simply want to create it... 10:36
konobi RabidGravy: know if the supplies can apply backpressure?
pfortuny however, if I Lock.protect the open(), everything works.
RabidGravy there's a throttle
pfortuny, is the $fname or the resulting handle in a scope outside the tap? 10:37
konobi mmm... backpressure is a little different, a plain throttle wouldn't cover it
RabidGravy sure
pfortuny RabidGravy: the $fname is local to the tap, the resulting handle is necessarily (for me) defined outside 10:38
RabidGravy but throttle does have a mechanism for passing, "woaaaah back off for a bit messages" 10:39
pfortuny the handle (an array) is in the global scope of the file.
RabidGravy you almost certainly *do* need the lock then
pfortuny even if I am sure that the race will not happen? 10:40
sure as in "impossible"
(not "just pretty sure")
(not that I mind, certainly) 10:41
konobi RabidGravy: ah... supply gives a large file as fast as it can to the next thing, which encrypts, then that writes to disk
you'd end up blowing your stack without backpressure
10:42 domidumont left
RabidGravy anyway must awah to the supermarket 10:42
pfortuny thanks konobi & RabidGravy 10:43
10:43 domidumont joined 10:47 RabidGravy left, Shozan is now known as SHODAN, CurtisOvidPoe joined
CurtisOvidPoe Is there a module to pretty print data structures? 10:48
DrForr github.com/drforr/perl6-pp is a start.
Juerd Open for discussion and testing: github.com/rakudo/rakudo/pull/688
stmuk_ Data::Dump too
and probably one or two others 10:49
doc.perl6.org/language/faq#How_can_...ilar%29%3F 10:50
CurtisOvidPoe *resolve stage failed for Pretty::Printer: Project Pretty::Printer not found in the ecosystem)
stmuk_ its not in the ecosystem 10:52
DrForr Probably haven't added it.. or the usual broken META.info that I've been copying 'round.
DrForr looks.
yep, broken.
Fixed, pushed. Need to add .travis.yml as well. 10:54
jaffa4 this does not work push @tokens, [$(%starts_ends{$prevpos}),$prevpos, $_]; 10:55
r: print "hello" 10:56
camelia rakudo-jvm 6c0f93, rakudo-moar 775271: OUTPUT«hello»
10:56 brrt joined
jaffa4 r: my @d; push @d,[1,2,3]; push @d,[4.5.6]; say @d.perl; 10:56
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Malformed postfix call␤at /tmp/tmpfile:1␤------> 3my @d; push @d,[1,2,3]; push @d,[4.5.7⏏056]; say @d.perl;␤»
..rakudo-jvm 6c0f93: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Unsupported use of . to concatenate strings; in Perl 6 please use ~␤at /tmp/tmpfile:1␤------> 3my @d; push @d,[1,2,3]; push @d,[4.5.67⏏5]; say @d.perl;␤»
jaffa4 r: my @d; push @d,[1,2,3]; push @d,[4,5,6]; say @d.perl; 10:57
camelia rakudo-jvm 6c0f93, rakudo-moar 775271: OUTPUT«[[1, 2, 3], [4, 5, 6]]␤»
jaffa4 bug in the compiler 10:58
gfldex m: my @d; push @d,|(1,2,3); push @d,|(4,5,6); say @d.perl; 10:59
camelia rakudo-moar 775271: OUTPUT«[1, 2, 3, 4, 5, 6]␤»
gfldex jaffa4: nope
jaffa4 I mean in version This is perl6 version 2015.09 built on MoarVM version 2015.09
gfldex m: my @d; append @d,[1,2,3]; append @d,[4,5,6]; say @d.perl;
camelia rakudo-moar 775271: OUTPUT«[1, 2, 3, 4, 5, 6]␤»
gfldex or you use the right method 11:00
DrForr jaffa4: That's from what, October of last year? Maybe update if you can.
gfldex back then it was by design
jaffa4 there is no windows version
installer 11:01
DrForr Like I said, if you can. 11:02
stmuk_ in hindsight pre 6.c releases probably should have exited after Dec 25th :) 11:03
git-man-page-generator.lokaltog.net/ 11:06
11:12 perlawhirl joined, maslan joined
sortiz m: my @d; @d.push: 1,2,3,4; @d.push: 5,6,7,8; say @d; # Without unneeded ( nor [ 11:14
camelia rakudo-moar 775271: OUTPUT«[1 2 3 4 5 6 7 8]␤»
11:15 virtualsue joined 11:18 brrt left
perlawhirl hi perlers 11:19
11:19 ELBeavers joined
perlawhirl can i multi method with a role ?? What's the signature 11:19
m: role T { multi method t (Str) { "STRING" } }; my $i = 'foo' but T; say $i.t;
camelia rakudo-moar 775271: OUTPUT«Cannot call t(Str+{T}: ); none of these signatures match:␤ (Str+{T} $: Str, *%_)␤ in block <unit> at /tmp/16r7MKc6dH line 1␤␤»
gfldex m: role T { multi method t (Any:D \s:) { "STRING" } }; my $i = 'foo' but T; say $i.t; 11:23
camelia rakudo-moar 775271: OUTPUT«STRING␤»
CurtisOvidPoe Any critique of my Perl 6 style is welcome: gist.github.com/Ovid/43c131642b82f202d977
gfldex perlawhirl: ^^^
11:23 ELBeavers left
gfldex m: role T { multi method t (Str:D \s:) { "STRING" } }; my $i = 'foo' but T; say $i.t; 11:26
camelia rakudo-moar 775271: OUTPUT«STRING␤»
gfldex likely better
perlawhirl gfldex: thanks! playing around now 11:28
what does the \s: mean? 11:29
gfldex \s means binding and the : means "left of me is the invocant"
11:31 azawawi joined
azawawi hi 11:31
CurtisOvidPoe p6: sub foo {my @a = 1,2; my @b = 3,4; return @a, @b}; my (@c,@d) = foo(); say @c.perl 11:32
camelia rakudo-moar 775271: OUTPUT«[[1, 2], [3, 4]]␤»
CurtisOvidPoe How do I stop @c from slurping up both arrays?
azawawi github.com/azawawi/perl6-magickwan...-hello.pl6 # ImageMagick in Perl 6 :)
CurtisOvidPoe Or maybe I phrased that poorly. In my above example, I’d like @a to be assigned to @c and @b to be assigned to @d. What did I do wrong? 11:34
11:36 pfortuny left 11:37 cc9989 left
gfldex p6: sub foo {my @a = 1,2; my @b = 3,4; return $@a, $@b}; for foo() -> @c, @d { dd @c; dd @d } 11:38
camelia rakudo-moar 775271: OUTPUT«Array @a = [1, 2]␤Array @b = [3, 4]␤»
gfldex CurtisOvidPoe: i believe what you want is NYI
jnthn m: role T { multi method t() { "STRING" } }; my $i = 'foo' but T; say $i.t 11:39
camelia rakudo-moar 775271: OUTPUT«STRING␤»
jnthn Why the sig at all?
gfldex p6: sub foo {my @a = 1,2; my @b = 3,4; return $@a, $@b}; my ($c, $d) = foo(); dd $c, $d;
camelia rakudo-moar 775271: OUTPUT«Array $c = $[1, 2]␤Array $d = $[3, 4]␤»
CurtisOvidPoe Ah, thanks. That seems strange, but OK :)
jnthn sub foo {my @a = 1,2; my @b = 3,4; return @a, @b}; my (@c, @d) := foo(); say @c.perl; say @d.perl 11:40
m: sub foo {my @a = 1,2; my @b = 3,4; return @a, @b}; my (@c, @d) := foo(); say @c.perl; say @d.perl
camelia rakudo-moar 775271: OUTPUT«[1, 2]␤[3, 4]␤»
jnthn CurtisOvidPoe: Use binding to get "what a signature would do"
11:40 _mg_ left
CurtisOvidPoe jnthn: thank you :) 11:41
gfldex CurtisOvidPoe: in your example you could do: 11:42
m: sub n(){Nil}; while my $line = n() // False { say 'never ever!' }
camelia ( no output )
DrForr Oh, Olson, how do I hate thee. Let me count the ways: 0,1,1:15,2,3:30...
gfldex CurtisOvidPoe: why do you need the .defined? Can linenoise return something that is undefined but True? 11:45
11:46 CurtisOvidPoe left, CurtisOvidPoe joined
CurtisOvidPoe gfldex: I just used the code from the Linenoise synopsis. 11:46
11:47 cc9989 joined
sortiz jnthn, It's supposed that Failure.handled to be visible? According to doc.perl6.org/type/Failure#method_handled, yes but not in the current implementation. 11:49
gfldex CurtisOvidPoe: it will return Str (what is undef and False), see: github.com/hoelzro/p6-linenoise/bl....pm.in#L84 11:50
CurtisOvidPoe gfldex: thanks.
perlawhirl gfldex: why doesn't it work with arrays... 11:51
m: role T { multi method t (Any:D \s:) { "ARRAY" } }; my @i = ('foo') but T; say @i.t;
camelia rakudo-moar 775271: OUTPUT«Method 't' not found for invocant of class 'Array'␤ in block <unit> at /tmp/lHvjqNjdYF line 1␤␤»
perlawhirl m: role T { multi method t (Array:D \s:) { "ARRAY" } }; my @i = ('foo') but T; say @i.t;
camelia rakudo-moar 775271: OUTPUT«Method 't' not found for invocant of class 'Array'␤ in block <unit> at /tmp/RlWm9moWsq line 1␤␤»
gfldex you try to bind it to the wrong kind of container 11:52
@a <-- not Array but a container of type Array
m: role T { multi method t (Array:D $s:) { "ARRAY" } }; my @i = ('foo') but T; say @i.t
camelia rakudo-moar 775271: OUTPUT«Method 't' not found for invocant of class 'Array'␤ in block <unit> at /tmp/JFFWI2meqY line 1␤␤»
11:54 _mg_ joined
gfldex m: role T { multi method t () { "ARRAY" } }; my $i = Array.new('foo') but T; say $i.t 11:55
camelia rakudo-moar 775271: OUTPUT«ARRAY␤»
gfldex m: role T { multi method t () { "ARRAY" } }; my @a := Array.new('foo') but T; say @a.t
camelia rakudo-moar 775271: OUTPUT«ARRAY␤»
11:56 ely-se left
gfldex perlawhirl: i would be careful with mixin into array containers. To me that does not make sense. 11:56
or better it's unclear if you want to mixin into the Array or the values inside the array
11:57 pfortuny joined
perlawhirl gfldex: thanks. i'm not sure i would ever use it. just poking around in the some corners of the language 11:59
12:01 xpen left
CurtisOvidPoe Using gist.github.com/Ovid/43c131642b82f202d977, the first time I try to fetch data from a table with ~ 10K rows, it says “Cannot invoke this object (REPR: Null)(gen/moar/m-CORE.setting 3959 sink-all)”. The second time I try it (in the same session) it works just fine. Sound familiar to anyone? 12:01
12:03 ely-se joined 12:05 espadrine_ is now known as espadrine 12:11 cognominal left
moritz CurtisOvidPoe: vaguely familiar, but I don't know what's wrong 12:12
nor how to fix it :(
12:13 cognominal joined
jaffa4 when I use panda when I run './rakudobrew build-panda' it responds 'No version set' - 12:14
CurtisOvidPoe moritz: ah well. No worries. It will eventually sort itself out :)
jaffa4 could anyone tell me what is the solution?
ely-se jaffa4: "build panda", not "build-panda" 12:16
jaffa4 it makes no difference 12:17
12:17 kid51 joined
azawawi jaffa4: why are you doing a './rakudobrew' not 'rakudobrew' ? 12:18
moritz jaffa4: what verison of rakudo did you build?
jaffa4 I d not know
WHat should I?
moritz jaffa4: what did you type to build a rakudo, and when did you do it? 12:19
12:19 kaare_ joined
jaffa4 I get the same message 12:20
if I try to build rakudo
I have a version of rakudo which I installed outside of rakudobrew
12:21 perlawhirl left
moritz jaffa4: what do you type to build rakudo? 12:21
12:21 Skarsnik joined 12:23 FROGGS left
jaffa4 ok, it works for now 12:23
moritz jaffa4: what did you do to make it work? 12:24
jaffa4 not sure yet.. I am building rakudo through rakudobrew 12:25
that works
moritz you said earlier you get the same message when you build rakudo 12:27
oh, never mind
12:28 kid51 left, maslan left
azawawi moritz: github.com/azawawi/perl6-magickwan...-hello.pl6 # more fun with Perl 6 :) 12:28
moritz azawawi: :-) 12:29
Skarsnik azawawi, you dont need to 'sudo apt-get install libmagickwand-dev' that was the point of having the version number x)à avoiding the need to devel package x) 12:30
azawawi i know :)
Skarsnik And hello 12:31
azawawi and hi :)
gptrixie does not work on imagewand api btw :)
Skarsnik hm
how it fail? 12:32
12:32 brrt joined
azawawi unbox exception when i tried on /usr/include/ImageMagick/wand/magick-image.h 12:33
Skarsnik wow
that a weird error to have
12:34 brrt left
azawawi i think it has to do with the extern MagickBoolType func1, funct2, ...; block 12:36
12:37 pfortuny left
azawawi Skarsnik: also the module name needs to be clearer a bit... i did not know about your excellent work until i read the readme 12:37
12:41 lnrdo left
Skarsnik constant are compiled? 12:41
It keep me the value of a constant as it was at precomp time 12:43
moritz of course 12:44
they are constant after all :-)
12:45 virtualsue left
Skarsnik err 12:45
root@testperl6:/home/skarsnik/devel/perl6-gumbo# PERL6_GUMBO_LIB=/usr/local/lib/libgumbo.so.1 perl6 -I lib t/01_basic.t 12:46
1..5
Cannot locate native library '/usr/local/lib/liblibgumb.so':
DrForr "Constants aren't, variables won't, functions don't."
Skarsnik That a bit bad when the constant is like that: constat LIB = %*ENV<> || ('gumbo', v1) 12:47
DBIish has that btw
It should not be compiled when it depend on a variable x)
jnthn constant is defined as "evaluated at BEGIN time"
If you don't want that behavior, use my \foo = ... instead. 12:48
Skarsnik in my case it keep the value it has when compiled, not BEGIN
jnthn BEGIN time *is* compile time
That's the difference between BEGIN and INIT 12:49
Skarsnik I am pretty sure lot of people will fail for it
it should have a name more like static 12:50
ely-se lol 12:51
jnthn "I didn't bother to learn this language properly. Why doesn't it read my mind?"
ely-se because the motto is DWIM
jnthn Well, it does exactly what *I* mean :D
CurtisOvidPoe Why is it when I type “perl6 --doc Linenoise”, I get the warning “Could not open Linenoise. Failed to stat file: no such file or directory”. Linenoise *is* installed and I’m using it with no problem. 12:52
Skarsnik constant is not in the doc 12:53
gfldex Skarsnik: it is. It's just well hidden: doc.perl6.org/language/5to6-nutshell#constant 12:54
CurtisOvidPoe Or is that not the way to print documentation for the code?
Skarsnik jnthn, I think it will be confusing for people since script are not precompiled, only modules are
12:55 domidumont left
jnthn Skarsnik: At the moment, at least :) 12:55
Skarsnik and I need to patch DBIish x)
It's a wonder we did nave complain about that
but it's not really documented so.. 12:56
jnthn Yeah, needs documenting to mention it's BEGIN time
lunch &
Skarsnik hm, I can't export a sigilless var? 12:57
moritz you can't export and import vars 12:58
not in a way that does what you mean
Skarsnik well it work for constant x)
moritz but those are, well, constant 12:59
which is not *quite* the same as a variable
gfldex i didn't try but i believe you could export a Proxy, what will looks surprisingly like a variable 13:00
timotimo o/ 13:01
13:01 xpen joined
Skarsnik azawawi, hm, how did you run gptrixie on the header, it does not get pass gccxml for me 13:02
weird error btw x) 13:04
Term definition requires an initializer
at /home/skarsnik/devel/perl6-gumbo/lib/Gumbo/Binding.pm6:16
------> my \GUMBO_LIB⏏ is export = %*ENV<PERL6_GUMBO_LIB> || (
azawawi GCC-XML version 0.9.0 13:05
gcc 4.8.4
13:06 virtualsue joined
Skarsnik /usr/include/ImageMagick-6/wand/magick-image.h:26: error: 'WandExport' does not name a type 13:06
dalek c: 568651e | (Wenzel P. P. Peppmeyer)++ | doc/Language/terms.pod:
doc constant
timotimo Skarsnik: can you even use "is export" on a backslashed var like that?
13:06 xpen left
Skarsnik probably missing an include path 13:06
I try with our but it complain that it's not implemented 13:07
azawawi gptrixie /usr/include/ImageMagick/wand/magick-wand.h
13:07 ELBeavers joined 13:10 lnrdo joined
CurtisOvidPoe How do I view the documentation of an installed module? 13:10
13:11 maslan joined
Skarsnik perl6doc? 13:11
13:12 ELBeavers left
CurtisOvidPoe Skarsnik: there is no such command that I see. 13:12
I thought it was “perl6 —doc …”, but that’s not it either.
gfldex CurtisOvidPoe: NYI 13:13
Skarsnik I am pretty most of modules does not have a doc outside the readme.. so
CurtisOvidPoe 15 years and we still don’t have perldoc for Perl 6? o_O :)
Skarsnik azawawi, hm, I can't get it run. I tried on MagickWand.h 13:16
stmuk_ CurtisOvidPoe: p6doc should work as of earlier in the week 13:19
Skarsnik ok, I really need exclusion in gptrixie: -Functions: 2362
probably a lot from stdlib and pthread x)
stmuk_ assuming the module actually has pod in it .. otherwise App::ecoreadme might help 13:20
since it pulls the GH README often used
13:21 RabidGravy joined
CurtisOvidPoe I just successfully installed p6doc, got “Command not found”. Eventually found it in /Users/curtispoe/.rakudobrew/moar-nom/install/share/perl6/site/bin/p6doc, but when I run it on Linenoise, I get “Type check failed in binding %RESOURCES; expected Associative but got Nil” 13:23
stmuk_ you need to 'rakudobrew rehash'
13:24 _Vi left
CurtisOvidPoe stmuk_: thanks, that lets my shell find p6doc, but I still get the error message for ‘p6doc Linenoise’ 13:24
stmuk_ ok I'll try it
yeah I get the same error its probably because its nativecall not pure perl6 I'll take a look 13:26
CurtisOvidPoe Cheers :) 13:27
Skarsnik hm, it's actually annoying that the native trait has his param evaluated at compile time 13:28
13:29 virtualsue left 13:30 Ven joined
RabidGravy Skarsnik the same applies for all traits 13:30
Ven o/, #perl6
timotimo well, i can see reasons for properly setting up %*RESOURCES even when "just" extracting docs
RabidGravy I can't see how it could be much different to be hones
Skarsnik I really prefer to write my $LIB = %*ENV<PATH_TO_LIB> || ('foo', v1) and is native($LIB); that giving it a sub for it 13:31
RabidGravy constant
Skarsnik constant keep it compile time value
so it does not work
stmuk_ hmmm P6DOC_DEBUG=1 suggests a rakudo rather than p6doc bug :/
sortiz CurtisOvidPoe, for panda installed modules use: p6doc -I Module 13:32
Skarsnik I think I will add a nativecall-register-lib function to allow to just say is native xD
or something like that 13:33
ely-se m: subset Slug of Str where /<[a..z0..9-]>+/; my Slug $s = 'hello-world'; 13:34
camelia ( no output )
CurtisOvidPoe sortiz: is that a -l? If so, I just get a usage message, not the docs
13:35 jaffa4 left
stmuk_ sortiz: there is no -I option to p6doc 13:35
Skarsnik or nativecall-register-lib('foo', something the native trait take); and is native('foo'); what do you think RabidGravy? x)
stmuk_ I added -l sometime this week and it's intended to list modules 13:36
sortiz See: github.com/perl6/doc/commit/ac1285...1567781c43
stmuk_ the problem is that perl6 --doc as called by p6doc is errring 13:38
13:39 treaki joined
sortiz Sorry for the confusion, I see I not l 13:41
timotimo i suppose the problem is that we have to run --doc on the source (because of DOC phasers and such), but we also have to set up everything the same way it would've been set up when first installing the module (hence %RESOURCES)
that's ... a little difficulty that has to be addressed
Skarsnik damn, I can't export a constant that point to a sub? 13:42
13:42 lnrdo left
timotimo you ... can't? 13:42
Skarsnik m: module A { our constant B is export = sub { say foo}}; &B;
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/xPu2K8e6Lj␤Undeclared routine:␤ foo used at line 1␤␤»
Skarsnik m: module A { our constant B is export = sub { say "foo"}}; &B;
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/YICELtfcp9␤Can't use unknown trait 'is export' in a sub declaration.␤at /tmp/YICELtfcp9:1␤ expecting any of:␤ rw raw hidden-from-backtrace hidden-from-USAGE␤ pure default DEPRECATED i…»
timotimo oh, huh 13:43
m: module A { our constant Foo is export = "hi"; }; import A; Foo;
camelia rakudo-moar 775271: OUTPUT«WARNINGS for /tmp/9kfLz4aJI9:␤Useless use of constant value Foo in sink context (line 1)␤»
timotimo m: module A { our constant Foo is export = "hi"; }; import A; say Foo;
camelia rakudo-moar 775271: OUTPUT«hi␤»
timotimo m: module A { our constant Foo is export = sub :: { say "yo"; 1 }; }; import A; say Foo; 13:44
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/DvxhTtafIB␤Can't use unknown trait 'is export' in a sub declaration.␤at /tmp/DvxhTtafIB:1␤ expecting any of:␤ rw raw hidden-from-backtrace hidden-from-USAGE␤ pure default DEPRECATED i…»
MadcapJake Skarsnik: why do you need to export it, just put that sub right in Binding.pm6
timotimo seems like the "is export" lands in the wrong spot
Skarsnik because I don't want to deal with a sub name x)
[Coke] (licensing, same as) Please don't say "same as perl". Please explicitly say which license you're using for your user's sanity. 13:45
*users'
Skarsnik x)
ZoffixWin m: module A { my &x = sub { say "foo"}; our constant B is export = &x}; import A; B;
MadcapJake why not just `constant LIB = sub { %*ENV<PERL6_GUMBO_LIB> || ('gumbo', v1) }`
camelia rakudo-moar 775271: OUTPUT«WARNINGS for /tmp/UUbFPQw6Kt:␤Useless use of B symbol in sink context (line 1)␤»
Skarsnik I say same as Rakudo, see the LICENCE file x)
ZoffixWin m: module A { my &x = sub { say "foo"}; our constant B is export = &x}; import A; B(); 13:46
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/J_m33oEy6H␤Target type too complex to form a coercion type␤at /tmp/J_m33oEy6H:1␤------> 3onstant B is export = &x}; import A; B()7⏏5;␤»
azawawi github.com/azawawi/perl6-magickwan...-hello.pl6 # side-by-side original image + filtered image in Perl 6 :)
Skarsnik azawawi, give a screen shot!
azawawi pasteboard.co/Z6ccvsJ.png 13:48
Skarsnik MadcapJake, close but the native trait do not do magic when you give him a Callable (like transforming gumbo to libgumbo.so)
13:49 pierre-vigier joined
ZoffixWin twitches 13:49
Some people, I tell you... This person is complaining about my article saying one would still need to learn Perl 6 when I said "For the most part, there's no need to learn some new arcane thing to hack on the innards of the compiler." 13:50
azawawi ZoffixWin: what article? and hi :)
ZoffixWin azawawi, blogs.perl.org/users/zoffix_znet/20...erl-6.html 13:51
13:51 pierre-vigier left
ZoffixWin And the comment is on reddit: www.reddit.com/r/perl/comments/425..._6/cz7p5w8 13:51
MadcapJake Skarsnik: what about `do if %*ENV<PERL6_GUMBO_LIB> -> $lib { $lib } else { ('gumbo', v1) }`
13:52 pierre-vigier joined
Skarsnik hm 13:52
azawawi ZoffixWin: thx 13:53
Skarsnik m: use NativeCall; my \LIB = ('foo', v1); sub foo is native(LIB) {*};
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pffKOypNCU␤Can't use unknown trait 'is native' in a sub declaration.␤at /tmp/pffKOypNCU:1␤ expecting any of:␤ rw raw hidden-from-backtrace hidden-from-USAGE␤ pure default DEPRECATED i…»
Skarsnik that's weird?
DrForr "But...but.. if perl6 is written in perl 6 then I need to learn Perl 6 before I can learn perl 6...*boom*" 13:54
azawawi DrForr: :)
Skarsnik ok I am confused 13:55
[Coke] Skarsnik: as long as you include a copy of the LICENSE file, all is good. thanks. :)
Skarsnik m: use NativeCall; my $LIB = ('foo', v1); sub foo is native($LIB) {*};
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/v3FPVmq_Bi␤No appropriate parametric role variant available for 'NativeCall::Native'␤at /tmp/v3FPVmq_Bi:1␤»
ZoffixWin m: use NativeCall; my $LIB = ('foo', v1); sub foo is native(|$LIB) {*}; 13:56
camelia ( no output )
Skarsnik m: use NativeCall; my $LIB = ('foo', v1); say $LIB.WHAT; sub foo is native($LIB) {*};
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/H1V35R_si8␤No appropriate parametric role variant available for 'NativeCall::Native'␤at /tmp/H1V35R_si8:1␤»
stmuk_ the Linenoise pod isn't parsed by perldoc (p5) either
Skarsnik m: use NativeCall; my $LIB = ('foo', v1); say $LIB.WHAT;
camelia rakudo-moar 775271: OUTPUT«(List)␤»
Skarsnik hm 13:57
stmuk_ is "=begin pod" valid perl5 pod?
Skarsnik m: use NativeCall; my $LIB = ('foo', v1); say $LIB.WHAT; sub foo is native(('foo', v1)) {*};
camelia rakudo-moar 775271: OUTPUT«(List)␤»
Skarsnik I don't get why putting lib does not work 13:58
ZoffixWin stmuk_, yes
Skarsnik The role accept List
Peter- ZoffixWin, Looks like they didn't get the end of your post, if they had then their comment wouldn't make much sense
ZoffixWin Peter-, I think some people just like making noise.... 13:59
[Coke] reads www.infoworld.com/article/3024248/a...?nsdr=true which is positive but not quite right on some details. Woolfy++ for posting the biggest one in the comment section. 14:00
ZoffixWin stmuk_, though, the standard `perldoc` command doesn't parse it. That is, `=begin pod` is a valid POD command that says "This section is in format 'pod'", but perldoc isn't looking for 'pod' format sections. 14:01
(just tested)
stmuk_ ZoffixWin: that explains why perldoc can't parse Linenoise.pm although podchecker says its valid
Skarsnik gah now I get the $precomp-id error x) 14:02
MadcapJake m: use NativeCall; my @LIB = ('foo', v1); say @LIB.WHAT; sub foo is native(@LIB) {*}; 14:03
camelia rakudo-moar 775271: OUTPUT«(Array)␤»
14:03 _mg_ left 14:05 lnrdo joined
MadcapJake Skarsnik: it works with a @ sigil ^ 14:05
14:05 _Vi joined
Skarsnik hm, what can I do when precomp stop working? 14:07
stmuk_ turn it off and on again? 14:08
14:08 domidumont joined, virtualsue joined, pierre-vigier left, _mg_ joined, virtualsue left
dalek rl6-most-wanted: 6f5608a | RabidGravy++ | most-wanted/modules.md:
Rename CouchDB
 14:09
14:09 pierre-vigier joined, _mg_ left
Skarsnik I delete the lib/.precomp and now it complain that it can't find a dep 14:10
14:10 lnrdo left 14:11 lnrdo joined 14:16 lnrdo left, harrison_ joined 14:18 skids joined 14:20 FreezerburnV joined, FreezerburnV left 14:21 FreezerburnV joined
RabidGravy I concluded that renaming ^ was going to stop me worrying about it 14:24
Skarsnik ? 14:25
RabidGravy Seeing as I've been working very very slowly on it since April it needs to get fixed
14:29 mr-fooba_ left 14:30 ZoffixWin left, sjoshi left 14:31 sjoshi joined
azawawi what Perl 6 type can be convert to num64? 14:32
RabidGravy any that has a .Num as far as I know 14:33
14:35 harrison_ left
azawawi m: sub foo(num64 $x} { $x.say }; foo(10.0); 14:35
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/9laqE94AuM␤Malformed parameter␤at /tmp/9laqE94AuM:1␤------> 3sub foo(num64 $x7⏏5} { $x.say }; foo(10.0);␤ expecting any of:␤ constraint␤»
azawawi m: sub foo(num64 $x) { $x.say }; foo(10.0); 14:36
camelia rakudo-moar 775271: OUTPUT«This type cannot unbox to a native number␤ in sub foo at /tmp/Y5xGQjZ7vn line 1␤ in block <unit> at /tmp/Y5xGQjZ7vn line 1␤␤»
jnthn Needs to be a Num
arnsholt Try 10.0e0
jnthn m: sub foo(num64 $x) { $x.say }; foo(10e0);
camelia rakudo-moar 775271: OUTPUT«10␤»
arnsholt azawawi: Perl 6 decimal constants without exponent are rationals, not floating point objects 14:37
azawawi i see thx
timotimo "anything that has a .Num" isn't enough if the .Num doesn't end up getting called :)
jnthn Right, types in Perl 6 are (by default) constraints, not coercions :)
Ven m: class A {has Int $.x is rw;}; my A $a .= new; $a.x = 5; say $a.x 14:42
camelia rakudo-moar 775271: OUTPUT«5␤»
Ven ely-se: ^^
14:42 pierre-vigier left
ely-se m: class A { has Int $x; }; my A $a .= new; $a.x = 5; say $a.x 14:42
camelia rakudo-moar 775271: OUTPUT«Method 'x' not found for invocant of class 'A'␤ in block <unit> at /tmp/C85R195gQq line 1␤␤»
ely-se Ven: ^^ 14:43
ugh I missed the .
whatever
Ven $x is $!x there :P
14:43 Hathersage joined, arlenik_ joined
ely-se m: class A { has Int $.x; }; my A $a .= new; $a.x = 5; say $a.x 14:43
camelia rakudo-moar 775271: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at /tmp/GdKPVw4Y5X line 1␤␤»
jnap can someone point me at a good P6 regex guide? I'm playing with the DOM::Parser module that zoffxx started but getting in a bit over my head with the regex stuff. Honestly regex was never a strong point for me in P5 either :) 14:45
azawawi pasteboard.co/Z9TfaSL.png # charcoal effect in Perl 6 :)
14:45 ely-se left
awwaiid nice 14:46
jnap: doc.perl6.org/language/grammars most of the things about grammars are also about regexes, if that helps 14:47
jnap ok so that is still considered canonical then?
I wasn't totally sure :)
awwaiid hmm. but htat doc is too short
jnap maybe there's test cases in the p6 repo 14:48
jnthn There's doc.perl6.org/language/regexes too
awwaiid any place where doc.perl6.org is not canonical, it is a bug
jnap lol, but I'd never know :)
if the docs are right or not...
ok I'll start with that and complain if things don't seem to match up
MadcapJake docs are currently the most up-to-date source 14:49
14:50 harryirene joined 14:52 lnrdo joined 14:53 FROGGS joined, eyck joined 14:55 molaf joined, sufrostico joined 14:56 ELBeavers joined 14:58 dylanwh left 15:01 ELBeavers left, Hathersage left, lnrdo left, maslan left
Skarsnik RabidGravy, and other NC user. what do you think of gist.github.com/Skarsnik/3da8fbf2d7532ca06bdd 15:02
15:02 domidumont left, xpen joined 15:04 dylanwh joined
[Coke] I never saw an answer to what #perl6-release is for; it's for talking about how we're going to deal with releases going forward. (how to figure out which spec we're using. are we going to have a five-like model of dev & stable releases, or have a single rakudo compiler each month that has both dev & stable? How does the release process have to change? what do we do with roast after the initial christmas cut, 15:04
maybe we need to re-arrange how we store the version of the spec...)
FROGGS [Coke]: it is a on-topic channel about all of these things 15:05
15:06 domidumont joined
RabidGravy Skarsnik, seems sensible to me 15:06
[Coke] Are you just agreeing with me? Can't parse the intent of your send.
FROGGS [Coke]: I am agreeing yes
15:07 sjoshi left
FROGGS [Coke]: I opened this channel so we can discuss 2016.01 and all the issues that need to be resolved for it 15:07
stmuk_ . o O ( 2016.02 surely )
arnsholt Skarsnik: I'd prefer not to overload the native treat to take both a registered name and library name
[Coke] stmuk_: it's not too late... but it's probably almost too late. 15:08
FROGGS yeah
[Coke] and that's fine. I'd rather the next release be solid.
15:08 xpen left, lokien_ joined
Skarsnik arnsholt, hm, why? 15:08
arnsholt It's an unnecessary ambiguity in the arguments of the trait 15:09
RabidGravy maybe it would be better if it returned some "handle" object
stmuk_ BTW I'm considering releasing an unofficial MSI (probably just 2015.12 + panda ) this weekend unless anyone is very opposed
arnsholt Yeah, that could work
Then you could use multidispatch to distinguish is native(Str) and is native(Handle) 15:10
gfldex stmuk_: you can only improve the situation
FROGGS you can also declare you orn 'is gumbo' trait that applies the 'is native' trait
own*
arnsholt Oooh, hadn't thought of that. That's a pretty good idea too
RabidGravy alternatively if the register could be made to load the dll then the arguments can be omitted altogether 15:11
Skarsnik the is native trait took a scalar, it the role behind call a sub that does something with this scalair x)
RabidGravy, The issue with that, I need to change code in nqp or more I think 15:12
[Coke] stmuk_: I am pretty sure we'll have a rakudo * by the end of the first week of february.
15:12 Ven left
[Coke] (for 2016.xx) that will include only non-spec-breaking bugfixes and improvements. 15:12
Skarsnik and for the handle solution. I am not sure how you can do that since the trait is solved at compile time 15:13
arnsholt Oh, that's true. But that means that *your* sketched solution won't work either, actually 15:14
Skarsnik it work
I just code it xD
stmuk_ [Coke]: OK
arnsholt If your implementation works, a Handle approach will work equally well 15:15
[Coke] I know nine is on vacation this week, I'm currently the only volunteer for executing releases, and I have a big dayjob thing I'm trying to close the deal on. :)
Skarsnik the is native trait just apply a parametred Native role to the sub. it's when you call the sub for the first time that the argument is really solved
arnsholt Oh, you're relying on that, of course. That sounds like a bad idea, TBH 15:16
Skarsnik arnsholt, I mean if you write my $handle = register-lib; is native($handle) it will fail since $handle is non typed at compile time
arnsholt Yeah, that's true. But I really don't think the way you're going about it is a very good idea either
Since it relies on something quite specific happening after compile-time. You can't actually sanity-check it at compile-time 15:17
Skarsnik sanity check what? x) 15:18
15:19 n0tjack joined 15:20 dylanwh left
jnthn fwiw, my expectation was that is native(SomeType) would work some day, where SomeType implements some API used to do library resolution. 15:20
arnsholt Yeah, that sounds more reasonable 15:21
Skarsnik I am not how you want to use that
15:22 dylanwh joined
|Tux| Shhhhh 15:23
csv-ip5xs 17.723 15:24
test 22.432
test-t 11.937
csv-parser 50.247
tux.nl/Talks/CSV6/speed4.html + tux.nl/Talks/CSV6/speed5.html
Skarsnik I am not what you mean. You can't solve library issue at compile time. And having to define a special type sound annoying to write 15:25
abraxxa Skarsnik: finally took the time to rebase my fork of DBIish, PR is updated now 15:26
Skarsnik nice :)
jnthn |Tux|: Fist under 12? :) 15:27
Uh, first!
Skarsnik: No, the point is that you pass the object to the trait at compile time, but it can call methods on it at runtime
15:27 ely-se joined 15:28 ely-se left 15:30 bowtie left
azawawi Skarsnik: any luck with that bug for gptrixie & imagewand header files? 15:31
Skarsnik I don't have the same bug 15:32
MadcapJake Skarsnik: you could build the module file on install and place the absolute path that way.
15:33 dylanwh left
Skarsnik All I want is to have my user be able to set and ENV to change the library path 15:33
I can do that if I pass a sub to the native trait
15:33 dylanwh joined
Skarsnik but its sucks 15:33
MadcapJake why does it suck?
azawawi www.imagemagick.org/api/magick-image.php # Gonna write a parser to take out documentation/header and then feed it to gptrixie :) 15:34
Skarsnik because it's ugly?
15:34 dylanwh left 15:35 bowtie joined, dylanwh joined
azawawi "Once ImageMagick version 7 is released, we will continue to support and enhance version 6 for a minimum of 10 years." 15:36
:)
ImageMagick v7 is expected on 1st Q 2016. 15:37
Skarsnik I am updated the gist with a libhandle type 15:38
but not sure if that could work
15:39 sammers left, Zero_Dogg left, Zero_Dogg joined, jme` joined, Zero_Dogg left, Zero_Dogg joined
Skarsnik azawawi, try smething like: perl6 -I lib bin/gptrixie --all /usr/include/ImageMagick-6/wand/magick-wand.h -I /usr/include/ImageMagick-6/ -I /usr/include/i386-linux-gnu/ImageMagick-6/ ? 15:40
15:44 jack_rabbit left
azawawi Skarsnik: started it in gptrixie git repo... still parsing :) 15:45
Skarsnik take like a minute
15:46 n0tjack left
Skarsnik because there is lot of stuff included 15:46
azawawi Skarsnik: "doing magic" on ImageMagick :)
Skarsnik huhu
azawawi died in functions... printed constants 15:47
|Tux| jnthn, YES! 15:48
15:48 laz78 joined
Skarsnik This error I guess? 15:49
Method 'name' not found for invocant of class 'GPT::Class::PointerType'
in sub resolve-type at /root/piko/cpp/gptrixie/lib/GPT/DumbGenerator.pm6 line 71
15:50 domidumont left 15:51 sammers joined 15:52 stmuk_ is now known as stmuk, virtualsue joined
azawawi Skarsnik: yup 15:52
15:53 dylanwh left
Skarsnik I don't get this error (seems like a bt error or something) 15:53
azawawi bt = bluetooth? :) 15:54
or backtrace :)
Skarsnik backtrace
if you want it fail on a $type ~~ Something && $type.name; but if I displa $type ~~ Something, it false x) 15:57
15:57 dylanwh joined
azawawi enables Mac OS X Travis CI testing for ImageMagick 16:02
Skarsnik oh, it's new? 16:03
azawawi Skarsnik: travis-ci.org/azawawi/perl6-magick.../104123346 16:05
16:06 AlexDaniel joined
Skarsnik Oh it was a BT error 16:09
the issue is on the next line
16:10 mr-foobar joined
Zero_Dogg Is there any method that I can use to wait for a command started with IO.run to finish, other than busy-waiting? 16:11
16:12 lnrdo joined
jnthn Zero_Dogg: .close the handle(s) you opened 16:12
uh, output ones
iirc
That blocks until process exit 16:13
Skarsnik should I repport?
m: gist.github.com/Skarsnik/6bff05010aa08d1b7869
camelia rakudo-moar 775271: OUTPUT«Method 'piko' not found for invocant of class 'Any'␤ in block <unit> at /tmp/bnUfbGZxf_ line 4␤␤»
Skarsnik it should point to line 5
ugexe closing the run handle too early sometimes seems to cause segfaults of its own
jnthn Skarsnik: It's part of a statement starting on line 4, and line reporting is at statement level iirc 16:14
Zero_Dogg jnthn: If I might need to read .err, and don't redirect .out and .in though?
jnthn Zero_Dogg: You'd only have to close the ones you said you wanted 16:15
At least, that's how I remember it working... :) 16:16
ugexe: Reported somewhere, I assume?
16:17 khw joined
Skarsnik jnthn, yes that what I figured that it stop at the whole statement, but should I repport at LTA? It cost lot of time debugging for nothing xD 16:17
+me 16:18
16:19 lnrdo left
jnthn Skarsnik: You can, it may not be too hard to fix 16:19
ugexe jnthn: dunno. i dont send reports myself unless i have a better idea of what to point to though
Zero_Dogg jnthn: but I won't have any to close. If I don't redirect in/out, those aren't present, and If I might need to read .err, I can't close that
stmuk Stage parse : 183.238 (maybe I should have spent more than 150 euro-buck-quids on a windows build box)
jnthn Zero_Dogg: But if you're reading from it then those reads block
Zero_Dogg jnthn: ah, right
jnthn: thanks 16:20
16:20 cc9989 left 16:22 FreezerburnV left, arlenik_ left 16:23 abraxxa1 joined, abraxxa left 16:25 sftf joined 16:29 skyl4rk joined
Skarsnik azawawi, update gptrixie, it work fine now (and have fun, it give me nearly 5000 lines) 16:32
16:33 mr-foobar left 16:36 harryirene left 16:40 maslan joined
azawawi runs "zef --force install App::GPTrixie" 16:40
16:41 cc9989 joined 16:45 salv0 left 16:46 virtualsue left 16:49 nakiro left 16:52 psy_ joined, psy_ left 16:53 psy_ joined 16:55 hankache joined 16:56 zakharyas left
hankache hello #perl6 16:57
Juerd Hello hankache
hankache hi Juerd
16:58 abraxxa1 left 17:00 leont joined 17:04 cc9989 left, telex left 17:06 telex joined
azawawi Skarsnik++ # thx :) 17:08
Skarsnik excluding std/pthread stuff is my todo list x) 17:09
17:09 lokien_ left
Skarsnik I will need it for c++ that a simple include <string> is like 30 additionnals headers x) 17:10
17:10 maslan left
azawawi hmm, MacOS X travis CI containers are faster to boot/run vs linux precise builds 17:14
17:15 mcmillhj left, mcmillhj joined 17:16 xtreak joined, nadim joined 17:25 jack_rabbit joined, Calibellus joined 17:26 sjoshi joined 17:28 xtreak left
RabidGravy Oooh I didn't know there were MacOS containers 17:31
17:31 domidumont joined
RabidGravy is it a fairly standard OS build? 17:32
azawawi can one depend on more than version in a native trait e.g. is native('libfoo', v1, v2) etc? 17:34
sortiz azawawi, Nop. 17:35
17:35 domidumont left
RabidGravy no, but it would be quite cool if you could, I don't think there's a particularly good story around that 17:35
azawawi and if one removes the version, one gets a warning...
RabidGravy yes
azawawi what's the point of having symbolic links / ABI compatiblity 17:36
geekosaur ?
sortiz Unless you use a Str with a full path, so no version is checked.
geekosaur the .so without the version selects what version you link against at compile time; it's not an ideal solution but it works
sortiz Not when a Callable is involved.
s/Not/NOr/ 17:37
17:37 lnrdo joined
geekosaur runtime, i's not that comon to be able to assert wih certainty that you can hande multiple major versions, unless the library developer is unclear on the concept o API versioning 17:37
17:38 domidumont joined
RabidGravy azawi, if want to support more than one version I guess you could create a subroutine that tests for the particular versions in much the same way as LibraryCheck checks for the library and returns the one it finds 17:38
azawawi, ^ 17:39
typing too fast
azawawi im already looking at it :)
RabidGravy cool
azawawi e.g. i have libmagickwand4 and libmagickwand5 to support which are in the 6.x range (precise...trusty) 17:40
geekosaur wants :api which could in theory take a range 17:42
still need to put that whole thing together
azawawi docs.travis-ci.com/user/languages/perl6 # still has rakudobrew build-panda
geekosaur (see irc discussion ca. 12 hours ago, and list email from early to mid December) 17:43
azawawi geekosaur: yup i remember reading that irc log
17:45 leont left, jack_rabbit left 17:48 TimToady left, ely-se joined 17:49 kanishka left 17:50 _Vi left, TimToady joined
ely-se Oh boy, writing a type checker in Perl 6. 17:54
17:54 Actualeyes left 17:57 leont joined 18:02 laz78 left 18:06 perl6newbee joined 18:07 molaf left
moritz \o/ a rather well-regarded German IT print magazine just asked me if I want to write for them about Perl 6 18:16
I did that in 2009 once already
(iX by heise.de)
RabidGravy GWAN! DO IT!
m: role Foo { method bar() { say "jjsjs"; }}; my @a does Foo; @a.bar; # is there a way of doing this for an @!attribute without getting jiggy with a trait? 18:20
camelia rakudo-moar 775271: OUTPUT«jjsjs␤»
moritz RabidGravy: I will :-) 18:21
18:22 _nadim joined
RabidGravy yay! 18:24
18:25 nadim left, sftf left, molaf joined
RabidGravy the dumbest solutions are best: 18:26
m: class Foo is Array { method bar() { say "jjsjs"; }}; class Bar { has @.a is Foo; }; Bar.new.a.bar
camelia rakudo-moar 775271: OUTPUT«jjsjs␤»
18:26 lnrdo left, _nadim left 18:28 musiKk joined
[Coke] moritz: nice to have articles written by people who actually know what's going on. :) 18:29
moritz [Coke]: sounds like you're not talking about me :-)
ok, so now I have to summarize 7 years of Perl 6 development. Yikes. 18:30
RabidGravy shit happened, it's groovy and works now
18:32 ELBeavers joined
stmuk does linenoise work on windows? 18:35
18:36 ELBeavers left
hoelzro stmuk: yes 18:40
I just tested it with the latest rakudo as of wednesday
ely-se m: subset T of Str
camelia ( no output )
ely-se does that default to where { True }?
hoelzro ely-se: I believe so 18:41
stmuk I see Cannot locate symbol 'linenoiseSetCompletionCallback' in native library 'C:\rakudo\share\perl6\site\resources\54D0D684007FC0E71CAE12C86E5266F79B48B13A.dll' with 2015.12
hoelzro m: subset T of Str; my S $s = ''
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5===␤Type 'S' is not declared␤at /tmp/lQ2AhStefU:1␤------> 3subset T of Str; my S7⏏5 $s = ''␤Malformed my␤at /tmp/lQ2AhStefU:1␤------> 3subset T of Str; my7⏏5 S $s = ''␤␤»
hoelzro m: subset T of Str; my T $s = ''
camelia ( no output )
hoelzro stmuk: dammit, Hotkeys had that same problem...which version of Windows are you on?
FROGGS and I also found that at the SPWS, before I knew Linenoise worked on windows 18:42
stmuk windows 10 .. I'm using a v old VS2010 I admit
hoelzro hmm
I have it working on win10...but using mingw
18:43 addison_ joined
PerlJam moritz: write for them one time about Perl 6 or regularly? 18:43
moritz PerlJam: one time 18:44
PerlJam I was so hoping for the latter :)
moritz perlgeek.de/blog-en/perl-6/perl-6-in-2011.html "for example the mandelbrot fractral generator used to take 18 minutes to run on a machine of mine, and now takes less than 40 seconds" 18:45
18:45 maslan joined
moritz the fractal generator now runs in 2.4s 18:45
(on a machine that's from 2006 or so, so it's not just hardware :-) 18:46
18:48 Djibril joined
moritz and adding two "int" annotations speeds it up to 1.4s 18:48
PerlJam That might make a good story for others to support jnthn's grant. 18:50
18:50 Djibril left
sjn is looking at rosettacode examples 18:54
rosettacode.org/wiki/Multiline_shebang#Perl_6 doesn't work any more :-\
18:54 pochi_ is now known as pochi
RabidGravy moritz for reference it runs in 0.64 seconds here 18:54
this is a five year old i7 18:55
18:55 addison_ left
RabidGravy tiny weeny little mandelbrot though 18:57
[Coke] points at twitter.com/cokefloats/status/6906...4424042500
PerlJam [Coke]++ 18:58
[Coke] PerlJam++
[Coke] got an email, a text, and an alert on his watch from perljam's retweet. :) 18:59
mspo [Coke]: you are subscribed
19:01 addison__ joined 19:03 maslan left, kmel joined 19:04 leont left 19:06 kmel left, hankache left, kmel joined 19:07 kmel left, Actualeyes joined 19:09 hankache joined 19:10 molaf left
RabidGravy [Coke], retweeted (twice for good measure ;) 19:14
gfldex S11 claims: It is also possible to re-export the imported symbols: 19:16
use Sense :EXPORT;
is that implemented?
19:17 autarch1 joined
moritz I don't think so 19:20
PerlJam gfldex: see github.com/perlpilot/p6-Test-Class...ass.pm6#L6
moritz ok, please help me brainstorm: what's the key points of development in Perl 6 since 2009? 19:22
nom/MOP, moarvm, concurrency, Graphemes come to mind
19:23 laz78 joined
PerlJam those are the big ones 19:23
GLR too 19:24
RabidGravy Oooh, PerlJam++ it never occurred to me that I could do that
19:25 ely-se left 19:28 perl6newbee left 19:31 Amendil left
jnthn moritz: After learning how not to build a Perl 6 compiler, we finally wrote one in a rightish way :) 19:43
jnthn considers nom the turning point
19:44 maslan joined
moritz jnthn: I'm going to quote you on that :-) 19:45
azawawi github.com/perl6/roast/blob/master...INET.t#L25 # Maybe broken since on linux 19:46
m: say $*DISTRO.name; 19:48
camelia rakudo-moar 775271: OUTPUT«opensuse␤»
azawawi so how does one determine that he's running on darwin, linux, ...etc 19:49
jnthn m: say $*KERNEL.name
camelia rakudo-moar 775271: OUTPUT«linux␤»
jnthn Maybe the test file should use that 19:50
sjn \o
hankache o/
moritz oh, I guess in 2009 we didn't have much gradual typing either, did we? 19:51
I mean, compile time checks of types and stuff
azawawi moritz: so this is basically skipped on linux github.com/perl6/roast/blob/master...INET.t#L25 19:53
RabidGravy If I wanted to re-export a trait would I do something like "sub EXPORT { { '&trait_mod:<is>' = &trait_mod:<is>.dispatcher }} 19:54
19:55 maslan left 19:58 hankache left 19:59 pmurias joined
pmurias is the type system we have in Perl 6 really gradual typing? it seems like we don't really have a static type system just type checks 19:59
jnthn moritz: No, we didn't have much that relies on interseting BEGIN time pre-nom :)
pmurias: Certainly 20:00
pmurias: The natives are handled in a very static way
timotimo well, hopefully we'll have a lot more compile-time type checking at some point, but ... with the way our method calls are complete wildcards ... 20:01
arnsholt Like much of type-related discussion, I guess it depends on what you mean by types and such
jnthn I don't expect to statically check method calls in general
OO is muchly about late binding. 20:02
The lexical = early bound, object = late bound distinction is fairly fundemental in Perl 6.
pmurias arnsholt: a lot of the gradual typing seems to be about having both a "typical" static type system and mixing it with dynamic typing 20:03
timotimo right. well, method calls are late-bound in C++, too, but the way derived classes are strongly restricted in what their method signatures and return types may be is A Thing
pmurias Perl 6 seems more about sprinkling a bunch of type check to check if things are sane
arnsholt pmurias: Right. Perl 6 comes at it from the other direction, adding types to a dynamic system 20:04
Whether those two things should have different names is a fair question, I guess =)
jnthn I think both are fair to call gradual typing 20:05
timotimo so, i was wondering: if we have something like AClassLiteral.new( :named-arg ); would it be sensible to do the "does the constructor we end up calling actually take that named?" check then? 20:07
20:09 autarch1 left, sjoshi left
jnthn Usually not 'cus the default new just delegates to bless 20:12
20:13 darutoko left, ely-se joined
timotimo well, if we know the default is in use, we can go introspecting the BUILDALLPLAN 20:13
that'll conveniently also tell us if there's custom BUILD methods involved 20:15
pmurias it would be fun if at some point it was possible to import a static type system from CPAN 20:17
jnthn pmurias: Should be possible, especially once the slang stuff lands. 20:18
20:20 skyl4rk left
sjn likes "optional typing" 20:21
ely-se doesn't 20:22
too hard to tell what can be relied on 20:23
RabidGravy is there any prior-art in the ecosystem or elsewhere for "Array that gets a promise from the objects it stores and deletes any when the promise is kept"? 20:24
masak RabidGravy: you want the array to *automatically* delete elements? 20:25
RabidGravy yeah
masak ...why?
I mean, usually I prefer my data structures to stay still when I'm not looking at them 20:26
moritz jnthn: uhm, from reading the git logs in the nqp and MoarVM repos, it looks like you started MoarVM in 2012, and JVM interop in Rakudo landed later, in 2013. Is that right? Or am I missing something?
moritz surprised
masak moritz: that's how I remember it
pmurias ely-se: isn't it pretty clear in Perl 6 what can be relied on?
ely-se no. 20:27
nor in other "optionally typed" languages like TypeScript and Dart
moritz masak: I guess my surprise stems from the fact that I learned about MoarVM much later
timotimo masak: sounds more like he wants just a collection, not necessarily an ordered one 20:28
[Coke] note that the moarvm repo was private for a while.
20:28 khw left
jnthn moritz: Yes, MoarVM was develoepd in secret for a while 20:28
'cus I'd no idea how it'd go :) 20:29
RabidGravy well I have an object which has a list of other contained objects, the objects can be created by name, but it seems natural to have a .delete method on each object
jnthn And wanted to design in at my own pace and in peace. :)
Skarsnik azawawi, what is the name of imagemagic lib?
RabidGravy and because the object has been deleted it shouldn't be in the containing object any more
timotimo imagemagick, usually
jnthn And if I came up with nothing interesting, then I'd not have disappointed everyone. :) 20:30
Happily, everything worked out well. :)
Skarsnik I mean the so files )
moritz jnthn: aye :-)
jnthn bbt :) 20:31
timotimo gnite jnthn
azawawi Skarsnik: you mean magickwand or core lib?
Skarsnik hm libmagickcore-6.q16-2: /usr/lib/i386-linux-gnu/libMagickCore-6.Q16.so.2 20:32
look likeit a case of the need for pkg-config probably
RabidGravy so in terms of pseudo-code I want to be able to do "my $d = $f.things[0]; ..... ; $d.delete; # and removed from $f.things
ely-se that sounds like quite a terrible design
20:32 rindolf left
ely-se you're coupling a value to an array it happens to be an element of 20:33
20:33 khw joined
masak jnthn++ # MoarVM, and I'm not sure I've given karma for that yet, but it's definitely deserved 20:33
ely-se cyclic dependences like these are often a sign of unnecessary complexity
timotimo actually, the element can notify multiple arrays of its deletion wish
moritz RabidGravy: maybe you should give us a broader view of what you want to achieve
well, thinking more about it, if you structure it as an element that can be deactivated, and views on arrays that consider only active elements, it doesn't sound too crazy anymore 20:35
20:35 cpage_ left
timotimo just make sure iteration over the array makes sense when things get kicked out 20:35
20:36 khw left
RabidGravy It's a list of databases in couchdb, having the .delete makes sense on the representation of the individual db, I just want to save having to refetch them all 20:37
moritz RabidGravy: does it need to be an ordered list?
RabidGravy (which would require signalling the higher level object in some other way anyway ) 20:38
no
moritz if talk about "pools" instead of "arrays", things sound less spooky to me :-)
*if you
RabidGravy yeah sure 20:39
moritz and no, I don't think there's something premade for it 20:40
Skarsnik any though on plop2 gist.github.com/Skarsnik/3da8fbf2d7532ca06bdd ? 20:44
20:45 firstdayonthejob joined
RabidGravy actually, thinking about it, it's more of a pain in the arse than it's worth as you'd have to over-ride push append, shift and all the key access 20:47
azawawi m: say $*VM.config<dll>
camelia rakudo-moar 775271: OUTPUT«lib%s.so␤»
RabidGravy goes for a brute force method
moritz "when in doubt, use brute force" 20:48
ely-se that's *not* what she said
20:49 _Vi joined
moritz ely-se: no, what did she say? :-) 20:49
ely-se "use Perl 6"
Skarsnik azawawi, there is a $*DISTRIB I think 20:50
moritz ely-se: fairynuff
Skarsnik hm, what was the syntax to say what an arg on main is for? pod why? like #| hello? 20:54
20:54 Amendil joined 20:55 laz78 left
azawawi travis-ci.org/azawawi/perl6-magickwand # build passing on travis ci linux and osx :) 20:55
RabidGravy azawawi++ # go! go! go! 20:56
Skarsnik nice
20:57 espadrine left
Skarsnik azawawi, hm major version for me is 2 20:58
well it's imagemagick3 on debian stable, probably why 20:59
*graphicmagick
or this one? libmagickwand-6.q16-2: /usr/lib/i386-linux-gnu/libMagickWand-6.Q16.so.2
azawawi this one 21:01
this is a wrapper for magickwand C-based API
www.imagemagick.org/script/magick-wand.php 21:02
Skarsnik you can do use NativeCall :ALL; return guess_library_name((LIB, v4)) in your lib stuff x) 21:04
azawawi i see... thanks
Skarsnik if you don't want to bother doing the plateform translation
azawawi but is it a stable API? 21:05
21:05 maslan joined
Skarsnik well originaly it exportable for me to write test x) 21:06
21:06 xpen joined
Skarsnik You can try $*VM.platform-library-name($libname.IO, :version(v5).Str; it will probably not be changed 21:08
21:08 fireartist joined 21:09 pi1 left 21:10 xpen left 21:11 pi1 joined
azawawi Skarsnik: if you think about it, Travis CI/Appveor are good tool to test on various platforms provided you're pushing commits. When the commit rate slows and core API changes, it is bound to fail without notifying the module author 21:12
ugexe it has to be changed
RabidGravy I'm going to have make an exception called TeenSpirit so I can have a "throws-like { ...}, TeenSpirit;"
Skarsnik xD
azawawi One needs to restart CI builds every 1 day at least to have good testing results vs latest 21:13
s/latest/rakudo latest/
21:14 Begi joined
Begi Hi. I've a question : If I write a second module for Gravatar, will it be accecpted ? 21:14
Skarsnik Gravatar?
Begi to generate avatar from an email 21:15
a simple API
azawawi wonders how fast is the average fail test report time on cpan testers when there is a breaking change in one of the module's dependencies / Perl
RabidGravy Begi, why would anyone? After all there are four JSON parsers and five HTTP clients, the more the merrier ;-)
Skarsnik Begi, probably ask the author of the first if you want a different API? 21:16
still no breaking UA into 2 modules?
Begi Sorry, I've not time. I'll look that later. Thansk and good night #per6 ! 21:18
21:18 Begi left 21:19 domidumont left 21:20 lnrdo joined
ely-se m: class A { has B $.x; }; class B { } 21:22
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5===␤Type 'B' is not declared␤at /tmp/I_ABsAbGzH:1␤------> 3class A { has B7⏏5 $.x; }; class B { }␤Malformed has␤at /tmp/I_ABsAbGzH:1␤------> 3class A { has7⏏5 B $.x; }; class B { }␤␤»
ely-se any nice way to do this without moving definition of B before definition of A?
timotimo just have class B { ... } in front of class A 21:25
that's how you declare stubs in p6
ely-se ah ok 21:26
21:27 laz78 joined
ely-se then I get a redeclaration error at the non-stub definition 21:27
timotimo then you didn't spell it "..." 21:28
m: class B { ... }; class A { has B $.x; }; class B { }
camelia ( no output )
ely-se I absolutely did. 21:30
timotimo that's strange
can you show the code?
hopefully it's not a nasty bug :S 21:31
21:31 ajr_ joined
ajr_ In P5 "die message\n" kills the program without specifiying the line number. Where can I find the P6 equivalent? 21:32
21:35 Timbus left
timotimo note "oh no!"; exit(1); 21:35
Skarsnik ely-se, if it's in two separate file it does not work x)
masak ajr_: what timotimo said. I usually define a subroutine for that.
21:36 Timbus joined
moritz m: class X::WithoutLineNumber is X::AdHoc { method Str { $.payload } }; die X::WithoutLineNumber.new(payload => "message") 21:38
camelia rakudo-moar 775271: OUTPUT«message␤ in block <unit> at /tmp/5qEPIvzW2V line 1␤␤»
moritz m: class X::WithoutLineNumber is X::AdHoc { method gist { $.payload } }; die X::WithoutLineNumber.new(payload => "message")
camelia rakudo-moar 775271: OUTPUT«message␤»
moritz that way it's still catchable.
21:42 maslan left
ajr_ sorry, badly constructed question; I want to RTFM about die, though it looks as though timotimo's answer may be what I want. 21:43
timotimo die is for throwing an exception that can be caught, not for exiting the program :)
ely-se found the PHP programmer 21:44
moritz doc.perl6.org/language/exceptions
timotimo what do you mean, ely-se?
ely-se timotimo: in PHP, die exits the program 21:45
timotimo oh
21:45 pmurias left
ely-se this is ironic doc.perl6.org/type/X%3A%3AIO%3A%3ADoesNotExist 21:45
I think it should stay that way forever
mspo that could be the 404 page 21:46
SmokeMachine___ is there a way to declare a class attribute as lazy? 21:49
ajr_ Thanks, moritz.
ely-se m: Cool.new.perl.say; Cool.new.say 21:50
camelia rakudo-moar 775271: OUTPUT«Cool.new␤Cool.new␤»
ely-se m: Cool.new.sqrt
camelia rakudo-moar 775271: OUTPUT«Cannot call Numeric(Cool: ); none of these signatures match:␤ (Mu:U \v: *%_)␤ in block <unit> at /tmp/4OIdDp0BYO line 1␤␤»
ajr_ ely-se "die" does that in P5 and P6 - it's just the unwanted line number I was trying to suppress. 21:51
dalek c: d0dae21 | moritz++ | doc/Language/exceptions.pod:
Mention way to suppress backtrace
 21:52
21:52 kaare_ left
ajr_ A whole class to do that in P6 seems like a complicated regression. 21:52
sortiz In my module, I want travis uses "trusty" (via dist), but in combo with "language: perl6", "precise" are used! Any ideas? 21:53
Skarsnik this a sub trait to emulate croak x)
why whould you want to die without the backtrace?
masak ajr_: you don't need a whole class. 21:54
ajr_: just two statements, possibly wrapped in a convenience subroutine.
ajr_ Maybe I'm misreading your 16:37 message? 21:55
PerlJam SmokeMachine___: you mean you want to delay initialization until first use? 21:57
SmokeMachine___ PerlJam: yes, and if never used, never initialized...
ajr_ Sorry, masak, I meant moritz's
SmokeMachine___ PerlJam: just like perl5 moose has' lazy "adverb"... 21:58
PerlJam SmokeMachine___: I think people have done this with a Proxy that replaces itself with the appropriate value. I don't know of a better way off hand. 21:59
22:00 skids left
masak ajr_: I'm only now seeing what moritz wrote. still what timotimo wrote only requires two statements, and what I wrote only requires a sub. 22:01
22:01 addison__ left 22:04 wtw left 22:09 ELBeavers joined, geraud joined
ajr_ The note/exit construction does what I want. The subtle change in "die" will definitely be a topic for 5->6 training. 22:10
22:13 ELBeavers left
RabidGravy PerlJam, to be honest I've gone with the "has $!foo; method foo() is rw { if not $!foo.defined { ... } $!foo }" route in a lot of places 22:14
22:15 roguelazer left, roguelazer joined
RabidGravy there's scope for putting that in a trait I guess 22:15
22:16 TEttinger joined, roguelazer left 22:17 roguelazer joined
timotimo we've had that in core for a few days, then it was supposed to be ripped out and put into a module (literally, you could have 1:1 copy-pasted teh code) but nobody did it 22:17
RabidGravy I have also succeeded with a method !foo in the past
timotimo, I must have missed it as I'm sure I would have done :) 22:19
timotimo "i'm sure i would have (missed it)" or "i'm sure i would have (made that module)"?
RabidGravy I would have made the module
timotimo gimme a minute. 22:20
bdc66e79daa19d54e2467e65bb97fedb52e08b6e 22:21
a rakudo commit
github.com/rakudo/rakudo/commit/bd...db52e08b6e
22:22 addison_ joined
[Coke] posted something to the Perl 6 group on FB about "where's the release". If I get some positive feedback, I'll probably just CNP it to blogs.perl.org 22:23
(gist.github.com/coke/21c70b50e3cce8d88347 if you want to read it but don't have FB)
timotimo could you rename gistfile1.txt to gistfile1.md so it soft-wraps the text? 22:24
22:24 sufrostico left 22:26 lnrdo left
RabidGravy timotimo¸ that looks entirely doable 22:27
timotimo good luck!
[Coke] timotimo: done
timotimo TYVM
[Coke] timotimo: redone 22:28
timotimo ah, yes
i think it'd be good to spell "use v6.c" (or something) out in point 4
otherwise people may think they have to "use ThatOneFeature; use ThatOtherFeature; no NotThatFeature;" 22:29
cool. i really like your text 22:30
[Coke] updatd 4
Skarsnik Should NC start having version? 22:31
timotimo that's what i was wondering about, too
[Coke] can it without the 6.c spec changing?
[Coke] heads out
timotimo well, nativecall isn't part of the spec test suite, is it? ;) 22:32
but we might want to extend versioning out to it anyway
Skarsnik NC is lacking too much (IMO) to be tagged like 6.c
but having a version could be nice 22:33
timotimo yeah, NC is amazing, but there's still a ways to go.
Skarsnik The issue with NC is... (for 6.c stuff) what is tested in nqp/moar that is here for NC
NC.pm is small 22:34
timotimo that's true
Skarsnik You will that funny, but lot of NC missing/undef stuff make mad when I work on gptrixie 22:37
like "How I should generate this? There is 2-3 way to solve it"
I am not even sure how I should handle ref x)
22:39 addison_ left 22:40 yqt joined
Skarsnik good night #perl6 22:42
22:43 addison_ joined
masak 'night, #perl6 22:45
22:48 addison_ left
timotimo gnite Skarsnik 22:49
22:50 sufrostico joined
flussence m: use NativeCall; sub MVM_sha1(OpaquePointer $tc, str $str) is native('moar') { * }; say &MVM_sha1() 22:58
camelia rakudo-moar 775271: OUTPUT«NativeCall: Consider adding the api version of the library you want to use, sub foo is native(moar, v1)␤Calling MVM_sha1() will never work with declared signature Expected: :(NativeCall::Types::Pointer $tc, str $str)␤ in method CALL-ME at /home/ca…»
flussence some fun LHF: figure out how to make the first error go away ;)
timotimo doesn't matter, you'll never get a $tc for the currently running program 23:01
if you do, hmm.
23:02 musiKk left 23:03 azawawi left 23:06 mr-foobar joined
Skarsnik m: use NativeCall; sub MVM_sha1(OpaquePointer $tc, str $str) is native('libmoar.so') { * }; say &MVM_sha1() 23:07
camelia rakudo-moar 775271: OUTPUT«Calling MVM_sha1() will never work with declared signature Expected: :(NativeCall::Types::Pointer $tc, str $str)␤ in method CALL-ME at /home/camelia/rakudo-m-inst-1/share/perl6/sources/075EFE4B4CDAAF73190194EA876F81A1F128D1A2 line 278␤ in block <…»
23:09 leont joined
fireartist Any ideas why `prove -e perl6 -lr t` isn't loading from my './lib' directory? I'm trying to run tests on a github project 23:10
Skarsnik PERL6LIB=lib prove ... 23:11
fireartist thanks, I'll try that!
RabidGravy or prove -e "perl6 -Ilib" ...
depending on taste 23:12
isn't there a prove6 in the house nowadays?
23:12 M-Illandan left
fireartist PERL6LIB=lib is working - I can tell by how long it's taking :-) 23:12
all tests pass - I've just written my first perl6 code since the pugs project was a thing! 23:14
23:16 kid51 joined
ely-se m: if(True) { say 'hi'; } 23:18
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5===␤Word 'if' interpreted as 'if()' function call; please use whitespace instead of parens␤at /tmp/m_y8fGCE1Q:1␤------> 3if7⏏5(True) { say 'hi'; }␤Unexpected block in infix position (two terms in a row)␤at /tmp/m_y8fGCE1Q:1…»
ely-se m: if (True) { say 'hi'; }
camelia rakudo-moar 775271: OUTPUT«hi␤»
ely-se does Perl 6 have any reserved words?
timotimo hm 23:20
sometimes it's hard to call something "class"
without the parser seeing broken class definitions everywhere
m: my \class = 10; say class;
camelia rakudo-moar 775271: OUTPUT«10␤»
llfourn m: class class {}; say class.new;
camelia rakudo-moar 775271: OUTPUT«class.new␤»
llfourn :)
timotimo m: sub class; say class;
camelia rakudo-moar 775271: OUTPUT«5===SORRY!5=== Error while compiling /tmp/23XS4i05HE␤A unit-scoped sub definition is not allowed except on a MAIN sub;␤Please use the block form.␤at /tmp/23XS4i05HE:1␤------> 3sub class;7⏏5 say class;␤»
timotimo m: sub class { }; say class; 23:21
camelia rakudo-moar 775271: OUTPUT«Nil␤»
ely-se I was wondering how best to do reserved words in a grammar so I took a look at STD.pm6 only to find it doesn't do reserved words
timotimo yeah, it doesn't
23:21 addison_ joined
ely-se I don't like to list them over and over again 23:22
llfourn m: Perl, Cool and has class {} with Version, 6
camelia ( no output )
ely-se so I'll see if I can use some BEGIN trickery of some sort
well I don't think I have to actually 23:23
I only have to list them once
23:23 BenGoldberg joined 23:25 ely-se left, Skarsnik left
RabidGravy I knew it would work, I changed the name of the couchdb module to Sofa and I've already got more useful stuff done on it than since I started it in April 23:28
23:31 skids joined 23:32 leont left
timotimo good work 23:35
RabidGravy anyway beddy byes for Rabid 23:38
toodles people
23:41 fireartist left 23:42 woodruffw left 23:43 RabidGravy left 23:46 woodruffw joined, woodruffw left 23:47 woodruffw joined 23:49 addison_ left
ajr_ Is there something native in P6 equivalent to "curl", or should I be looking at ways to execute in a subshell? 23:49
AlexDaniel there is a bunch of modules for this… 23:50
does anybody know what is the best one?
I remember some did not work with https 23:51
llfourn well RabidGravy just updated: github.com/sergot/http-useragent
I wonder if anyone has written a promise based one 23:53
AlexDaniel there's also github.com/shoichikaji/perl6-HTTP-Tinyish and github.com/azawawi/perl6-net-curl
llfourn and LWP::Simple while we're at it
AlexDaniel yeah 23:54
23:55 pierre-vigier joined
ajr_ I think I need to execute some commands in a subshell anyway. Pointer to docs, please? 23:57
AlexDaniel ajr_: doc.perl6.org/routine/run
ajr_: or if you insist on shell then ｢shell｣ or qx 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!