»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by diakopter on 14 April 2013.
pmichaud I'm not too surprised that using QRPA doesn't significantly change parsing speed; the big improvements are in shift/unshift, which the grammar engine doesn't do a lot of. 01:04
japhb_ pmichaud, mostly just push/pop? 01:07
pmichaud yup. 01:09
and those aren't significantly different from RPA
github.com/perl6/nqp/commit/b13610905d # ouch, PLEASE don't give QRPA any METHODS!
better is to find the places that are using .push and get them to do nqp::push instead. 01:10
japhb_ I get the feeling there is some miscommunication happening. There seems to be some confusion throughout the day about .push versus nqp::push, and when each is used. 01:11
(That may be an incorrect impression.)
pmichaud one does .push on higher-level sixmodel-based structures. 01:12
one does nqp::push on low-level vm-specific stuff
so, on a QAST node, use .push 01:13
on a QRPA PMC, use nqp::push
RPA provided both .push method and nqp::push (vtable) forms, which was okay at the time, but now we really need to be more strict about usage I think. Also, the vtable form is incredibly slow. 01:14
sorry
the vtable form is okay, the method form on a (Q)RPA is incredibly slow.
because it creates an inferior runloop
japhb_ wonders if there is some slow in NQP or Rakudo because of .push used on (Q)RPAs 01:15
pmichaud I don't mind if QPRA keeps the METHODS on a very temporary basis to assist with converting, but I don't want them there for long.
(and yes, I see that they were added in a branch, not the master, but I just wanted to point out that they're heading in the wrong direction :) 01:16
I don't know if there's some slow in NQP/Rakudo because of .push on (Q)RPAs... I've been trying to eliminate them as much as possible over time. 01:17
Obviously they're not all gone, though, if there was a need to add them to QRPA in order to make it work. 01:18
japhb_
.oO( That's one of the duties of project leads: pointing out when things are headed the wrong way ... )
japhb_ is not-so-secretly hoping one of the remaining (Q)RPA .push locations turns out to have been a hot spot, and we get a visible win fixing it 01:19
[Coke] hacking on release: ! Couldn't find module or a distribution Time::y2038 () 01:30
hurm. might be due to networking issue. 01:32
timotimo pmichaud: thanks very much about the clarification on .push! 01:33
i'm glad i put the nqp::push changes i made thus far into a branch so that i can now (tomorrow!) salvage them
is there a good way to collect places where .push has been called "wrongly"? as in source locations? but maybe not die with a stacktrace, but continue running and collect occurences, so i could fix them in bulk? 01:34
[Coke] should nto have volunteered.
timotimo .o(also, if the .push method is incredibly slow, maybe there'll be a bit more improvements to be had still!) 01:35
pmichaud timotimo: add a METHOD push to QRPA that throws an exception immediately. 01:42
and maybe gives a stacktrace that lets you find out where it's being called wrongly. :)
or, if you want it to keep running, have METHOD push do some sort of logging w/stacktrace and then you'll have them all collected. 01:43
there really shouldn't be that many, though.
timotimo i believe there were a whole bunch in rakudo 01:45
anyway, i'll see tomorrow.
pmichaud .push on RPA? 01:46
that'd be... interesting.
anyway, again, it's not .push/.pop that are slow on RPA, it's .shift and .unshift
although .push and .pop are definitely slower than nqp::push and nqp::pop
I bet I could benchmark those :-) 01:47
japhb_ :-) 01:48
colomon has URI working again. 01:53
japhb_ colomon++ 01:54
gtodd Europeans - it is too late to be working !! :-) 01:58
diakopter FROGGS_: hey it's lats
late
gtodd wonders if yoleaux has TZ time for bed feaures 01:59
labster r: gist.github.com/labster/aaa720c8659549616cd0 02:03
camelia rakudo 203e3a: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/VuMMNiyebU:1␤------> https⏏://gist.github.com/labster/aaa720c865954␤ expecting any of:␤ argument list␤ postfix␤ infix stopper␤ infix or meta-infix␤ st…
labster camelia doesn't like to run gists now? 02:04
anyway, that bit of code is mysteriously parsefailing on me with role composition, and I don't know why. 02:05
diakopter labster: it never cohld run private ones I think
labster oh, right 02:06
labster r: gist.github.com/labster/5417609 02:07
camelia rakudo 203e3a: OUTPUT«===SORRY!===␤Unable to parse class definition␤at /tmp/COCPCaaRlw:7␤------> my class IO::Path is Cool does IO::⏏FileTestable {␤ expecting any of:␤ scoped declarator␤ generic role␤»
[Coke] are we still creating an NQP release at the same time as the rakudo release? 02:08
diakopter have to if it's needed 02:10
er
is that a tautology or a truism 02:11
[Coke] ah, just because we tag a release of nqp doesn't mean we have to require it.
diakopter yeah, but why bother; nobody uses nqp releases except rakudo downloaders 02:13
labster class nqp::release { has $to if need Ed; }
Okay, if I swap the location of the last two classes in my gist, it works, but neither depends on each other. 02:15
labster is confused
diakopter wonders how many end up using --gen-nqp anyway
labster I used it once, but then I needed to gen a new parrot anyway.
diakopter would hou mail me the gist url i can't click links here now 02:17
nm 02:18
I can use the log
pmichaud [Coke]: standard is to create a release of nqp and use it for the rakudo release, yes. 02:19
thus the 2013.04 rakudo release should be using the 2013.04 nqp release.
[Coke] ok. 02:21
pmichaud gist.github.com/pmichaud/5417664 # timings, RPA vs QRPA, and native calls versus METHOD calls 02:26
using .push and .pop for RPA is 20x-50x slower than using nqp::push or nqp::pop 02:27
[Coke] "Surely someone has written a script to automate cutting a release". :P 02:35
colomon another proto question: is there a way to specify that a parameter may be @ or %? 02:42
[Coke] but not other things that would fit in $ ? 02:43
colomon well, I think anything that allowed those two choices would be okay. 02:44
[Coke]++ # that did it. 02:49
I'm starting to have serious doubts that Algorithm::Diff ever worked...
japhb_ pmichaud, that level of timing difference is ... scary 02:53
Some of the modules seem to have been written speculatively.
colomon does if (&keyGen eq &default_keyGen) have a chance of working correctly? 02:57
r: my &a = -> $a { say $a; }; say ~&a;
camelia rakudo 203e3a: OUTPUT«␤»
[Coke] I do not have permission to upload files to rakudo.org 03:10
dalek kudo/nom: c984cfe | coke++ | docs/ChangeLog:
Updating changelog for release prep
03:16
kudo/nom: 9dadcf2 | coke++ | docs/release_guide.pod:
label release
kudo/nom: c06a447 | coke++ | docs/announce/2013.04.md:
add new release announcement
kudo/nom: 5566844 | coke++ | tools/build/NQP_REVISION:
[release] bump NQP revision
rakudo/nom: 88221cf | coke++ | VERSION:
rakudo/nom: [release] bump VERSION
japhb_ [Coke], shouldn't the release guide say to push to nqp before rakudo ...? 03:17
dalek p: 891075b | coke++ | VERSION:
bump VERSION to 2013.04
[Coke] it was stuck due to git branch silliness. 03:19
japhb_ ah
[Coke]++ # April releases! 03:20
Who volunteered for Rakudo Star?
(Or am I imagining that?)
[Coke] no one volunteered for star that I recall. 03:23
tangentstorm so do you need someone to do it? 03:26
[Coke] we don't always do a star release.
tangentstorm I will do it if you tell me how :D
oh
[Coke] tangentstorm: I'm pretty sure it'll be in teh sar repo under something like docs/release_guide.pod
japhb_ Although there have been enough module fixes in the last couple days that it seems like a good idea, *assuming* we haven't regressed on any. 03:27
tangentstorm so it's supposed to release in 2 days? 03:32
pmichaud anytime between now and end-of-april, but only if there aren't any module failures. 03:36
however, it would be good to find/fix the module failures.
I was on the hook for the March Star, and would like to do April, but I'm unreliable for any specific tasks at the moment. :-( 03:37
tangentstorm 2013-04-18 Rakudo #63 "Albany" (Coke)
did you just put your name there by default, [Coke] ?
pmichaud that's for the compiler, not Star
star's release instructions are at github.com/rakudo/star/blob/master...-guide.pod 03:38
japhb_ tangentstorm, he just did that one. :-)
[Coke] kicks of a test/build... on OSX which has known test failures on spectest. coke starts over on feather.
tangentstorm oh i missparsed the "star repo" part. :) 03:39
pmichaud Albany++
japhb_ has been building the new Rakudo on linux/x86-32 03:40
Parrot and NQP built so far, working on CORE.setting 03:41
pmichaud I think I want "Austin" for the 2013.05 release. 03:42
in honor of yapc::na 2013, which has a *ton* of Perl 6 related talks, hackathons, and the like. 03:43
[Coke]: let me know when it's safe to modify release_guide.pod for that.
tangentstorm Okay I will volunteer.
For Star I mean. 03:44
pmichaud [Coke]: did you get permission to rakudo.org yet? 03:45
tangentstorm is it okay if i use emacs instead of vi though? ;D 03:46
labster heresy!
pmichaud tangentstorm: as long as we don't know about it, yes. :D
tangentstorm i'll keep it on the down low 03:47
japhb_ Rakudo built fine, rebootstrapping panda 03:48
Nom- emacs? pfft ... I use Textmate : 03:51
:P
vim when I have nothing but a terminal 03:52
pmichaud I sometimes use cat(1) as my editor. :-P
japhb_ considers App::EDLIN 03:53
... but only for a moment
panda and my usual modules all installed fine 03:54
labster reinstall and repanda works fine here too. 03:56
[Coke] pmichaud: nope 03:57
pmichaud [Coke]: do you have a public ssh key? 04:00
(that I can use to enable login to rakudo.org)
one used with github is fine, fwiw
tangentstorm is writing a new kind of editor but will stick with ema--... whatever he's using now... for the time being :)
pmichaud afk # sleep 04:17
pmichaud gist.github.com/pmichaud/5418133 # slightly better benchmark of RPA vs QRPA and native vs METHOD 04:22
labster managed to break RESTRICTED.setting while CORE.setting works fine: Null attribute map for P6opaque in class 'IO' 04:27
[Coke] arglebargle, feather so slooooow
(for spectesting)
Heather hi 04:32
labster hi 04:34
[Coke] WTF is make stresstest? 04:51
ah, runs 2 more tests, thank goodness, I can run those 2 by hand. 04:53
moritz spectest + a few files that are known to take very long 04:53
Heather how to regex string in % % ? 05:10
r: say '%a%' ~~ /%/w+%/
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name␤at /tmp/Tl1AmXJcAG:1␤------> say '%a%' ~~ /%⏏/w+%/␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ prefix or term␤ pre…
Heather r: say '%a%' ~~ /%/ 05:11
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter % (must be quoted to match literally)␤at /tmp/T3rpWFrrrG:1␤------> say '%a%' ~~ /⏏%/␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ p…
Heather r: say '%a%' ~~ /\%/
camelia rakudo 88221c: OUTPUT«「%」␤␤»
Heather r: say '%a%' ~~ /\%/w+/ 05:11
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/5TRHZQ82Wo:1␤------> say '%a%' ~~ /\%/⏏w+/␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ statement end␤ statement modifier␤ …
sorear r: say "this is my %foo%".comb(/\%<(\S+)>\%/)
camelia rakudo 88221c: OUTPUT«foo␤»
Heather sorear thank you ) 05:12
sorear but what's comb ?
r: say '%a%' ~~ /\%<(\S+)>\%/
camelia rakudo 88221c: OUTPUT«「a」␤␤»
sorear Heather: Extracts all matches, returns them as list of strings 05:13
Heather sorear I've got it as given pattern when /\%<(\S+)>\%/ 05:13
sorear how can I use it there? 05:14
Heather I see, as "$/" 05:15
Heather sorear but what if there is a list... 05:16
then @/ ?
labster still $/
it would have a list in $/[0] 05:17
Heather labster I need to subst all this matches
labster isn't [0] a first element?
labster that's a different matter entirely...
why not just run a substitution, like $pattern ~~ s:g/'%' (\S+) '%'/{ foo(~$0) }/ 05:22
Heather labster what...
is it
r: say $pattern ~~ s:g/'%' (\S+) '%'/{ foo(~$0) }/
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Variable '$pattern' is not declared␤at /tmp/nTolrB5eav:1␤------> say $pattern⏏ ~~ s:g/'%' (\S+) '%'/{ foo(~$0) }/␤ expecting any of:␤ postfix␤»
Heather r: say '%a%' ~~ s:g/'%' (\S+) '%'/{ foo(~$0) }/
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Undeclared routine:␤ foo used at line 1␤␤»
Heather r: '%a%' ~~ s:g/'%' (\S+) '%'/{ say (~$0) }/ 05:23
labster you can't substitute unless it's a container, like $var... raw strings are immutable
camelia rakudo 88221c: OUTPUT«a␤Cannot modify an immutable value␤ in sub infix:<=> at src/gen/CORE.setting:12888␤ in block at /tmp/KRosAE5Diq:1␤␤»
labster tx camelia :)
Heather r: my $x = '%a%'; x ~~ s:g/'%' (\S+) '%'/{ say (~$0) }/
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Undeclared routine:␤ x used at line 1␤␤»
Heather r: my $x = '%a%'; $x ~~ s:g/'%' (\S+) '%'/{ say (~$0) }/ 05:24
camelia rakudo 88221c: OUTPUT«a␤»
Heather r: my $x = '%a% + %b%'; $x ~~ s:g/'%' (\S+) '%'/{ say (~$0) }/
camelia rakudo 88221c: OUTPUT«a␤b␤»
labster r: my $x = '%a%'; $x ~~ s:g/'%' (\S+) '%'/{ uc (~$0) }/; say $x
camelia rakudo 88221c: OUTPUT«A␤»
Heather r: my $x = '%a% + %b%'; $x ~~ s:g/'%' (\S+) '%'/{ subst(~$0, 'zzz') }/; say $x 05:25
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Undeclared routine:␤ subst used at line 1. Did you mean '&substr'?␤␤»
Heather r: my $x = '%a% + %b%'; $x ~~ s:g/'%' (\S+) '%'/{ ("-> ~$0") }/; say $x 05:26
camelia rakudo 88221c: OUTPUT«-> ~a + -> ~b␤»
Heather ah
Cool
labster thanks !
labster after the first slash, we have the pattern, which captures \S+ in $0. the second part is the string to substitute there. But we're using a block inside a string "{ do stuff() }", to make it do anything.
[Coke] have kicked off last run of stresstests, but it's not going to finish for another hour. I'll do the final push tomorrow. 05:27
Heather labster it's something awesome ) 05:31
Heather labster how to say that string myst start with xxx in regex? 05:33
labster Heather: $string ~~ /^ xxx/
Heather labster thank you! 05:34
Heather r: say dc("A") 05:51
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Undeclared routine:␤ dc used at line 1. Did you mean '&lc', '&uc'?␤␤»
Heather r: say bc("A")
camelia rakudo 88221c: OUTPUT«===SORRY!===␤Undeclared routine:␤ bc used at line 1. Did you mean '&lc', '&uc'?␤␤»
Heather r: say lc("A")
camelia rakudo 88221c: OUTPUT«a␤»
FROGGS morning 06:31
[Coke]++ 06:32
Heather I want this github.com/potyl/perl-Gtk3-WebKit/...browser.pl in perl6 ) 06:44
sorear Heather: start with github.com/sorear/niecza/blob/mast...browser.pl 06:45
Heather sorear it's CLR ... 06:46
sorear I mean I can write it 10 times faster and beautier in F# :D
Heather I know I will try servo :D 06:50
FROGGS .ask jnthn Any objections about giving nqp a bunch of install directories like nqp/vendor/site/home ? 06:55
yoleaux FROGGS: I'll pass your message to jnthn.
hoelzro haha, yoleaux?
that's hilarious
(also, good morning)
FROGGS what does it mean? 07:00
(also, good morning :o)
mst it's a pun You Only Live Once 07:01
s/pun/pun on/
FROGGS ahh, thanks mst 07:02
just thought it might be more than yolo 07:03
lizmat morning #perl6 07:44
diakopter 'marn 07:45
hoelzro morning 07:46
diakopter heh 07:54
tadzik hello 08:13
Heather tadzik hi 08:35
kresike hello all you happy perl6 people 08:37
lizmat is glad that there aren't any unhappy perl6 people around 08:39
FROGGS *g* 08:42
hi all
labster hi FROGGS
kresike FROGGS, o/ 08:44
FROGGS hi labster, kresike 08:45
lizmat rn: say (a => 1) 08:46
camelia rakudo 88221c, niecza v24-37-gf9c8fc2: OUTPUT«"a" => 1␤»
lizmat rn: say :a<1>
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Excess arguments to gistcat, unused named a␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (gistcat @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1366 (say @ 7) ␤ at /tmp/9inWGs7rd2 line 1 (mainline @ 3) ␤ at /home/p6eval/ni…
..rakudo 88221c: OUTPUT«␤»
FROGGS rn: say :a(1) 08:47
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Excess arguments to gistcat, unused named a␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (gistcat @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1366 (say @ 7) ␤ at /tmp/6cLimTmHeE line 1 (mainline @ 3) ␤ at /home/p6eval/ni…
..rakudo 88221c: OUTPUT«␤»
lizmat I feel that the error in Niecza is correct, and silentness of rakudo in this case is not helpful for beginning Perl 6 rogrammers
FROGGS hmmm
r: say :a(1).WHAT 08:48
camelia rakudo 88221c: OUTPUT«(Pair)␤»
FROGGS r: say :a(1).key
camelia rakudo 88221c: OUTPUT«a␤»
FROGGS r: say :a(1).kv
camelia rakudo 88221c: OUTPUT«a 1␤»
Heather rn: CLR::("System.Console").Write("hi"); 08:49
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: CLR objects may not be used directly in safe mode␤ at /tmp/wbJnhbz1BX line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4299 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4300 (module-CORE @ 583) ␤ at…
..rakudo 88221c: OUTPUT«===SORRY!===␤Combination of indirect name lookup and call not (yet?) allowed␤at /tmp/RIRCuVBC1p:1␤------> CLR::("System.Console")⏏.Write("hi");␤ expecting any of:␤ argument list␤»
lizmat (found this while preparing a lightning talk about Perl 5's fat comma to Perl6' pairs
Heather can't use CLR :( 08:50
FROGGS Heather: safe mode 08:51
Heather I see
but I wonder why CLR ~~ unsafe
diakopter just because it's too hard to filter out all the possibly destructive ones
Heather diakopter well, I guess not much harder than in perl 08:52
FROGGS is somebody aware of nqp's dynamic variables for runtime? I mean, are there any yet available? 08:53
arnsholt NQP has dynamic variables, yes
FROGGS arnsholt: I know that much :o)
arnsholt Not sure what you mean by "for runtime" 08:54
FROGGS nqp: say($*OS);
camelia nqp: OUTPUT«␤»
FROGGS are there any "installed" ?
nqp: say($?FILES); 08:55
camelia nqp: OUTPUT«Null PMC access in get_string()␤current instr.: '' pc 768 (src/stage2/gen/NQPCORE.setting.pir:355) (src/stage2/NQPCORE.setting:645)␤»
FROGGS nqp: say($?LINE); 08:56
camelia nqp: OUTPUT«Null PMC access in get_string()␤current instr.: '' pc 768 (src/stage2/gen/NQPCORE.setting.pir:355) (src/stage2/NQPCORE.setting:645)␤»
FROGGS nqp: say(%*COMPILING<%?OPTIONS>);
camelia nqp: OUTPUT«␤»
diakopter nqp-jvm: say($?LINE);
camelia nqp-jvm: OUTPUT«Lexical '$?LINE' not found␤ in <anon>␤ in <anon>␤ in eval␤ in evalfiles␤ in command_eval␤ in command_line␤ in MAIN␤ in <anon>␤ in <anon>»
jnthn %*COMPILING exists while things are compiling 09:19
yoleaux 06:53Z <FROGGS> jnthn: Any objections about giving nqp a bunch of install directories like nqp/vendor/site/home ?
FROGGS hmmm
FROGGS jnthn: what I want to do is make v5 installable via panda, and for that it would be handy if there was a nqp-home, which panda could install to 09:20
since nqp won't ever load modules from perl6 land 09:21
jnthn FROGGS: Is there anything that stops you installing things to the place that Rakudo itself gets installed to? 09:27
Perl6::Grammar etc. are NQP modules. 09:28
FROGGS jnthn: that is what I do currently, v5's make install does that for example 09:29
but I'm not sure how panda chooses the install path
jnthn FROGGS: Maybe see how perl6-debug gets installed? It must figure out a path not too differently from how you need to... 09:37
tadzik perl6-debug installs to a regular locatin 09:38
tadzik it's just altering the build process, not the install process 09:38
altering the install process is not currently possible with panda :(
FROGGS right: github.com/jnthn/rakudo-debugger/b...r/Build.pm
and since you use a custom binary, it is not the same like use'ing a module from perl6, which loads nqp-modules 09:39
*same as
jnthn Oh, and panda already knows where to put a binary? 09:40
FROGGS I mean, it is not ideal anyway if a module-author could tell panda that it wants to be installed in a specific location
well, it gets locally made into bin/..., and panda puts all from bin/* into home/bin for example 09:41
jnthn Well, Panda's job is to install Perl 6 modules really, so it's not so surprising installing NQP with it doesn't quite work out...
FROGGS ya, but I guess we'll need that anyway 09:42
tadzik wouldn't v5 want to be a core-core thing eventually?
FROGGS there will be other modules which are directly ontop nqp
tadzik: no idea
tadzik as the last resort we can make panda respect Makefiles again
it used to do that. "If there's a makefile, 'make test install' and we're done here" 09:43
FROGGS possible, but not ideal
tadzik aye 09:44
FROGGS I'd like to see that if I'm not rootish, that panda installs modules to [nqp|perl]/home
tadzik explain: [nqp|perl]/home?
lizmat thinks that v5 is a core core thing now already :-) 09:44
FROGGS but I dont know how to tell panda that a module (a file) is an nqp module rather than an perl6 one
tadzik there's no way to tell panda that 09:45
but!
you could adjust CUSTOM_LIB somehow
that's what panda uses as an install location
but how to do that from inside the installation process, I don't know
FROGGS subclassing?
tadzik subclassing what? 09:46
FROGGS the installer class, and overwrite method install, set custom lib, call SUPER...
moritz EVERYTHING
tadzik that's one of the ideas
FROGGS *override
tadzik that's how Build.pm works I think
FROGGS ya
jnthn Well, a .nqp file extension is a good "this is NQP" hint :) 09:47
tadzik "is there a Build.pm? Is it a Panda::Builder? Good, load it and use it instead"
haha
FROGGS and having nqp recognizing nqp/vendor/site/home is not a big deal
tadzik I'd like to ship panda2 soon 09:47
jnthn I've already changed all the NQP files in the jvm-support branch to be Foo.nqp.
lizmat nr: say $_:p for 11..20
camelia niecza v24-37-gf9c8fc2: OUTPUT«===SORRY!===␤␤Variable $_:p is not predeclared at /tmp/zm6ZJV_zQt line 1:␤------> say ⏏$_:p for 11..20␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p…
..rakudo 88221c: OUTPUT«11␤12␤13␤14␤15␤16␤17␤18␤19␤20␤»
FROGGS jnthn: hmmm, but it compiles to bpc anyway, no? 09:48
*pbc
tadzik it's not complete, but maybe someone can either fix the remaining bits, or at least plan ahead for a new architecture
(or tell me that the new one is very wrong)
it's a least a lot less painful to reuse
lizmat nr: say $_ :p for 11..20
camelia rakudo 88221c: OUTPUT«===SORRY!===␤You can't adverb that␤at /tmp/fAmrvfd7eE:1␤------> say $_ :p⏏ for 11..20␤ expecting any of:␤ pair value␤»
..niecza v24-37-gf9c8fc2: OUTPUT«===SORRY!===␤␤You can't adverb that at /tmp/MZUDT_IvnT line 1:␤------> say $_ :p⏏ for 11..20␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p6eval/niec…
FROGGS tadzik: is there a roadmap, or at least some doc that tells what's missing?
tadzik FROGGS: no, there's just a 90%-working code on my laptop 09:48
FROGGS k
tadzik I think only the bootstrap is not working 09:49
or maybe the installer itself too ;)
FROGGS tadzik: just spit it out to github :o)
and lizmat, don't use smilies in code :P
jnthn FROGGS: Yes, true. 09:50
FROGGS jnthn: and I installed these, not the nqp's
jnthn FROGGS: May want both so the .pbc can be re-created in the future, just as we do with Perl 6 modules 09:51
FROGGS jnthn: ya, good point
so if panda would decide about the install location depending of the filename *before* compilation, it would work 09:52
or we have to create special module-subdirs like nqp-blib and perl6-blib :o) 09:53
lizmat suddently notices that S02 talks about "fatarrow" rather than "fat comma" 10:20
fat comma seems to be a more common term than fatarrow. 10:21
Does anybody know why TimToady used "fatarrow" rather than "fat comma" ?
moritz it's not a comma anymore in Perl 6 10:22
FROGGS r: my %h = 1, 2; say %h
camelia rakudo 88221c: OUTPUT«("1" => 2).hash␤»
FROGGS r: my %h = 1 => 2; say %h
camelia rakudo 88221c: OUTPUT«("1" => 2).hash␤»
FROGGS looks like it is still commaish 10:23
at least when assigning to a hash
Timbus :I thats because it is a hash. which contains pairs 10:24
moritz but not in other contexts
foo(a => 1) is totally different from foo('a', 1) 10:25
FROGGS well, that is the same in perl5
you need to assign %var = @_ to get a hash
moritz in p5 it's the same, but in p6 it's totally different 10:26
and afaict lizmat asked about S02
lizmat indeed… this is not about functionality, but nomenclature
so: consensus is that "fat comma" is Perl 5, whereas "fatarrow" is Perl 6? 10:27
jnthn The semantics/meaning are different, and imho highlighting that through a different name is a good idea.
tadzik I think fat arrow makes more sense, because fat comma... isn't a comma
FROGGS and => doesnt look like a comma at all :o) 10:28
tadzik my point exactly :)
huf it does if you're high on perl5 :)
lizmat well, that's what "=>" is called in Perl 5
huf but yeah, dont call it that ;)
tadzik .u comma
yoleaux U+002C COMMA [Po] (,)
lizmat .u arrow 10:30
yoleaux U+02C2 MODIFIER LETTER LEFT ARROWHEAD [Sk] (˂)
U+02C3 MODIFIER LETTER RIGHT ARROWHEAD [Sk] (˃)
U+02C4 MODIFIER LETTER UP ARROWHEAD [Sk] (˄)
lizmat no fatarrow yet
tadzik .u front-facing baby chick 10:31
yoleaux U+1F425 FRONT-FACING BABY CHICK [So] (🐥)
huf dammit, not another one
lizmat .u camel
yoleaux U+1F42A DROMEDARY CAMEL [So] (🐪)
U+1F42B BACTRIAN CAMEL [So] (🐫)
tadzik .u side-facing baby chick
yoleaux No characters found
lizmat no camelia :-(
tadzik .u back-facing baby chick 10:32
yoleaux No characters found
diakopter .u onion
tadzik stupid. Why is there front-facing then
yoleaux No characters found
tadzik perhaps PILE OF POO was more necessary
FROGGS hehe 10:34
tadzik I mean, onions? Eww, gross. 10:36
FROGGS .u Peril 10:41
yoleaux No characters found
FROGGS .u Geek
yoleaux No characters found
FROGGS had expected 'U+04FC OILY HAIRED FAT GEEK [Po]' 10:42
.u chick 10:43
yoleaux U+1F414 CHICKEN [So] (🐔)
U+1F423 HATCHING CHICK [So] (🐣)
U+1F424 BABY CHICK [So] (🐤)
FROGGS .ask TimToady Should the grammar switch to Peril5 too if it detects a 'use 5.10', rather just for a 'use v5' ? 11:54
yoleaux FROGGS: I'll pass your message to TimToady.
nwc10 Peril 5 :-) 11:57
mst dons his Peril Sensitive Sunglasses 11:59
lizmat it's just the German's way of pronouncing Perl 12:00
timotimo personally, i pronounce it "pörl" 12:01
lizmat says purl 12:02
which in Dutch is the same as pörl in German (or close to it, anyway)
timotimo i'm convinced they end up sounding the same, yeah 12:03
Patterner germans shout "PÖRL!" 12:12
lizmat unless they live in Perl, we found out
timotimo is there documentation on how to manipulate stage0? 12:14
moritz make bootstrap-files 12:16
timotimo is it that easy? :)
lizmat rn: my $a= 'foo'; my $b= :$a; say $b.WHAT
camelia rakudo 88221c, niecza v24-37-gf9c8fc2: OUTPUT«(Pair)␤»
timotimo moritz: either i'm not doing it right or i miscommunicated what i wanted to do 12:17
lizmat rn: my $a= 'foo'; my $b= :$a; say $b
camelia rakudo 88221c, niecza v24-37-gf9c8fc2: OUTPUT«"a" => "foo"␤»
timotimo in my source code for nqp i removed the method versions of the QRPA push and friends, but it's still being used in the stage0
i suppose i'll have to put the methods back, compile a full nqp and then recreate the bootstrap files to not use the method forms any more?
jnthn Correct. 12:18
moritz unless you want to edit several dozen thounds lines of PIR code :-)
timotimo i would prefer not to have to do that :) 12:19
jnthn Another thing maybe worth checking: that you can build Rakudo on Parrot with the latest jvm-support branch. It works for me; I'd apprecaite a check of it from a non-Windows platform. 12:21
nwc10 NQP revision 2013.04 required (currently 2013.03-77-g635e421). 12:26
last I checked earlier today
jnthn grr
nwc10 IIRC before that the spectest was exploding quite spectacularly
moritz echo 2013.03-77-g635e421 > nqp/tools/build/NQP_REVISION # to shut up the version check 12:27
nwc10 I didn't say - I assumed it was in hand, or a spurious thing to say
jnthn Odd, I got a working Rakudo out of it.
nwc10 me repeats 12:28
^/
jnthn I guess an NQP release has been tagged now, though, so I'm keen to merge it soon.
[Coke] tagged. need privs to upload to rakudo.org 12:28
jnthn (Though gotta do $dayjob stuff for the next hours...)
nwc10 moritz: it's not that
moritz [Coke]: need a public SSH key for that
nwc10 hell Pm 12:29
moritz [Coke]: I can give you access when I know your pubkey
nwc10 moritz: s!nqp/!
gah
moritz: s!nqp/!! I think
nwc10 blames the network 12:30
moritz nwc10: depends on your directory layout
nwc10 it can't be the beer :-)
[Coke] moritz++ 12:34
release files uploaded.
pmichaud Coke++ 12:36
good morning, #perl6
moritz good am, pm
jnthn o/ pmichaud
pmichaud: On the .push thing, NQP on JVM actually supports it by declaring an NQPArray type with the VMArray REPR. 12:37
pmichaud: It has a method push($x) { nqp::push(self, $x) }
dalek kudo/nom: de2080d | pmichaud++ | docs/release_guide.pod:
Propose "Austin" for the 2013.05 release name.
jnthn pmichaud: So no nested runloop.
pmichaud: Anyway, that's why I didn't end up removing all of them in the porting work. 12:38
[Coke] email sent. 12:38
(ah, much easier to *announce* compiler releases, anyway.) 12:39
pmichaud jnthn: no problem. do you agree that we'd still be better of with nqp::push for the vm-level objects, though? 12:40
[Coke] snickers at his tired self from last night.
"* various optimizations in the optimizer and the runtime
optimizations!? in the OPTIMIZER!? WHAT IS GOING ON!?
pmichaud in other words, the note about optimizations to the optimizer is sub-optimal. 12:41
:-)
jnthn pmichaud: Well, it'll be faster to use nqp::push
pmichaud jnthn: right.
jnthn pmichaud: Other thing we need to be careful with these days: the nqp::push_[sin]
pmichaud and in parrot's case, a lot faster :-)
jnthn pmichaud: Parrot's QRPA has coercive semantics. 12:42
pmichaud: 6model semantics are constraining. That is, you can do nqp::push_s on an nqp::list_s, not on an nqp::list.
pmichaud I don't think there's much code where we use push/unshift to coerce a native value. 12:43
most of the stuff written with nqp::push pre-dates the natives work
moritz
.oO( push'in now is a [sin] )
jnthn pmichaud: Yeah, it's not a huge issue, just one of those places where you can write code that works on the Parrot backend at the moment, but that will fail in JVM land. 12:44
There's plenty of stuff you can do the other way round, mind...
nqp: say(42.HOW.name(42))
camelia nqp: OUTPUT«Can only use get_how on a SixModelObject␤current instr.: '' pc 43 ((file unknown):40) (/tmp/DsDfb0thPj:1)␤»
jnthn nqp-jvm: say(42.HOW.name(42))
camelia nqp-jvm: OUTPUT«BOOTInt␤»
jnthn :)
pmichaud eventually I suspect we'll be able to tighten these things up a fair bit.
jnthn *nod*
nwc10 jnthn: pasta.test-smoke.org/484 12:45
jnthn wtf...
moritz [Coke]++ # rakudo release 12:46
jnthn nwc10: Thanks, will see if I can reproduce this evening :)
nwc10 jnthn: exmaple pasta.test-smoke.org/485 12:47
pmichaud I'm building nqp-parrot from the jvm branch now, will test rakudo on it in a bit. 12:48
(have to wait to clone the gargantuan parrot repo first)
nwc10 jnthn: also pasta.test-smoke.org/486 12:48
surely you already have a clone of that? 12:49
jnthn nwc10: Curious. I don't have an immediate guess of what that's about...
nwc10 me neither :-) 12:50
tadzik oh, apparently I contributed to Rakudo this month :)
pmichaud oh, apparently I didn't. :) 12:51
jnthn Both Carl Masak *and* Carl Mäsak contributed too... :) 12:51
lizmat thinks an honorary mention would be in place
jnthn
.oO( But we all know they's a bunch of clones... )
pmichaud perhaps masak++ has also been upgraded to an entangled form of existence? 12:52
bbiab 12:53
moritz of six clones one two have contributed to rakudo? :-) 12:55
diakopter read that as clowns
colomon far more than two clowns have contributed to rakudo! 12:56
lizmat some of them even twitter 12:57
pmichaud pmiclown, clownomon, carlown masak, moritz clownenz, and jnthn
moritz I mean masak clones, of course :-) 12:59
lizmat n: my %foo = (:a<1>,:b<2>); say %foo.val("a")
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Unable to resolve method val in type Hash␤ at /tmp/otiOJ1cOfB line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4299 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4300 (module-CORE @ 583) ␤ at /home/p…
lizmat wonders what %foo<a> desugars to for instructional purposes 13:00
moritz at_key in rakudo
pmichaud %foo<a> is %foo.{'a'}
jnthn %foo.postcircumfix:<{ }>('a')
lizmat thanks! 13:01
moritz and for a single key, the postcircumfix then re-dispatches to at_key
[Coke] jnthn: we need to update the toolto fold carl back into a single person. 13:02
... because who is in their right mind when cutting a release to check that. Grumblefutz. 13:03
pmichaud release_guide, item 5: Please check the result manually for duplicates and other errors.
pmichaud :-P 13:04
lizmat for instructional purposes, I was hoping this would have worked: 13:04
rn: my %foo = (:a<1>,:b<2>); say %foo.at_key("a", p => 1)
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Excess arguments to Hash.at_key, unused named p␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (Hash.at_key @ 1) ␤ at /tmp/DRhg3WkuPS line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4299 (ANON @ 3) ␤ at /home/…
..rakudo 88221c: OUTPUT«1␤»
[Coke] pmichaud: in my defense, I feel asleep accidentally at least twice while trying to cut the release last night. :| 13:05
lizmat I think I'll use that nonetheles, for instructional purposes (lightning talk about adverbial pairs coming up at DPW)
jnthn lizmat: :p is an argument to postcircumfix:<{ }>. at_key is specifically for getting *one* key.
lizmat I get that
jnthn lizmat: The idea is that if you're implementing your own hashish types, you only need implement at_key 13:06
lizmat but for Perl 5 people, postcircumfix:<{ }> is very scary
aha, ok, good to know
timotimo so many invocations to replace ... :|
moritz hashish types? :-)
lizmat plenty of those in Amsterdam :-)
jnthn *lol*
lizmat and don't even start about hashish pipes!
jnthn Didn't mean *that* :P
lizmat: But it's very general syntax 13:07
jnthn lizmat: But yeah, you don't normally ever write that. 13:07
LylePerl_ hi 13:11
lizmat agree, but Perl 5 people need to hang something to hang new concepts on
LylePerl_ Some of you are probably aware that jnthn and masak come to Bristol last year to give a Perl6 talk and workshop 13:12
timotimo hang some hooks on the wall
LylePerl_ this year our event clashes with YAPC::NA
and so masak and jnthn aren't available
would anyone else be interested? 13:13
details of the event are here: bristol.itmegameet.co.uk
timotimo oh, damnit 13:17
lizmat alas, I have the same clash :-( otherwise I certainly would have come, potentiallly even to give a (beginners) Perl 6 presentation
timotimo i made some change that causes rakudo to use up all of my ram when trying to parse the core setting
and i can't tell what's wrong >_<
my diff looks sane, i.E. never turned a shift into an unshift by accident or anything like that
jnthn had a nice time at LylePerl_++'s event last year. :) 13:18
pmichaud might be caught in an exception loop
timotimo oh, damn, that could be
how do i figure out such a thing? :(
jnthn timotimo: Try --ll-exception
timotimo will that trigger a backtrace if perl6 gets killed by oom? :| 13:20
moritz no, but the reason might be an exception from within the exception handler
also, exceeding a certain stack frame depth causes an exception 13:21
and then the exception handler needs a call frame
causing parrot to throw the original exception again
LylePerl_ we had an open bar after the event last year ;) 13:22
pmichaud gist.github.com/pmichaud/5420322 # jnthn, I also get large number of spectest fails when using jvm-support version of Rakudo on Parrot 13:23
jnthn pmichaud: Looks like same set nwc10++ got, give or take 13:24
pmichaud likely.
diakopter what do you guys think about 2 alc drink tickets per person at the yapc banquet - and those who want to drink more can get tickets from those who have extra 13:26
[seems to work well at some business events I've been to] 13:27
timotimo jnthn: i'm at 1gb of parrot in my swap now, when will i actually get an exception? >_< 13:28
pmichaud timotimo: I said it *might* be an exception loop. :-) 13:29
diakopter er, to be clear, "guys" means guy-ettes, too 13:30
jnthn diakopter: Depends, is the alc all like Bud Light? :P
.oO( Sir, my water tastes funny... )
timotimo that's quite a few that fail, but amazingly also a bunch that seem to pass :D 13:31
diakopter www.utexas.edu/cee/tcc/downloadzone...ervice.pdf
all our drinks are $6, so includes cocktails, it seems
note the "premium" beer
timotimo is list.push(%foo) somehow special for rpa?
jnthn oh ffs Bud Lite actually is on the list :P
I WAS JOKING!!
diakopter don't worry, I'll bring some Natural Light for you
diakopter nudges pmichaud 13:32
jnthn diakopter: Anyway, it sounds sensible.
diakopter Natural Light is several steps below Bud Light
moritz agrees
pmichaud diakopter: am I being nudged about the alc question? I don't know if it'll work or not; I tend to be one that doesn't always use my tickets. :) 13:33
diakopter unfortunately we are clearly prohibited from bringing ANY outside food/drink...
to that venue
pmichaud: no, the Natural Light reference :)
diakopter has never used a ticket... 13:34
pmichaud: I was just hoping you could back me up on Nat Light's reputation, even if you hadn't had it yourself
pmichaud diakopter: I don't think I'd ever heard of Natural Light, tbh. But sure, I suspect its reputation isn't very high. 13:35
diakopter hahaa
LylePerl_ timotimo: thanks for considering it 13:36
pmichaud 13:31 <timotimo> is list.push(%foo) somehow special for rpa?
shouldn't be special, should be the same as nqp::push(list, %foo) 13:37
just be sure that 'list' is really an RPA. :-)
(or QRPA)
timotimo hm. my @params := $block<placeholder_sig> || ($block<placeholder_sig> := []); - seems so 13:45
nwc10 diakopter: suggest asking the Frankfurt lot what they budgeted, as the paid-for didn't *quite* run out, but it wasn't beer 13:47
diakopter this is cocktails, beer, and wine
nwc10 that was apfelwein
colomon cider! 13:49
colomon starts fantasizing about cracking open that bottle of JK's Scrumpy....
timotimo --trace doesn't work, does it? 13:50
probably since the problem occurs before parsing is even finished, so i'd need --meta-trace
nwc10 aye, cidah!
nwc10 finds that menu troubing. Heineken has positioned itself as "premium" 13:59
jnthn nwc10: I think premium here means "in a blind test, you would not confuse it with water" :P 14:00
diakopter yeah, probably
nwc10 mmm yes, I would not fonuse Heineken with water. 14:01
I think I'd prefer the water.
assuming that it's potable.
nwc10 (we're back to that one) 14:02
while we're flogging dead memes, lunch today had tentacles. Chance of contamination with horsemeat - pretty much zilch :-)
sergot Hi! o/ 14:03
Is there any way to fix "failed to load libpcre" problem?
moritz configure parrot --without-pcre
sergot Other than --without-pcre (or somelike)
:)
moritz fix pcre lookup in parrot :-) 14:04
sergot Oh.. That's not for me I think. :(
PerlJam sergot: why do you want pcre?
sergot PerlJam: What is it really for? :P
PerlJam: I just don't wanted to have any errors like that. :P 14:05
moritz I think deinstalling pcre headers could also make the error go away 14:06
PerlJam then --without-pcre gets rid of that error :)
PerlJam sergot: but pcre isn't needed for NQP or Rakudo to my knowledge, which is why I was asking. 14:06
sergot PerlJam++ 14:07
moritz++
thanks! :)
kresike bye folks 14:17
PerlJam never sees kresike say anything other than "bye" 14:20
moritz PerlJam: there's a roughly corresponding number of "hello"s :-)
PerlJam moritz: any other correspondence? 14:21
moritz a bit
daxim I switched my unicode dingus from fileformat.info to codepoints.net, example page: codepoints.net/U+1F63B 14:22
tell me what you think about it
moritz kinda likes it 14:23
pmichaud daxim: first glance -- looks very nice
lots of good detail there
moritz just misses "follow this character on twitter", and "people who liked this character also liked ..." :-) 14:27
daxim "also mentioned on G+ Fake Unicode Consortium" 14:31
lizmat just finished two lightning talks at DPW: "Fat Comma to Adverbial Pairs" and "Is Perl 6 'Perl'?" 14:42
PerlJam lizmat: Is Perl 6 'Perl'? ;) 14:49
moritz PerlJam: any(True, False)
lizmat True
moritz nr: say so any(True, False)
camelia rakudo de2080, niecza v24-37-gf9c8fc2: OUTPUT«True␤» 14:50
moritz lizmat, rakudo, niecza and me agree :-)
lizmat BTW, this also goes for Perl 4..5
not sure whether Perl 1..3 would be considered Perl nowadays
[Coke] r: say "{1..6.pick} is the best perl." 14:55
camelia rakudo de2080: OUTPUT«1 2 3 4 5 6 is the best perl.␤»
[Coke] r: say "{1..6.pick(1)} is the best perl."
camelia rakudo de2080: OUTPUT«1 2 3 4 5 6 is the best perl.␤»
PerlJam heh
[Coke] O_o.
geekosaur parens?
lizmat rn: say "{(1..6).pick} is the best perl." 14:58
camelia niecza v24-37-gf9c8fc2: OUTPUT«1 is the best perl.␤»
..rakudo de2080: OUTPUT«3 is the best perl.␤»
lizmat submits bug
:-)
n: say "{1..6.pick} is the best perl." 14:59
camelia niecza v24-37-gf9c8fc2: OUTPUT«1 2 3 4 5 6 is the best perl.␤»
lizmat so why are the parens needed? 15:00
PerlJam to listify the range looks like 15:03
(or is that parcelify?)
[Coke] oh, it's picking -the range- and then stringifying the whole thing. 15:04
duh. perljam++
PerlJam yep
[Coke] r: say "{1..6.flat.pick(1)} is the best perl."
camelia rakudo de2080: OUTPUT«1 2 3 4 5 6 is the best perl.␤»
[Coke] r: say "{1..6.list.pick(1)} is the best perl."
camelia rakudo de2080: OUTPUT«1 2 3 4 5 6 is the best perl.␤»
[Coke] now explain that. ;P
PerlJam flat may suffer the same problem with ranges.
r: say (1..6.flat).WHAT; 15:05
camelia rakudo de2080: OUTPUT«(Range)␤»
jnthn 1..6.pick is 1 .. 6.pick 15:06
PerlJam oh, precedence 15:07
jnthn Method calls bind tight.
PerlJam indeed
jnthn I find it helpful to always imagine infixes as having spaces around them, even if I don't actually type them.
PerlJam jnthn++ 15:08
r: say 1..6.flat 15:09
camelia rakudo de2080: OUTPUT«1..1␤»
PerlJam I don't get that one.
Nom- Alrighty... back to seeing what I can do about this rbx nqp compiler :)
PerlJam r: say 1 .. 6.flat
camelia rakudo de2080: OUTPUT«1..1␤»
PerlJam r: say 6.flat
camelia rakudo de2080: OUTPUT«6␤»
flussence r: say 6.flat.perl 15:11
camelia rakudo de2080: OUTPUT«(6,).list␤»
flussence it's effectively 1..(+[6])
PerlJam ah. interesting. 15:13
flussence that one trips me up a lot... maybe the parser should warn if it sees .. and . consecutively with no punctuation inbetween 15:14
lizmat something that just tripped me up: 15:15
rn: (say (1..6).pick), " is the best perl."
camelia rakudo de2080: OUTPUT«2␤»
..niecza v24-37-gf9c8fc2: OUTPUT«5␤»
lizmat just a parentheses open at the wrong place, but no warning that there is a value that just drops on the floor 15:16
daxim .u bottle
yoleaux U+1F376 SAKE BOTTLE AND CUP [So] (🍶)
U+1F37C BABY BOTTLE [So] (🍼)
daxim needs moar coke bottle 15:17
lizmat .u coke
yoleaux No characters found
daxim .u cola
yoleaux No characters found
daxim
www.catb.org/~esr/jargon/html/C/cokebottle.html 15:18
lizmat true, having a character for it, would make it a commodity, and thus useless as a trademark
.u camel
yoleaux U+1F42A DROMEDARY CAMEL [So] (🐪)
U+1F42B BACTRIAN CAMEL [So] (🐫)
lizmat too bad for O'Reilly :-)
daxim PHOTOSHOPPED MONKEY 15:19
Nom- So correct me if I'm wrong here... NQP compiles into QAST... which the JVM modifications then convert into a string stream which can be parsed by a Java class to create bytecode? 15:23
PerlJam "string stream"? 15:24
Nom- Well, yes... found the java class which parses input which appears to be simple strings... things like "+ class" 15:25
colomon r: say (0..10)[(1, 2, 3, 5, 6, 7).list.item]
camelia rakudo de2080: OUTPUT«6␤»
colomon r: say (0..10)[(1, 2, 3, 7).list.item]
camelia rakudo de2080: OUTPUT«4␤»
PerlJam Nom-: that doesn't jibe with my mental model, but then I also haven't looked into the details of how things work either :) 15:26
Nom- fair enough :) 15:27
Nom- It looks like, based on what I've toyed with so far, I can pretty easily start creating an AST in rbx using the Rubinius::Compiler classes 15:28
How closely things match is a bit unknown... the compilers are very Ruby centric, but it might be 'close enough'
masak oh hai, #perl6 15:29
hoelzro ahoy masak 15:30
PerlJam greetings masak \o/
masak Juerd++ # photoshopped monkey
so, I'm kinda stoked about github.com/perl6/mu/wiki/perl6-wor...p-may-2013 -- I hope y'all are, too.
I want to spend parts of the weekend drawing up a more exact agenda for the thing. 15:31
PerlJam photoshopmonkeys.com/
hoelzro I'm pretty excited
what kind of stuff will be covered?
masak hoelzro: things I know how to explain about programming that I consider worth knowing, filtered through the lens of Perl 6 / Perl 5. 15:32
hoelzro hmm
masak I expect there to be interesting, deep discussion during those hours, of concepts that we all care about but don't often put into words. 15:33
jnthn Nom-: It's a textual format 'cus text is the easiest thing to throw over a boundary. When I have time, I'll probably get rid of it.
masak kind of like a live c2.com
jnthn Nom-: During the cross-comp phase you have a process boundary. Once you self host, that goes away 15:34
Nom- jnthn: Yeah, figured... I wouldn't mind seeing the raw output to see if it's close enough to what I can parse, but i'll be damned if I can work out something that can do that ... scouring makefiles now 15:35
jnthn Nom-: Yeah, I didn't want the epic hassle of generating the bytecode format itself. 15:37
Nom- Is there anything still in that branch which will generate the output which runtime/.../JASTToJVMByecode.java ? 15:39
masak hoelzro: could you elaborate a bit on your "hmm"? it felt like a bit of a drop from "I'm pretty excited". :P 15:40
timotimo here come some embarassing commits 15:41
dalek p/qrpa_experiment: af36e92 | (Timo Paulssen)++ | src/ (12 files):
search&replace nqp::list -> nqp::qlist; change .push to nqp::push etc
nqp/qrpa_experiment: 491eec9 | (Timo Paulssen)++ | src/ (6 files):
nqp/qrpa_experiment: so many search and replace. probably too many.
hoelzro well, I'm just hoping to learn some of the neat Perl 6 tricks I occasionally see here =)
timotimo oh, saved by the bell :D
PerlJam masak: I'm pretty excited!
masak: hmm!
;-)
hoelzro masak: an example of a "thing you know about programming" would be welcome =)
Juerd masak: :)
FROGGS timotimo: does it still compile? 15:43
masak hoelzro: <masak> I want to spend parts of the weekend drawing up a more exact agenda for the thing.
FROGGS (and work?)
masak hoelzro: could you maybe ask me on Sunday? :)
timotimo FROGGS: nqp compiles, perl6 compiles, but it can't compile the setting (it will spend all its time eating up memory before parsing has finished)
hoelzro if I remember =)
masak hoelzro: I have a pretty good idea myself, but I don't want to pre-commit on a bad representation of anything.
hoelzro fair enough
timotimo if you're able to figure out where it infiniloops, that would be splendid.
timotimo and some time in the future, i shall squash all those useless revert commits and friends 15:43
FROGGS timotimo: need to get 'require' done for v5 :/ 15:44
timotimo sure.
masak that said, I'm really happy/flattered that 11 people have signed up just to come to hear me blabber about something unannounced. 15:51
if there's something you'd *like* to hear about/discuss/go through, it's definitely not too late to request stuff.
when I teach, I tend to do less of features/syntax, and more of concepts. or rather, the concepts drive the features/syntax. 15:52
lizmat maybe there is a set of Synopses that all people should be familiar with, other than (All) ? 15:53
PerlJam masak: Could you do a complete GEB-esque treament of programming? ;) 15:54
masak lizmat: I'd say so. S01-S06, and S12. basically all the ones with Apocalypses.
lizmat ack 15:55
masak PerlJam: even though I have no idea what you mean, that sounds like a nice goal for the day. :)
masak PerlJam: are we meant to dress up as Greek demigods and turtles? 15:55
PerlJam that's okay, I'm not sure what I mean either ;)
masak or will we alternate between academic treatment of recursion and silly prose? 15:56
PerlJam isn't that what we do everyday anyway?
lizmat dinner at DPW&
masak .oO( with Perl 6, every day is Christmas. with #perl6, every day is GEB. ) 15:56
[Coke] .u lizmat 15:58
yoleaux No characters found
masak .u les_miserables 15:59
yoleaux U+005F LOW LINE [Pc] (_)
U+0061 LATIN SMALL LETTER A [Ll] (a)
U+0062 LATIN SMALL LETTER B [Ll] (b) 16:00
[Coke] .meow 16:05
yoleaux edgecats.net/cats/catselfscare.gif
masak the bottle noise was an inside job! 16:10
Nom- blah i'm lost... will look again in the morning :( 16:44
pmichaud timotimo: now that I've had a chance to see the diff, I'm not so sure I want to switch all $rpa.push(...) to be nqp::push($rpa, ...), at least not in the NQP sources. It doesn't read as well. 16:49
pmichaud I think I might live with the .push overhead for non-hotpaths, if only to make the code more p6-ish 16:50
[Coke] r: raw.github.com/coke/p6-lingua-en-s...yllable.pm 17:08
camelia rakudo de2080: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/4Sfco1fWCD:1␤------> https⏏://raw.github.com/coke/p6-lingua-en-syll␤ expecting any of:␤ argument list␤ postfix␤ infix stopper␤ infix or meta-infix␤ st…
[Coke] (should we allow that to work?)
FROGGS why not? 17:15
moritz maybe we should allow any URLs that return text/plain content type 17:22
FROGGS and do not exceed X MB 17:27
[Coke] that seems dangerously reasonable.
dalek : 7ae7577 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
steal `require`-mechanism from rakudo
17:36
labster r: class Foo::One { }; my class Foo is Foo::One { }; class Foo::Two is Foo::One { }; 18:28
camelia rakudo de2080: OUTPUT«===SORRY!===␤'Foo::Two' cannot inherit from 'Foo::One' because it is unknown.␤at /tmp/TZ9M7g2JKq:1␤------> ␤»
labster is that supposed to happen?
moritz no. 18:30
colomon r: class Foo::One { }; class Foo::Two is Foo::One { }; 18:34
camelia rakudo de2080: ( no output )
colomon oooo, funky 18:34
masak what's so oooo funky? :) 18:35
sorear injecting Foo as a class makes rakudo lose the package of the same name 18:36
colomon that. though I hadn't figured out what was going on with it. sorear++ 18:37
[Coke] r: class Foo::One { }; class Foo is Foo::One { }; class Foo::Two is Foo::One { };
camelia rakudo de2080: ( no output )
masak <hoelzro> well, I'm just hoping to learn some of the neat Perl 6 tricks I occasionally see here =)
[Coke] even better, it's the 'my'.
masak hoelzro: oh, I wouldn't rule that out.
hoelzro: "things I know about programming" is fairly tied to "some of the neat Perl 6 tricks you occasionally see here" :) 18:38
colomon oh, does that actually make sense?
n: class Foo::One { }; my class Foo is Foo::One { }; class Foo::Two is Foo::One { };
camelia niecza v24-37-gf9c8fc2: OUTPUT«===SORRY!===␤␤Unhandled trait Foo::One for this context at /tmp/FN3PuYDsfX line 1:␤------> oo::One { }; class Foo::Two is Foo::One ⏏{ };␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting…
colomon n: class Foo::One { }; class Foo is Foo::One { }; class Foo::Two is Foo::One { };
camelia niecza v24-37-gf9c8fc2: ( no output )
colomon it's not a bug, it's a feature.
[Coke]++ 18:39
masak hoelzro: it's just that... even Perl 6, cool as it is, is just a substrate. I program in order to make the computer do useful things. and I wouldn't attend an evening course in French "hoping to hear some of the cool French sounds I occasionally hear". :) I would do so in order to be able to talk to the waiter in his own tongue. 18:46
if you see what I mean.
labster Yeah, it's definitely the 'my'. I was trying to figure out why src/core/IO.pm doesn't compile with perl -c, when the setting builds (ignoring the stub classes, of course).
r: gist.github.com/labster/5417609 18:48
camelia rakudo de2080: OUTPUT«===SORRY!===␤Unable to parse class definition␤at /tmp/ENJUV2W32I:7␤------> my class IO::Path is Cool does IO::⏏FileTestable {␤ expecting any of:␤ scoped declarator␤ generic role␤»
labster the same bug happens in reverse here.
colomon n: gist.github.com/labster/5417609 18:51
camelia niecza v24-37-gf9c8fc2: OUTPUT«===SORRY!===␤␤Pathed definitions require our scope at /tmp/VlaOJPYP_u line 1:␤------> my role IO::FileTestable ⏏{ }␤␤Pathed definitions require our scope at /tmp/VlaOJPYP_u line 3:␤------> class IO::Handle does IO…
masak heh, "Pathed".
japhb_ .tell daxim re: irclog.perlgeek.de/perl6/2013-04-19#i_6723489 , it's strange that when you mouse over a small glyph image, it zooms without improving resolution (I end up looking at a blur of large pixels). Otherwise, very nice indeed. 18:55
yoleaux japhb_: I'll pass your message to daxim.
dalek : 69273e3 | (Tobias Leich)++ | Makefile:
added "make uninstall"
19:02
: 565c60d | (Tobias Leich)++ | STATUS.md:
updated STATUS, the QRPA changes can be ignored
jnthn If you declare a "my" something it hides anything we may otherwise find via GLOBAL 19:14
FROGGS but is something inherit from a global, even if this something is 'my', this should not affect the global, right? 19:15
jnthn I'm not exactly sure what's happening, or if it's right, but... 19:16
class Foo::One { }; my class Foo is Foo::One { };
The first one installs a GLOBAL::Foo::One for sure.
FROGGS r: class A { }; my class B is A { }; class C is A { } 19:17
camelia rakudo de2080: ( no output )
jnthn The second installs a lexical Foo that hides it
FROGGS hmmm, so it is about nested namespaces too
so it might be right
jnthn When you write Foo::One it really means find Foo, .WHO it to get the Stash associated with it, then look up One in that hash. 19:18
FROGGS labster: so doing a my class IO::Path will hide an global IO
jnthn I originally didn't support my Foo::Bar { } at all as I could see the can of worms it could open, as if packages needed any more of those. Then somebody (probably moritz++) quickly pointed out to me that if I did that, we couldn't declare things like my X::OhNoes { } in CORE.setting. 19:21
So we got it, and the semantics are that if there's no lexical X already in the declaring scope, it creates one, and then the rest of the path has the usual semantics. 19:23
FROGGS note to myself: perl5 has an implicit namespace if there is no package declarator 19:38
so if it require-s a script, I just have to put its symbols into main::, and lookup these by default 19:41
diakopter posted my first ever significant-size thing on perlmonks: perlmonks.org/?node_id=1029571 19:55
masak diakopter: I enjoyed reading that. thanks. 20:00
Tene diakopter: "flood algorithm"? 20:06
masak Tene: many people redundantly solving the same problem in different ways. 20:08
a kind of Darwinian problem solving.
Tene Ah.
masak I think "flood fill" comes from a TimToady SoTO.
the one with cards describing various Perl 6 people.
flussence I've heard the term "rampaging horde syndrome" used in a similar way... :D 20:09
PerlJam with the CPAN way you get the benefit of darwinism, but you also suffer the tyranny of choice. 20:11
it all goes back to the waterbed theory of "language" complexity :) 20:12
masak my personal solution to the tyranny of choice is: listen to people who know shit. in IRC channels and at conferences. use the modules they enthuse about. 20:13
after I started doing that, things got *much* simpler.
but it essentially means that in order to use CPAN well, one has to be a Perl community member.
PerlJam a very active member
your work-a-day coder who doesn't hang out on IRC is at a severe disadvantage. 20:14
masak yes. they'd have to, like, take a Perl course at Edument or something.
(where I distil the required knowledge and sell it to them at a reasonable price.)
I dunno, I find that a fair deal. 20:15
(by which I think I mean, on balance I prefer the darwinism and the tyranny of choice to pre-chewed and pre-vetted) 20:17
PerlJam yeah, I'm pretty much the same.
what would be nice would be some sort of community best practices (I don't know if I mentioned that here or on #perl the other day) that were kept up-to-date by those people who are in-the-trenches using and building stuff. 20:21
sort of like cpan-ratings but with more of "use module X because ..." and "don't use module Y because ..." and a lot of "I use X, Y, and Z because ..."
masak yeah. 20:22
I tend to find that sort of info "between the lines", not online, but at confs, in talks, in the hallway track, and in the pub in the evening.
PerlJam exactly. I'd like to bring that out into the open where more eyes can see it :) 20:23
masak goodspeed. 20:24
as with all derived information, the challenge will be to pump tuits into it to keep it up-to-date.
but certainly not impossible; just adequately challenging ;)
PerlJam yep. 20:25
But now it's time for me to attend a talk on ocean going robots
PerlJam &
dalek rl6-roast-data: 4995a5f | coke++ | / (4 files):
today (automated commit)
21:05
masak skims through rosettacode.org/wiki/N-queens_problem and decides he dislikes all of the solutions posted 21:22
except perhaps the Rascal solution. 21:23
all of the other ones contain cruft or take their eye off the ball in some way.
diakopter i.imgur.com/INVnQtp.jpg just now. ftw-doze 22:08
masak "anything that can crash, will crash as t -> \infty" 22:09
FROGGS hmmmm, I can't have several v5 scopes atm :/ 22:26
lue never understood why (La)TeX's abbreviation of infinity was infty (as opposed to inf) 22:29
adu_ how do I prepend to a list? 22:33
flussence @l.unshift 22:34
adu r: say 1 + [2, 3]
camelia rakudo de2080: OUTPUT«3␤»
adu r: say [1] ~ [2, 3] 22:35
camelia rakudo de2080: OUTPUT«12 3␤»
adu r: say [1] _ [2, 3]
camelia rakudo de2080: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/EuFz3zi_Gq:1␤------> say [1] ⏏_ [2, 3]␤ expecting any of:␤ postfix␤ infix stopper␤ infix or meta-infix␤ statement end␤ statement modifier␤ stat…
FROGGS r: say (1, 2).unshift(3)
camelia rakudo de2080: OUTPUT«Cannot call 'unshift'; none of these signatures match:␤:(Any:U \SELF: *@values, Mu *%_)␤ in method unshift at src/gen/CORE.setting:1461␤ in block at /tmp/mXJJZARmOd:1␤␤»
FROGGS r: say Array.new(1, 2).unshift(3)
camelia rakudo de2080: OUTPUT«3 1 2␤»
adu r: say [1, 2].unshift(3)
camelia rakudo de2080: OUTPUT«3 1 2␤»
FROGGS right
adu thanks :)
FROGGS <--- is slow at typing right now 22:36
adu r: say [1, 2].cat([3, 4])
camelia rakudo de2080: OUTPUT«No such method 'cat' for invocant of type 'Array'␤ in block at /tmp/HCXARUed6W:1␤␤»
FROGGS r: say [1, 2] | [3, 4]
camelia rakudo de2080: OUTPUT«any(1 2, 3 4)␤»
FROGGS hmmmm, there is an operator for that 22:37
masak 'night, #perl6. see you tomorrow.
FROGGS night masak
adu r:say(list([1, 2] & [3, 4]))
r: say(list([1, 2] & [3, 4])) 22:38
camelia rakudo de2080: OUTPUT«all([1, 2], [3, 4])␤»
FROGGS r: say ([1, 2] | [3, 4]).flat
camelia rakudo de2080: OUTPUT«any(1, 2, 3, 4)␤»
FROGGS but I guess there is a nicer way
and you see that this is a junction 22:39
adu yes
FROGGS r: say ([1, 2] | [3, 4]).flat.WHAT
camelia rakudo de2080: OUTPUT«(Junction)␤»
adu r: say [1, 2].push([3, 4]) 22:40
camelia rakudo de2080: OUTPUT«1 2 3 4␤»
adu yey
FROGGS ahh, of course
[Coke] rnp: my $a = (1,2,3); say $a.WHAT; 22:45
camelia rakudo de2080, niecza v24-37-gf9c8fc2: OUTPUT«(Parcel)␤»
..pugs: OUTPUT«Array()␤»
[Coke] according to RT #66304, that should be a Seq. 22:46
I think that's probably a fossil.
sorear r: Seq 22:47
camelia rakudo de2080: OUTPUT«Cannot look up attributes in a type object␤current instr.: 'gimme' pc 284714 (src/gen/CORE.setting.pir:125032) (src/gen/CORE.setting:6046)␤called from Sub 'sink' pc 293821 (src/gen/CORE.setting.pir:128675) (src/gen/CORE.setting:6354)␤called from Sub 'MAIN' pc 381 (…
sorear wat
I was going to say that I think Seq is gone, but ... this is a new bug? 22:48
r: Sq
camelia rakudo de2080: OUTPUT«===SORRY!===␤Undeclared name:␤ Sq used at line 1␤␤»
FROGGS r: my $x = Seq
camelia rakudo de2080: OUTPUT«Cannot look up attributes in a type object␤current instr.: 'gimme' pc 284714 (src/gen/CORE.setting.pir:125032) (src/gen/CORE.setting:6046)␤called from Sub 'sink' pc 293821 (src/gen/CORE.setting.pir:128675) (src/gen/CORE.setting:6354)␤called from Sub 'MAIN' pc 381 (…
sorear r: &x
camelia rakudo de2080: OUTPUT«===SORRY!===␤Undeclared name:␤ &x used at line 1␤␤»
sorear r: &Seq
camelia rakudo de2080: OUTPUT«===SORRY!===␤Undeclared name:␤ &Seq used at line 1. Did you mean 'Seq'?␤␤»
sorear ah, the Seq type object does not like being sunk
FROGGS r: Seq.new() 22:49
camelia rakudo de2080: ( no output )
diakopter r: &(Seq)
camelia rakudo de2080: OUTPUT«Cannot look up attributes in a type object␤current instr.: 'gimme' pc 284714 (src/gen/CORE.setting.pir:125032) (src/gen/CORE.setting:6046)␤called from Sub 'sink' pc 293821 (src/gen/CORE.setting.pir:128675) (src/gen/CORE.setting:6354)␤called from Sub 'MAIN' pc 381 (…
timotimo adu, FROGGS, the operator to concatenate lists is "," 22:49
r: my @a = 1, 2, 3; my @b = 100000; say (@a, @b).perl 22:50
camelia rakudo de2080: OUTPUT«(Array.new(1, 2, 3), Array.new(100000))␤»
timotimo r: my @a = 1, 2, 3; my @b = 100000; say (@a, @b).flat.perl
camelia rakudo de2080: OUTPUT«(1, 2, 3, 100000).list␤»
diakopter sorear: looks liks you're right
timotimo r: my @a = 1, 2, 3; my @b = 100000; my @c = (@a, @b); say @c.perl; # see how assigning to a @variable will flatten the list automatically for you
camelia rakudo de2080: OUTPUT«Array.new(1, 2, 3, 100000)␤» 22:51
diakopter like
FROGGS timotimo: ya, and it makes sense :o) 22:52
it is just to late...
timotimo sorry :(
i was watching the new doctor who episodes i missed
adu timotimo: ooooo
adu <3 dr.who 22:53
diakopter oh oops didn't see the orig; sry
adu how do I add properties to a Match object? 22:57
[Coke] sorear: this is somethin b used to do. 22:57
adu $<name> = … isn't working, I get "Cannot modify an immutable value"
FROGGS timotimo: I meant it is too late to proper use my mind... :o) 22:58
so, dont worry about missing an SLA :o) 22:59
[Coke] r: my $a = 1,2,3; say $a<1>.WHAT 23:00
camelia rakudo de2080: OUTPUT«WARNINGS:␤Useless use of "," in expression "my $a = 1,2,3" in sink context (line 1)␤(Failure)␤»
[Coke] r: my $a = (1,2,3); say $a<1>.WHAT
camelia rakudo de2080: OUTPUT«(Failure)␤»
jnthn Hm, I didn't think Seq existed any more...
[Coke] r: my $a = (1,2,3); say $a[1,2].WHAT
camelia rakudo de2080: OUTPUT«(Parcel)␤»
jnthn Well, sleep time...should have for JVM porting tuits tomorrow. :)
'night, #perl6 23:01
FROGGS jnthn: sleep well!! :o)
[Coke] jnthn: it's in the spec in a few places, and about a dozen tests. 23:01
[Coke] .ask pmichaud if we can get a spec ruling on if we should rip out Seq based on your understanding of the current list situation. 23:02
yoleaux [Coke]: I'll pass your message to pmichaud.
[Coke] r: Coke]: I'll pass your message to pmichaud. [19:12] [[Coke]( 23:12
camelia rakudo de2080: OUTPUT«===SORRY!===␤Unexpected closing bracket␤at /tmp/WFPZ2rYKSr:1␤------> Coke⏏]: I'll pass your message to pmichaud. [␤»
[Coke] r: $_ = 1; my $tracker = ''; for 12 -> $a { if $_ == 1 { $tracker ~= "1 : $_|"; $_ = 2; } }
camelia rakudo de2080: OUTPUT«Null PMC access in find_method('Stringy')␤ in block at /tmp/HWQMO6B4wa:1␤ in method reify at src/gen/CORE.setting:5737␤ in method reify at src/gen/CORE.setting:5632␤ in method gimme at src/gen/CORE.setting:6053␤ in method eager at src/gen/CORE.setting:6032␤ …
FROGGS r: for 12 -> $a { "$_" } 23:13
camelia rakudo de2080: OUTPUT«use of uninitialized value of type Any in string context in block at /tmp/YZOVXI9pzj:1␤␤»
FROGGS r: $_ = 1; for 12 -> $a { "$_" } 23:14
camelia rakudo de2080: ( no output )
FROGGS r: $_ = 1; for 12 -> $a { my $x ~= "$_" }
camelia rakudo de2080: ( no output )
FROGGS r: $_ = 1; for 12 -> $a { my $x ~= "$_|" }
camelia rakudo de2080: ( no output )
FROGGS hmmm
r: $_ = 1; for 12 -> $a { if 1 { my $x ~= "$_|" } } 23:16
camelia rakudo de2080: OUTPUT«Null PMC access in find_method('Stringy')␤ in block at /tmp/5WQEd1mmQC:1␤ in method reify at src/gen/CORE.setting:5737␤ in method reify at src/gen/CORE.setting:5632␤ in method gimme at src/gen/CORE.setting:6053␤ in method eager at src/gen/CORE.setting:6032␤ …
FROGGS r: $_ = 1; for 12 -> $a { if 1 { my $x ~= "$_" } }
camelia rakudo de2080: OUTPUT«Null PMC access in find_method('Stringy')␤ in block at /tmp/bHA_n88VaK:1␤ in method reify at src/gen/CORE.setting:5737␤ in method reify at src/gen/CORE.setting:5632␤ in method gimme at src/gen/CORE.setting:6053␤ in method eager at src/gen/CORE.setting:6032␤ …
FROGGS r: $_ = 1; if 1 { my $x ~= "$_" }
camelia rakudo de2080: ( no output )
FROGGS r: $_ = 1; for 12 -> $a { if 1 { "$_" } }
camelia rakudo de2080: OUTPUT«Null PMC access in find_method('Stringy')␤ in block at /tmp/Bify_dvuSE:1␤ in method reify at src/gen/CORE.setting:5737␤ in method reify at src/gen/CORE.setting:5632␤ in method gimme at src/gen/CORE.setting:6053␤ in method eager at src/gen/CORE.setting:6032␤ …
[Coke] FROGGS: RT# 117677 if you golf it more.
FROGGS it only comes if there is an if within a loop 23:17
r: $_ = 1; for 12 -> $a { unless 0 { "$_" } }
camelia rakudo de2080: OUTPUT«Null PMC access in find_method('Stringy')␤ in block at /tmp/2A9OMuXLSv:1␤ in method reify at src/gen/CORE.setting:5737␤ in method reify at src/gen/CORE.setting:5632␤ in method gimme at src/gen/CORE.setting:6053␤ in method eager at src/gen/CORE.setting:6032␤ …
FROGGS r: $_ = 1; while $_++ < 2 { unless 0 { "$_" } }
camelia rakudo de2080: ( no output )
FROGGS r: $_ = 1; until $_++ > 3 { unless 0 { "$_" } } 23:18
camelia rakudo de2080: ( no output )
pmichaud Coke: rip out Seq. 23:37
yoleaux 23:02Z <[Coke]> pmichaud: if we can get a spec ruling on if we should rip out Seq based on your understanding of the current list situation.
pmichaud If we end up needing Seq, it won't look like what's in the spec. 23:37
or I can certainly put it back.
pmichaud I'm afk again. 23:40