»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.
rjbs Ah, another channel I forgot to make irssi autojoin after I reboot. 00:10
dalek blets: ea9bc96 | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
fix utf in B and update smartmatch table
 01:05
blets: 40e7bd4 | (Herbert Breunung)++ | docs/appendix-b-grouped.txt:
lost one op
 01:07
diakopter r: my Int $x = Int(9); say $x.WHAT 02:22
camelia rakudo ef73eb: OUTPUT«(Int)␤»
diakopter r: my Int $x = Int(9/1); say $x.WHAT
camelia rakudo ef73eb: OUTPUT«(Int)␤»
diakopter uh-huh.
sorear what surprises you 02:23
diakopter
.oO( to the pyromaniac, everything looks like flamebait... )
<-- the pyro 02:24
diakopter toddr: WAT. 02:24
sorear: well, what shouldn't surprise me?
sorear diakopter: stuff working as designed 02:25
unless I'm missing somethin
diakopter n: my Int $x = Int(9/1); say $x.WHAT
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Unable to resolve method postcircumfix:<( )> in type Int␤ at /tmp/2awbbvbVqZ 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…
diakopter n: my Int $x = (9/1).Int; say $x.WHAT 02:26
camelia niecza v24-37-gf9c8fc2: OUTPUT«(Int)␤»
sorear Int(whatever) is a coercion, just like c++ and javascript
NYI in niecza
diakopter er. I'm a dummy. The .WHAT isn't supposed to be there 02:27
*facepalm*
diakopter r: my Int $x = Int(9); say $x; 02:27
camelia rakudo ef73eb: OUTPUT«9␤»
diakopter r: my Int $x = Int(9/1); say $x
camelia rakudo ef73eb: OUTPUT«9␤»
diakopter oh.
ergh. misreading/mistyping. 02:28
rn: my Associative $spot .= new; $spot 02:30
camelia niecza v24-37-gf9c8fc2: ( no output )
..rakudo ef73eb: OUTPUT«Cannot modify an immutable value␤ in method dispatch:<.=> at src/gen/CORE.setting:1057␤ in any at src/gen/Metamodel.nqp:1352␤ in block at /tmp/L3OLJ03SIp:1␤␤»
diakopter no clue who's right.. sorear, halp? 02:31
sorear Neither one of them. 02:32
diakopter should it give some different error?
sorear rn: my role Assoc { method at_key() { ... } }; Assoc.new
camelia rakudo ef73eb: OUTPUT«Method 'at_key' must be implemented by Assoc because it is required by a role␤ in any at src/gen/Metamodel.nqp:2207␤ in any compose_method_table at src/gen/Metamodel.nqp:2198␤ in any apply at src/gen/Metamodel.nqp:2216␤ in any at src/gen/Metamodel.nqp:2356␤ …
..niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Method 'at_key' must be implemented by 'Assoc' because it is required by role 'Assoc[curried][...]'␤ at /tmp/MoF3Td_sMk line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4299 (ANON @ 3) ␤ at /home/p6eval/niecza/li…
diakopter hunh. 02:33
sorear Associative is a role with required methods. "my Associative $spot .= new" tries to pun it, and should fail at that point.
sorear p: Associative.new 02:33
camelia pugs: OUTPUT«*** No such subroutine: "&Associative"␤ at /tmp/hmBR1NWIRT line 1, column 1 - line 2, column 1␤»
sorear mildew: Associative.new 02:34
perlito: Associative.new
diakopter rn: say (role { }).new
camelia rakudo ef73eb: OUTPUT«<anon>.new()␤»
..niecza v24-37-gf9c8fc2: OUTPUT«ANON.new(...)␤»
diakopter okay.. 02:35
rn: say (role { method new { ... } }).new 02:36
sorear looks correct
camelia niecza v24-37-gf9c8fc2: OUTPUT«Unhandled exception: Method 'new' must be implemented by 'ANON' because it is required by role 'ANON[curried][...]'␤ at /tmp/MnogE2C3Mu line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4299 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/COR…
..rakudo ef73eb: OUTPUT«Method 'new' must be implemented by <anon> because it is required by a role␤ in any at src/gen/Metamodel.nqp:2207␤ in any compose_method_table at src/gen/Metamodel.nqp:2198␤ in any apply at src/gen/Metamodel.nqp:2216␤ in any at src/gen/Metamodel.nqp:2356␤ in…
sorear also looks correct, but the error messages could be improved
diakopter k 02:36
nqp-jvm: say((role boo { method new() { ... } }).new) 02:37
camelia nqp-jvm: OUTPUT«Cannot stringify this␤ in FAILGOAL␤ in blockoid␤ in method_def␤ in routine_declarator:sym<method>␤ in !protoregex␤ in routine_declarator␤ in term:sym<routine_declarator>␤ in !protoregex␤ in term␤ in termish␤ in EXPR␤ in statement␤ in statementlist␤ in block…
diakopter hehee
nqp: say((role boo { method new() { ... } }).new) 02:38
camelia nqp: OUTPUT«Unable to parse expression in blockoid; couldn't find final '}' at line 2, near "... } }).n"␤current instr.: 'panic' pc 14721 (src/stage2/gen/NQPHLL.pir:5232) (src/stage2/gen/NQPHLL.nqp:279)␤»
diakopter oh
nqp-jvm: say((role boo { method new() { nqp::die } }).new)
camelia nqp-jvm: OUTPUT«Operation 'die' requires 1 operands␤ in <anon>␤ in <anon>␤ in compile_op␤ in as_jast␤ in as_jast␤ in <anon>␤ in compile_all_the_stmts␤ in as_jast␤ in as_jast␤ in <anon>␤ in compile_all_the_stmts␤ in as_jast␤ in as_jast␤ in <anon>␤ in compile_all_the_stmts␤…
diakopter nqp-jvm: say((role boo { method new() { nqp::die('') } }).new)
camelia nqp-jvm: OUTPUT«Method 'new' not found␤ in <anon>␤ in <anon>␤ in eval␤ in evalfiles␤ in command_eval␤ in command_line␤ in MAIN␤ in <anon>␤ in <anon>»
diakopter nqp: say((role boo { method new() { nqp::die('') } }).new) 02:38
camelia nqp: OUTPUT«Method 'new' not found for invocant of class 'Sub'␤current instr.: '' pc 56 ((file unknown):44) (/tmp/c67uRzCaPo:1)␤»
diakopter hm, ok
diakopter r: my int @array; @array = 5,6; 02:50
camelia rakudo ef73eb: OUTPUT«No such method 'STORE' for invocant of type 'Int'␤ in block at /tmp/wEtJI29Lly:1␤␤»
diakopter rn: class Sub { }; say(5); sub foo { say 55 }; say &foo.WHAT 03:00
camelia niecza v24-37-gf9c8fc2: OUTPUT«5␤(Sub)␤»
..rakudo ef73eb: OUTPUT«===SORRY!===␤Can not bind attribute '$!signature' declared in class 'Code' with this object␤»
jnthn (from backlog) roles in NQP don't pun into classes (and mostly likely never will get that feature) 06:13
teaching & 06:22
FROGGS morning pals 06:45
diakopter slap gninrom
FROGGS :o) 06:46
rindolf FROGGS: morning. 07:26
FROGGS: what's up?
cognominal :) seen in magit.el: (error "Don't push a detached head. That's gross.") 07:28
FROGGS rindolf: $work 07:29
rindolf FROGGS: ah.
masak cognominal: :) 08:29
morning, #perl6
masak ooh, seems we're live: www.reddit.com/r/perl/comments/1cwh..._readable/ 08:32
I should probably make a blog post, too.
cognominal morning, * 08:33
arnsholt o/ 08:37
mathw o/ 08:40
masak o/ 08:44
mathw Today is the day my new laptop comes. I'm spending a lot of time staring down the road looking for the big brown UPS van 08:47
How is everyone else?
sorear good!
mathw :) 08:49
masak yes, good indeed. 08:51
tadzik hello hello 09:01
mathw: eh, you're tempting me again and again :) 09:02
mathw hehe 09:10
It's not deliberate, I'm just excited 09:11
tadzik I can understand that
mathw As yet another vehicle enters the street which !~ Van::Delivery::UPS 09:14
masak surely you mean !~~ 09:17
mathw probably 09:20
I'm not exactly concentrating very well this morning
Also been writing far too much C# lately 09:21
masak :)
mathw: I'm gaining respect for C# day by day.
mathw: especially the things Microsoft Research are doing with the language.
mathw it has some very, very cool stuff in it 09:22
masak yeah.
mathw: if you haven't looked into Rx yet, you probably should.
mathw In my last job once we started using Reactive Extensions and some stuff built on that, UI programming became glorious 09:23
masak heh :)
mathw It's not so obviously useful in the new job which is all web stuff
masak you know of it, then. good.
mathw heck yes :)
I'm a bit of an FRP enthusiast
so I simply had to check it out
and then today I discovered Elm 09:24
which is a Haskell-derived FRP solution for building client-side web stuff
masak ooh 09:29
mathw Very interested in that kind of thing right now, since I find JavaScript so unpleasant 09:33
although AngularJS let me build a basic pure-client-side web app with reasonable ease (especially since I actually wrote it in CoffeeScript :)) 09:34
spider-mario CoffeeScript has its own problems, though 09:35
mathw true
but it was preferable to JavaScript for that particular project
spider-mario I don’t like its variable scoping, for example
mathw what don't you like about it? 09:36
spider-mario the lack of a keyword for declaring variables
I don’t like that in Python and Ruby either, and that’s actually one of the reasons why I prefer Perl 09:37
superficial as it sounds.
hoelzro spider-mario: that's one thing I strongly prefer about Perl as well 09:38
its scoping rules make sense 09:39
FROGGS and the topping are the dynamic variables... 09:40
mathw yeah I do like having explicit variable declarations
I like it even better if they're subject to static type checking, which CoffeeScript does not give me
masak mathw: I teach a lot of JavaScript. it *is* a pleasant language, if you navigate the reefs a bit. and if you're prepared to lift much of the complexity stack yourself. 09:41
mathw: check out KnockoutJS if you haven't already.
it's not a framework like AngularJS, but it's a nice implementation of the Observer pattern. 09:42
mathw I've heard about it 09:44
I'll take a look
masak in my course, I usually demo it through learn.knockoutjs.com/ -- it's a nice intro. 09:45
mathw just did the first lesson 09:49
looks really nice
might re-implement my AngularJS thing in it to see how they compare :)
masak I would be interested in hearing about your findings. I haven't used AngularJS at all. 09:53
my impression from a distance is that KnockoutJS is more "library" and AngularJS is more "framework". 09:54
mathw maybe
masak (with no value judgement beyond that. slightly different niches, is all) 09:55
mathw heh 09:57
I like their test data
meal options: standard (sandwich), premium (lobster), ultimate (whole zebra)
not sure that'd fit on an airline meal tray
masak this is why it's important to keep your seat upright while eating. 09:59
to fit the whole zebra on the tray table.
mathw isn't it so the person behind you can fit the zebra on their tray table? 10:04
masak yes, of course. 10:05
I was assuming the usual axiom of golden-rule transitivity here.
daxim airplane seats should be arranged in a circle so that everyone can kick the seat in front to annoy its occupant 10:07
masak make the plane donut-shaped, while we're at it ;) 10:08
mathw That works for helicopters 10:10
if you put iwngs on the outside and spin it fast enough it should take off
you might need some fairly elaborate systems to help the pilot see where it's going though
masak the passengers will be so delighted by the sensation of flying very fast in a tight circle, that they will all order the zebra.
mathw but there's only room on board for twenty-four zebras! 10:11
masak of course, you could use some kind of ball-bearing design, so that the insides are actually stationary...
tadzik bull-bear desin
moritz full-bear design? 10:26
tadzik All black and brown and covered in hair
daxim with the camelia fair! 10:27
masak moritz! \o/ 10:33
masak moritz: #masakism is a thing now. could you ilogger2 it? :) 10:33
moritz masak, no just _ilbot it :-) 10:36
moritz [X] done 10:40
masak moritz++ 10:48
(for those who don't know what #masakism is about -- here's the context: gist.github.com/masak/5431185 ) 10:53
mathw hehe 10:58
elegant code in PERL???
you must be sane, sir :)
My colleagues don't believe it's possible 10:59
masak .oO( well, maybe not if you code in PERL, not Perl... ) 11:00
moritz that's when you start calling them cow-orkers :-)
mathw that was emphasis, not error :P
masak moritz suggests a new mailing list: perl6-freezers. 11:04
for people interested in bringing the spec to a frozen state.
moritz which would be like p6l, but without all the design astronautism 11:05
mathw would there still be yak shaving? 11:06
masak it could be a platform for good work like the one lizmat++ is doing.
cognominal What is the rationale behind the systematic use of named parameters in its A{I instead of positional ones? 11:56
masak cognominal: parse failure near "in its A{I". could you rephrase? 12:00
cognominal: are you mauvaise-fenetring again? :)
hoelzro is jnthn around?
masak no, he's at $dayjob. 12:01
hoelzro =(
I have a question about NativeCall, but it can wait
brrt don't wait 12:10
lizmat masak: I didn't realise that I'm working towards having the spec in a frozen state ?
hoelzro well, it's regarding a very...odd behavior in a module I'm writing that uses NativeCall 12:11
nwc10 lizmat: I read it as "you're working towards something similarly useful"
ie "aiming *at* something" and "trying hard not to go in circles or get distracted"
lizmat I see… ok I guess I am… :-)
away again& 12:12
cognominal oops 12:13
What is the rationale behind the QAST API systematic use of named parameters instead of positional ones? 12:14
FROGGS cognominal: I'd say that if you have quiet a bunch optional parameters then these should be named ones 12:16
so you dont have to pass 10 positionals, if ten usecases exist which want to set one (named) argument 12:17
cognominal ok, I get the logic 12:20
FROGGS cognominal: but that is just my opinion...
cognominal contrary to what I said there are positional parameter to constructors but they represent child nodes 12:21
reading github.com/perl6/nqp/blob/master/d...t.markdown
masak lizmat: what nwc10 said. in order to be able to freeze the spec, we must align it with reality/implementation, and remove fossils and stuff. 12:24
lizmat: as opposed to "hey, I've got this wonderful idea that really should be spec!" 12:25
cognominal and this new constructor is inherited by derived nodes github.com/perl6/nqp/blob/master/s...de.nqp#L12
I did not know that .new could be inherited. 12:26
nwc10 masak: effectively, also somewhat triage it?
in terms of "too hard to implement before launch", "must be implemented before launch", "would be nice, but not essential"
and concentrate on the middle
masak that will almost have to be part of perl6-freezers' mission. 12:27
triage comes very hard in these circles.
it's one of our biggest weaknesses.
since we don't really have a central authority. the closest we have is a set of core members. 12:28
nwc10 I think a chunk of it *can* be done at consideration of implementation 12:29
"too hard to implement" is relatively easy to tag
and then it's between the other two
masak yeah. 12:33
the spec already has "Conjectural" markers for some things which are either not very well-explored or deemed to hard for 6.0.0
cognominal Also there is "performance wise irrealistic"
pmichaud good morning, #perl6 12:34
cognominal Should we use github comments for that? I think one can comment line by line?
FROGGS hi pmichaud
cognominal ...or not?
hi pmichaud 12:35
pmichaud I also wonder if at some point we should consider the 6.0.0 "spec" in the same vein as an Internet STD -- i.e., it comes after implementations, not before.
masak yeah.
pmichaud and for a really interesting exercise, create the spec from the roast tests. 12:36
(since they officially define the language, at least according to S01)
also, I still think we should separate "Synopses" and "spec" 12:37
i.e., the Synopses have conjectural markers, the 6.0.0 minimal specification might not have them.
cognominal I discover pulse : github.com/perl6/nqp/pulse 12:38
FROGGS cognominal: github's line comments are cool, but one might loose the overview over issues 12:40
nwc10 there is an option not on the drop-down menu github.com/perl6/nqp/pulse/yearly
pmichaud Last year I had some preliminary designs for a toolset that would be able to easily pick out sections of the synopses to produce specific views of them, and still keep them in sync with changes.
e.g., an extract that contains only the sections that Rakudo implements, or (in this conversation) the 6.0.0 spec. 12:41
nwc10 I like the graphs on ohloh better www.ohloh.net/p/nqp
pmichaud oh, is ohloh not up-to-date on its stats? 12:42
pmichaud fixes.
spider-mario I like ohloh comments such as “with a well-commented source code” 12:43
cognominal FROGGS: in fact, I don't remember how they work. Searching. 12:44
pmichaud hopefully the ohloh stats will be up-to-date soon. 12:45
checking other projects
masak raiph: www.reddit.com/r/perl/comments/1cwh..._readable/ -- "via IRC and email". I'm actually not planning to use email. 13:26
gtodd how to automagically build rakudo using parrot-5.3? twiddle with Configure.pl/config.default? add --gen-parrot=5.3.0 when I run perl ./Conffgure.pl ?? 13:39
or wait for official change to occur :-) 13:40
gtodd /usr/local/bin/perl Configure.pl --gen-parrot=RELEASE_5_3_0 --gen-nqp --prefix=$HOME/perl6 seems to work 13:45
arnsholt Awesome module idea of the moment: (De)serialize arbitrary objects to/from a graph database using MOP introspection 13:46
gtodd sort of like that ZODB thingie in python? 13:51
arnsholt Yeah, that looks pretty close to what I really want instead of ORM systems 13:53
gtodd hmm speed could happen :-) 13:54
argh ... always errors with new parrot build 13:55
[Coke] are you on OS X? 13:56
gtodd freebsd 14:15
discussing a bit with Util in #parrot
gtodd [Coke]: Using Configure.pl with --gen-parrot=RELEASE_5_3_0 to build everything there's some breakage trying to build pod files 14:25
gtodd [Coke]: the build cd's into parrot/docs and tries to make packfile-c.pod and ops/bit.pod but does something strange 14:27
when trying to run" /usr/local/bin/perldoc -ud packfile-c.pod ../src/packfile/api.c
gtodd from rakudo's Configure.pl it cd's into parrot/docs and then it seems we try to build docs for "/usr/local/bin/perldoc" and "-ud" and "packfile-c.pod " instead of building "packfile-c.pod" from ../src/packfile/api.c using "/usr/local/bin/perldoc -ud ... as if the ARGS were shifted wrong somewhere 14:28
[Coke]: that is all :-) 14:29
census that section of my code is Time::HiRes::sleep(0.01*int(rand(50)+100*$num)) #$num is the last two digits of the birth year 14:54
open (my $n, '<', 'file.txt') or die "Cannot open file.txt $!"; while ($<n>) {next unless $_; $name = $; last; } close $n; 14:55
oops
forget that!
i was writing notes to myself :( 14:56
wrong window
rjbs seen TimToady 15:34
diakopter .seen TimToady 15:36
yoleaux I saw TimToady 11 Apr 2013 17:07Z in #perl6: <TimToady> thanks
jnthn evening o/ 16:17
diakopter hi
jnthn++
census hi!
jnthn ...what for? :)
diakopter jnthn++ 16:19
for karma!
moritz general awesomeness in Perl 6 context
census :)
#perl6++
jnthn :) 16:20
Last day of teaching tomorrow, and on Thursday I can spend the day on Perl 6 things :) 16:21
moritz \o/ 16:22
jnthn Things I learend today: 1) Boeing 717s exist. 2) They are still being flown. 3) I will fly on one next month. o.O
tadzik :D 16:23
jnthn And 2...propeller planes...
nwc10 where are you going to?
jnthn KOK
jnthn (Which is the airport in Kokkola, which is in the north of Finland.0 16:24
The place name looks terribly like koka kola... :)
jnthn wonders how many people just hit google maps up :P 16:26
[Coke] gtodd: does your build dir have a space or some other non-alpha-non-ascii char? 16:29
cognominal Boing 717 dates from 1995. Is that so old for a plane?
gtodd [Coke]: no 16:30
[Coke] r: say 2013-1995 ~ " years, YES"
camelia rakudo ef73eb: OUTPUT«18 years, YES␤»
gfldex Kokkola seams to be the right place to go if one needs cobalt or zinc, or wants to go to the only town in a region 16:30
gtodd [Coke]: and it works with parrot RELEASE_5_2_0
jnthn cognominal: huh, I thought from the number it would be much older... 16:31
Like, the one that came before the 727, which in turn came before the 737.. 16:32
gtodd jnthn: you may still be able to ski at that time 16:32
jnthn So basically, version numbers are useless indicators of any kind of anything. :)
diakopter well they didn't all go in order
[Coke] gtodd: ok. any chance you can open a ticket with parrot/parrot on github?
gtodd will do if it's not just me :-) ... just trying to make sure it is not an new perl or old groff perldoc related thing there was a perldoc UTF8 bug in 5.16 16:36
rt.cpan.org/Public/Bug/Display.html?id=77465 not sure why that would break with perldoc on parrot 5.3 but not with parrot 5.2 though 16:38
cognominal imagine perl A300-600T, the perl cargo compiler for big programs, also named beluga. btw, we did spot an airbus beluga with jnthn when we went back and forth between Toulouse and Blagnac in the way to Albi with copious indirecttions 16:39
gtodd [Coke]: building parrot all by itself: cd parrot ./perl Configure.pl ;make ... works fine ... but "make docs" then breaks but then (inside parrot) ... cd docs ; make then errm works fine ? 16:42
boggle
cognominal from the front, the beluga has an odd hydrocephalic look www.youtube.com/watch?v=UwLI4QDRM-A
jnthn time to find some dinner, bbl & 16:48
diakopter jnthn: 717-200 entered service long after (1999) the 767. 717-100 was the KC-135 stratotanker (1957) 17:08
GlitchMr > grep('a', 'a', 'b', 'c', 'd'|'a', 'e', 'a', 'abc').perl 17:28
("a", any("d", "a"), "a").list
Smartmatching is so interesting
moritz what did you expect? 17:29
GlitchMr I didn't expected it to work with subroutine. 17:30
Looks like Perl 5 brought me some confusion.
I always was doing grep * ~~ /something/... without even knowing it uses smartmatching.
moritz the dir() function internally uses something like readdir().grep(none('.', '..')) 17:37
which is a case where junctions, smartmatching and grep just work together really well 17:38
jercos 'lo gentlefolks. So basically either I'm completely confused about perl6 regex, or rakudo 2013.02 is doing something odd... My regex is /^ \d+ 'd' \d+ $/, which in my mind should match a die class string like "6d6". 17:56
But that helpfully informs me there's a "Missing block" just before the second \d
tadzik the error must be somewhere else 17:57
r: '4d20' ~~ /^ \d+ 'd' \d+ $/; say $/ 17:58
camelia rakudo ef73eb: OUTPUT«｢4d20｣␤␤»
tadzik jercos: can you paste the entire code somewhere?
jercos shoot, you're right, the regex before it used a perl5-style case insensitive flag >.> 17:59
thanks :p
I spent entirely too much time trying to "fix" that regex.
masak std: /foo/i
camelia std 86b102f: OUTPUT«===SORRY!===␤Unsupported use of suffix regex modifiers; in Perl 6 please use prefix adverbs at /tmp/eGtUaCElo5 line 1:␤------> /foo/i⏏<EOL>␤Other potential difficulties:␤ Unsupported use of /i; in Perl 6 please use :i at /tmp/eGt…
jercos Which is odd, given I didn't get that error... 18:01
masak r: /foo/i 18:02
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Unsupported use of /i; in Perl 6 please use :i␤at /tmp/zMn4NtUW2Y:1␤------> /foo/i⏏<EOL>␤»
isBEKaml r: /foo/i
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Unsupported use of /i; in Perl 6 please use :i␤at /tmp/MZpCsXwdQ9:1␤------> /foo/i⏏<EOL>␤»
jercos well, for educational purposes, here's the code :p pastebin.com/3QzNqYsx
or rather, the not-code, as evidenced by that random mess of variables that don't exist any more. 18:03
cognominal jercos, you will discover that if you plan ahead, you are better moving to a grammar and the associated actions 18:04
masak r: given "42" { when /^'0x'?<[0..9a..f]>+?$/i { say "OH HAI" } } 18:05
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Unable to parse regex; couldn't find final '/'␤at /tmp/JQO87X2ZKm:1␤------> '0x'?<[0..9a..f]>+?$/i { say "OH HAI" } ⏏}␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ s…
flussence r: /^'0x'?<[0..9a..f]>+?$/i
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Regex not terminated␤at /tmp/NClLoxBYvm:1␤------> /^'0x'?<[0..9a..f]>+?$/i⏏<EOL>␤ expecting any of:␤ statement list␤ prefix or term␤ prefix or meta-prefix␤ single quotes␤ infix …
flussence r: /^'0x'?<[0..9a..f]>+?$ /i
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Unsupported use of /i; in Perl 6 please use :i␤at /tmp/e2iVRXhhFC:1␤------> /^'0x'?<[0..9a..f]>+?$ /i⏏<EOL>␤»
jercos cognominal: could you clarify that?
masak cognominal: I agree in general and in this particular case.
flussence the $/ confuses it
masak cognominal: er, "disagree"/
cognominal: grammars are very useful, but they're not a given until you reach a certain level of complexity. 18:06
jercos Oh, never mind, I gotcha. Sorry, I got woken up at 6AM by some random router going down, I'm quite out of it.
cognominal plan ahead is de.webfail.com/2d891ddaad4
masak: anyway, to learn. It's better to start with regexen instead of full fledged grammars 18:07
jercos cognominal: currently straight-over porting a piece of perl5 code with some small changes in idiom as part of my journey into learning perl6.
So while a grammar might be the correct tool for the job, it's a bit out-of-scope for this project :p 18:08
cognominal jercos, that's a good way to start.
I tend to want to go too fast and to fall flat on my face. 18:09
masak cognominal: saying "you need grammars, not regexes" is like saying "you need classes/objects, not nested data structures". it's true some of the time and not true some of the time. 18:11
jercos Heh, well, I hear that :)
this nice self-pacing is a change in pace. 18:12
dalek rl6-roast-data: 2aff1db | coke++ | / (4 files):
today (automated commit)
 18:20
diakopter r: my \無 := Mu; say 無 19:00
camelia rakudo ef73eb: OUTPUT«(Mu)␤»
jnthn back 19:28
[Coke] jnthn: are we there yet? are we there yet? are we there yet? ;) 19:46
FROGGS hehe
jnthn :P 19:47
No, but I did find non-lager beer tonight :)
FROGGS so the driver has beer, ehh? 19:49
O.o
FROGGS fastens his seat belt
japhb imagines a bar that will serve any beer except lager 19:53
tadzik that's my future bar 19:54
jnthn japhb: I've heard of some that overprice lager to put off that kind of clientelle :)
tadzik . o O ( A lager walks into a bar. "We don't serve your kind here" )
FROGGS how do one walk into a "beyond all repair?" 19:55
geekosaur clearly not a sysadmin there :p 19:58
tadzik haha 20:04
dalek blets: 7b796e4 | (Herbert Breunung)++ | docs/appendix-a-index.txt:
added rify to A
 20:06
dalek kudo/jvm-support: 61e1e2a | jonathan++ | src/ (3 files):
Prepare for getting rid of Parrot property use.

We use them in a few places to hang things off of code objects, but there's no reason they can't just be in an attribute.		 20:23
kudo/jvm-support: ab9a836 | jonathan++ | src/Perl6/ (2 files):
First pass at eliminating setprop usage.

Gets rid of them all, but unfortunately breaks the build.
mykle taps microphone ... 20:31
FROGGS ya, we can hear you 20:32
mykle howdy. i have a perl6 problem & no idea where to start with it ...
[Coke] this is an excellent place to start. 20:33
what's wrong?
FROGGS well, one option is to tell us :o)
mykle it looks like on OS X, the Rakudo compiler i built today can't open internet sockets.
and/or something else is wrong with HTTP::Client
i installed it with panda, but the tests didn't run, 20:34
FROGGS there is an open parrot ticket for IO::Socket::INET issues :/
mykle aha. where's the ticket system live?
[Coke] for parrot, github.com/parrot/parrot/issues
mykle newbie question: rakudo uses parrot? 20:35
[Coke] Yes, it's built on top of nqp, which is built on parrot.
nqp was just recently ported to also run on the JVM.
FROGGS github.com/parrot/parrot/issues/942
[Coke] (rakudo hasn't quite caught up to that) 20:36
mykle when i built rakudo and ran the tests, i had a few test failures. is that normal?
t/spec/S32-num/is-prime.t ................................. Failed 6/35 subtests
[Coke] on OS X, unfortunately, yes. 20:36
is-prime is common. I have others related to ICU that don't happen to everyone. 20:37
mykle those were the only errs.
[Coke] (it is hard to fudge rakudo tests for just one platform>)
mykle meta-q: is rakudo the right way to run perl6 on os x?
[Coke] it's a way. You can also try niecza which runs on mono, or pugs that runs on haskell. rakudo is probably the most complete implementation at this point. 20:38
FROGGS niecza isnt bad though, but pugs is pretty outdated 20:39
[Coke] mykle: I don't actually see a ticket for the is-prime issue. If you send an email to [email@hidden.address] it'll open a ticket. (be sure to include your OS version and the output from the failed test.)
prn: say is-prime(1234567)
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«False␤»
..pugs: OUTPUT«*** No such subroutine: "&is-prime"␤ at /tmp/w25Y1ZwIbw line 1, column 5 - line 2, column 1␤»
[Coke] e.g. 20:40
mykle will do, much thanks.
froggs, i found this on github, not much else that matched: github.com/parrot/parrot/issues/942
is that what you meant?
[Coke] mykle: he actually gave you that URL about 24 lines ago. ;)
mykle doi! 20:41
thx.
FROGGS :o)
so, yes
mykle much thanks, everybody. 20:42
ohbtw: is there a make command to just run one test & get its output? all i did was "make test" for the whole suite. 20:43
flussence you can make path/to/test.t too
FROGGS jnthn: is it possible that nqp::loadbytecode() behaves differently if I invoke it again for the same file?
mykle: make sometests TESTFILES=t/spec/S11-modules/require.t 20:44
FROGGS this has the cool fudgy-stuff too 20:44
jnthn FROGGS: Sure, we don't duplicately load moduloes. A bunch of stuff in ModuleLoader takes care of that.
FROGGS hmmmm, sad that is 20:45
jnthn: I am currently working on: { use v5 }; { use v5 }
mykle cool. the first recipe seems to give all the output. 20:45
FROGGS I mean, if the module isnt precompiled, I could just re-eval it
jnthn: I'd like to do the same for precompiled modules 20:46
jnthn FROGGS: Sounds like you want to be writing an EXPORT sub or so, which is certainly run every load. 20:47
FROGGS: Uh, every use I mean.
FROGGS ohh, nice
grondilu Please suggest a better way of coding SHA-256: rosettacode.org/wiki/Talk:SHA-256
FROGGS jnthn: and EXPORT subs are working too for nqp modules? 20:56
mine doesnt get called :/ 20:57
jnthn Oh...
No, just Rakudo at present.
FROGGS hmmm
so, back to the beginning *g*
jnthn May not be hard to update it to do it though. 20:58
FROGGS can you point me at the right place?
do_import?
skids grondilu: I implemented SHA-256 in Sum already. 20:59
jnthn That sounds like it...I'd check where Rakudo does it, as I think things are factored similar
Search for &EXPORT or so
FROGGS k, thanks!
ya, do_import
skids grondilu: All the Sum stuff is designed such that it should not in theory take it all into memory, but last I checked rakudo did due to some extra unwanted laziness. 21:00
dalek blets: 0866fd0 | (Herbert Breunung)++ | docs/appendix-a-index.txt:
added .iterator to A
 21:04
jnthn
.oO( iterator doesn't start with A... )
 21:05
skids decommute
dalek blets: e76261a | (Herbert Breunung)++ | docs/appendix- (2 files):
added .infinite to A
 21:12
lichtkind jnthn: i ment the tablet index A
lue hello o/ 21:23
jnthn 'night, #perl6 21:28
colomon \o
masak 'night, jnthn.
colomon masak: does Text::CSV only read files? 21:32
masak good question.
it's years since I wrote it.
masak checks
colomon I'm not seeing anything else in there.
masak yeah, it only reads files. 21:33
it doesn't feel all that unreasonable that it should be able to write them, too.
colomon not a big deal, but I've just taken "use Text::CSV" out of my code. ;)
masak heh :)
adding that functionality would mean thinking a bit about how to fit it into the current API, though.
colomon yes 21:34
FROGGS r: module A { sub EXPORT { say 1 } }; import A; 21:35
camelia rakudo ef73eb: ( no output )
FROGGS r: module A { sub EXPORT { say 1 } }; import A <x>;
camelia rakudo ef73eb: ( no output )
FROGGS masak: jnthn said that EXPORT sub are always called when I use a module... but locally this doesnt work... do you have a one-liner for me? 21:36
subs*
masak 'fraid not. 21:40
maybe look at the spectests?
FROGGS searching for EXPORT shows nothing useful 21:41
it only checks that things are inside GLOBAL::Module::EXPORT::... after importing
masak I'm not sure I've heard about an EXPORT sub before...
yeah.
that sounds more familiar. 21:42
FROGGS r: use Test <x>;
camelia rakudo ef73eb: OUTPUT«===SORRY!===␤Error while importing from 'Test': no EXPORT sub, but you provided positional argument in the 'use' statement␤»
FROGGS no EXPORT sub...
lue though arglist case of use was still NYI. Huh. 21:44
FROGGS okay, declaring a sub EXPORT($@a) { ... does it, but putting it in a module-scope doesnt work 22:08
diakopter FROGGS: our scope? 22:25
FROGGS maybe
diakopter er
method EXPORT or sub EXPORT
FROGGS sub
lichtkind moritz: ping 22:32
FROGGS: du kannst doch bisserl p5 ? :)
FROGGS lichtkind: a bit :o) 22:35
lichtkind FROGGS: i just rmembered that we have to wrench moritz p6doc into the module Perl6::Doc and subsequently into grok and btw updates the tablets entries too 22:36
grondilu rn: subset Prime of Int where &is-prime; say 7 ~~ Prime 22:39
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«True␤»
japhb rn: subset Prime of Int where &is-prime; say 12 ~~ Prime 22:50
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«False␤»
japhb (Well, at least they got those two right.)
flussence
.oO( it'd be nice if I could do things like «$doc = require 'whatever.pod'; foo($doc.WHY)» )
 22:51
masak 'night, #perl6 22:56
FROGGS gnight masak
flussence ooh, this is a horrible hack but it works... 22:58
my $doc = eval slurp('whatever.pod') ~ "\n$=pod" 22:59
diakopter rn: subset Prime of Int where &is-prime; say 2**57885161-1 ~~ Prime 23:00
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«(timeout)»
diakopter rn: subset Prime of Int where &is-prime; say 2**521-1 ~~ Prime 23:01
flussence (well, it worked when that line was manually edited into the file... not quite as good in code though)
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«True␤»
diakopter rn: subset Prime of Int where &is-prime; say 2**607-1 ~~ Prime
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«True␤»
FROGGS jnthn++ # that EXPORT hint solved my problem 23:02
diakopter rn: subset Prime of Int where &is-prime; say 2**1279-1 ~~ Prime
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«True␤» 23:03
diakopter rn: subset Prime of Int where &is-prime; say 2**2203-1 ~~ Prime 23:05
flussence /facepalm... of course it won't work without escaping the $... 23:05
camelia niecza v24-37-gf9c8fc2: OUTPUT«(timeout)»
..rakudo ef73eb: OUTPUT«True␤»
census hi! is anybody university affiliated here? 23:07
lichtkind was there a reason to call the method "bounds" and not "list" 23:08
FROGGS gnight all 23:10
lichtkind sleep well FROGGS
dalek blets: 861d018 | (Herbert Breunung)++ | docs/appendix-a-index.txt:
added .bounds to A
 23:16
census good night FROGGS 23:17
grondilu diakopter: I wrote this Prime subset as a way to suggest a syntactic alternative to dealing with prime numbers in Perl6. I was not questioning the semantics, actually. I actually have no idea on how &is-prime works. 23:34
diakopter I thought it used native libraries
in both rakudo and niecza
japhb One of the probabilistic methods, I believe 23:36
(I don't recall if the test count was properly tuned to have very low failure likelihood across the entire range of Int that Rakudo could handle on real hardware, but the fact that it times out on very large inputs seems to indicate there is size-based tuning of some sort.) 23:38
diakopter it's timing out on the multiplication, I'm sure 23:39
er, exponentiation
rn: 2**57885161-1
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«(timeout)» 23:40
BenGoldberg IIRC, is-prime is spec'ed to use rabin-miller 23:42
err, miller-rabin
japhb diakopter, yes, that is quite possible -- and means we still don't know if the is-prime test is properly tuned yet. (I don't have the tuits to research that at the moment.)
diakopter what do you mean tuned 23:45
BenGoldberg r: say 5.expmod(7,13) 23:46
camelia rakudo ef73eb: OUTPUT«8␤»
BenGoldberg Is perl6's &is-prime is using exponentiation, rather than expmod? 23:47
japhb A probabilistic test needs to run a certain number of tests to reduce the likelihood of incorrect result below some threshold P. But only the user of the test can say what P should be, so it is my understanding that many libraries make P (or equivalently, the iteration count) one of the inputs to the test library. 23:48
BenGoldberg rn: 2.expmod(57885161, 3)
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: ( no output )
BenGoldberg rn: say 2.expmod(57885161, 3)
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«2␤»
BenGoldberg rn: say 2 ** 57885161 % 3 23:49
camelia rakudo ef73eb, niecza v24-37-gf9c8fc2: OUTPUT«(timeout)»
BenGoldberg If perl6 were to automatically turn a**b%c into a.expmod(b,c), there's a chance that the problem would be avoided 23:50
japhb BenGoldberg, note that the miller-rabin implementations may actually be in C code. 23:53
So Perl 6 is having to exponentiate the huge number and pass it off to a black box -- no opportunity to rewrite. 23:54

Powered by Cro and the Raku® Programming Language.

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

Thank you!