»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | YAPC::NA talk submissions due end of Month. Have you submitted your talk?
Set by toddr_ on 18 March 2013.
00:11 pmurias left 00:14 cosimo_ left 00:15 BenGoldberg left, cosimo joined 00:16 dmol left
dalek : b2e16c8 | (Tobias Leich)++ | / (3 files):
regexes are working now
 00:20
timotimo_ just like that o_O 00:21
00:23 kst joined
FROGGS *g* 00:24
gnight
00:27 raiph left
masak 'night, #perl6 00:31
00:33 xinming joined 00:36 xinming_ left 00:39 denisboyun_ left 00:41 denisboyun_ joined
kurahaupo FROGGS: was it intentional to switch leading '0' from decimal to octal? 00:47
00:53 denisboyun_ left
sorear kurahaupo: leading 0 is octal in perl 5 00:54
00:55 raiph joined 01:04 labster joined
kurahaupo sorear: I'm well aware of that; ditto for C & C++. I'd also question removing the warning/recommendation to use '0o' instead. 01:11
01:11 anuby joined 01:15 Targen joined
sorear kurahaupo: FROGGS is modifying Rakudo's parser into a perl 5 parser, so using p5 syntax rules is kind of a given, no? 01:19
01:20 thou left, hypolin joined 01:42 stevan_ left 01:43 stevan_ joined 01:49 kurahaupo left 01:57 kurahaupo joined 02:04 fgomez left 02:10 thou joined 02:16 swarley-freenode is now known as swarley 02:22 daniel-s_ joined 02:26 daniel-s left 02:27 census left 02:28 PacoAir left 02:31 thou left, thou joined 02:32 cognominal left 02:36 sqirrel_ joined, FROGGS_ joined 02:38 thou_ joined, cognominal joined, thou left 02:39 toddr_ joined 02:40 FROGGS left, sqirrel left
kurahaupo sorear++ thanks for the explanation 02:42
02:42 toddr_ left 02:46 am0c joined 02:47 lustlife joined 02:55 fgomez joined 02:58 thou joined, thou_ left 03:01 c1sung joined 03:05 orafu left, thou left, thou joined 03:08 orafu joined 03:16 preflex_ joined 03:17 preflex left, preflex_ is now known as preflex 03:25 thou left, thou joined 03:28 colomon joined 03:44 timotimo_ is now known as timotimo 03:46 thou_ joined, thou left 03:49 bruges_ left, bruges joined 03:56 labster left 04:03 skids left 04:06 thou joined, thou_ left 04:28 thou left, thou joined 04:32 thou left 04:35 thou joined, kurahaupo left 04:47 thou_ joined, thou left 04:49 rindolf joined 04:54 kurahaupo joined 05:03 dwoldrich joined 05:08 thou joined, thou_ left 05:21 kurahaupo left 05:24 xinming left 05:26 xinming joined, denisboyun_ joined 05:28 thou_ joined, thou left 05:29 SamuraiJack joined 05:31 kaleem joined 05:35 kaare_ joined 05:39 arlinius left 05:40 SamuraiJack left 05:41 SamuraiJack joined 05:43 pyjery joined 05:44 cognominal left 05:46 dmol joined 05:48 thou joined, thou_ left 05:49 odoacre joined
dalek kudo/extend-dump: 0cfabe7 | (Geoffrey Broadwell)++ | src/core/ (2 files):
Remove/clean up debugging output a bit
 06:00
kudo/extend-dump: 334a90b | (Geoffrey Broadwell)++ | src/core/ (4 files):
Improve DUMP output a bit for Bool, Parcel, Range, and a few misc cases
kudo/extend-dump: b91487f | (Geoffrey Broadwell)++ | src/core/ (2 files):
Minor whitespace cleanups
kudo/extend-dump: eddcaa5 | (Geoffrey Broadwell)++ | src/core/Mu.pm:
Add comment with desugared version of $attr.get_value(self), which jnthn believes might fix an edge case later
kudo/extend-dump: ac1c9b4 | (Geoffrey Broadwell)++ | t/02-rakudo/dump.t:
First attempt at tests for DUMP, starting with simple literals
kudo/extend-dump: b15d183 | (Geoffrey Broadwell)++ | src/core/Mu.pm:
Comment out debug output in DUMP routines in Mu.pm
kudo/extend-dump: d3e1010 | (Geoffrey Broadwell)++ | src/core/Mu.pm:
Begin implementing indented DUMP output; Mu.pm changes only for first pass
06:09 thou_ joined, thou left 06:10 arlinius joined 06:11 kaleem left, kaleem joined 06:26 dmol left 06:29 thou joined, thou_ left 06:30 kaare_ left 06:31 kaare_ joined 06:40 pupoque joined 06:47 am0c left 06:50 thou_ joined, thou left 06:53 kaare_ left, kaare_ joined 06:55 thou joined, thou_ left 07:00 thou left, thou_ joined 07:07 daxim joined 07:12 thou joined, thou_ left 07:13 Targen left 07:16 diakopter joined 07:28 fgomez left 07:32 thou left 07:33 thou_ joined 07:36 daniel-s_ left, daniel-s_ joined 07:38 pupoque left 07:41 thou joined, thou_ left 07:54 domidumont joined 07:55 daniel-s_ left 07:57 daniel-s_ joined 07:58 kaare__ joined 08:00 kaare_ left 08:01 labster joined 08:02 thou_ joined, thou left 08:05 domidumont left 08:06 domidumont joined 08:08 kaare__ left, kaare__ joined 08:10 thou_ left, thou joined 08:12 rindolf left, rindolf joined 08:13 kaare__ left 08:14 thou left 08:15 thou joined 08:19 drbean left, rindolf left, rindolf joined 08:20 drbean joined 08:25 rindolf left, rindolf joined 08:32 woolfy joined 08:33 thou left, thou joined 08:34 daniel-s_ left, daniel-s_ joined, mmcleric left 08:45 Heather joined
Heather Hello #Perl6 ! 08:45
moritz \o Heather 08:46
08:46 pyjery left
Heather is writing own Code Style Guide to fit own Code Style ... 08:46
labster Hi Heather
Heather labster hi
labster I read your blog, Heather++
08:46 thou left
Heather labster huh, cool :) 08:46
08:46 thou joined
labster It makes me want to write my own blog, just to show that Perl 6 has users. 08:47
Heather labster usually I lazy to write stuff to blog :)
labster me too
moritz labster: please do
Heather take a look github.com/Heather/Beauty-and-Freedom !!! It's brand new CODE STYLE ! It need users
*needs
moritz convincing people to adapt a new code style is even harder than convincing them to adapt a new programming language :-) 08:48
Heather moritz people must understand that Beauty and Freedom can be better then standard :P that's my point 08:49
*than
Heather always use 'then' wrong
labster I'm not even sure what style I'm using :/ 08:50
Heather labster then use Beauty and Freedom :D it will fit anything >_<
- If you're not agreed with some terms or don't like some Code Style you're Free to use Custom style you like 08:51
- If you desired to use Beauty-and-Freedom Style Guide you can also mix it with any other Style Guides you like
...
\o/
labster moritz: I will write something, once I get this patch to File::Spec out
08:54 denisboyun_ left 08:57 thou left, kivutar joined, thou joined
labster I just realized I have the first working version of IO::Path.filesystem (though I call it IO::Path::More.volume). 09:00
moritz labster: that might be an extension that should go into core 09:01
Heather labster btw if you liked my blogpost feel free to G+ and Twit it :D
labster It needs File::Spec to work, which means that at least a portion of File::Spec needs to go into core. And... it needs work. 09:02
moritz labster: when I wrote IO::Path I was aware that at some point, we need volumes, but since I don't have any windows system to test, I left it out for the first version
FROGGS_ labster: do you have a solution for case-insensitivity (insanity?) testing? 09:03
labster FROGGS_: What I submitted should work, though it's not the most elegant thing. I keep thinking I should move it to Unix.pm, but there are issues with that, mainly with everything subclassing from Unix. 09:06
09:06 thou left, thou joined
labster moritz: Did you write the spec for IO::Path, or just the implementation? Because if it's the spec, I have some issues to discuss. 09:07
moritz labster: both
lizmat morning #perl6 09:08
Heather lizmat hey
moritz \o lizmat 09:09
labster hi lizmat
lizmat FROGGS_: I think there should be a presentation about your Perl 5 in Rakudo work at the YAPC::NA
would you be able to give such a presentation?
hi all!
tadzik hello lizmat 09:11
labster moritz: ok. first of all, I think "filesystem" should be renamed to "volume". Filesystem implies any(<zfs ntfs fat32 ...>) to me, which isn't what we'd be returning. Volume is the P5 File::Spec terminology, and it seems to make more sense to me. 09:12
moritz labster: agreed 09:13
labster Next is stringification. What was the logic behind having File::Path.Str return just the basename? I have to say, I was surprised to see it return that rather than the whole path. 09:15
FROGGS_ lizmat: in north 'murica?
moritz labster: the reason was making dir() backwards compatible
labster: in retrospect that's probably a bad idea
so we need a deprecation notice, and wait 'til after the release 09:16
09:16 user____ joined
lizmat FROGGS: yes, in Texas even! 09:17
FROGGS_ lizmat: I'm afraid I don't have the money to get there :/
labster I was just looking around at File::Find, and File::Find::Result objects returns the whole path when stringified too.
09:18 thou left, user____ left
moritz ah yes, but File::Find isn't the only user of dir() 09:18
tadzik yeah, I was imitating Perl 5's File::Find and find(1)
09:19 am0c joined
labster that was more of a similar behavior issue but I guess you're already convinced :) 09:20
lizmat FROGGS_: lack of money is fixable, lack of free time to travel and do a presentation is not 09:22
moritz labster: yes, I'm quite boring to talk to. You don't even have to ramp up all your arguments to convince me :-) 09:23
labster: will you patch the specs?
labster huh, do I even have access to them? 09:24
moritz does github.com/perl6/specs/ say "Read-Only access" next to the URL?
09:24 daniel-s_ left
FROGGS_ lizmat: out of curiosity, how do I fix that? :o) 09:25
moritz should be read-write
lizmat FROGGS_: you could be sponsored
FROGGS_ that is ... interesting ... 09:26
labster moritz: I do have access 09:27
.oO ( The path from hacker to language implementor is surprisingly short around here ) 09:28
09:29 DarthGandalf left
moritz labster: iirc you submitted a pull request to the ecosystem repo, so I put you into the 'perl6' team, which we use for nearly all repos in the 'perl6' organization 09:30
lizmat afk for ~ 2.5 hours&
moritz because we're far too lazy to do any fine-grained access control
so, everybody who submits a new module gets write access to the specs repo too :-) 09:31
labster see, laziness is a virtue
moritz and roast, and evalbot, and perl6-examples, and most of the other repos
masak greetz, #perl6 09:32
moritz \o masak
labster moritz: also, please take a look at IO::Path::More, and see if there's anything there you feel should be part of the spec. Probably not stuff like .find
moritz labster: URL to the module/repo?
labster github.com/labster/p6-IO-Path-More 09:33
moritz the is_absolute and absolute and relative methods might be worth importing 09:34
(though as is-absolute)
maybe method parent
labster I get lazy and use underscores because I'm trying to port perl 5 09:36
moritz that's fine for a module, but not for core :-) 09:37
09:44 jaldhar joined
labster Oh, oh! forgot the last issue 09:45
r: my $p = '.'; say $p.=path.WHAT for ^6;
p6eval rakudo a23458: OUTPUT«(IO::Path)␤(Str)␤(IO::Path)␤(Str)␤(IO::Path)␤(Str)␤»
09:45 fhelmberger joined
labster .path alternates object types, which is strange. I don't know what to do about that. 09:46
moritz finding a better name for the string representation of a path would help 09:47
09:48 daniel-s_ joined
labster Str? fullpath? 09:48
09:49 bonsaikitten is now known as DrEeevil
moritz +1 09:49
just Str will be fine, I think
labster And IO::Path.path returns itself, I suppose. 09:50
moritz aye 09:51
09:58 jaldhar left
FROGGS_ labster: you have commit bit now for File::Spec 09:59
feel free to refactor it if you think it makes sense
10:00 DarthGandalf joined
labster FROGGS_: Thanks. I was about to ask for it, but I got sidetracked on the specs. 10:00
FROGGS_ labster: I'm really sorry that I can't help you with it right now 10:01
labster no, please keep working on perl5, that's awesome
10:01 jaldhar joined
FROGGS_ labster++ # I love you 10:01
masak :) 10:03
labster hugme: hug FROGGS_ 10:04
hugme hugs FROGGS_
FROGGS_ *g*
Heather UPDATE #PERL6 SET USER = USER + 1 10:06
masak there's a nice warm atmosphere in here today :) 10:07
Heather masak that's Stockholm syndrome
tadzik :D
It's funny because masak is not that far from Stockholm 10:08
Heather >_<
masak heh.
10:09 hypolin left
masak Heather: so you're saying Perl 6 somehow holds us hostage, and we're starting to have sympathy for it? 10:10
Heather masak yep :P
Osama Ben Camelia 10:11
labster .oO (Camelia with a long beard?) 10:12
10:12 anuby left
Heather labster ... Richard Ben Camlia ? >_< 10:12
Heather always imagine only one person when hear something about bread 10:13
tadzik the gingerbread man! \o/ 10:14
labster Oh wow, it is getting late here in California (π AM). I'll think about the specs tomorrow.
night everyone
Heather labster good night 10:15
Heather wants more MMO & RP elements on #perl6 o_O 10:16
masak Heather: um. strangelyconsistent.org/blog/perl-6-is-my-mmorpg 10:17
10:17 SamuraiJack left
Heather masak yes I saw it :) 10:17
masak also, github.com/masak/crypt
Heather well "SAW" is not what I wanted to say
masak how to play it? 10:18
tadzik Heather: you can help me write Quest :)
although it's not quite RPG, and definitely not MMO :)
masak Heather: should be just running bin/crypt in that repo. 10:19
10:19 SamuraiJack joined
Heather tadzik quest? 10:19
masak Heather: oh, and it has a dependency (Adventure::Engine), but it should all install through panda.
errand &
Heather masak I use CRYSTAL MAIDEN ! 10:20
-_-'
masak so first I will generate an gentoo ebuild for it ) 10:21
10:21 labster left 10:23 eiro left
tadzik Heather: github.com/tadzik/Quest 10:23
it's supposed to be a fun and addictive game
it's not, yet ;)
10:24 eiro joined
Heather for now MMO part is must have part of any game, even roguelike 10:24
tadzik I still prefer single-player games 10:25
this way I can stop playing whenever I want and no one gets angry
Heather well, I don't play single-player games even if they are very cool, I feel it boring
tadzik I see 10:26
Heather tadzik github.com/tadzik/Quest/pull/1 10:28
it's strange... 10:31
how github highlights it?
tadzik Heather: merged, thanks 10:32
Heather tadzik well, but it doesn't helped
I was thinking about making github to highlight the file
tadzik wrong file mode maybe? 10:33
10:33 dwoldrich left, daniel-s_ left
Heather I guess ... 10:33
10:33 daniel-s_ joined
Heather tadzik github.com/tadzik/Quest/pull/2 10:36
this way
tadzik Heather: hilight it how? 10:38
Heather tadzik github.com/Heather/Quest/blob/master/bin/quest
strange? 10:39
tadzik github is failing saying that " Heather 17 minutes ago Shebang " is last commit 10:40
tadzik hmm, it just thinks it's Perl 6
erm, Perl 5
why, it is the last commit
Heather tadzik no, github.com/Heather/Quest/commits/master
tadzik I'm confused. What's the problem? 10:41
I've only just merged pull-2
Heather good :) 10:42
now it highlights
but as perl5 seems like, yes
tadzik : what means # I'm a bad person ? ) 10:43
tadzik I probably wrote something ugly in the code 10:45
Heather tadzik why? :)
tadzik where did you find that? 10:46
Heather can't do anything awhile rust recompiling. 95% RAM 35% Swap
tadzik in bin file beginning, clear sub 10:47
tadzik ayes
because it's not Pure-perl in any way
10:47 am0c left
tadzik it feels hacky, I'd prefer it gone and replaced with a sane solution 10:47
ANSI code for example 10:48
but when I tried an ANSI code it didn't work quite well
10:50 tgt joined 11:04 denisboyun_ joined 11:19 toddr_ joined 11:20 immortal joined 11:21 bonsaikitten joined 11:23 shachaf_ joined, erkan left, takesako__ left, huf left, yx left, bloonix left, GlitchMr left, `patch` left, sitaktif1 left, robinsmidsrod left, pjcj left, pnu_ left, Pleiades` left, DrEeevil left, shachaf left, anocelot left, toddr left, Pleiades` joined, `patch` joined, perigrin joined, immortal left, immortal joined, bonsaikitten left, bonsaikitten joined, Pleiades` left, Pleiades` joined, huf joined, masak joined, mst joined, nemesys joined, anocelot joined, Util joined 11:24 pjcj joined, denisboyun_ left, patspam joined, GlitchMr- joined, Hor|zon joined, pnu joined, bloonix joined, BooK joined, robins joined, sitaktif1 joined, takesako__ joined 11:25 robins is now known as robinsmidsrod
masak well, that was fun. 11:25
11:26 LordV joined 11:28 yx joined, mangala left, Gothmog_ left, integral_ left, clkao left, PerlJam left, krunen left, Ayiko left, Tene left, xalbo left, sergot left 11:31 mangala joined, Gothmog_ joined, integral_ joined, clkao joined, PerlJam joined, krunen joined, Ayiko joined, Tene joined, xalbo joined, mikemol joined, ponbiki joined, sergot joined, snarkyboojum joined, mikec__ joined, jrockway joined 11:32 cibs_ joined
Heather masak ? 11:33
masak Heather: a netsplit.
11:33 census joined
Heather hm... 11:34
FROGGS_ beer with us? np 11:44
dalek : aecb96f | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
add support for unless+else constructs
 11:51
11:52 cibs joined
Heather bear with us 11:52
and balalaika 11:53
11:55 cibs_ left 11:57 GlitchMr- is now known as GlitchMr
dalek : c1d7375 | (Tobias Leich)++ | t/spectest.data:
skip base/lex.t for now
 11:58
12:05 PacoAir joined
dalek : d13e413 | (Tobias Leich)++ | test.pl:
we have better tests than this
 12:13
: b0d3beb | (Tobias Leich)++ | Makefile:
add fulltest target
: 2c5ad30 | (Tobias Leich)++ | .gitignore:
ignore *.v5 files
Heather could be fun but because of Perl and ML I want to visit some arabic country alike Egypt ... 12:15
12:17 yx left
masak Heather: I fail to see the connection. "because of Perl and ML" you want to visit some Arabic country? 12:19
is this about readability again? I thought we talked about this. :P
Heather masak It's about camel in logo 12:20
... 12:21
12:26 yx joined 12:27 kaleem left
lizmat is anybody here going to the QA Hackathon in Lancaster in ~ 2 weeks? 12:34
(as in non-lurkers)
12:37 kaleem joined 12:42 PacoAir left 12:45 PacoAir joined
lizmat takes that as a no :-) 12:49
[Coke] lizmat: ooh. if I was on one of my increasingly less rare trips to Leeds, I'd try to swing by. 12:52
lizmat the reason I'm asking, is that there is apparently going to be some serious CPAN hacking going on there
and I was wondering whether it would make sense to make sure that Perl 6 modules / distributions could be handled as well 12:53
either as a multi-version distribution (like with my Devel::MaintBlead support module) or in some other wat 12:54
dalek : 19b9cc8 | (Tobias Leich)++ | lib/Perl6/P5Actions.pm:
<blorst> -> <sblock>, this allows to use BEGIN blocks
 12:55
12:56 xinming left 12:57 xinming joined
[Coke] argh. "where did panda install itself, and why do I have to keep changing my path to find it." 12:57
FROGGS_ std: use v5; print -d "." 12:58
12:58 Heather left
p6eval std 86b102f: OUTPUT«ok 00:01 52m␤» 12:58
12:58 crab2313 joined
[Coke] modularizes his .bashrc slightly so he can change a variable pointing at a parrot revision to track where panda got to 13:00
FROGGS_ [Coke]: panda should be bundled with the compiler release, and it should be always there and always working and always something and ... 13:01
[Coke] I'm not dealing with releases. I'm building from source. 13:02
perhaps that is my problem.
FROGGS_ then panda must be made when rakudo gets recompiled 13:03
[Coke] ecosystem question: do we care if the ecosystem name doesn't match the module name? I see some that match except for a - instead of a ::, for example. 13:04
FROGGS_ hmmm, at least I dont care
[Coke] FROGGS_: Yes. every time I reinstall rakudo, I have to eithe rebootstrap panda OR completely reinstall it, depending on whether the underlying parrot changed.
tadzik [Coke]: what do you mean by the ecosystem name? 13:07
[Coke] as it appears on modules.perl6.org (the name in the META.info)
tadzik also, I've been thinking about reworking the installation process a bit, so these sort of things are not a problem anymore 13:08
FROGGS_ ++tadzik
tadzik ie imagine 'panda install Foo' saying "seems like you recompiled rakudo recently, do you want me to rebootstrap all the things?"
also getting rid of the PATH insanity we have now
FROGGS_ that would be sweet 13:09
[Coke] That would be nice.
I'm trying to figure out how to get panda to let me install "YaMl"
(or rather, correct me helpfully)
tadzik marry it to Text::Levenshtein
or even not
we could add a fallback, so if there's no such module it tries comparing lc($arg) to map { *.lc } @modules 13:10
then a "did you maybe mean $other-thing?"
[Coke] yes, that was going to be the first step, then maybe TL.
I'm trying to suss out where that would go, though. 13:11
tadzik there can be some controversial issues with that, as moritz commented on the GH ticket 13:13
moritz suggesting stuff is fine
autocoreccting not
because then you might put slightly wrong names into some automated script, and once a module exists with that slightly different name, you suddenly install a different module 13:14
but I think I wrote that already :-)
[Coke] moritz: yes, that's what we said.
"calm down." 13:15
moritz ok, great :-)
masak a program should never *ever* automatically correct faulty input.
if you said it already, it bears repeating. :)
[Coke] ok, I'm not working on this, then. :P 13:16
masak may I suggest extremely nice error messages instead?
the right way to go ahead from the idea of automatic correction -- every single time through history -- is to pimp the error messages to say "did you mean YAML?" 13:17
[Coke] That's what I was *going* to do, yes.
13:17 crab2313 left
[Coke] Holy crap, people. 13:17
masak sorry :/
[Coke] "I wish to make perl 6 nicer" "You're going to do it wrong!" 13:18
masak [Coke]: recently, after our nicer levenstein variable error messages, there have been suggestions that Rakudo automatically correct the source code. I guess I'm a bit touchy about the thought of that.
[Coke]: didn't mean to get it all over you.
13:21 dayangkun joined
[Coke] Is there a way to get the 6POD out of an installed module ? 13:21
I wish to look at documentation from the command line.
FROGGS_ masak: automatically correct variables... and that wasn't a joke? 13:22
tadzik [Coke]: perl6 --doc /path/to/module.pm 13:24
maybe p6doc is capable of doing that
[Coke] tadzik: "ew"
(see also complaints about "where did panda get installed" 13:25
tadzik ah
[Coke] something like perl6 --doc Module::Name would be spiffier.
masak FROGGS_: it was a suggestion by someone who hadn't seen the consequences of such a suggestion. I forget who.
tadzik it's even specced
I think it was me :)
masak FROGGS_: I think it's even been suggested two or three times recently, by different novices.
arnsholt timotimo: I can't reproduce your Zavolaj problem (it's also been reported as Zavolaj #23) =(
FROGGS_ masak: ohh, I just remember me suggesting that as a joke :o)
masak heh. 13:26
tadzik ah, in the source code
arnsholt So I'm guessing it's probably something GC-related
FROGGS_ masak: I'm a guy who would tike to force 'strict' for Perl 5, even for one-liners... I would never like to have source-code-autocorcection 13:27
[Coke] ... Did anyone every implement Perl6::Format for perl6 ? 13:28
*ever
13:28 fgomez joined
[Coke] github.com/mathw/form/ ? 13:28
masak yes, that's the one effort I know about. 13:30
it's a couple of years sleeping now.
mathw_: ping :)
[Coke] modules.perl6.org is written in perl5 ? 13:32
tadzik yes
feather.perl6.nl/~sergot/modules/ is all Perl 6
[Coke] ... so it's already been ported to perl6, but the main site is using the p5 version? 13:33
tadzik yeah
also, the latter has a per-module pages and tests results
[Coke] it also hasn't been updated in a while. 13:34
tadzik yeah 13:35
sergot: ping
[Coke] Is sergot's version going to become the main version at some point? blocking on anything?
tadzik not sure 13:36
[Coke] sees that p6prove is also under started. 13:37
er, *underway
which then makes the last word useless. ;) 13:38
tadzik yep
13:39 kaleem left 13:40 skids joined
lizmat wonders where panda is specced? 13:41
13:43 bluescreen10 joined
hoelzro morning #perl6 13:46
13:46 cibs_ joined
tadzik lizmat: it's not, not really 13:47
actually, Pies is supposed to be a spec and Panda the implementation
13:51 cibs left
[Coke] what do folks think about porting individual pure p5 modules vs. the v5 work going on rcently? 13:51
nwc10 a) long(er) term the ported version useful 13:52
b) short term the ported version means more code written in Perl 6, which is also useful
in the middle, having a way to automatically convert stuff is useful
lizmat I see the v5 work as important for darkpan 13:53
nwc10 oh yes, that too
lizmat porting CPAN modules (be they pure perl or not) is a task that will need to be done as needed
and while doing that, will make a good selection of useful CPAN modules in the process 13:54
anyway, that's my .02
masak I think something using PPI could get pretty far converting Perl 5 to Perl 6. 13:57
in fact, ooh! *that's* what my first CPAN module is going to be. obviously.
the idea being: if Perl::Critic can do a good job finding stuff to improve, a converter module can do a decent job sidegrading code to Perl 6, using known automatic conversions. 13:59
and it's also completely fine for it give up in some cases, or suggest improvements that can only be realistically carried out by a human. 14:00
14:01 mtk left
colomon Hmmm... porting tests well should be a priority, no? 14:02
lizmat so, as a sort of "ufo" that would build an annotated Perl 6 source code version?
14:02 kaare__ joined
masak colomon: that's a good point. 14:03
colomon: and maybe smaller scripts.
lizmat: :) 14:04
14:05 bluescreen10 left 14:08 cibs joined
raiph <hoelzro> where are the various dispatchers ... <FROGGS> in nqp's C land 14:09
timotimo well, the multiple dispatch is now implemented in nqp, too
or ... "only"
raiph hoelzro: what timotimo said 14:10
arnsholt timotimo: I'm not having luck reproing your bug, but it's been reported on GitHub as well 14:11
timotimo read that, thanks
14:11 cibs_ left
timotimo Rakudo-2003.03 - hehe, wat? :) 14:12
masak blast from the past! 14:13
lizmat is considering porting Sys::RunAlone to Perl 6
nice small module 14:14
14:14 bluescreen10 joined, mtk joined
timotimo is that when you only want to have one instance of your program? 14:14
lizmat yup
timotimo sounds nice
PerlJam that's an odd name though. "run alone"? without any friends? 14:15
tadzik like libunique
14:15 xilo left
tadzik which is now GtkApplication, apparently 14:16
PerlJam especially given a widely known term for "only one instance" -- singleton.
lizmat "WARNING Unique is now in maintenance mode, and its usage is strongly discouraged"
PerlJam (or, perhaps I'm just curmudgeonly todya :)
colomon masak: I'm now reading up on PPI.... 14:17
masak PPI is great.
Alias++ at his best, solving unsolvable problems.
timotimo arnsholt: in #22 ("how to store binary data in a buffer"), you refer to lots of plumbing having to be done; is that refering to vmarray?
lizmat Sys::RunAlone is much, much simpler
it just attempts to lock the source file at INIT time, and exits if it can't 14:18
tadzik clever 14:20
lizmat I can't take all the credit: "Inspired by Randal Schwartz's mention of using the DATA handle as a semaphore
on the London PM mailing list."
that was back in 2005 :-) 14:21
14:23 Psyche^ joined 14:27 Patterner left, Psyche^ is now known as Patterner 14:32 am0c joined 14:33 xilo joined 14:34 cibs_ joined 14:38 cibs left 14:42 tgt left 14:44 crab2313 joined
colomon masak: hmmm, corollary to my previous comment: porting p5 testing modules early on should probably be a priority. 14:57
arnsholt timotimo: Yeah, that's the vmarray stuff 15:01
15:02 bonsaikitten left
masak colomon: would be interesting to know which ones are the most used p5 testing modules. 15:02
colomon masak: yes, it would. 15:03
masak colomon: as long as it's just Test::Simple and Test::More, we're already set.
15:06 bluescreen10 left, bonsaikitten joined, bluescreen10 joined
tadzik and Test::Fatal 15:07
timotimo Test::Fatal is the one where it deletes your hard drive if the tests fail? 15:08
masak seems fair to me. 15:09
colomon Is that like the old product that would eject your floppy disk at high speed if it detected any signs of a system crash? 15:14
colomon suspects that ad was in Creative Computing circa April 1982.....
timotimo ha, that's cute 15:15
also i imagine painful
arnsholt timotimo++ # Backtrace from -g compiled Parrot/NQP
I <3 you a little bit now =)
timotimo :) 15:21
didn't even realise there was nothing wrong with the backtrace until now 15:22
arnsholt And there is an NQP frame in there: #21 is the invoke handler 15:29
Oh, could you add to the bug thread or the Gist if the backtrace is from the tweaked test file with loop, or the plain one?
(I'm guessing with loop?)
lizmat wrt to porting Test modules, Test::Effects ? 15:31
masak lizmat: nice. didn't know about that one before. 15:32
lizmat was exposed to it again at the SPW by the closing keynote of TheDamian 15:33
this would also mean Test::Trap
15:35 am0c left
timotimo it's from the one with the loop, yes 15:36
do you also want a stacktrace from without the loop?
arnsholt Yeah, do that as well. Might come in handy 15:38
15:46 thou joined
arnsholt But from the bottom of the stack trace it looks like something weird's happening during GC 15:46
*sigh* I really need to get my testing thing running 15:47
15:48 am0c joined
timotimo oh 15:48
another thing
using -Ilib instead of -Iblib/lib i get no segfault for the non-loop version
[Coke] finds retester-daoswald.dotcloud.com/ in his feedly. perl5 regex online debugger. 15:49
timotimo and for the loop version as well
15:52 thou left
timotimo arnsholt: can you make any sense of that? 15:53
serialisation and gc interoperating in a strange way?
15:55 thou joined
moritz the two most "fun" type of problems to debug interacting 15:55
timotimo how does one properly differentiate fun as in "perl6 is -Ofun" and "when playing dwarf fortress, you have lots of opportunities to have Fun"? is it just the capitalisation? 15:57
16:02 SamuraiJack_ joined, SamuraiJack left 16:04 bbkr joined
[Coke] We've had a p6 project manager in the past. Is anyone currently filling that roel? 16:07
*role?
16:12 domidumont left
timotimo isn't that pm's role or something? 16:12
given pmichaud { say $_ ~~ Perl6::Project::Manager } 16:13
16:13 Targen joined
[Coke] I wouldn't have guessed that, no. 16:17
tadzik: what's left todo on github.com/tadzik/Template-Mojo ? 16:21
(also, why is it Template::Mojo instead of 5's Mojo::Template ?) 16:22
masak "A leader is best when people barely know that he exists" -- Lao Zi
16:29 dmol joined 16:31 Chillance joined, fhelmberger left
arnsholt timotimo: Odd. If I -Iblib/lib or -Ilib it doesn't matter on my end 16:34
16:35 domidumont joined
PerlJam Coke: I don't understnad "p6 project manager". We don't have a "Perl 6 project" as such. We have a "Rakudo project" and a "Niecza project" and so forth though. And I'd say we have a "Perl 6 product", but that's something different. 16:38
[Coke]: who was the past perl6 project manager? 16:39
masak how many have we had?
_sri [Coke]: i like the name, keeps Mojo::Template free for the official port 16:42
PerlJam google tells me it was obra 16:43
(I never would have guessed that :)
masak that sounds vaguely familiar. 16:44
PerlJam From Aug 2007 -- www.theperlreview.com/Interviews/je...00708.html 16:46
It's so weird reading that now :)
[Coke] jesse is the one I remember. Allison just mentioned she was doing the job for a while. 16:54
PerlJam: some (myself included) would argue that that's part of the problem. :) 16:55
PerlJam [Coke]: I might just agree with you.
[Coke] though I think rakudo is "winning" through sheer force of jnthn++.
PerlJam
.oO( jnthn is Charlie Sheen? )
 16:56
arnsholt More like jnthn is the Chuck Norris of compiler hacking =D 16:57
masak I work with him. I can confirm he is Chuck Norris. 16:59
17:00 kaleem joined 17:01 daxim left
[Coke] ponders something project schedule like with no dates so we can document (and potentially agree) on what direction we're heading in. 17:02
masak that's partly what the planned production-readiness site is meant to do. 17:04
kind of like a compiler comparison chart, but focusing on "how far along are we with these desirables that would enable us to say 'yep, it's ready for serious use'" 17:05
arnsholt The Zavolaj fails in the recent past are really motivating me to whipping up this automated testing stuff I talked about recently as well 17:07
We really need to get something like that running 17:08
masak +1 17:09
[Coke] wasn't "emmentaler" the answer to that question?
masak yes.
[Coke] what's the current state of that?
masak sergot++ got something up and running. 17:10
tadzik linked to it earlier today in the backlog.
but it hasn't been updated in a while.
arnsholt Does emmentaler test the compilers as well?
17:16 flussence joined 17:18 kaleem left
masak arnsholt: don't think so. I see why that would be nice-to-have too, especially on multiple platforms. but it feels like a separate thing from testing the module ecosystem. 17:19
17:19 domidumont1 joined
FROGGS_ it should be kept separate, cpantesters only cares about modules too 17:20
you could treat perl as a module though
(on the websites)
arnsholt True, true 17:21
[Coke] arnsholt: I'm already testing the compilers:
17:22 domidumont left
[Coke] github.com/coke/perl6-roast-data 17:22
(3 of them, anyway)
ah. so emmentaler is not just "test the modules", but is also the answer to "dogfood modules.perl.org?" 17:23
masak seems so. 17:24
17:24 thou left
flussence g'morning, * 17:25
masak o/ 17:26
17:26 thou joined
[Coke] trips over something in java 1.7 and wonders if he's expecting something he shouldn't 17:27
raiph Heather: I glanced at cm. the MAIN sub has me wondering. do you know about automatic usage/help features? and using multi subs to process various command line options? perlcabal.org/syn/S06.html#Declarin...subroutine
flussence hoelzro: ping? perl6-posix/META.info has a stray whitespace at «[ "NativeCall " ]» and panda doesn't like that.
arnsholt What I'm thinking of is something to automatically build Rakudo and NQP (at least for the time being) with --gen-parrot and against Parrot HEAD, and test a selection of modules 17:28
raiph (for others reading along, i'm talking of github.com/Heather/CrystalMaiden/b...ter/bin/cm )
arnsholt At least test Zavolaj, since it's pretty intertwined with the compiler guts, and I tired of randomly finding out it's broken because of some internals change I missed =)
[Coke] arnsholt: I think the "selection of modules" testing is handled by sergot's emmentaler project. 17:29
arnsholt Right. I should probably look at emmentaler first
[Coke] checks to see which version of parrot he's testing rakudo with daily. 17:30
[Coke] is just using --gen-parrot
[Coke] suspects that's appropriate for roast tests.
tadzik [Coke]: I think Mojo::Template is a bad name :) 17:31
esp. in our ecosystem, where it has nothing to do with Mojo, but a lot to do with Templates
[Coke] and once Mojo exists? 17:32
tadzik then we'll see
_sri | [Coke]: i like the name, keeps Mojo::Template free for the official port
sounds good to me
[Coke] ... who is going to do the "official" port if not you? ;) 17:33
tadzik _sri, of course :)
[Coke] and why is this version not it? ;)
tadzik I'm only doing unofficial ports
[Coke] (_sri) ... ah. well that's fine, then!
tadzik of Dancer, of Mojo::Template... ;)
tadzik decomutees 17:34
dalek rlito: 4f7a85e | (Flavio S. Glock)++ | src5/lib/Perlito5/X64/Assembler.pm:
Perlito5::X64::Assembler - Operand set_disp8, set_disp32
 17:36
17:47 tgt joined 17:59 sqirrel_ left
hoelzro good * #perl6 18:20
colomon \o 18:21
18:21 spider-mario joined
hoelzro can anyone help me understand what '<O('%methodcall')>' does? 18:23
(in Grammar.pm)
arnsholt Sets the precedence of the thingy 18:24
Which the OPP (in EXPR) uses to sort out grouping
PerlJam precedence and associativity, n'est pas? 18:25
arnsholt And associativity, that's true 18:26
hoelzro where is the O token/rule defined?
and more importantly, where is the code that actually handles calls? 18:28
masak hoelzro: in HLL::Grammar
hoelzro ah ha
masak (in nqp)
hoelzro masak: was that for O, or call handling? 18:29
hoelzro is looking at that callsame bug
FROGGS_ and in Perl6::Grammar, search for Perl6::Grammar::O(
Perl6::Grammar line 2885
masak hoelzro: that was for O 18:30
hoelzro ah, ok
so it seems that callsame + routine traits doesn't work because Routine.can('postcircumfix:<( )>') is undefined
18:32 thundergnat joined
hoelzro so I think a "dummy" implementation should be present so callsame can do its job 18:32
PerlJam wonders if there are any other "unimplemented" operators like that.
masak hoelzro: I think you're simply the first to try and overload invocation like that. it makes a lot of sense to do it -- it's just not been done before. 18:34
hoelzro I'm guessing calling semantics are implemented in Parrot land
masak: yay, I'm first =)
PerlJam hoelzro: that means you need to blog about it ;)
masak not sure the calling semantics are that tied to Parrot these days.
hoelzro ;)
I'm guessing it's done in a dispatcher somehow 18:35
thundergnat Good morning (UGT) perl6
hoelzro my current technique relies on inserting nqp::say and recompiling =/
thundergnat: ahoy!
PerlJam hoelzro: does monky patching not work so that you don't have to endure the long compile? 18:36
hoelzro PerlJam: good question; haven't tried
PerlJam heh, "monky patching" ... that's what happens on perlmonks :)
masak thundergnat! \o/ 18:37
thundergnat masak \o\
errr... \o/
masak /o/ 18:38
PerlJam hoelzro: I bet you can Routine.HOW.add_method() at the top of your script before you add the trait.
18:38 crab2313 left
masak hoelzro: it just struck me... did you explore doing '$r()' rather than 'callsame' as a workaround? 18:38
thundergnat I had mentioned a few days ago that some infinite lists have extremely slow .first methods. Well, I put up a gist exploring the boundary of the problem and a possible fix. gist.github.com/thundergnat/5265633
PerlJam masak++ oh, good idea
masak hoelzro: need to pass in the right args too. but that's a detail. 18:39
thundergnat It may not be the best way to go about it, but it exposes the problem better at least.
hoelzro masak: no, I did not...
18:39 colomon left
hoelzro good call! 18:39
masak: infinite recursion =/ 18:41
[Coke] thundergnat++ 18:42
masak hoelzro: oh! right, 'does'...
18:42 kaare__ left
dalek rl6-roast-data: 64a9dd8 | coke++ | / (3 files):
today (automated commit)
 18:43
raiph hoelzro: fwiw, afaik the dispatcher is now in nqp (not C)
18:43 kaare_ joined
thundergnat [Coke]: I even went so far as to recompile Rakudo with those methods added and it still passes roast... but it might be attacking the problem from the wrong direction. 18:43
[Coke] niecza still has 18 failures. (which as I recall colomon couldn't duplicate - anyone else want to do a spec test run on niecza and see if they can find some tests to fudge?)
hoelzro I'm not entirely clear which dispatcher does what
ex. jnthn's work and src/Perl6/Metamodel/Dispatchers.pm 18:44
[Coke] thundergnat: pmichaud is probably the person to weigh in on that.
thundergnat [Coke]: tru
*e
18:46 SamuraiJack_ left
masak can anyone explain to me why twitter.com/AlgebraFact/status/317...8282331137 is true? 18:47
18:47 mmcleric joined
masak I mean, those are the only two groups of order four I know of, up to isomorphism. 18:47
but that doesn't mean it's true/proven. :)
18:48 mtk left
hoelzro it's the *binder* that jnthn ported to NQP, right? 18:48
18:48 mtk joined
masak .oO( full of women ) 18:49
hoelzro haha 18:50
18:50 colomon joined
raiph hoelzro: ah. as you were then. :) 18:55
hoelzro: a commit comment by jnthn: "Rip out most of the C-based multi-dispatcher. Only leftover bits are the compile-time dispatch analysis used by the optimizer, which still needs porting." 18:57
hoelzro I see
where is the dispatcher source located?
I could only find src/Perl6/Metamodel/Dispatchers.pm
raiph github.com/rakudo/rakudo/commit/61...ace011b457 ? 19:00
hoelzro I just want to find the code that implements postcircumfix:<( )> =( 19:03
arnsholt That should be somewhere in Routine or something 19:04
Lemme grep for a 'sec
hoelzro: Try src/core/Routine.pm:85 19:05
Oh, wait. No
moritz r: say (sub () { }).can('postcircumfix:<( )>') 19:08
p6eval rakudo a23458: OUTPUT«␤»
19:09 am0c left
moritz iirc there's some 6model stuff which handles invocations without having to go through postcircumfix:<( )> 19:09
arnsholt Oh, right. That's probably it 19:10
raiph r: say nqp::can(sub () { }, ('postcircumfix:<( )>') 19:11
p6eval rakudo a23458: OUTPUT«===SORRY!===␤Unable to parse expression in argument list; couldn't find final ')'␤at /tmp/D7uzLWyAWK:1␤------> :can(sub () { }, ('postcircumfix:<( )>')⏏<EOL>␤ expecting any of:␤ postfix␤»
raiph r: say nqp::can(sub () { }, 'postcircumfix:<( )>')
p6eval rakudo a23458: OUTPUT«0␤»
raiph gotta run, but github.com/rakudo/rakudo/blob/nom/...OTSTRAP.pm ? 19:12
moritz afaict BOOTSTRAP just calls that method, not installs it 19:14
arnsholt hoelzro: I'm completely stumped. Sorry =/ 19:21
hoelzro yeah, me too =/ 19:24
masak this may be the *weirdest* Perl 6 code I've ever written: gist.github.com/masak/5266081 19:25
(but, man do I love Perl 6)
also, it works. :)
19:26 shachaf_ left, shachaf_ joined 19:27 shachaf_ is now known as shachaf
masak but I think that answers my own question about Z_4 and Z_2 X Z_2 being the only two groups of that size. 19:31
perigrin safari doesn't like your infix operator. 19:34
unless ... you're intentionally using an empty box?
masak no, it's a cute little 'x' here.
perigrin blames mobile safari then.
masak .u ⨯
yoleaux U+2A2F VECTOR OR CROSS PRODUCT [Sm] (⨯)
masak yoleaux: yo! :) 19:35
perigrin yeah works fine in prompt.app ... fails miserably in mobile safari.
chrome doesn't like it either so I blame webkit or whatever the iOS HTTP widget is. 19:36
flussence .u × 19:38
yoleaux U+00D7 MULTIPLICATION SIGN [Sm] (×)
flussence I guess 2A2F is the more technically correct one 19:39
19:40 kivutar left
lue hello world o/ 19:41
masak .u 2A2F 19:43
yoleaux U+2A2F VECTOR OR CROSS PRODUCT [Sm] (⨯)
masak flussence: oh! point. will change. 19:44
that's the one I was looking for. missed it somehow.
19:48 labster joined 19:58 dmol left 20:21 xinming left 20:26 thirdm joined
sorear .u TENSOR 20:27
yoleaux No characters found
awwaiid masak, what is that FIRST business?
diakopter .u 2297 20:30
yoleaux U+2297 CIRCLED TIMES [Sm] (⊗)
diakopter eh
sorear .u 2299
yoleaux U+2299 CIRCLED DOT OPERATOR [Sm] (⊙)
sorear masak: 00D7 has a unicode alias of "cartesian product" 20:31
20:33 xinming joined 20:37 cognominal joined
sorear 2297 "tensor product" may also be appropriate.. 2A2F is rather not 20:38
diakopter o_O 20:39
masak awwaiid: it's a nice way to put a "header" into a print loop. 20:42
awwaiid so those happen onthe first iteration only? 20:44
fancy
masak yep.
PerlJam hence the name :) 20:45
awwaiid the all-caps threw me I guess. used to perl5 all-caps labels being strewn about
PerlJam masak: why didn't you use a FIRST block? Accident of implementation? or you just like the repeated FIRST? 20:46
colomon seems to recall desperately wanting FIRST a few years back, but has never actually used it since it actually got implemented.
sorear awwaiid: it's all caps because it's viewed as a generalization of BEGIN/CHECK/INIT/END
awwaiid interesting
lue still likes now - BEGIN now for execution time 20:47
PerlJam awwaiid: we call those things "phasers" because they are all about when things are executed
awwaiid I tend to discourage/avoid BEGIN/CHECK/INIT/END whenever possible
sorear "phasers" is the term we use now, because they correspond to a generalized notion of "execution phase changes"
PerlJam (in what phase of execution)
awwaiid absorbs twice as fast
sorear after a phaser you can have a block or statement, a syntactic category that we named "blast" because puns 20:48
masak PerlJam: just like the repeated FIRST.
awwaiid excellent priorities, sorear 20:49
sorear awwaiid: don't credit me, they were already blasts when I joined 20:53
20:57 colomon left 20:59 kaare_ left 21:01 fgomez left 21:02 fgomez joined, rindolf left 21:03 skids left 21:05 bruges left 21:07 bruges joined
arnsholt sorear: I think Rakudo calls them blorst, these days =) 21:08
21:08 fgomez left
sorear arnsholt: when i joined STD called them blasts and Rakudo called them blorsts. Still not fixed? :( 21:09
arnsholt Not AFAIK
21:09 fgomez joined
PerlJam blorst still fits ... block or statement 21:10
:)
masak I think 'blast' won. despite someone's touching defense on p6l of 'blorst'. 21:12
21:15 kaare_ joined 21:19 rindolf joined
hoelzro ok, I think I may have figured out how this works 21:20
sort of.
I think it's routine_def (or whatever it's called)
that's at least the part that associates the QAST block with the Perl 6 Block object, I think 21:21
21:22 colomon joined 21:23 spider-mario left
hoelzro ah ha! 21:24
found it!
BOOTSTRAP.pm line 555
21:26 thou left 21:27 thou joined 21:33 thirdm left, thirdm joined 21:34 thirdm left
PerlJam hoelzro: how does that line relate to your problem? 21:34
PerlJam doesn't immediately see it.
hoelzro PerlJam: well, it and 556 show me that Parrot has an "invoke" hook that gets bound to whatever's in the $!do attribute 21:35
this is from digging around in other places, too
21:43 am0c joined 21:47 domidumont1 left, gdey joined
timotimo moritz: any new ideas WRT EXPORT_hash? 21:49
hoelzro (EXPORT improvements)++ 21:51
masak .oO( IMPORT exprovements ) 21:56
hoelzro so, it seems the QAST block gets associated with the code object at compile time... 21:58
...but I'm not sure where it actually gets compiled to bytecode 21:59
22:01 xilo left
hoelzro wow, this is hard =/ 22:02
I should really be doing my Coursera stuff.
or visiting with family.
22:03 lustlife left, thou left
hoelzro moar typo fixes! github.com/rakudo/rakudo/pull/119 22:04
22:05 thou joined 22:07 rindolf left
labster Okay, I've never done a commit to the specs yet, so how does this IO::Path look? gist.github.com/labster/5267241 22:09
22:09 woolfy left, kaare_ left
hoelzro ok, I haven't figured out how to fix the callsame bug (yet) 22:15
but I *do* have a workaround: my $clone = $r.clone; $r does role { method postcircumfix:<( )>(|) { say 'before'; my $result = $clone(); say 'after'; return $result; } }
22:19 xilo joined 22:27 thou left 22:29 thou joined 22:35 skids joined
masak hoelzro: nice. 22:36
awkward, but better than no workaround. :)
'night, #perl6 22:47
dalek kudo/nom: b176ca8 | (Rob Hoelz)++ | src/Perl6/Metamodel/BOOTSTRAP.pm:
More typo fixing
 22:48
kudo/nom: 25a9d6d | (Tobias Leich)++ | src/Perl6/Metamodel/BOOTSTRAP.pm:
Merge pull request #119 from hoelzro/nom

More typo fixing
ecs: 6960683 | (Brent Laabs)++ | S32-setting-library/IO.pod:
Change IO::Path stringification, add new methods.

  .path is now a no-op (so Str.path.path still returns IO::Path)
  .contents is now documented - lists the directory
  .is-absolute, .is-relative, .absolute, .relative added		 22:53
22:55 revdiablo joined, thou left 22:56 thou joined 23:13 PacoAir left
[Coke] rn: ('A' ... 'a').say 23:22
23:22 bluescreen10 left
p6eval niecza v24-35-g5c06e28: OUTPUT«A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a␤» 23:22
..rakudo 25a9d6: OUTPUT«(timeout)»
FROGGS_ rn: ('A' ... 'Z').say 23:24
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«A B C D E F G H I J K L M N O P Q R S T U V W X Y Z␤»
FROGGS_ rn: ('_' ... 'a').say 23:25
p6eval rakudo 25a9d6: OUTPUT«(timeout)»
..niecza v24-35-g5c06e28: OUTPUT«_ ` a␤»
FROGGS_ rn: say '_'.ord
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«95␤»
FROGGS_ rn: say 'a'.ord
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«97␤»
23:25 tgt left
[Coke] rakudo: my %h=('a'=>1,'b'=>2); my $h_ref=\%h; say $h_ref.{'a'} 23:29
p6eval rakudo 25a9d6: OUTPUT«(Any)␤»
23:30 tgt joined
[Coke] masak: please check rt.perl.org/rt3/Ticket/Display.html?id=76462 - I think this is probably correct autoviv behavior now. 23:30
r: eval "{\n\n\n\n\'}"
p6eval rakudo 25a9d6: OUTPUT«===SORRY!===␤Confused␤at /tmp/p_KRuLhO3q:1␤------> eval "{\n\⏏n\n\n\'}"␤ expecting any of:␤ postfix␤»
dalek : a71daf8 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
let -d be parsed as term:filetest rather that an unknown sub with prefix

I added *a lot* debug stuff for this, which can be turned on by setting $DEBUG.		 23:33
FROGGS_ r: say '.'.IO.d 23:39
p6eval rakudo 25a9d6: OUTPUT«True␤»
dalek : e0e5402 | (Tobias Leich)++ | lib/Perl6/P5 (2 files):
dispatch -e "foo" to "foo".IO.e
 23:48
23:50 fgomez left
[Coke] rn: +(23..2300000000) 23:50
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«(timeout)» 23:51
[Coke] rn: 0..^3.3
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: ( no output )
[Coke] rn: say 0..^3.3
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«0..^3.3␤»
pmichaud good evening, #perl6 23:52
[Coke] rn: say 0..^3.3.list
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«0..^1␤»
[Coke] rn: say (0..^3.3).list
p6eval rakudo 25a9d6, niecza v24-35-g5c06e28: OUTPUT«0 1 2 3␤»
[Coke] pmichaud: Allo.
FROGGS_ hi pmichaud
[Coke] wonders if +(23..2300000000) should emit a number quickly. 23:53
FROGGS_ well, it could cheat 23:54
[Coke] I am looking at a patch which makes it cheat, aye.
pmichaud [Coke]: I suspect it wouldn't be too hard to manage, no.
er
[Coke] RT# 82312. seeing if it breaks anything... 23:55
pmichaud if there's an easy + reliable way to emit a number quickly, I'd go for that :)
[Coke] ok. I'll get this applied.
pmichaud not sure if Numeric needs a :D invocant there 23:56
[Coke] test for that? +Range ? 23:57
> +Range 23:59
Cannot look up attributes in a type object

Powered by Cro and the Raku® Programming Language.

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

Thank you!