»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!
Set by sorear on 4 February 2011.
[Coke] yawns. 00:33
[Coke] remains blocked on partcl-nqp for now. 00:39
dalek rl6-roast-data: 083a5ab | coke++ | / (4 files):
today
 00:40
rl6-roast-data: 00bc2db | coke++ | / (3 files):
today
rl6-roast-data: c0d51ce | coke++ | p (2 files):
today
rl6-roast-data: bcf5da5 | coke++ | p (2 files):
today
rl6-roast-data: 8cd608d | coke++ | p (2 files):
today
rl6-roast-data: 770f64f | coke++ | p (2 files):
today
rl6-roast-data: e004f76 | coke++ | / (4 files):
today
TimToady ?""" 01:18
TimToady that would be my kitty, who doesn't know Perl 6 very well yet 01:21
colomon std: ?""" 01:28
p6eval std cfe8d43: OUTPUT«===SORRY!===␤Confused at /tmp/9yM1_q_E9N line 1:␤------> ?""⏏"␤ expecting any of:␤ POST␤ infix or meta-infix␤ infixed function␤ postcircumfix␤ postfix␤ postfix_prefix_meta_operator␤ postop␤ statement modifier
..loop␤Parse failed␤…
MikeFair_ waves to #perl6. 02:15
colomon o/ 02:53
sorear o/ 02:54
tipdbmp r: "abc".chars 04:27
p6eval rakudo d1f4dc: ( no output )
tipdbmp r: say "abc".chars
p6eval rakudo d1f4dc: OUTPUT«3␤»
tipdbmp r: say "за̀пра".chars 04:28
p6eval rakudo d1f4dc: OUTPUT«6␤»
tipdbmp How do you count the characters of a string like the above (not counting the stress mark as a character)? 04:31
sorear r: say "за̀пра".ords
p6eval rakudo d1f4dc: OUTPUT«1079 1072 768 1087 1088 1072␤»
sorear r: say "за̀пра".comb.grep(/<!:Mark>/)
p6eval rakudo d1f4dc: OUTPUT«з а ̀ п р а␤»
sorear n: say "за̀пра".comb.grep(/<!:Mark>/)
p6eval niecza v21-1-ga8aa70b: OUTPUT«з а ̀ п р а␤»
sorear n: say "за̀пра".comb.grep(/<!:M>/) 04:32
p6eval niecza v21-1-ga8aa70b: OUTPUT«з а ̀ п р а␤»
sorear n: say "за̀пра".comb.grep(/<:Gc(!/M/)>/)
p6eval niecza v21-1-ga8aa70b: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/lib/CORE.setting line 1287 (warn @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 266 (Mu.Str @ 15) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting…
sorear n: say "за̀пра".comb.grep(/<:ccc(0)>/)
p6eval niecza v21-1-ga8aa70b: OUTPUT«===SORRY!===␤␤Cannot parse number: A␤ at /home/p6eval/niecza/lib/CORE.setting line 1435 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3525 (ANON @ 11) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3527 (NumSyntax.str2num @ 5…
sorear n: say "за̀пра".comb.grep(/<:ccc<0>>/)
p6eval niecza v21-1-ga8aa70b: OUTPUT«===SORRY!===␤␤Cannot parse number: A␤ at /home/p6eval/niecza/lib/CORE.setting line 1435 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3525 (ANON @ 11) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3527 (NumSyntax.str2num @ 5…
sorear theoretically .chars will do it. 04:33
TimToady we don't have NFG yet
tipdbmp Oh, ok.
TimToady but 5 would be the "correct" answer by spec
sorear n: say "за̀пра".comb.grep(/<:ccc("0")>/) 04:37
p6eval niecza v21-1-ga8aa70b: OUTPUT«з а п р а␤»
sorear n: say "за̀пра".comb.grep(/<:ccc("0")>/).elems 04:38
p6eval niecza v21-1-ga8aa70b: OUTPUT«5␤»
sorear r: say "за̀пра".comb.grep(/<:ccc("0")>/).elems
p6eval rakudo d1f4dc: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 2, near "(\"0\")>/).e"␤»
sorear \o/ 04:39
tipdbmp =) 04:42
sorear of course this will be several orders of magnitude slower than .chars :( 04:46
Su-Shee good morning everyone. 06:25
scott__ welcome 06:30
sorear o/ 06:36
FROGGS good morning 07:07
brrt 'o 07:11
sorear o/
scott__ o/
dalek osystem: ee40694 | ronaldxs++ | META.list:
Add Enc::MIME::Base64 module

See end of README.md for explanation of why Perl 6 wants another MIME Base64 encoder.		 07:22
osystem: b5a2b54 | (Aliaksandr Zahatski)++ | META.list:
Merge pull request #13 from ronaldxs/patch-1

Add Enc::MIME::Base64 module
tadzik haha, I just did the same thing :) 07:28
dalek rlito: 8b845c5 | (Flavio S. Glock)++ | / (3 files):
Perlito5 - javascript: fix context in eval and in anon sub
 08:03
kresike good morning all you happy perl6 people 08:21
tadzik good day kresike-y
kresike hello tadzik o/ 08:22
masak morning, #perl6
kresike hello masak o/
moritz \o masak 08:26
\o tadzik, kresike
sorear o// 08:26
kresike hello moritz o/ 08:33
moritz just created his first object in the RIPE test database 08:35
dalek rlito: da14ca5 | (Flavio S. Glock)++ | / (4 files):
Perlito5 - update Makefile; bump perl version to 5.14.1
 08:36
Su-Shee moritz: so, you're going to bring down the european internet until noon today? :)
moritz Su-Shee: I'm trying hard, yes :-) 08:38
no, it's just the test database so far
Su-Shee moritz: I just wanted to make you nervous. :) 08:40
moritz and suddenly I have to think of www.youtube.com/watch?v=_y36fG2Oba0 08:41
abraxxa i've upgraded rakudo from 2011.01 to 2012.08 and now the line $play_again = $input eq 'j'; result in the error Cannot assign into a PMCNULL container in block at
hoelzro that's quite an upgrade 08:42
moritz r: my $input = 'foo'; my $play_again = $input eq 'j'; say 'alive'
p6eval rakudo d1f4dc: OUTPUT«alive␤»
moritz abraxxa: seems to work here... maybe show more context?
abraxxa: and delete any old .pir files (from precompiled modules) 08:43
Su-Shee well if that's all he's got as a error with an upgrade of a year...
moritz well, it's the *first* error :-)
abraxxa gist.github.com/806301
line 81
it was the first time i touched perl6 and never had time to do so since then 08:44
was a coding contest against a friend who does c#
the original c# code of a co-student had over thousand lines!
moritz it doesn't even show me the first prompt here :* 08:45
:(
Su-Shee abraxxa: "Willkommen bei EuroMillionen"? What are you going to pay for help? :))
isn't there just a ; missing in the line above? 08:47
moritz shouldn't be necessary, because the previous line ends with }
Su-Shee ok.
abraxxa when i comment out line 81 i get Null PMC access in get_bool() in block at ./euromill.pl6:82 08:49
moritz abraxxa: are you sure you're using the right perl6 binary 08:50
?
so many Null PMC accesses are very uncharacteristic of rakudo 2012.08
abraxxa [ahartmai@ahartmai-nb:~/perl6$ ]$ which perl6 -> /home/ahartmai/perl6/rakudo-2012.08/install/bin/perl6
sorear r: gist.github.com/806301
moritz what does perl6 --version say?
p6eval rakudo d1f4dc: OUTPUT«(timeout)Willkommen bei EuroMillionen!␤use of uninitialized value of type Str in string context in block at /tmp/NVkV7LiIcX:34␤␤use of uninitialized value of type Str in string context in block at /tmp/NVkV7LiIcX:34␤␤use of uninitialized value of type Str in st… 08:51
sorear n: gist.github.com/806301
abraxxa This is perl6 version 2012.08 built on parrot 4.4.0 revision RELEASE_4_4_0
p6eval niecza v21-1-ga8aa70b: OUTPUT«(signal XFSZ)Willkommen bei EuroMillionen!␤Wollen sie einen Quicktipp abgeben? j/n: Wollen sie einen Quicktipp abgeben? j/n: Wollen sie einen Quicktipp abgeben? j/n: Wollen sie einen Quicktipp abgeben? j/n: Wollen sie einen Quicktipp abgeben? j/n: Wollen si…
moritz wtf is signal XFSZ?
sorear File size limit exceeded
I did not know you had a disk quota on p6eval 08:52
moritz it's per-process
set in Evalbot::Executer or so
not globally
sorear I guess it just kept spewing messages to the output file until it got to 15M 08:53
.oO( proof of niecza speed )
 08:54
abraxxa so the syntax hasn't change somewhere since i wrote it? 08:56
sorear doesn't look like it 08:59
but I do see one questionable bit
until (my $input = prompt("Bitte geben sie Nummer $_ ein (1-50): ")) ~~ any @allowed_numbers && $input !~~ any @numbers {}
until (my $input = prompt("Bitte geben sie Stern $_ ein (1-9): ")) ~~ any @allowed_stars && $input !~~ any @stars {}
moritz precedence? 08:59
sorear that is parsed as $input ~~ any( @allowed_numbers && $input !~~ any @numbers ) 09:00
a function call without parens swallows everything to the right
moritz ... depending on the precedence of the stuff on the right :-)
sorear I am fairly certain that there is no such dependance 09:02
ah, 'and'
abraxxa the code runs until after the prompt in line 81 09:07
is the repeat while loop ok?
sorear repeat while loops are still valid 09:08
moritz sorear: and, or, feed operators, statement modifiers
sorear that particular repeat while is legal syntax 09:09
abraxxa line 80 is different from the same lines above by not using my before $input
sorear if you fix the precedence thing what happens? 09:10
abraxxa the error remains the same 09:11
sorear you may have found a nom regression./ 09:15
you could try my compiler :D
FROGGS sorear: you'are the only one hacking on niecza?
moritz now please reduce the example to something minimal that still exposes the bug, and submit as a bug report 09:17
moritz somehow prompt doesn't work at all on my machine :( 09:18
UncleFester6 After a few small fixes my new Enc::MIME::Base64 module now installs OK with panda ... 09:19
tadzik awesome 09:20
abraxxa paste.scsys.co.uk/207402 09:21
sorear FROGGS: no, colomon has been contributing a lot lately
and others have done so in the paste
abraxxa that's a minimal test case 09:24
moritz: how should perl6 bugs be submitted? can i transform it to a test case for the testsuite? 09:25
moritz abraxxa: per email to [email@hidden.address]
abraxxa: and yes, a test case would also be nice 09:26
abraxxa is the pasted code sufficient?
moritz no
it contains way too much irrelevant code
abraxxa you might want to add this info to the channel topic
moritz remove everything from it that you don't need to reproduce the bug
abraxxa moritz: i'm talking about this one paste.scsys.co.uk/207402
sorear wonders if that can still be simplified 09:28
e.g. do you really need the outer loop?
or the ~~ any('j', 'n') ?
think about this in terms of features
abraxxa works without the loop
sorear making your test case not use junctions would be a big simplification
abraxxa works also in a block 09:29
moritz does it need prompt?
abraxxa: oh, I didn't see that paste. Much better.
sorear sleep&
abraxxa paste.scsys.co.uk/207406 09:30
moritz abraxxa++ 09:31
abraxxa moritz: no prompt for you ;)
so the error message pointed to the right line
rakudo++
moritz r: my $r; repeat { $r = False } while $r 09:32
p6eval rakudo d1f4dc: OUTPUT«Cannot assign into a PMCNULL container␤ in block at /tmp/uNv11_2WYi:1␤␤»
dalek rlito: 6807ab1 | (Flavio S. Glock)++ | README-perlito5:
Perlito5 - add a note about javascript minification
moritz r: my $r; repeat { } while $r
p6eval rakudo d1f4dc: OUTPUT«Null PMC access in get_bool()␤ in block at /tmp/3zlZmjjYKh:1␤␤»
moritz ooh, different error message :-)
abraxxa looks like a boolification problem
moritz the problem is that likely that $r isn't set up properly
so it's either a problem with assignment to it, or boolification, depending on what happens first 09:33
abraxxa++
abraxxa: do you want to submit the bug, or should I?
abraxxa if that last code snippet is ok for submission i'll do it 09:34
moritz it is
abraxxa sent 09:36
moritz r: my $r; repeat { $r = 1 } while False; 09:38
p6eval rakudo d1f4dc: ( no output )
moritz r: my $r; repeat { $r = 1 } while $r 09:39
p6eval rakudo d1f4dc: OUTPUT«Cannot assign into a PMCNULL container␤ in block at /tmp/cDgkya3ZJR:1␤␤»
moritz so it's dependent on $r being reference in the while condition
moritz TIL that the RIPE database stores poems 10:16
estrai is there an easy way to tell a class to accept only the attributes it knows about (as in defined) in the constructor? 10:40
tadzik not that I know of 10:41
it should be possible to write MuX::StrictConstructor, like Moose has it :)
flussence make a stricter constructor that just passes through to default .new? 10:42
(is the default .new even reachable once you define another one?) 10:43
jnthn nextsame 10:44
arnsholt Something like method new(:$arga!, :$argb!) { nextsame; } should do it I think
jnthn aye
arnsholt Or submethod, if it shouldn't be inherited
jnthn role CheckedNew { method new(*%args) { nextsame; 10:45
oops
flussence r: class A { has $.foo = 0; only method new(:$foo) { nextsame } }; my $a = A.new(:foo(5), :wuz(3)); 10:46
p6eval rakudo d1f4dc: ( no output )
estrai that's awesome, didn't know you can override .new like that
flussence that lack of error seems wrong 10:47
jnthn oh...
methods get an implicit *%_
flussence r: class A { has $.foo = 0; only method new(:$foo) { die if %_; nextsame } }; my $a = A.new(:foo(5), :wuz(3));
p6eval rakudo d1f4dc: OUTPUT«␤ in method new at /tmp/ygSHTT47yY:1␤ in block at /tmp/ygSHTT47yY:1␤␤»
flussence that sounds better 10:48
jnthn r: class A { has $.foo = 0; only method new((% (:$foo)) { nextsame } }; my $a = A.new(:foo(5), :wuz(3));
p6eval rakudo d1f4dc: OUTPUT«===SORRY!===␤Missing block␤at /tmp/dY2PiB6Ds7:1␤»
jnthn grr
r: class A { has $.foo = 0; only method new(*% (:$foo)) { nextsame } }; my $a = A.new(:foo(5), :wuz(3));
p6eval rakudo d1f4dc: OUTPUT«Unexpected named parameter 'wuz' passed in sub-signature␤ in method new at /tmp/UjfWbX8HUQ:1␤ in block at /tmp/UjfWbX8HUQ:1␤␤»
jnthn That also does it :)
flussence jnthn++
jnthn but really, you can write a role that pulls in a new method that does the validation.
(against the object attributres) 10:49
estrai what would be the check I could use?
estrai eval likely, right? 10:51
buubot_backup estrai: ERROR: syntax error at (eval 20) line 2, at EOF
jnthn do self.^attributes to introspect the object's available attributes. 10:54
Only look for those that have an accessor
estrai that's cool but still involves some manual work to remove sigils, is there any other method to retrieve only the attribute names? 10:58
jnthn Don't think so 10:58
otoh, it's only a call to .substr :)
jnthn and you're factoring it out into a role so it's not so much work 10:59
(as in, can re-use the effort)
estrai yy, just checking :) thank you 11:01
daxim someone with a reddit account please reply: www.reddit.com/r/perl/comments/zi3e...l6/c6533z7 "You're probably thinking of [November](www.perlfoundation.org/perl6/index.cgi?november) by viklund and masak." 11:11
sergot hi o/ 12:45
colomon \o 13:31
SmartViking If I don't know perl to begin with, should I focus in learning perl6 or perl5? 13:34
I'm asking for opinions because I have none 13:35
hoelzro SmartViking: Perl6 still isn't considered "production ready"
tadzik SmartViking: what do you want to do with it?
hoelzro Perl 5 is still alive and kicking, and will be for a while
SmartViking tadzik: Just for my own amusement
moritz SmartViking: if you want results fast, Perl 5 is probably better. If you want to learn a nicer language, Perl 6 is better :-)
hoelzro and learning Perl 6 after Perl 5 shouldn't be a problem
tadzik SmartViking: Perl 6 is good for own amusement :)
SmartViking Hm ok now I've got a better idea, thanks :) 13:36
PerlJam SmartViking: so ... which are you going to put effort into learning? :) 13:43
SmartViking PerlJam: I think I'm going to learn perl 6 :) 13:43
PerlJam SmartViking: excellent! 13:44
SmartViking PerlJam: :)
[Coke] SmartViking: welcome aboard. Let us know if you have any questions. 13:51
SmartViking [Coke]: Thanks, I will :)
moritz github fail: cuts off two characters from an overlong message, and replaces it by three dots 14:01
[Coke] moritz: wow, that sounds like something *I* would do. 14:02
PerlJam sounds like a one of the standard rails routines. 14:04
moritz I've even seen that once with a printer for bank statements 14:07
the last page contained something along the lines of "please remove all the pages printed so far, more forthcoming" 14:08
and then there was one more :-)
Woodi rn: sub set( $a ) { say $a.encode.bytes }; set( 3 );
p6eval rakudo d1f4dc: OUTPUT«No such method 'encode' for invocant of type 'Int'␤ in sub set at /tmp/Dzez5aMpI5:1␤ in block at /tmp/Dzez5aMpI5:1␤␤»
..niecza v21-1-ga8aa70b: OUTPUT«Unhandled exception: Unable to resolve method encode in type Int␤ at /tmp/h4O2ThGwCf line 1 (set @ 4) ␤ at /tmp/h4O2ThGwCf line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4138 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting lin…
Woodi it's nqp ? 14:09
jnthn encode is a method on Str
Woodi but $a is Perl varible 14:10
jnthn wat?
Woodi so Int2Str conversion should be automatic I think
moritz not every possible method 14:11
Woodi or it is perl6 here :)
jnthn That's only true for things we explicitly put in Cool.
moritz only for those specced to be present in class Cool
jnthn I'm not sure encode belongs there.
Woodi k, understud
moritz I'm sure it doesn't belong there
jnthn :)
[Coke] crud. I just did a git stash on a branch, did a bisect, checkout'd master, git stash pop... conflicts. is there a way to undo the git stash pop? (are the stash commits stored somewhere I could get a pristine copy?) 14:28
moritz [Coke]: check with 'git stash show' if the stash is still in the stash list 14:30
[Coke]: if yes, you can just 'git checkout .' to reset the working copy to the repo, which undoes the stashing (but discards all other changes you made)
hoelzro git stash drop shouldn't mutate the stash if there are conflicts
[Coke] ah, even better.
moritz [Coke]: or did you want to find the commit from which you did the 'git stash'? 14:31
moritz is a bit confused 14:32
[Coke] no, I wanted the stash back so I could re-apply it cleanly. I just freaked out when I got the conflict. 14:35
all set now, danke. 14:36
moritz++ hoelzro++ 14:38
jnthn goeteborg & 15:11
kresike bye folks 15:27
SmartViking What does the //= operator do? Is it different from =? 15:46
hoelzro SmartViking: //= is "defined or" 15:47
PerlJam SmartViking: $a //= $b is the same as $a = $a // $b
diakopter if the lhs is not defined, it assigns the rhs to the lhs
SmartViking Ok thanks 15:48
What does it mean that it's not defined? 15:49
daxim that the variable has no value
SmartViking daxim: But the variable does exist?
daxim yes
SmartViking Ok thanks
PerlJam SmartViking: when you say "my $foo;" $foo will be undefined until you assign a value to it. 15:50
TimToady this is often used with array elements to make a cache 15:50
SmartViking PerlJam: Ah that makes sense
TimToady in that case, the array exists, but the element won't exist 15:50
SmartViking I'm liking perl 6 though. Seems like a language that will allow me to glu stuff together quickly 15:54
glue*
And get stuff done 15:55
PerlJam That's pretty much the definition of "Perl" 15:55
pmurias jnthn: how do I hook it a js backend to nqp?
SmartViking I've used Python for some years and while it's great, it aimes to be so 'clean' and canonical, it's hard to run a system command for example, and regex isn't that straighforward either which I've heard it is in perl 16:01
PerlJam SmartViking: python tries to shoe-horn regex into strings. Having a little more first-class syntax for them helps. 16:03
SmartViking PerlJam: Yeah. Python isn't that great with regex from the stance of the developer in my experience, it's always been a bit of a hazzle 16:06
SmartViking Excuse my verbosity, bad personality trait 16:10
TimToady verbosity is not a problem here, as long as you're interesting :) 16:21
we find that we all have complementary personality defects here, so as long as it contributes to the overhealth of the culture and the technology, we like defective people here 16:23
*overall health
SmartViking Great :) 16:23
PerlJam SmartViking: and if you're not careful, someone may hug you. 16:25
SmartViking PerlJam: I'll watch my back :)
TimToady there are a few personality defects that we try to discourage, such as the kinds of behaviors that reinforce the Male Only ethos prevalent among geeks
TimToady basically, it's fine to be weird here, as long as it's not an excuse to be antisocial 16:27
SmartViking What does '~~ 'do? 16:29
PerlJam SmartViking: ~~ is smart match.
TimToady in which context? in normal expressions, it's a smartmatch
SmartViking TimToady: Yeah normal expressiosn 16:30
TimToady it's something a little different in regexes, but not yet implemented, actually, I think
TimToady smartmatching is described in S03:3552 16:32
SmartViking Thanks 16:33
TimToady afk & 16:34
sorear good * #perl6 17:01
felher o/ sorear 17:05
jnthn pmurias: Probably similar-ish to Rakudo. 17:24
pmurias: I suspect that QAST::Compiler will want to become QAST::Compiler::Parrot or so, and then write a QAST::Compiler::JS or so. That's how the integration will want to look eventually, I think. 17:25
TimToady nr: proto A(Int, Int) { {*} } 17:28
p6eval niecza v21-1-ga8aa70b: OUTPUT«Potential difficulties:␤ &A is declared but not used at /tmp/bVYHvmhseb line 1:␤------> proto A⏏(Int, Int) { {*} }␤␤» 17:28
..rakudo d1f4dc: OUTPUT«===SORRY!===␤{*} may only appear in proto at line 2, near " }"␤»
TimToady um, that *is* a proto 17:29
jnthn Hm :)
moritz TimToady: no, it's inside a block inside a proto :-)
jnthn What it really wants to tell you is "we only support proto foo(...) { * }"
TimToady and the idea was that we could put stuff in the proto with the {*}, before and after
like, caching
jnthn Yeah. I even implemented and use that in NQP
Kinda funny that NQP has a feature that Rakudo doesn't :) 17:30
TimToady (you may recognized that as the proto for an ackermann function :)
jnthn If somebody want to steal NQP's handling of it, go for it. Otherwise I'll try and get to it when I have a tuit supply again :)
jnthn is currently preoccupied with the talk he's giving tomorrow that ain't really written yet :) 17:30
TimToady of course, niecza's failure mode is to just return Whatever.new there... 17:31
so I guess I prefer rakudo's failure mode on sorear++'s principle of failing early :)
TimToady jnthn: no rush, just didn't want it slipping past into 6.0.0 :) 17:33
dalek rlito: 52de322 | (Flavio S. Glock)++ | / (7 files):
Perlito5 - javascript: add "Javascript2" backend for optimization experiments
 17:35
jnthn TimToady: Yeah, but adding the things I already figured out how to is kinda LHF ;) 17:42
TimToady will it handle: proto A(Int $m, Int $n) { (state @)[$m][$n] //= do {*} } 17:43
jnthn I don't immediately see any reason why not 17:44
TimToady I suppose a question is whether it requires the 'do' there 17:45
jnthn nqp: proto sub foo() { my $x := {*}; say($x) }; multi sub foo() { 42 }; foo() 17:50
p6eval nqp: OUTPUT«42␤»
jnthn Not in NQP, apparently :)
felher (state @) is an persistent array with an empty name?
jnthn yeah
felher k, thnx :) 17:51
sorear It's only persistent relative to the current instance of the current sub
for the most part state variables act like normal variables lifted 1 sub up
jnthn thinks of it as "one per closure clone" 17:52
PerlJam well, you could bind it to a name with a wider scope
sorear sub foo { return sub { push (state @), $_ } } # each time foo is called, a new state @ is created
jnthn aye
felher I see. Thanks :)
TimToady then the other question is, if {*} is just a magic term, can it be used where other blocks are expected, such as loop {*} 17:55
or would that have to be written {{*}} ? 17:57
[Coke] that looks like another way to do something macroish. 18:12
TimToady well, kind of, except it's standing for a dispatcher, not an AST 18:17
it's similar to {YOU_ARE_HERE} in the setting, which also has the brackets to remind you that it's a subscope, but also allow -n to be implemented as: for lines() {YOU_ARE_HERE} 18:18
TimToady that's a little closer to a macro, since the setting could construed as a large macro around your compilation unit 18:20
maybe if we treat that as an actual unquasi, and generalize those to any rule in the grammar, we'll replace {YOU_ARE_HERE} someday with ¤blockoid:UNIT or some such 18:25
and it would be one less special case
I doubt we'd do anything similar to {*} though, since 1) it's not an AST, and 2) it needs to be huffmanly short 18:26
TimToady thinking of the setting as a macro means you could conceivably have a setting that composes multiple compilation units, which is a bit mind blowing 18:28
sirrobert is there any current work going on for a DBI-type module? 18:39
I saw DBIish, but the last commits were 4-5 months ago
moritz is there anything in particular you want from it? 18:40
sirrobert Yes... but I can't remember what it was that wasn't working for me at the moment ... =)
moritz please open a ticket the next time you remember
sirrobert sure thing 18:41
I'll try to re-implement DBI and see what I hit
DBIish
moritz my Perl 6 tuits have decreased, but they are still > 0)
sirrobert heh good
I can try to do some stuff on it too
if I understand it =)
moritz why would you try to reimplement it?
sirrobert I meant re-integrate 18:42
not re-implement
moritz oh
:-)
sirrobert heh
sirrobert moritz: huh, it integrated easily. I guess either I didn't know enough p6 a couple months ago or the architecture changes made it "just work" 19:09
moritz sirrobert: good to hear :-) 19:09
sirrobert: what backend are you using?
sirrobert mysql 19:11
sirrobert I think I'll start a light-weight ORM project 19:12
I didn't see one in the modules list
arnsholt Yeah, that one's a TODO
moritz fwiw I consider the postgresql backend the most robust one
it does proper prepared statements (which the mysql backend doesn't) 19:13
sirrobert huh 19:14
I've never used postgres
sirrobert I'll see if switching makes sense for me 19:14
moritz you should. It's quite a nice database
arnsholt Postgres can be a bit odd at first when coming from MySQL, but on the whole it's awesome 19:15
moritz at least if you start with something newish
sirrobert my old boss was into it (though the team used mysql)
moritz if you have a huge body of mysql data and SQL, switching won't make sense at all
sirrobert nah, it's a small set
arnsholt There was a pretty good post on HN comparing MySQL and Postgres the other day
sirrobert most everything is programmatic (p6-wise right now) and uses adapters 19:16
HN?
arnsholt Hacker News
arnsholt news.ycombinator.com/ 19:16
sirrobert looking
found it; thanks =) 19:17
reading
benabik this? news.ycombinator.com/item?id=4495377
sirrobert I had found news.ycombinator.com/item?id=4387407
but yours looks better, benabik =)
sirrobert pretty cool 19:28
(postgres)
I'll have to take a practical look too, but I like the theoretical emphasis.
arnsholt I think the post you found is pretty reasonable 19:31
The data truncation stuff in MySQL does make a certain kind of sense when you see the DB as "where some app stores its data" 19:32
But if you're in the "lots of apps sharing a data source" it's pretty crazy
sirrobert nod 19:38
[Coke] wonders if we are anywhere near an oracle or ms sql integration 20:10
sirrobert is there native big-int support in rakudo? 20:12
[Coke] "built in" is probably better than "native" when mentioning ints. Yup. 20:14
nr: say 2**100
p6eval rakudo d1f4dc, niecza v21-1-ga8aa70b: OUTPUT«1267650600228229401496703205376␤»
[Coke] nr: say 2**1000
p6eval niecza v21-1-ga8aa70b: OUTPUT«10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196…
..rakudo d1f4dc: OUTPUT«1071508607186267320948425049060001810561404811705533607443750388370351051124936122493198378815695858127594672917553146825187145285692314043598457757469857480393456777482423098542107460506237114187795418215304647498358194126739876755916554394607706291457119647768654…
sirrobert cool
how about base conversions? 20:15
hex to dec, etc?
[Coke] <nod>
sirrobert is there a page somewhere with code examples?
[Coke] r: say :08<10>
p6eval rakudo d1f4dc: OUTPUT«8␤»
sirrobert ooh
[Coke] r: say :08<13>
p6eval rakudo d1f4dc: OUTPUT«11␤»
[Coke] r: say :13<22>
p6eval rakudo d1f4dc: OUTPUT«28␤»
sirrobert wow, it's fast too 20:16
[Coke] sirrobert: there's docs.perl.org, the synopsis, and the spec tests (perl6/roast on github). that's where I usually look
sirrobert I mean, fast enough for me =)
r: say (:10<0xa111ff4efc4c11e18d67080027f5e34b>) ** 10
p6eval rakudo d1f4dc: OUTPUT«2023715812404703679235203663383420543796506403783669618127906032009103508390627719708547814169269664260282428553784381864262027334858930200053064046901569273192257097336014118586511821526010272249171977304148848604687935662228390091491975276477058903425186245122413…
sirrobert thanks, that's excellent 20:17
colomon just figured out a $work bug which had been plaguing him for about six weeks... 20:28
sirrobert w00t =)
colomon something incredibly simple once I figured it out, of course. :) 20:32
sirrobert heh, was playing with randomly-chosen numbers to test things... hit a very weird thing (bug?) 20:36
r: say 24.base(20)
p6eval rakudo d1f4dc: OUTPUT«14␤»
sirrobert r: say 24.base(24)
p6eval rakudo d1f4dc: OUTPUT«10␤»
sirrobert r: say 24.base(36)
p6eval rakudo d1f4dc: OUTPUT«O␤»
sirrobert 0???
but then I realized it was 'O' not 0 =) 20:37
geekosaur hee 20:40
sirrobert =)
geekosaur was just commenting about that kind of thing...
[Coke] . 21:04
Circlepuller r: say 24.base(12) 21:05
p6eval rakudo d1f4dc: OUTPUT«20␤»
Circlepuller r: say 24.base(16)
p6eval rakudo d1f4dc: OUTPUT«18␤»
Circlepuller r: say 16.base(24)
p6eval rakudo d1f4dc: OUTPUT«G␤»
Circlepuller r: say 9.base(16) 21:06
p6eval rakudo d1f4dc: OUTPUT«9␤»
nwc10 Is the ; mandatory after a sub declaration?
ie
std: sub foo {} foo
p6eval std cfe8d43: OUTPUT«===SORRY!===␤Strange text after block (missing comma, semicolon, comment marker?) at /tmp/GpwdzYOEaw line 1:␤------> sub foo {}⏏ foo␤ expecting any of:␤ horizontal whitespace␤ infix or meta-infix␤ infixed function␤ statement
..mo…
benabik I think it's not required if there is a newline.
nwc10 ah OK.
benabik This is more difficult to demonstrate in IRC.
Circlepuller like reg perl
nwc10 that's valid perl 5
Circlepuller ^ 21:07
diakopter p6eval takes a ␤ as a newline
std: sub foo {}␤ foo
p6eval std cfe8d43: OUTPUT«ok 00:00 42m␤»
nwc10 can I do this? 21:08
p5: sub foo {} foo
no, just tumbleweed
Circlepuller no, go to #perl for that
diakopter eval: sub foo {} foo
buubot_backup diakopter: No output.
diakopter eval: sub foo {die} foo 21:09
buubot_backup diakopter: ERROR: Died at (eval 20) line 1.
benabik diakopter++
nwc10 aha thanks 21:10
will it crash with interesting pack templates? :-)
benabik nwc10: We thank you for not abusing our bots. 21:12
nwc10 thanks for the help 21:14
masak good morning, #perl6! 21:20
sorear masak++ 21:21
sorear good #perl6, masak 21:21
jnthn evening o/ 21:22
with free wine
oops, wrong window
masak :P 21:24
jnthn Wrong window is consequence of very laggy connection rather than wine :P 21:25
masak just keep telling yourself that. :P 21:28
masak is on a train, and still gets window changing right!
jnthn
.oO( Wine Is Not an Excuse? )
 21:29
masak :P 21:30
flussence
.oO( I want a loose-precedence "." so I don't need to bounce on the shift/9/0 keys as often... )
 22:47
diakopter std: maybe { } maybe { } 22:53
p6eval std cfe8d43: OUTPUT«===SORRY!===␤Strange text after block (missing comma, semicolon, comment marker?) at /tmp/tiHeHpA6hN line 1:␤------> maybe { }⏏ maybe { }␤ expecting any of:␤ horizontal whitespace␤ infix or meta-infix␤ infixed function␤
..stateme…
diakopter TimToady: std doesn't like S17's maybe
chained, at least
TimToady std: maybe { }; maybe { } 23:06
p6eval std cfe8d43: OUTPUT«ok 00:00 42m␤»
TimToady looks fine to me
std: maybe { }␤ maybe { } 23:07
p6eval std cfe8d43: OUTPUT«ok 00:00 42m␤»
TimToady there are no blocks in Perl 6 that are special-cased not to need a semicolon when followed by something on the same line
we absolutely do not distinguish built-ins from user-defined code at that point 23:08
so S17 is just wrong 23:09
dalek ecs: a9f9c60 | larry++ | S17-concurrency.pod:
uncuddle maybes to avoid need for ;

discovered by diakopter++		 23:12
diakopter here's my number, so call me maybe 23:16
TimToady just added a fancy version to rosettacode.org/wiki/Ackermann_function#Perl_6 that allows us to calculate A(4,2) 23:24
also shows off use of Unicode
diakopter aww; my poor font 23:25
lots o boxes
sorear TimToady: what do we think of S06:1135 these days? 23:26
TimToady it was an attempt to out-dwim Ruby on magic callback syntax, but should probably be marked as conjectural 23:28
TimToady we do seem to be doing pretty well without it 23:29
diakopter how long does A(4,2) take 23:31
TimToady 2.2 seconds in rakudo, 3.1 in niecza, so rakudo wins on this one 23:34
diakopter 3.1s; that's a good length for profiling 23:35
TimToady without the extra cheating terms, A(4,2) chewed up multiple gigs before I killed it
(even with the caching) 23:36
((which I wish I could write in the proto, but oh well))
timotimo (state @)[𝑚][𝑛] //= 𝔸(𝑚,𝑛) - that's a really nice trick (and the copypaste totally botched it i think)
TimToady (((actually, I wish "is cached" did that for me)))
diakopter is memoed? 23:37
TimToady we're trying to speak English, not CSese
timotimo agreed.
TimToady though in this case, an explicit cache with integer indexes is likely to beat an automatic hash-based cache 23:39
and "is cached" can't know whether the integers are going to be compact or sparse
even if declared Int
nor can it know without help that the integers never go negative 23:40
I suppose if it were declare sub A(Uint8, Uint8) or some such, it could figure it out 23:41
diakopter rn: say $_[Nil].WHAT 23:42
p6eval niecza v21-1-ga8aa70b: OUTPUT«Parcel()␤»
..rakudo d1f4dc: OUTPUT«use of uninitialized value of type Nil in numeric context in block at /tmp/r9MvqOmGer:1␤␤Any()␤»
TimToady but then we'd not be feeding it back into the cache with results larger than 256, which wouldn't work
timotimo Uint8? wow, you're optimistic :))
perl6 won't be that amazing, even when finished
diakopter :P 23:43
TimToady (Uint8, Uint) would maybe work
TimToady since we feed A() to itself only in the second argument 23:43
diakopter n: say Inf[0-1] 23:45
p6eval niecza v21-1-ga8aa70b: OUTPUT«Any()␤»
TimToady to a first approximation, though, "is cached" is always going to be hash based
timotimo: but we do have uint8 specced, and Uint8 is just a boxed one of those :) 23:46
diakopter rn: say 1[1]
p6eval rakudo d1f4dc: OUTPUT«Index out of range. Is: 1, should be in 0..0␤ in method gist at src/gen/CORE.setting:9831␤ in sub say at src/gen/CORE.setting:7161␤ in block at /tmp/HYCAiGKW2x:1␤␤»
..niecza v21-1-ga8aa70b: OUTPUT«Any()␤»
timotimo no, i meant, having values up to 256 for the arguments is optimistic ;) 23:47
uh, up to 255 of course 23:48
diakopter timotimo: heh; nothing will be amazing when Perl 6 is finished
TimToady oh, okay, then make it Uint3 or so :) 23:50
timotimo perl6 getting finished will redefine amazing for quite a long time, no?
TimToady and it'll still be amazing :)
timotimo i'm already amazed! 23:51
TimToady iterestingly, the RC "fixed width" font isn't really quite, notice how the mathematical italic m and n are different widths 23:52
TimToady and the double-struck A is a bit wider 23:52
at least in my firefox rendering 23:53
TimToady ponders whether "iterest" should be a word or not
flussence TimToady: I've noticed a few "monospace" fonts aren't, too. Inconsolata isn't and that's one of the ones I like :( 23:58
timotimo you were the one font he liked and you BLEW IT! 23:59
flussence (that case is worse, because it's got varying widths in *latin-1* characters...)

Powered by Cro and the Raku® Programming Language.

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

Thank you!