»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by masak on 28 November 2015.
timotimo ah, no, it's just some code that gets run on startup that uses postfix:<++> i suppose 00:00
diakopter_______ m: my int64 $x = 2**63; my int64 $y = -2**63; say $x + $y; say $x - $y;
camelia rakudo-moar 20718a: OUTPUT«0␤0␤»
diakopter_______ timotimo: ^^
Zoffix m: class テスト { method モー { say "モー" } }; テスト.new.モー 00:01
camelia rakudo-moar 20718a: OUTPUT«モー␤»
Zoffix \o/
00:03 n0tjack left
timotimo neato 00:03
zengargoyle m: say 二十一 + 二十二; 00:07
camelia rakudo-moar 20718a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/W6v6ySaXzM␤Undeclared routines:␤ 二十一 used at line 1␤ 二十二 used at line 1␤␤»
ilmari .u 二十一
yoleaux U+4E00 <CJK Ideograph, First> [Lo] (一)
00:09 lucasb left
zengargoyle m: for 一..十 -> { .say } 00:09
camelia rakudo-moar 20718a: OUTPUT«5===SORRY!5===␤Function '十' needs parens to avoid gobbling block␤at /tmp/4fXQZ5oB8D:1␤------> 3for 一..十 -> { .say }7⏏5<EOL>␤Missing block (apparently claimed by '十')␤at /tmp/4fXQZ5oB8D:1␤------> 3for 一..十 -> { .say }7⏏…»
rjbs Hi! 00:10
Configure.pl loads tools/lib/NQP/Configure.
pm, which requires perl 5.10.
but instead of saying "use v5.10" it dies horribly on syntax errors
What kind of PR would you prefer? A better use line, or a fix for the syntax?
(I'd prefer to supply the former.)
dalek kudo/nom: 5a9e0b1 | timotimo++ | src/core/ (2 files):
throw out a few expensive ++ in hot-ish pieces of code
Zoffix rjbs, do whatever you think is the best solution and submit that :) 00:13
I doubt a core P5 hacker can mess it up :P
rjbs Hm. Failing to pass the nqp/moar/build/probe.pm test on solaris with gnucc. 00:14
Zoffix :(
zengargoyle how big a thing is the 5.10 needing thing?
rjbs zengargoyle: I'm not sure. At minimum, it uses //
00:15 mephinet joined
rjbs I think it's fair to require 5.10. That's system perl for years now. 00:15
zengargoyle not sure if they started w/ 5.008 and just never noticed using 5.10isms
dalek href="https://modules.perl6.org:">modules.perl6.org: a21fbfb | (Zoffix Znet)++ | / (2 files):
Fix decoding JSON on raw bytes (flussence++) (Closes #48)
zengargoyle guess unless somebody wants to go back and make sure everything is 5.8 clean forevermore... 5.10 is probably ok. 00:17
timotimo hm. what does CentOS have in its old-ish version?
zengargoyle 5.8 in like CentOS 5.x IIRC
Zoffix ew
zengargoyle 5.8.8 thou i think. 00:18
but CentOS/RedHat Perl is usually borked in some way or another even if recent. :)
timotimo TimToady: i'm a bit surprised you called it STATEMENT_LIST and not STATEMENT-LIST 00:19
Zoffix :O
00:21 donaldh left
TimToady it's still really an internal name, hence _ 00:21
uruwi Why can't I return a lowercase int from a method? 00:22
zengargoyle wonders if i should even think about trying to compile for Solaris 10 i86pc with Sun compilers. 00:23
Hotkeys er 00:24
dalek href="https://modules.perl6.org:">modules.perl6.org: 8d3f016 | (Zoffix Znet)++ | lib/ModulesPerl6/Controller/Root.pm:
Mention in docs that /total is now definitely used by www.modulecounts.com/ lest someone decides to toss it
uruwi m: my int a {return 1;}; say a;
camelia rakudo-moar 20718a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CFu_sK2OdF␤Malformed my␤at /tmp/CFu_sK2OdF:1␤------> 3my int7⏏5 a {return 1;}; say a;␤»
uruwi m: sub int a {return 1;}; say a;
camelia rakudo-moar 20718a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/_NzOpX4_5n␤Did you mean to write "my int sub a" or put "returns int" before the block?␤at /tmp/_NzOpX4_5n:1␤------> 3sub int a7⏏5 {return 1;}; say a;␤»
uruwi m: int sub a {return 1;}; say a;
camelia rakudo-moar 20718a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/AwzZCao5q0␤Two terms in a row␤at /tmp/AwzZCao5q0:1␤------> 3int7⏏5 sub a {return 1;}; say a;␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ stateme…» 00:25
uruwi m: sub a returns int {return 1;}; say a;
camelia rakudo-moar 20718a: OUTPUT«1␤»
Zoffix m: sub a returns int {return 1;}; say a.WHAT;
camelia rakudo-moar 20718a: OUTPUT«(Int)␤»
TimToady moritz: s/wroth/worth/ 00:27
"I am wroth a lot." --Donald Trump
00:30 Psyche^ joined
TEttinger "I am become death, destroyer of worlds" --a GOTO label in a java function I wrote 00:33
00:34 Psyche^_ left
timotimo m: my int $a; say $a.WHAT 00:35
camelia rakudo-moar 5a9e0b: OUTPUT«(Int)␤»
timotimo uruwi: the .WHAT causes it to be boxed, i believe
[Coke] m: my Int a = 3; say a ~~ int; 00:38
camelia rakudo-moar 5a9e0b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/34R5ARKA1b␤Malformed my (did you mean to declare a sigilless \a or $a?)␤at /tmp/34R5ARKA1b:1␤------> 3my Int a7⏏5 = 3; say a ~~ int;␤»
[Coke] m: my Int \a = 3; say a ~~ int;
camelia rakudo-moar 5a9e0b: OUTPUT«False␤»
[Coke] m: my int \a = 3; say a ~~ int;
camelia rakudo-moar 5a9e0b: OUTPUT«Type check failed in binding; expected int but got Int␤ in block <unit> at /tmp/IStRAFK0ll:1␤␤»
rjbs Should this be fatal, or just "things won't be as fast": Configuring native build environment ................... JIT isn't supported on i86pc-solaris yet. 00:43
timotimo definitely not fatal 00:44
rjbs So, what I end up with is:
probing whether your compiler thinks that it is gcc Can't compile simple gcc probe, so something is badly wrong at build/probe.pm line 92
00:44 molaf left
rjbs both gcc and cc report: gcc (GCC) 3.4.6 00:44
timotimo huh
rjbs I don't see any option for verboser reporting.
timotimo perhaps probe.pm is easy enough to look into? 00:45
timotimo has to go now
rjbs yeah, working on that 00:46
so far, ust Command failed (status 512):
diakopter_______ TEttinger: I thought java didn't have goto
rjbs woah
TEttinger diakopter: technically, no. but it does have labeled break
and it uses the same syntax for labels as GOTO
rjbs Okay, surely this is a problem with my local install. I'll sort tha tout.
diakopter_______ rjbs: well it should at least hint to use --no-jit 00:47
rjbs maybe it isn't obvious; the error I get is: probing whether your compiler thinks that it is gcc cc: installation problem, cannot exec `as': No such file or directory 00:48
diakopter_______ TEttinger: oh, yes, upwards goto... yeah 00:51
TEttinger C# has actual goto, which is kinda interesting that of all the modernizing changes to java they made, they also add goto 00:52
geekosaur rjbs, there's an OS package you need to install for that; which solaris version? 00:54
00:54 diakopter____ joined
rjbs illumos, I'll pkg in build-essential later tonight 00:54
gtodd rjbs: you're using illumos!? 00:55
rjbs Should I be ashamed? :)
gtodd haha no
00:55 diakopter_______ left
rjbs Pobox runs on SmartOS, but I told a coworker I'd prove out building rakudo on illumos for him. 00:56
gtodd it's some good stuff rbac zfs etc etc
geekosaur would actually have been able to help with actual solaris, being that he supports a couple customers that still use it extensively >.>
(I had a little smartos box installed here at one point, but it was too little; 2GB doesn't cut it for zfs root >.> )
gtodd geekosaur: true ... it always "feels" small because I boot it from a USB key 00:57
00:59 n0tjack joined, AndyDee left 01:02 njmurphy_ left 01:04 n0tjack left 01:07 yeahnoob joined 01:15 BenGoldberg joined 01:20 diakopter____ left 01:22 diakopter___ joined 01:39 znpy` joined 01:40 shamu_ left
TimToady pity we didn't name NativeCall something like "Nativity"... 01:41
gtodd hehe
and article describing it as part of the advent could use the title in a pun-like way ... 01:42
an article
01:43 diakopter___ left
geraud How could I check that $something does a My::Role? with the smartmach operator? 01:43
.oO(there is no pun-like way, only a pun-hate way)
geraud: that works
m: say Hash ~~ Associative
camelia rakudo-moar 5a9e0b: OUTPUT«True␤»
geraud hum... 01:45
01:46 ShimmerFairy joined 01:49 BenGoldberg left
rjbs So, moarvm's build seems to always pass -mt on Solaris, even if using gcc. 01:49
Then, it ends up trying to build moar's jit code, but it seemed to say it wasn't possible... 01:52
01:53 dayangkun joined
rjbs Can I use --moar-option somehow? Doesn't seem documented. 01:56
rjbs uses --moar-option=--no-jit 01:57
psch rjbs: if youre working on compiling for a new platform i'd recommend two things: (1) build things in separate steps and (2) for building moar, maybe join #moarvm and complain there if you stumble 01:59
02:00 n0tjack joined
rjbs I'm mostly trying to just get this done ASAP because it's a favor and not something actually helpful to me. ;) 02:00
psch rjbs: Configure.pl should have checked out everything for you already, which means you have nqp/ below your rakudo clone and nqp/MoarVM for moar
rjbs: oh, nevermind then :)
psch needs to get some rest anyway o/ 02:01
rjbs undefined symbol cfmakeraw :-/
ni ni
02:02 uruwi left 02:04 cygx joined, n0tjack left
cygx rjbs: you need a more recent libuv version (cf github.com/libuv/libuv/commit/672b...8d70a75df7 ) 02:05
probably worth a moarvm ticket 02:06
02:08 csd_ left
cygx 'night o/ 02:09
02:09 cygx left
ShimmerFairy using modules: still broken 02:12
skids Yeah. Also panda rebootstrap.pl (easy fix, use ~$_ on the complaining line, but then dies on trying to install already-installed File::Find) 02:15
02:16 sufrostico left
ShimmerFairy skids: by using modules being broken, I mean trying anything that hasn't been "installed" (I guess). See rt.perl.org/Public/Bug/Display.html?id=126765 02:17
skids: so I have no clue about if/how rebootstrap.pl is broken, I haven't gotten that far :P
skids Oh, I don't have that problem, I install in ~/.local and only run from one account (agree it's a problem of course.) 02:19
geraud glad to see I'm not the only one having troubles with panda. Although, I'm struggling with bootstrap.pl believing CompUnit::Repository is a function.
02:21 kaare_ joined
ShimmerFairy skids: to anyone who might suggest I should just change install location, I can only say "remove the local/ bit and you'll have what every single package manager will end up doing" :) 02:24
02:24 FROGGS_ joined
lucs Um, panda fails for me too. 02:25
skids Yeah, pretty much. My last laptop I was using /usr/local/
lucs Is the culprit panda or rakudo?
skids I guess it depends on what fails with an exit code when reinstallation fails, whether it should, and how panda should react when it does. 02:26
lucs Oookay :)
skids (but the "expected Str but got CompUnit::Local::Installation" one is a panda problem.) 02:27
02:28 FROGGS left, uruwi joined, Herby_ joined
Herby_ Good evening, everyone! 02:28
skids :/ so much indirection 02:31
Herby_ I'm having an issue with a simple 02:32
"spurt" example
when I append a file, I can't get it to recognize a newline character
uruwi Anyone care to help me?
Apparently this class's new method doesn't accept a named parameter for w
(Unexpected named parameter 'w' passed) 02:33
[Coke] Herby_: newlines are special. 02:34
Herby_ Coke, how would I spurt append "blah blah \n another line"?
[Coke] Zoffix: it would be nice if, when searching on docs, if I could type "spurt<enter>" and if only one thing came back for spurt, have it autoselect and go. 02:36
Herby_: checking.
Herby_ I swear I've looked :)
spurt "datafile.txt", "blah blah \n another line", :append;
that just puts everything on the same line
[Coke] Herby_: "stuff".IO.spurt("lots of \n stuff", :append); 02:37
llfourn uruwi: it looks like it should. FWIW I've also had odd errors like that, they were usually indicative of some other problem...
[Coke] ^^ that works fine here.
Herby_: what version of rakudo are you using?
Herby_ 2015.11?
skids What's .IO.nl-out set to?
llfourn uruwi: if you can reproduce that bug it would be awesome 02:38
Herby_ skids, I'm not quite sure how to check that
[Coke] m: say "stuff".IO.nl-out
camelia rakudo-moar 5a9e0b: OUTPUT«␤␤»
skids Herby: just say it
Herby_ coke, actually I'm on 2015.09
[Coke] Herby_: that's ancient. :) 02:39
Herby_ I get an error when I say that
skids Yeah it wasn't invented then. 02:40
Herby_ "Method 'nl-out' not found for invocant of class 'IO::Path'
so I need to update my rakudo?
Hotkeys might want to update
Herby_ another dumb question. I'm on Windows 7. Is there an easier way to update Rakudo other than uninstall, reinstall?
AlexDaniel m: say gather for ^10 { take take take take ^10 .pick }
camelia rakudo-moar 5a9e0b: OUTPUT«(2 2 2 2 1 1 1 1 4 4 4 4 1 1 1 1 1 1 1 1 7 7 7 7 2 2 2 2 9 9 9 9 6 6 6 6 5 5 5 5)␤»
Hotkeys Herby_: if you can get visual studio 2010 02:41
via a method of your choosing
you can use rakudobrew
Herby_ Will Visual Studio 2013 work, or is it special to 2010?
Hotkeys you can try it
it wouldn't build for me
i've heard the same from others 02:42
Herby_ ok, and worst case scenario just uninstall rakudo and reinstall?
the new version that is
Hotkeys yeah
rakudobrew is a script
you just run 'rakudobrew build moar' from the visual studio command prompt 02:43
and it builds the latest version from the repo
AlexDaniel m: say gather { take gather { take gather { take gather { take 42 } } } }; # that's fun
camelia rakudo-moar 5a9e0b: OUTPUT«((((42))))␤»
Hotkeys (or you can specify releases like 2015-11)
.oO(with curli in this state maybe we shouldn't be advising people to brew)
Herby_ ok, let me take a look
skids or figure out a triple that is safe
ShimmerFairy Huh, on my system using perl6's mkdir will respect my umask, despite having a 0o777 default parameter (even making it explicit). Is that intended behavior or not?
skids isn't umask enforced by the OS? 02:45
02:46 eternaleye is now known as M-eternaleye
ShimmerFairy skids: I honestly don't know how that all works, so beats me :) 02:46
geekosaur the mode is passed to the OS system call, which applies the umask 02:47
skids yeah you have to chmod to defy your process umask.
02:47 M-eternaleye is now known as eternaleye
ShimmerFairy skids: alright, that modifies my complaint in the bug report somewhat :) 02:48
AlexDaniel is there any recursion limit?
skids not judging from some of the spins I've accidentally created :-) 02:49
Herby_ another beginners question... 02:53
uruwi Oh, stray semicolon
Herby_ the rakudo website only has msi for 2015.09 it looks like: rakudo.org/downloads/star/
uruwi I'd wait for the 2015.11 msi's 02:54
Herby_ okie dokie
02:55 dayangkun left
dj_goku anyone having issues building moar-nom on osx? 02:56
err it looks like there might be an issue with panda: Type check failed in assignment to $prefix; expected Str but got CompUnit::Repository::Installation
Herby_ well thanks for the help!
02:56 Herby_ left
skids dj_goku: yes, on the complaining line, use ~$_ instead of $_ 02:57
Hotkeys dj_goku it fails for me on windows too 02:59
03:01 n0tjack joined
skids Or well, maybe that's not the right fix. 03:01
m: %*CUSTOM_LIB.perl.say
camelia rakudo-moar 5a9e0b: OUTPUT«Failure.new(exception => X::Dynamic::NotFound.new(name => "\%*CUSTOM_LIB"), backtrace => Backtrace.new)␤»
dj_goku skids: there are a few new commits on panda. 03:02
ShimmerFairy m: say %*CUSTOM_LIB 03:03
camelia rakudo-moar 5a9e0b: OUTPUT«Dynamic variable %*CUSTOM_LIB not found␤ in block <unit> at /tmp/BCtwCyLBjn:1␤␤Actually thrown at:␤ in block <unit> at /tmp/BCtwCyLBjn:1␤␤»
ShimmerFairy The funny thing is that a recent commit claims to fix this.
dj_goku skids: weird, made the change and re-ran rakudobrew build moar and it is moving along. 03:04
skids Yeah the fix dslodges things but only by making a failure go unnoticed. 03:05
dj_goku ahh failed further down the script.
skids Reinstalling File::Find?
Hotkeys you can just rakudobrew build panda no?
saves you from having to rebuild rakudo
03:06 n0tjack left 03:08 cdg left 03:11 noganex_ joined 03:12 kid51 left 03:13 noganex left 03:14 uruwi_ joined 03:15 uruwi left 03:16 kaare_ left
dj_goku I just nuked moar-nom and all is good now. 03:17
03:18 uruwi_ left
dj_goku FROGGS_: did you ever finish the testers dancer changes for the unicode issue we discussed? 03:19
03:20 uruwi joined
skids Oh I guess that ~$_ fix was actually working -- hrm what are the conditions under which %*CUSTOM_LIB populates? 03:24
dj_goku home => (Any), perl => inst#~/.rakudobrew/moar-nom/install/share/perl6, site => inst#~/.rakudobrew/moar-nom/install/share/perl6/site, vendor => inst#~/.rakudobrew/moar-nom/install/share/perl6/vendor 03:25
oh 03:26
diakopter np. XD 03:27
skids m: use Pod::To::Text; %*CUSTOM_LIB.perl.say 03:30
camelia rakudo-moar 5a9e0b: OUTPUT«{:home(Any), :perl(CompUnit::Repository::Installation.new('/home/camelia/rakudo-m-inst-1/share/perl6')), :site(CompUnit::Repository::Installation.new('/home/camelia/rakudo-m-inst-1/share/perl6/site')), :vendor(CompUnit::Repository::Installation.new('/home/…»
skids There we go. %*CUSTOM_LIB only exists if you actually load something.
ShimmerFairy I was going to suggest that perhaps we should revert the CURI merge what with all the problems, but as far as I can tell it was done as a rebase. What the fuck. 03:31
(Please tell me I'm just missing the merge commit, and that I'm wrong about it being a rebase.)
dj_goku m: say [1.3, 2.8, 32323423.4, 4.0] 03:32
camelia rakudo-moar 5a9e0b: OUTPUT«[1.3 2.8 32323423.4 4]␤»
ShimmerFairy skids: huh, my :home is unset as well. 03:34
skids In retrospect it might have been best to have made a 2015.11 star before the CURLI merge. I'm sure the pain will subside.
03:34 dayangkun joined
ShimmerFairy decides to direct her frustration into writing a rant post :P 03:35
03:35 vendethiel joined
dj_goku timotimo: around? 03:36
diakopter ShimmerFairy: better than a rant post would be a thorough explanation of the pain points and missed expectations, so when nine wakes up, he can calmly absorb the new information and best be able to use his limited time to fix them quickly :D 03:38
ShimmerFairy: I say this because the complaints/symptoms are currently spread quite widely across the irc log and I'm not sure anyone's gathered them all together 03:40
(perhaps you're exactly a right person to collate them?)
ShimmerFairy diakopter: sure, I'm not in the business of writing incoherent ALL-CAPS RAGE, but I'm not in the mood to play nice anymore :) . So, in other words, I'm not gonna mince words or try and point out the positives (esp. since I can't _get_ to the potential positives)
diakopter: essentially, I'd like said post to be informative, but also indicative of the displeasure I feel.
diakopter well, displeasure is a problem of improper expectations, IFF it was feasible/plausible/reasonable to have had appropriate expectations felt :p 03:42
in this case, primary head branch of rakudo should still be considered quite unpredictable on most platforms/systems, even a few weeks before release 03:43
ShimmerFairy diakopter: I think the issue boils down to not nearly enough testing pre-actually-rebase. I'm not sure how that couldn't been achieved without an actually-rebase, aside from thinking about the old alpha/b/etc. days, so I'm not going to act like there could've been way more testing or something.
diakopter no; it was a rushed rewrite, a necessary down-to-the-wire evil 03:44
AlexDaniel “Unhandled exception: failed to load library 'dynext/libperl6_ops_moar.so'” huh?
diakopter ShimmerFairy: not nearly enough testing, when all resources are donated/volunteered/sponsored, will always be the case 03:45
03:45 znpy` left
ShimmerFairy diakopter: so that's why I'm not going to write things with SEETHING RAGE; partly because that's not who I am, and partly because I'm too good at remembering the limitations in our development process :) 03:45
zengargoyle did that ~/.perl6 location get configurable?
dalek osystem: 7b74507 | bluebear94++ | META.list:
Add Math::Random
osystem: d6da07d | (Zoffix Znet)++ | META.list:
Merge pull request #100 from bluebear94/master

Add Math::Random: github.com/bluebear94/Math-Random
ShimmerFairy (I do think having more people use the Universal Default Install Location™ would be nice, but alas MoarVM thinks it shouldn't do that default.)
zengargoyle actually shares a rakudobrew amongst multiple users...
diakopter *boggles*
zengargoyle they're both me, but ... 03:47
03:47 uvtc joined
diakopter *giggles* 03:47
03:48 uvtc left, yqt left
zengargoyle i have $WORK name to make ssh/email easy and zen me for other stuff, and install stuff on one machine and rsync to other machines (vs building on each machine for each user) 03:48
03:49 uvtc joined
zengargoyle have a dread that it's going to be a PITA now. 03:49
uvtc Oooh, nice. Tried rakudobrew this afternoon, and `rakudobrew build-panda` left me searching for where `panda` got installed, but just tried again now and it's there in ~/.rakudobrew/bin!
Juerd I was going to "rakudobrew build moar" before I went to bed, but it failed. 03:50
I'm just dropping this here in case it's useful to anyone
Type check failed in assignment to $prefix; expected Str but got CompUnit::Repository::Installation
in sub MAIN at rebootstrap.pl:13
in block <unit> at rebootstrap.pl:8
Not investigating it because I'm off to bed. Good any(night, $localtime), *
skids Yeah, actually moar succeeded it was panda that failed.
uvtc Is `panda install Linenoise` still the preferred way to get regular keyboard shortcuts working in the repl? 03:52
zengargoyle uvtc: probably, but i'm liking not using Linenoise and just using rlwrap.
it wraps better, handles unicode, keeps history across sessions, does completions on words it has seen. 03:53
uvtc zengargoyle, how do I install rlwrap and make it work with perl6?
dj_goku I was just going to ask the same thing. :D 03:54
zengargoyle you on a linux system? don't know if there's rlwrap for windows... # rlwrap == GNU readline wrapper 03:55
uvtc zengargoyle, yes, Debian Testing.
zengargoyle apt-get install rlwrap
uvtc Ooof. Right. :)
Thanks, zengargoyle
dj_goku brew install rlwrap for osx
zengargoyle then: rlwrap perl6 # for the basic usage. Note, i believe you have to not install Linenoise (or remove the Linenoise.pm module for it to work right) 03:56
dj_goku uvtc: rlwrap perl6
zengargoyle: whoops
uvtc zengargoyle, thanks! Running system updates, but will try that after a few minutes. 03:57
zengargoyle i've tweaked a bit...: rlwrap -m -M .pl6 -r -O'^> ' -pBlue -S'> ' --no-warnings -R perl6
dj_goku lol 03:58
zengargoyle which makes Ctrl-^ to edit a line in $EDITOR work nicer... there's still a bit of borkedness since the REPL is line oriented.
03:58 Guest51288 left
zengargoyle multiple lines from $EDITOR each get their own '> ' prompt sorta inserted in the output vs all running as one big chunk. 03:59
diakopter are REPLs used to prototype ideas about little scripts? or to explore methods/syntax while learning? or what?
zengargoyle both/all 04:00
skids only finds them useful in a debugger context
diakopter do you frequently turn repl transcripts into persistent script files?
or is it more for practice/memory 04:01
flussence on a related note, perl6 supporting “”‘’ quotes is a godsend for shell one-liners, because bash doesn't notice them *at all* :)
uvtc zengargoyle, rlwrap works great. Thanks again!
zengargoyle is bad for working with giant one-liners, then dropping them into $EDITOR and formatting/saving them. :)
but with rlwrap, at least there's a history file you can copy/edit a bit.
do that quite often with BASH work... futz around and then go back and delete the stuff that didn't work. :P 04:02
diakopter ok, so it's more about using a line editor to approximate a multi-line editor where you can iteratively run it up to a certain point only
geekosaur often does "desk calculator" type things and quick checks
04:03 n0tjack joined
diakopter geekosaur: checking your memory of something? or checking whether something works the way you imagined it, in the course of problem solving? 04:04
mspo doc.perl6.org/language/$_
geekosaur either. or someone pops up in a channel asking for help and I see if I can reproduce it etc
zengargoyle for bigger stuff, i'm a fan of mini-continuous-integration. make the script with a MAIN('test') that does tests, or just outputs.
geekosaur in any case it's not something I do development in 04:05
zengargoyle and use inotify to run it after every save. edit in editor, watch other window for results.
skids Really they are *very* useful inside debuggers, and worth it just for that, so worth supportig in general.
zengargoyle i *think* hoelzro?? is pondering having readline integration support built in. 04:07
04:07 n0tjack left
flussence last I heard, the main obstacle there is readline's license 04:08
zengargoyle but it's some sort of licensing issue actually including it. but use it if you have it is fine.
diakopter an insurmountable license to be bundled by default, for sure.. but it would be nice if a module worked seamlessly
zengargoyle yeah, that. i think the hooking in part wouldn't be that difficult. 04:09
i wrote a little filter for rlwrap to remove the extra '> 's, but it's sorta hacky. plenty of filter examples come with the rlwrap install down in /usr/shar/doc/rlwrap or such. 04:11
04:17 AlexDaniel left 04:26 scottj joined
bzipitidoo In doc.perl6.org/language/grammars under Action objects, how does the example work? The class seems to have a a "make" subroutine or function, but the object calls "made". 04:27
TimToady made is the incoming ast data from the leaves, make is how you make a new outgoing ast node 04:28
bzipitidoo ast?
TimToady abstract syntax tree
skids abstract syntax tree
TimToady what you get from a parser :) 04:29
ugexe are bin scripts and resources not considered part of a Distribution object?
TimToady well, if you get fancy; by default you just get the match object tree
04:29 balazs left
TimToady ugexe: I dunno if we've thought much beyond modules 04:30
ugexe what i mean is .install gets passed ($distribution, %resources, %scripts), but to me those seem like they would go in the Distribution object so i wonder if im overlooking something 04:32
diakopter I dunno if TimToady's looked at the new precompile architecture XD
TimToady I do :)
I only read jnthn's gist design 04:33
diakopter er, compunit repo installation architecture, I meant 04:34
bzipitidoo hm... all I get when I try to change the example is "No such method 'made' for invocant of type 'Any'" 04:37
TimToady you have to call it on a Match object that matched 04:38
(and that already did a make)
04:41 vendethiel left 04:55 keix joined 04:56 skids left 04:57 anaeem1 joined 05:04 n0tjack joined, cdg joined, csd_ joined 05:06 anaeem1 left, anaeem1 joined, anaeem1 left 05:08 n0tjack left 05:09 csd_ left 05:23 khw left 05:28 davmillar joined 05:32 uvtc left
MadcapJake Does the new function composition operator `o` support being written as unicode `∘`? 05:46
ShimmerFairy MadcapJake: well, for every unicode-needing operator, we add another operator that uses ASCII only, so infix:<∘> gets defined along with another infix:<o> that just calls the unicode one 05:47
MadcapJake but it does work?
ShimmerFairy MadcapJake: of course it does :) 05:48
MadcapJake just wondering because i looked over the gh repo, and various release notes and nowhere does it mention a unicode version of that op 05:49
ShimmerFairy m: sub infix:<Γ>($a, $b) { $a + $b }; constant &infix:<G> := &infix:<Γ>; say 1 Γ 2; say 1 G 2; # one way of doing it, that AFAIK lets them truly be synonyms
camelia rakudo-moar 5a9e0b: OUTPUT«3␤3␤» 05:50
ShimmerFairy MadcapJake: well, o is easier to type for most I imagine, and I think (but I don't speak for everyone) that both versions are typically thought of as the same operator, so "the o operator" implies all possible spellings. 05:51
MadcapJake m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = f o g; &h(2); 05:52
camelia rakudo-moar 5a9e0b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/TOpYsRXBLt␤Undeclared routine:␤ o used at line 1␤␤»
MadcapJake m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = f ∘ g; &h(2); 05:53
camelia rakudo-moar 5a9e0b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/enmliclFsH␤Preceding context expects a term, but found infix ∘ instead␤at /tmp/enmliclFsH:1␤------> 3 1 }; sub g($x) { $x + 10 }; my &h = f ∘7⏏5 g; &h(2);␤»
ShimmerFairy m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = &f ∘ &g; &h(2);
camelia ( no output )
ShimmerFairy m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = &f ∘ &g; say &h(2);
camelia rakudo-moar 5a9e0b: OUTPUT«13␤»
gfldex at the bottom of github.com/rakudo/rakudo/blob/nom/...rs.pm#L655
ShimmerFairy MadcapJake: you need to use &f form to refer to the routine as an object, otherwise it's seen as a call :) 05:54
MadcapJake oh yeah :P
gfldex: thanks! That's helpful! 05:55
m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = &f o &g; &h(2);
camelia ( no output )
MadcapJake m: sub f($x) { $x + 1 }; sub g($x) { $x + 10 }; my &h = &f o &g; say &h(2);
camelia rakudo-moar 5a9e0b: OUTPUT«13␤»
MadcapJake I just had to check for my own sanity :D
05:57 raiph left
MadcapJake Just getting some of the latest additions into my perl6 highlighter, thanks for the help! G'night all! 06:02
06:02 MadcapJake left 06:03 Actualeyes1 joined, xfix joined 06:04 n0tjack joined
hoelzro zengargoyle: I have Readline integration on my TODO list, but ENOTUITS 06:09
06:09 n0tjack left 06:14 b7j0c joined 06:25 darutoko joined 06:27 bjz joined 06:31 bjz left
ShimmerFairy to whoever handles the advent site's look: it could really use a search feature :) 06:44
06:45 geraud left 06:47 bjz joined 06:49 bzipitidoo left 06:53 FROGGS_ left 06:55 domidumont joined 07:00 domidumont left 07:01 domidumont joined 07:03 quester joined, quester left, quester joined 07:05 davmillar left, n0tjack joined 07:08 CIAvash joined 07:10 n0tjack left
ShimmerFairy .tell [Coke] looking at gist.github.com/coke/63c7bbb82f075c7bb1f5 , there's a spurious Lue line that needn't be there (line 356 of output), and neither the so-called aliases nor the svn-y email for ShimmerFairy should be there either :) 07:10
yoleaux ShimmerFairy: I'll pass your message to [Coke].
07:12 b7j0c left 07:15 vendethiel joined 07:17 kjs_ joined
dalek kudo/nom: ec41c2f | (Ricardo Signes)++ | Configure.pl:
Configure ends up requiring perl v5.10.0 for // op
kudo/nom: ba106f8 | FROGGS++ | Configure.pl:
Merge pull request #607 from rjbs/config-version

Configure ends up requiring perl v5.10.0 for // op
07:27 FROGGS joined, kjs_ left
Woodi hallo #perl6 :) 07:28
nine .tell FROGGS panda's rebootstrap is obsolete. You can now upgrade your rakudo and keep all installed distributions (including panda)