»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! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
Limbic_Region ;-) 00:00
TimToady yowsers, not even my browser has the \o/ character...
rjbs TimToady: It's one of the new Unicode 6 Emoji, I believe.
TimToady: You know about PILE OF POO, right?
I've been waiting for someone to write a PILE OF POO operator for p6... but then we'd end up with hyperpoo -- but maybe a poo-reduce would help with that problem. 00:01
TimToady Limbic_Region: rosettacode.org/wiki/Pattern_matching is really about red-black trees
Limbic_Region chuckles and steps AFK &
TimToady - I am looking at AVLs, red-black trees, etc - all of the self-balancing binary trees 00:02
TimToady that's a problem with rosettacode; it's not well indexed by what you really want to look up
in fact, an implementation of red-black could be very useful in P6
Limbic_Region TimToady - have you seen eternallyconfuzzled.com/jsw_home.aspx 00:03
sjohnson what's baby C? sounds cute
Limbic_Region all of the source is C, but the commentary is very understandable
I tend to implement based on plain english explanations anyway 00:04
and now I really must go AFK &
sorear has anyone played with NDP for Perl 6?
TimToady Natural Data Processing?
sorear nested data parallelism
TimToady please leave a pointer, I have to go too. 00:05
afk &
sjohnson learns baby C
flussence I want to learn C some day, but I don't want to learn autotools... 00:07
00:07 jferrero left, leprevost joined
flussence actually I'd prefer D first, it looks nice 00:08
sjohnson get 2nd ed K&R C book 00:11
and kiss autotools goodbye!
00:15 lichtkind left, aloha joined 00:30 patspam left 00:35 justatheory left 00:45 leprevost left
colomon red-black trees, eh? 00:54
01:00 justatheory joined 01:25 risou joined 01:27 leprevost joined 01:30 aindilis joined
ash_ my only problem with D is they took everything in every other language and threw it into C++ and renamed it D, but it just has a HUGE amount of reserved words and stuff, but they aren't really what i would call organized as as thought out as they could be 01:35
there is only so much you can tack on before you need to reconsider your overall design
01:35 fod left 01:40 meppl left 01:46 Chillance joined 01:55 qw3rty joined 01:57 Limbic_Region left
qw3rty hi 02:03
02:04 kid51 joined
qw3rty i'm still thinking about the fraction problem i posted here yesterday: for 1..60 -> $n { say "n=$n"; for 1..$n { $n %% $^a ?? say "\t$a/" ~ $n/$a !! 0 } } 02:05
02:05 astinus left
qw3rty it generates fractions with the condition that numerator * denominator are equal to $n 02:05
i saw that you can put constrained data types.... how could i do that program with constraints? (using "where", etc) 02:06
ash_ rakudo: sub f($a where { $a > 10 }) { say $a }; f 15; f 10; 02:07
p6eval rakudo a820a4: OUTPUT«15␤Constraint type check failed for parameter '$a'␤ in 'f' at line 22:/tmp/oIh6f50G5L␤ in main program body at line 22:/tmp/oIh6f50G5L␤»
ash_ first one passes, the 2nd complains about the constraint 02:08
02:08 astinus joined
ash_ constraints don't have to be where clauses though 02:09
rakudo: multi f(Int $a) { say "Int" }; multi f(Str $a) { say "Str" }; f 1; f 'a';
p6eval rakudo a820a4: OUTPUT«Int␤Str␤»
ash_ those are constraints that are based off of type 02:10
qw3rty rakudo: sub f($a, $b, $c where {$a * $b == $c}){ say "$a/$b" }; f(2,5,10) 02:11
ash_ rakudo: say <a b c>[1..2]
p6eval rakudo a820a4: OUTPUT«2/5␤»
rakudo a820a4: OUTPUT«bc␤»
02:12 astinus left
qw3rty rakudo: sub f($a, $b, $c where {$a * $b == $c}){ say "$a/$b" }; f(2,5,11) 02:13
p6eval rakudo a820a4: OUTPUT«Constraint type check failed for parameter '$c'␤ in 'f' at line 22:/tmp/kKFmll5TzE␤ in main program body at line 22:/tmp/kKFmll5TzE␤»
02:13 astinus joined
qw3rty cool 8) 02:13
ash_ so what were you trying to do again? 02:16
02:17 leprevost left
qw3rty irclog.perlgeek.de/perl6/2010-09-23#i_2855389 02:17
that
02:18 astinus left
qw3rty but taking advantage of everything perl6 has to offer :) 02:18
of which i know little...
yet 02:19
02:19 astinus joined, astinus left, astinus joined 02:20 kid51 left, cdarroch left
qw3rty mmm the link does not seem to work... well, there is the problem: 02:21
here's the original problem for the line i posted "Create an ordered list of fractions which product of the numerator and denominator equals a specific integer n.
ex: n=60
1/60, 2/30, 3/20, 4/15, 5/12, etc....
then do it with a list of integers."
02:22 ggoebel joined 02:24 astinus left, astinus joined
ash_ rakudo: sub fn($n) { for 1..$n -> $a { for 1..$a -> $b { say "$b/$a" if $a * $b == $n } } }; fn(60); 02:27
p6eval rakudo a820a4: OUTPUT«(timeout)␤4/15␤3/20␤2/30␤»
ash_ that doesn't seem optimal 02:28
but it does give the correct answer
if you run it on your local machine it won't time out
rakudo: sub fn($n) { for ^$n -> $a { for ^$a -> $b { say "$b/$a" if $a * $b == $n } } }; fn(60);
p6eval rakudo a820a4: OUTPUT«(timeout)␤4/15␤3/20␤2/30␤»
qw3rty ok i'll try it 02:30
02:30 synth left 02:31 rgrau_ left
ash_ sub fn($n) { for ^$n -> $a { for ^$a -> $b { say "$b/$a\n$a/$b" if $a * $b == $n } } }; fn(60); 02:31
rakudo: sub fn($n) { for ^$n -> $a { for ^$a -> $b { say "$b/$a\n$a/$b" if $a * $b == $n } } }; fn(60); 02:32
p6eval rakudo a820a4: OUTPUT«(timeout)␤5/12␤12/5␤4/15␤15/4␤3/20␤20/3␤2/30␤30/2␤»
ash_ sub fn($n) { for 1..$n -> $a { for 1..$a -> $b { say "$b/$a" if $a * $b == $n } } }; fn(60); 02:33
hmm
you cant do ^$n because that skips $n, so you do need 1..$n
i still think its checking some of those cases more than once
which means its sub optimal
02:35 astinus left
ash_ rakudo: sub fn($n) { for ($n.sqrt.floor)..$n -> $a { for 1..$a -> $b { say "$b/$a" if $a * $b == $n } } }; fn(60); 02:35
p6eval rakudo a820a4: OUTPUT«(timeout)␤4/15␤3/20␤2/30␤»
ash_ there, that doesn't check as many cases with repeats
still i bet you can do better
02:36 astinus joined 02:37 synth joined 02:38 leprevost joined
qw3rty mmm can i make a construct like "@num=1..60; @denom=1..60; ... ; where any(@num) * one(@denom) == 60" ... ? 02:44
ash_ rakudo: sub fn($n) { map { ($n/$_) ~ "/$_" }, grep { $n %% $^a }, 1..(60.sqrt.ceiling) }; fn 60 02:46
p6eval rakudo a820a4: ( no output )
ash_ rakudo: sub fn($n) { map { ($n/$_) ~ "/$_" }, grep { $n %% $^a }, 1..(60.sqrt.ceiling) }; say fn(60)
p6eval rakudo a820a4: OUTPUT«60/130/220/315/412/510/6␤»
ash_ rakudo: sub fn($n) { map { ($n/$_) ~ "/$_" }, grep { $n %% $^a }, 1..(60.sqrt.ceiling) }; say ~fn(60)
02:46 astinus_ joined, nymacro joined
p6eval rakudo a820a4: OUTPUT«60/1 30/2 20/3 15/4 12/5 10/6␤» 02:46
02:46 astinus left
ash_ that doesn't do hardly any re-calculating 02:46
qw3rty hehe,... cool
ash_ i can't think of how to get it more effective than that... 02:47
if you need the other half of the list (the 1/60, etc...) you can just take the list and flip it
qw3rty thanks... i will study that line :) 02:48
ash_ rakudo: sub fn($n) { map { ($n/$_) ~ "/$_" }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; say ~fn(60)
p6eval rakudo a820a4: OUTPUT«60/1 30/2 20/3 15/4 12/5 10/6␤»
ash_ hehe, thats a bit smaller, change 1.. to ^
^ is up to in perl6
rakudo: say ^60
p6eval rakudo a820a4: OUTPUT«01234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859␤»
ash_ if you use it as a prefix to a numeric that is
you only have to check up to the sqrt of N in your chase, anything beyond that won't matter or will have one of those values in it already 02:49
hence the $n.sqrt.ceiling
sqrt of 60 is 7.something so i only check 0...7, find which ones divide evenly into 60, then those are either numerators or denominators for your list of ratios 02:50
fractions, whatever you want to call them
that gives you half the list, one sec, i'll find the other half of it real fast 02:51
02:53 justatheory left
ash_ rakudo: sub fn($n) { map { ($n/$_), ($_/$n) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; { (.numerator ~ '/~ .denominator).say } for fn(60) 02:54
p6eval rakudo a820a4: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
02:55 astinus_ left
ash_ rakudo: sub fn($n) { map { ($n/$_ ~"/$_"), ("$_/" ~ $n/$_) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .say for @(fn(60)); 02:56
p6eval rakudo a820a4: OUTPUT«60/1␤1/60␤30/2␤2/30␤20/3␤3/20␤15/4␤4/15␤12/5␤5/12␤10/6␤6/10␤»
02:56 stkowski left
ash_ that gives you both sides, if you want them sorted i can do that too 02:56
qw3rty i still dont understand "map" completely, the 3rd parameter ... is that like a "for" in c? 02:58
02:59 ggoebel left
ash_ ya 02:59
rakudo: map *.say, 1...5
p6eval rakudo a820a4: OUTPUT«No candidates found to invoke for method 'map' on object of type 'Array'; available candidates have signatures:␤:(Mu : &block;; *%_)␤␤ in 'map' at line 1758:CORE.setting␤ in main program body at line 22:/tmp/wvrwYDNU6v␤» 03:00
ash_ lol
rakudo: map { .say }, <a b c>
p6eval rakudo a820a4: ( no output )
ash_ i am fail right now geeze
rakudo: map { .say }, <a b c> 03:01
p6eval rakudo a820a4: ( no output )
ash_ try.rakudo.org/?input=map%20{%20.sa...20b%20c%3E
try that
map takes a block and a list as params, it applies the block to the list and collects the results 03:02
rakudo: say map * + 1, 1 ... 3
qw3rty mmm... "Bool::True" ??
p6eval rakudo a820a4: OUTPUT«No candidates found to invoke for method 'map' on object of type 'Array'; available candidates have signatures:␤:(Mu : &block;; *%_)␤␤ in 'map' at line 1758:CORE.setting␤ in main program body at line 22:/tmp/u97xtySAV7␤»
ash_ map doesn't like whatevercode
rakudo: say ~map { $_ + 1 }, 1..4 03:03
p6eval rakudo a820a4: OUTPUT«2 3 4 5␤»
ash_ that took 1, 2, 3, 4 and added one, then collected the results, so you end up with an array that contains (2, 3, 4, 5) 03:04
whatever the block returns, the map collects the result 03:05
it printed Bool::True because say returned True in the sample i made up (with <a b c>) 03:06
qw3rty my @list = map { .say }, <a b c d>; @list.perl.say;
03:06 Guest7724 joined
qw3rty rakudo: my @list = map { .say }, <a b c d>; @list.perl.say; 03:06
p6eval rakudo a820a4: OUTPUT«a␤b␤c␤d␤[Bool::True, Bool::True, Bool::True, Bool::True]␤»
qw3rty well now i got something to experiment with, thanks ash 4 your patience 03:07
good night!
p6 is amazing :)
ash_ rakudo: my @list = map { $_ + 1 }, 3..4; say ~@list;
p6eval rakudo a820a4: OUTPUT«4 5␤» 03:08
qw3rty my @list = map { .say }, <a b c d>; @list.perl.say;
my @list = map { $_ ~ '-' }, <a b c d>; @list.perl.say; 03:09
rakudo: my @list = map { $_ ~ '-' }, <a b c d>; @list.perl.say;
p6eval rakudo a820a4: OUTPUT«["a-", "b-", "c-", "d-"]␤»
qw3rty rakudo: my @list = map { ~$_ }, <a b c d>; @list.perl.say;
p6eval rakudo a820a4: OUTPUT«["a", "b", "c", "d"]␤» 03:10
ash_ functional programmers will use map in most places you have a for loop
qw3rty rakudo: my @list = map { +$_ }, <1 2 3 4>; @list.perl.say;
03:10 Guest7724 left
p6eval rakudo a820a4: OUTPUT«[1, 2, 3, 4]␤» 03:10
qw3rty rakudo: my @list = map { $_ }, <1 2 3 4>; @list.perl.say;
p6eval rakudo a820a4: OUTPUT«["1", "2", "3", "4"]␤»
ash_ prefix:<+> will convert stuff to a number 03:11
rakudo: my @a = <a b c>; say +@a; # on an array, its the length
or number of elements in the array 03:12
p6eval rakudo a820a4: OUTPUT«3␤»
ash_ rakudo: say (+"1.23").WHAT # converted the string to an Num
p6eval rakudo a820a4: OUTPUT«Num()␤»
ash_ rakudo: sub postfix:<!> (Int $a) { [*] 1..$a }; say 5! 03:13
p6eval rakudo a820a4: OUTPUT«120␤»
ash_ btw, thats how you can add new operators to perl6, like that is the factorial operator
qw3rty oh yes i saw that in a screencast 03:14
well... is late here got to go sleeping... 03:15
thanks for all again, bye!
03:15 qw3rty left 03:36 ash_ left 03:43 astinus joined 03:48 astinus left 03:57 dduncan joined, dduncan left 03:59 DemoFreak left, envi^home joined 04:05 DemoFreak joined 04:08 astinus_ joined, astinus_ left, astinus_ joined 04:12 astinus_ left, astinus_ joined 04:17 astinus_ left 04:24 astinus_ joined, astinus_ left, astinus_ joined 04:34 alester joined 04:35 Raynes left
shortcircuit TimToady: I'm hoping to use Semantic MediaWiki to be able to do better indexing (based on, e.g. ideas invoked explicitly and implicitly in a task description), but that's going to have to wait until we resolve SMW performance issues. :-| 04:46
04:47 flatwhatson joined 04:48 Raynes joined 04:51 rgrau_ joined 04:53 rgrau_ left, rgrau_` joined 05:11 IllvilJa left, leprevost left 05:20 alester left 05:22 Su-Shee joined
sorear good * #perl6 05:22
Juerd: ping 05:27
05:45 plobsing left 05:58 kaare joined, kaare is now known as Guest41340 06:09 nymacro left 06:18 kjeldahl joined 06:19 wtw joined 06:28 dual left
moritz_ good morning 06:30
sorear hello moritz_
dalek ast: 885149f | moritz++ | S02-builtin_data_types/bool.t:
[bool.t] test stringification more precisely
 06:32
sorear Do [+] and [\+] guarantee a specific order of execution? 06:35
dalek kudo: 5e7b432 | moritz++ | build/PARROT_REVISION:
bump PARROT_REVISION for testing
moritz_ sorear: I'm curious, how can [\+] work without order of execution? 06:43
06:47 aindilis left 06:48 Intensity left, aindilis joined 06:49 mberends joined, jaldhar left, jaldhar joined 06:51 Intensity joined
sorear moritz_: suboptimally - on a sequential machine 06:51
moritz_ thinks he never met a machine where out-of-order-execution would make it any faster 06:54
06:56 cotto joined
sorear [\+] ^($big**2) = my @a = ^($big**2); my @b = @a.natatime($big); my @c = [+]<<@b; my @d = [\+]<<@b; return @c <<+>> @d; # look, O(sqrt N) [\+] on a machine with O(sqrt N) processors; I think there's also an O(log^2 N) recursive version 06:56
06:56 btrace joined
dalek odel: 12f5afe | mberends++ | java/compiler/ (6 files):
[java/compiler] incomplete set of files very similar to dotnet/compiler
 06:58
moritz_ doesn't understand at all
btrace hi,every
moritz_ coolnamehere.com/geekery/rakudo/index.html is quite nice
btrace ? 06:59
dukeleto btrace: howdy
sorear Hello btrace 07:00
btrace how can i get perl6' source?
dukeleto btrace: it is on github 07:01
btrace: http:/github.com/perl6
btrace: github.com/perl6 rather
btrace o ,thx~
sorear moritz_: [\+] (@a, @b) === (([\+] @a), (([\+] @b) »+» ([+] @a))) # Do you follow this identity 07:02
moritz_ sorear: yes 07:03
dukeleto btrace: github.com/rakudo/star
moritz_ sorear: but to me that's more additions than a straight evaluation
dukeleto btrace: that is where you can find a stable release of Rakudo Perl 6
moritz_ btrace: rakudo.org/how-to-get-rakudo 07:04
sorear moritz_: more additions, sure, but I've cut the dependency chain in half
dukeleto btrace: github.com/rakudo/star/downloads <-- if you don't want to use git
moritz_ there is not single "perl6" compiler, but several (just like there are multiple C and Fortran compilers). Rakudo is one of them.
sorear moritz_: if I have two processors, I can do the LHS and RHS in parallel
btrace thanks a lot~ 07:05
moritz_ sorear: and after that you still have to do the »+» [+] @a operator, which is just as expensive as doing the second half of the additional normally
sorear moritz_: »+» spreads over arbitrarily many processors very easily 07:06
moritz_ sorear: if you can prove that all infix:<+> multis in scope are associative, I think you're free to change the order of execution 07:07
07:08 btrace left
sorear moritz_: that's basically my question 07:08
moritz_: are [+] and [\+] aggregate operations that Perl 6 is allowed to assume are associative and side-effect-free? 07:09
07:09 cotto left
moritz_ no 07:09
07:10 cotto joined, jaldhar left 07:11 aindilis left, sftp left, yahooooo7 left 07:12 Ross joined, Ross left, Ross joined 07:13 yahooooo joined 07:14 DemoFreak left, DemoFreak joined 07:15 aindilis joined 07:16 [particle] joined 07:17 PacoLinux left 07:20 M_o_C joined 07:21 perigrin joined 07:22 cotto left 07:24 Ross left, Ross joined 07:26 jhuni left 07:34 nymacro joined 07:35 cotto joined 07:37 M_o_C left 07:41 fod joined 07:59 wamba joined 08:06 cognominal left, s1n left 08:09 s1n joined 08:15 IllvilJa joined 08:36 Benzino joined 08:37 satyavvd joined 08:39 cognominal joined 08:44 Benzino left 08:46 Benzino joined
Juerd sorear: pong 08:49
sorear Juerd: several people on #parrot have been pining for a community server to run irc bots, other minor things on 08:50
should I refer them to you?
(whiteknight and bacek at least) 08:51
08:53 tadzik joined 09:01 Guest6471 joined, Guest41340 left 09:02 Benzino left, dakkar joined, Ross left 09:03 kaare joined, kaare is now known as Guest72792
moritz_ rakudo: say 1 - 0.9**3 09:05
p6eval rakudo 5e7b43: OUTPUT«0.271␤»
Juerd sorear: Definitely; it sounds exactly like a job for feather
sorear: And feathers goals specifically include Parrot as well as Perl 6. 09:06
First line of the motd:
- This box is for Perl 6 development. This includes Parrot, and Pugs, and IRC,
09:07 pmurias joined
sorear hello pmurias 09:08
pmurias sorear: hi
phenny pmurias: 22 Sep 22:38Z <sorear> ask pmurias Does Mildew-Setting-SMOP normally require >3GB to build? It's failing both locally and on feather1 with ENOMEM during the gcc call (so not an infinite allocation bug in our code)
09:08 Guest6471 left
pmurias hmm 09:08
sorear possible gcc 4.4 just doesn't like 170kloc, 7MB source files?
pmurias maybe 09:09
i'll try to build that and see how much memory it takes on machine 09:10
sorear anyways, I've gotten TO the gcc step
which I think is a good sign :)
moritz_ sorear: re GCC not liking big source files, I kinda doubt that. It has to compile itself, and since it recursively follows all includes, it is quite huge itself 09:11
sorear (most of the issues were from 1. STD warns about unused function arguments now 2. 'viv' is no longer installed in @PERL5LIB)
pmurias so how did you deal with 2?
sorear I changed it to just 'use STD' 09:12
09:12 oyse joined
sorear and patched up a couple places where we were relying on viv's definitions 09:13
pmurias sorear: i could try to switch from gcc to emitting llvm code
s/code/IL/
09:14 Benzino joined
sorear moritz_: cpp output = 23MB, impressive number of casts 09:14
pmurias `i have gcc 4.4.4 09:15
and i don't think it takes very long to compile stuff 09:16
compared to what i think is the SSA transform in mildew
sorear well, not necessary "it takes a long time", but it needs a LOT of memory 09:18
pmurias what could solve that? 09:19
sorear ~sorear/public_html/offending{,_cpp}.c 09:20
on feather1
pmurias having the compiler output spit out into separate files
sorear if anyone wants to look at the files
pmurias looks
sorear that's the larger of the two 09:21
pmurias sorear: permission denied on the offending.c one 09:22
sorear fixed
pmurias the files to look as expected 09:24
* the file seems to look as i would expect it to
sorear: all the ops are translated into C statements so the file is quite big 09:25
09:28 tadzik left, IllvilJa left 09:29 proller joined, proller_ joined, proller left
pmurias sorear: gcc took 700m here 09:31
sorear: but it took much longer than i expected
09:32 rindolf joined
rindolf Hi all. 09:32
moritz_: perlgeek.de/blog-en/perl-5-to-6/09-....writeback - you have some POD junk at the end of the post. ("=for"). 09:33
sorear pmurias: what version?
moritz_ rindolf: thanks, I'll look into it 09:34
sorear also, compile options
rindolf moritz_: you're welcome.
moritz_: also - you say that "The comments on this blog post have been disabled; the comment form below will not work." - in that case, you should not put the form in the page at all. 09:35
pmurias sorear: gcc 4.4.4
moritz_ rindolf: I know; but I haven't got around to hacking conditionals into the template system
rindolf moritz_: what kind of templating system are you using? 09:37
moritz_: why didn't you use Template Toolkit?
moritz_ rindolf: it's a hacked blosxom blog. The default is the built-in s/(\$\w+)/eval $1/g or so template hack 09:38
rindolf: and I'm not a fan of Template Toolkit, for that matter 09:39
pmurias sorear: how can i get the compile options?
sorear not sure
haven't touched that
pmurias gcc trying to optimise code more on feather could cause the problem 09:40
but i'm not sure how to check that
09:42 daxim joined
rindolf moritz_: there's some POD junk here too - perlgeek.de/blog-en/perl-5-to-6/10-....writeback 09:42
moritz_ rindolf: I'm going through all the documents now
rindolf moritz_: ah, OK.
moritz_ rindolf: seems that POD needs a newline before =for, and I wasn't aware of that
rindolf moritz_: ah.
sorear moritz_: according to perlpod, POD tools that require the blank line are working off an obsolete spec 09:44
moritz_ sorear: I haven't updated my POD parser in quite some time, so that might well be the case 09:45
rindolf moritz_: perlgeek.de/blog-en/perl-5-to-6/10-....writeback - "not very exiting" should be "not very exciting"
moritz_: also "I'll write about lazy list" ==> s/$/s/ 09:46
moritz_: "exicting" there should be "exciting". 09:47
pmurias sorear: do you want to use mildew for anything, or are you just repairing the damage STD changes caused? (i was planning on playing with perlito today but can try to fix that problem if it's blocking you) 09:48
moritz_ huh, I don't seem to have the source for 10-where-we-are-now in my repo :/ 09:50
sorear pmurias: the latter
pmurias: since I caused the STD changes selfishly, I thought it only right 09:51
moritz_ rindolf++
rindolf moritz_: thanks. :-) 09:52
09:53 rindolf left
moritz_ the POD mess is now gone 09:53
pmurias sorear: if you commit your changes i can finish up the fix later on 09:57
09:57 kensanata joined 10:02 Benzino left
sorear pmurias: long since committed 10:02
sorear out
10:03 timbunce joined 10:06 nymacro left 10:26 dolmen joined 10:27 dolmen is now known as Guest67249 10:33 tadzik joined 10:35 Su-Shee left 10:36 kensanata left 10:44 nymacro joined 10:46 snearch joined 11:03 pmurias left 11:09 wamba left 11:19 Italian_Plumber joined 11:33 jaldhar joined 11:34 Kodi joined 11:41 Mirell joined 11:46 Ross joined, Ross left, Ross joined 11:51 Ross left 12:03 Ross joined, Ross left, Ross joined 12:11 Guest67249 left
moritz_ std: require Foo; 12:11
p6eval std : OUTPUT«ok 00:01 115m␤»
moritz_ stdbug
12:12 satyavvd left
moritz_ oh, not so 12:13
moritzbug
gfldex is it selffixing? 12:14
moritz_ no, I have to do it manually
gfldex :->
moritz_ rakudo: use Test <plan ok>;
p6eval rakudo 5e7b43: OUTPUT«===SORRY!===␤Confused at line 22, near "use Test <"␤» 12:15
12:16 patspam joined 12:21 masak joined
masak performs a 360 degree greeting 12:21
moritz_ returns it with a 2pi greeting
12:22 ruoso joined 12:23 juju joined
masak I thought of doing a steradian greeting, but I don't know if I can manage it in Earth's gravitation well. 12:24
s/gravitation/gravitational/
moritz_ I'm sure you can, in first (and second and third order) approximation :-) 12:25
juju hi. Is there someone in here is involved in the perl6.org website ?
moritz_ yes
juju well the "viv" STD-based reference parser used by various other compilers link is dead 12:26
masak oh! thanks, juju++
juju bye !
masak bye :) 12:27
12:27 juju left
moritz_ damn, didn't stay long enough to get a commit bit :-) 12:27
12:28 bluescreen joined
masak oh, didn't think of the commit bit thing. 12:29
masak--
moritz_ hugme still can't add people to the 'perl6' team :(
organizations don't have an API yet on github :(
Raynes They still haven't even completed the gist API. :( 12:30
Or the gist API documentation. If there is a meaningful difference. 12:31
12:31 icwiener joined 12:34 xabbu42 joined
moritz_ std: require Foo; Foo.WHAT 12:39
p6eval std : OUTPUT«===SORRY!===␤Undeclared name:␤ 'Foo' used at line 1␤Check failed␤FAILED 00:02 115m␤»
12:41 kjeldahl_ joined
flussence is there anything along the lines of "(namespace Foo { ... }).^methods"? 12:48
daxim rakudo: NaN.WHAT # joke of the day 12:49
p6eval rakudo 5e7b43: ( no output )
daxim qua?
flussence rakudo: say NaN.WHAT
daxim ah right
p6eval rakudo 5e7b43: OUTPUT«Num()␤»
flussence blame the IEEE for that one
12:50 stepnem left 12:52 drbean joined
moritz_ moritz@trudi:~/rakudo>echo 'module TestModule { our sub greetz { say 42 } }' > TestModule.pm 12:53
moritz@trudi:~/rakudo>./perl6 -e 'require TestModule; TestModule::greetz'
42
12:55 Kodi left 12:56 rgrau_` left
tadzik yay, require works? 12:57
moritz_ locally
the simplest form of require, that is :-)
tadzik can you pass a string to it? 12:58
moritz_ not yet
but should be not too hard to do
... at least I hope so :-) 12:59
tadzik that will make another (third)Warsaw Perl Monger happy :) 13:00
moritz_ I've now overriden the 'require' branch on github with my changes 13:03
tadzik oh, by the way. What can I do with P6 grammars so it will tell in which line of input there's a syntax error? 13:04
moritz_ there's no automated way yet :(
tadzik and what is the way?
moritz_ die $/.pos 13:05
tadzik or, how Rakudo/STD does it?
PerlJam tadzik: count the lines :)
moritz_ and then count the lines up to that position
tadzik PerlJam: :) moritz_, in actions?
moritz_ tadzik: wherever you determine the parse error
tadzik I thought grammars determine them 13:06
well, themselves
moritz_ token force_foo { <foo> || { die $/.pos } }
tadzik mhm
moritz_ 15:04 < moritz_> there's no automated way yet :(
tadzik is there a hope for an automated way?
moritz_ <aragon>There's always hope!</aragon>
masak these two tweets were so absurd that I thought the tweeter was a bot at first: twitter.com/brianrue/status/25378535881 twitter.com/brianrue/status/25378574647 13:09
but seems not.
tadzik wat
rjbs Odd!
PerlJam tough job market. Maybe his cab driver is a perl 6 hacker :) 13:10
masak 哈哈
tadzik (:
masak moritz_++ # Rakudo release 13:12
moritz_ thanks 13:13
pmichaud moritz_++
masak TimToady: re '[+/]' meaning 'reverse [\+] reverse'. I'm unable to decide if I think that is utterly brilliant or completely bonkers. :) which probably means we should try it. 13:14
moritz_ please don't
pmichaud ...what does that mean?!
masak :) 13:15
rakudo: say (<a b c d e> Zxx 0,1,0,1,0) # via TimToady, backlog 13:16
rakudo: say (<a b c d e> Zxx 0,1,0,1,0).Str
p6eval rakudo 5e7b43: OUTPUT«bd␤»
rakudo 5e7b43: OUTPUT«e e␤»
masak submits rakudobug
pmichaud looks like a variation of the take bug again 13:17
masak aye.
13:18 Patterner left
pmichaud rakudo: say (<a b c d e>)[(0,1,0,1,0).pairs.grep(?*).keys].Str # just curious 13:19
p6eval rakudo 5e7b43: OUTPUT«a b c d e␤»
pmichaud rakudo: say (0,1,0,1,0).pairs.perl
p6eval rakudo 5e7b43: OUTPUT«(0 => 0, 1 => 1, 2 => 0, 3 => 1, 4 => 0)␤»
pmichaud rakudo: say (0,1,0,1,0).pairs.grep(?*).perl 13:20
p6eval rakudo 5e7b43: OUTPUT«(0 => 0, 1 => 1, 2 => 0, 3 => 1, 4 => 0)␤»
pmichaud rakudo: say (<a b c d e>)[(0,1,0,1,0).pairs.grep(.value).keys].Str # just curious
p6eval rakudo 5e7b43: OUTPUT«Method 'value' not found for invocant of class ''␤ in main program body at line 22:/tmp/6cX1LeMToc␤»
pmichaud rakudo: say (<a b c d e>)[(0,1,0,1,0).pairs.grep(*.value).keys].Str # just curious
moritz_ rakudo: say 'abc' ~~ / foo <alpha>? || <alpha> /; say $<alpha>.WHAT
p6eval rakudo 5e7b43: OUTPUT«a␤Array()␤»
rakudo 5e7b43: OUTPUT«a b␤»
moritz_ is that correct?
pmichaud looks right to me. 13:21
moritz_ I kinda thougth that alternations reset match counting
so I expected the second branch to have a not-quantified $<alpha> match
pmichaud if a capture is quantified in any alternative, it's quantified in all.
moritz_ ok.
pmichaud S05:3242 13:23
oh, but there's a counter-example later on 13:24
hmmmm
moritz_ ./perl6 -e 'require "TestModule"; TestModule::greetz' 13:26
42
tadzik yay
moritz_ hm, maybe that's wrong
in perl 5, the require STRING form expects a file name, not a module name
13:27 Kodi joined
tadzik maybe it'll be good to change it 13:27
moritz_ curses under his breath
S11 says filename
tadzik mhm
Kodi Is IO-Socket-INET.t (or rather, IO-Socket-INET.rakudo) failing for anybody else? 13:28
phenny Kodi: 22 Sep 17:10Z <moritz_> tell Kodi to please wait with merging until after the release on Thursday
moritz_ wonders if a filename shouldn't require a :filename adverb or so
Kodi: I had one failure in my last test run
pmichaud moritz_: you may be correct that it's a bug (the <alpha> example) 13:29
moritz_ (btw smolder.parrot.org/app/projects/smoke_reports/5 can answer such questions for you too)
pmichaud I don't know why I'm remembering it as being the other way.
13:29 Psyche^ joined, Psyche^ is now known as Patterner
moritz_ Kodi: but it's not always failing here 13:30
13:30 PacoLinux joined
Kodi moritz_: It consistently fails here. Also, Rakudo is eating enormous amounts of memory, which might be the cause. 13:30
moritz_ Kodi: maybe related to recent parrot GC changes 13:31
Kodi Sounds likely.
I think I'll delay merging perl6scalar-cleanup until this is sorted out.
PerlJam pmichaud: where's the counter-example in S05? 13:32
moritz_ Kodi: are you on parrot HEAD?
Kodi The latest PARROT_REVISION, whichever SVN revision that is.
moritz_ that's pretty new, yes 13:33
masak submits moritz_' question that pmichaud thinks maybe is a bug
pmichaud PerlJam: S05:3289
13:34 astrojp left
Kodi Parrot HEAD is about 20 revisions ahead at the moment. 13:34
13:34 Guest72792 left
pmichaud rakudo: 'abc' ~~ / (a) xy | (.)+ /; say $0.WHAT; 13:34
p6eval rakudo 5e7b43: OUTPUT«Array()␤»
pmichaud rakudo: 'abc' ~~ / (a) .. | (.)+ /; say $0.WHAT; 13:35
p6eval rakudo 5e7b43: OUTPUT«Array()␤»
PerlJam pmichaud: that shows a non-quantified <file> in each branch of the alternation, ergo no array.
pmichaud oh, right.
*whew*
PerlJam: thanks for correcting my misreading of that. I should probably get a Dr Pepper or something to wake up. :) 13:36
13:37 plainhao joined
masak so, no bug. 13:38
masak un-submits
13:45 xabbu42 left 13:49 stepnem joined 13:52 molaf joined
masak tadzik: [backlog] it's ".comb" because "comb for" means "go through and keep what you're interested in" -- "chop" sounds destructive, and indeed it is. 13:54
13:55 IllvilJa joined, PacoLinux left, ash____ joined 13:58 jaldhar left, Kodi left
jnthn o/ 13:59
13:59 jaldhar joined
ash____ try.rakudo.org/?input=sub%20fn($n)%...@(fn(60)); 14:00
anyone know any faster ways of doing that?
tadzik I can't copy that :) 14:01
14:01 ash____ is now known as ash_gti
ash_gti rakudo: sub fn($n) { map { ($n/$_ ~"/$_"), ("$_/" ~ $n/$_) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .say for @(fn(60)); 14:01
p6eval rakudo 5e7b43: OUTPUT«60/1␤1/60␤30/2␤2/30␤20/3␤3/20␤15/4␤4/15␤12/5␤5/12␤10/6␤6/10␤»
ash_gti rakudo: sub fn($n) { map { ($n/$_ ~"/$_"), ("$_/" ~ $n/$_) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; say ~fn(60);
p6eval rakudo 5e7b43: OUTPUT«60/1 1/60 30/2 2/30 20/3 3/20 15/4 4/15 12/5 5/12 10/6 6/10␤»
ash_gti also, is there a way to print a Rat so it shows up as Numerator/Denominator? 14:02
rakudo: say (1/2).Rat; # always seem to be decimals 14:03
p6eval rakudo 5e7b43: OUTPUT«0.5␤»
ash_gti rakudo: say (1/2).^methods(:local);
p6eval rakudo 5e7b43:
..OUTPUT«newnudeperlBridgeBoolRatNumsuccpredACCEPTSRealIntComplexStrrealsisNaNabsexplnsqrtrootssignfloorceilingtruncateroundcisunpolarrandsinasincosacostanatansecaseccosecacoseccotanacotansinhasinhcoshacoshtanhatanhsechasechcosechacosechcotanhacotanhatan2Numericloglog10to-radiansfrom-radian…
tadzik rakudo: (1/2).Str.say
p6eval rakudo 5e7b43: OUTPUT«0.5␤»
masak rakudo: say (1/2).nude.perl
tadzik ash_gti: ↑
oh, noes
p6eval rakudo 5e7b43: OUTPUT«(1, 2)␤»
tadzik hrm
rakudo: (1/2).perl
rakudo: (1/2).perl.say 14:04
-_-
p6eval rakudo 5e7b43: ( no output )
rakudo 5e7b43: OUTPUT«1/2␤»
14:04 xabbu42 joined
masak rakudo: say (1/2).nude.perl # just do this 14:04
tadzik nah
p6eval rakudo 5e7b43: OUTPUT«(1, 2)␤»
14:04 xabbu42 left, xabbu42_ joined
masak jnthn: \o 14:04
ash_gti rakudo: sub fn($n) { map { ($n/$_), ($n/$_) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .nude.perl.say for @(fn(60).sort); 14:05
p6eval rakudo 5e7b43: OUTPUT«(10, 1)␤(10, 1)␤(12, 1)␤(12, 1)␤(15, 1)␤(15, 1)␤(20, 1)␤(20, 1)␤(30, 1)␤(30, 1)␤(60, 1)␤(60, 1)␤»
14:07 frodwith joined
ash_gti rakudo: sub fn($n) { map { (($n/$_)/$_), ($_/($n/$_) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .nude.perl.say for @(fn(60).sort); 14:07
p6eval rakudo 5e7b43: OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
ash_gti rakudo: sub fn($n) { map { (($n/$_)/$_), ($_/($n/$_)) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .nude.perl.say for @(fn(60).sort);
p6eval rakudo 5e7b43: OUTPUT«(5, 3)␤(3, 5)␤(15, 1)␤(1, 15)␤(12, 5)␤(5, 12)␤(15, 4)␤(4, 15)␤(20, 3)␤(3, 20)␤(60, 1)␤(1, 60)␤»
ash_gti rakudo: sub fn($n) { map { (($n/$_)/$_), ($_/($n/$_)) }, grep { $n %% $^a }, ^($n.sqrt.ceiling) }; .nude.perl.say for @(fn(60).flat.sort); 14:08
14:08 snearch left
p6eval rakudo 5e7b43: OUTPUT«(1, 60)␤(1, 15)␤(3, 20)␤(4, 15)␤(5, 12)␤(3, 5)␤(5, 3)␤(12, 5)␤(15, 4)␤(20, 3)␤(15, 1)␤(60, 1)␤» 14:08
ash_gti there, thats what i was looking for
still curious if it can be any faster
14:13 PacoLinux joined
jnthn masak: o/ 14:15
14:19 Mowah joined 14:20 Holy_Cow joined
flussence ash_gti: the list's symmetrical, so you can skip half the divides and do something fancy with .reverse 14:20
14:20 mberends left
flussence (not sure if it'd be faster though...) 14:20
ash_gti well, i already add everything twice with the map
14:22 arnsholt left
flussence wait... is that last output correct? I'm not sure what it's meant to do 14:27
ash_gti the last output is correct, its supposed to: "Create an ordered list of fractions which product of the numerator and denominator equals a specific integer n." 14:30
flussence (1, 15) for 60?
ash_gti that got simplified
flussence oh 14:31
ash_gti i think... one sec
flussence yeah, that makes sense
ash_gti ya, that was 2/30
simplified to 1/15 14:32
flussence I wonder if there's a way to prevent that...
moritz_ if you want to prevent it, don't use Rats
but for example Pair objects, or two-element lists
ash_gti a pair object sounds easy enough 14:33
can you sort pair objects?
masak moritz_: [backlog] wasn't there talk about an 'is associative' trait at some point? 14:34
moritz_ ash_gti: sure
masak: yes, there was talk. There's always talk. :-) 14:35
masak moritz_: I mean, mightn't it still be a good idea. then your "no" in the backlog as a reply to sorear's question could, it appears, be qualified.
14:35 leprevost joined
ash_gti although, it would probably be better to make the list in the right order the first time, and not have to sort it... 14:35
masak s:1st/\./?/ 14:36
moritz_ rakudo: say (3 => 4, 0 => 3, 0 => 2, 3 => 2).sort.perl 14:37
p6eval rakudo 5e7b43: OUTPUT«(0 => 2, 0 => 3, 3 => 2, 3 => 4)␤»
moritz_ sorts on key first, then on value
masak stumbles over c2.com/cgi/wiki?SufficientlySmartCompiler -- and realized he uses the phrase fairly often 14:39
14:39 wtw left
moritz_ masak: actually here you use "sufficiently smart lint utility" or so quite often :-) 14:41
masak aye.
I guess jnthn is working on raising the intelligence of Rakudo right now. 14:42
jnthn Actually right now I'm snacking on a chokladbollar...
masak :D
but I really do think it's realistic to build a smart lint utility on top of STD.
ash_gti what about a smarter repl? 14:43
jnthn But yes, I expect making a bunch more stuff available at compile time (which I will be doing) will mean that we can do a lot more checks. :-)
ash_gti that lets you do: if true { \n
jnthn
.oO( and a lot more Slovaks...don't want them feeling left out... :-) )
I expect detecting my Int $x = 'wtf'; will be easily detectable and whineaboutable, for example. 14:44
s/detecting /
moritz_ that's kinda hard to do at the moment :/ 14:46
14:46 justatheory joined
jnthn Right. 14:47
Thus why stuff must change.
masak ash_gti: I think that's a nice goal. 14:51
ash_gti i guess a smarter repl needs knowledge of what terminates a statement
masak and that's HLL-specific, and the REPL (I think) is outside of Rakudo and in PCT/nqp-rx. 14:53
it's approximately the same problem as with command-line option handling.
14:56 ggoebel joined
colomon ash_gti: not sure if this is clear in the above or not: 14:57
rakudo: say (1/2).perl
p6eval rakudo 5e7b43: OUTPUT«1/2␤»
colomon also 14:58
rakudo: say (0, 1/60 ... 1).perl
p6eval rakudo 5e7b43: OUTPUT«(0, 1/60, 1/30, 1/20, 1/15, 1/12, 1/10, 7/60, 2/15, 3/20, 1/6, 11/60, 1/5, 13/60, 7/30, 1/4, 4/15, 17/60, 3/10, 19/60, 1/3, 7/20, 11/30, 23/60, 2/5, 5/12, 13/30, 9/20, 7/15, 29/60, 1/2, 31/60, 8/15, 11/20, 17/30, 7/12, 3/5, 37/60, 19/30, 13/20, 2/3, 41/60, 7/10, 43/60, 11/15,
..3/4, …
14:59 sftp joined
colomon rakudo: say (0, 1/30 ... 1).perl # ought to fit, I think 15:00
p6eval rakudo 5e7b43: OUTPUT«(0, 1/30, 1/15, 1/10, 2/15, 1/6, 1/5, 7/30, 4/15, 3/10, 1/3, 11/30, 2/5, 13/30, 7/15, 1/2, 8/15, 17/30, 3/5, 19/30, 2/3, 7/10, 11/15, 23/30, 4/5, 5/6, 13/15, 9/10, 14/15, 29/30, 1/1)␤» 15:01
15:05 dual joined 15:07 amkrankruleuen joined
ash_gti colomon: ya, i figured out how to print rat's properly now 15:07
colomon ash_gti++ 15:08
15:09 Mowah left
colomon is frustrated at how much more work printing out a simple message is in C++. :( 15:11
15:12 leprevost left
colomon I mean, "<< endl" is twice as many characters as "say"! 15:12
tadzik "\n" is shorter 15:13
ash_gti you still need to << "\n" 15:14
although, endl isn't always what you want, since it flushes the buffer too 15:15
nymacro colomon: not to mention that if you don't want to import the std namespace, it is even more ;)
15:15 risou left
colomon nymacro: excellent point. 15:16
ash_gti: flushing the buffer is always what I want when I'm debugging. :)
nymacro: I think I got so sick of typing std::endl that I imported just endl into my namespace. :) 15:17
jnthn o/, bbl 15:20
15:28 mantovani left
hudnix rakudo: (0,5 ... 10).perl.say; (0,5 ... 11).perl.say 15:28
15:28 mantovani joined
p6eval rakudo 5e7b43: OUTPUT«(timeout))␤» 15:28
15:28 mantovani left
hudnix rakudo: (0,5 ... 10).perl.say 15:29
p6eval rakudo 5e7b43: OUTPUT«(0, 5, 10)␤»
hudnix rakudo: (0,5 ... 11).perl.say
p6eval rakudo 5e7b43: OUTPUT«(timeout)»
colomon hudnix: 0, 5 ... 11 is an infinite sequence in the latest standard
15:29 mantovani joined
colomon s/standard/spec/ 15:29
you have to hit the ending condition of the sequence exactly. 15:30
hudnix ok
colomon rakudo: (0, 5 ... * > 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 5, 10, 15)␤»
colomon rakudo: (0, 5 ... * >= 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 5, 10)␤»
colomon rakudo: (0, 5 ...^ * >= 11).perl.say 15:31
p6eval rakudo 5e7b43: OUTPUT«(0, 5, 10)␤»
colomon rakudo: (0, 5.5 ...^ * >= 11).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 11/2)␤»
hudnix what does the ^ do? 15:32
colomon ugh. cold medicine messing with my brainz.
masak hudnix: excludes the final value of the sequence.
rakudo: say 1..10; say 1..^10; 15:33
p6eval rakudo 5e7b43: OUTPUT«12345678910␤123456789␤»
colomon basically, the right-hand side is a test
when the test is true for the current value of the sequence, a ... sequence returns that last value
(and then ends)
and a ...^ sequence doesn't return it, but does end.
rakudo: (0, 1 ... 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤» 15:34
colomon rakudo: (0, 1 ...^ 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9)␤»
colomon rakudo: (0, 1 ... * > 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11)␤»
15:34 oyse left
colomon rakudo: (0, 1 ...^ * > 10).perl.say 15:34
p6eval rakudo 5e7b43: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
colomon rakudo: (0, 1.5 ... * > 10).perl.say 15:36
p6eval rakudo 5e7b43: OUTPUT«(0, 3/2, 3/1, 9/2, 6/1, 15/2, 9/1, 21/2)␤»
colomon rakudo: (0, 1.5 ...^ * > 10).perl.say
p6eval rakudo 5e7b43: OUTPUT«(0, 3/2, 3/1, 9/2, 6/1, 15/2, 9/1)␤» 15:37
sjn quick question: what methods do I use to find out what signature a method has? 15:38
15:38 snearch joined 15:39 snearch left, snearch joined, rgrau_` joined
nymacro sjn: &sub.signatures 15:39
moritz_ rakudo: sub foo($a, :$b){}; say &foo.signature.perl 15:40
p6eval rakudo 5e7b43: OUTPUT«:(Any $a, Any :b($b))␤»
TimToady but you can't necessarily name a method with &sub
nymacro It wasn't meant in the literal sense :P 15:41
15:41 patspam left
TimToady by you 15:41
I think sjn meant a literal method
in which case &sub is unlikely to work directly 15:42
nymacro my bad, I should have been clearer
15:42 patspam joined
hudnix rakudo: class A{method a($a,:$b){}};A.^can('a').signature.perl.say 15:42
p6eval rakudo 5e7b43: OUTPUT«Method 'signature' not found for invocant of class 'P6Invocation'␤ in main program body at line 22:/tmp/qpWKnKDkqf␤» 15:43
TimToady sorry, it's still too early here, and I'm cross... :)
hudnix rakudo: class A{method a($a,:$b){}};A.^can('a').perl.say
sjn rakudo: say round.signature.perl
p6eval rakudo 5e7b43: OUTPUT«Method 'perl' not found for invocant of class 'P6Invocation'␤ in main program body at line 22:/tmp/7Vb2uNNOBH␤»
rakudo 5e7b43: OUTPUT«Not enough positional parameters passed; got 0 but expected between 1 and 2␤ in 'round' at line 1979:CORE.setting␤ in main program body at line 22:/tmp/rKG_mshfo6␤»
15:43 rgrau_` left
hudnix rakudo: class A{method a($a,:$b){}};A.^can('a').WHAT.say 15:43
p6eval rakudo 5e7b43: OUTPUT«Method 'WHAT' not found for invocant of class 'P6Invocation'␤ in main program body at line 22:/tmp/DMPJzOQh8D␤» 15:44
hudnix rakudo: class A{method a($a,:$b){}};A.^can('a')
p6eval rakudo 5e7b43: ( no output )
TimToady or maybe I'm just grumpy, which is a more normal state of things...
hudnix rakudo: class A{method a($a,:$b){say "hi"}};A.^can('a').()
p6eval rakudo 5e7b43: OUTPUT«Not enough positional parameters passed; got 0 but expected 2␤ in 'A::a' at line 22:/tmp/USnTLi1ydM␤ in main program body at line 22:/tmp/USnTLi1ydM␤»
hudnix rakudo: class A{method a($a,:$b){say "hi"}};A.^can('a').(1,:b) 15:45
p6eval rakudo 5e7b43: OUTPUT«Not enough positional parameters passed; got 1 but expected 2␤ in 'A::a' at line 22:/tmp/bwcG_XDmGS␤ in main program body at line 22:/tmp/bwcG_XDmGS␤»
sjn ok, so what's the "simple" way of just exploring Perl 6 using introspection methods?
TimToady it's not supposed to be simple :) 15:46
sjn aw :-\
TimToady if it were simple, people would cheat...
sjn wishes exploration would be simple :-) 15:47
TimToady but yeah, it'd be nice to be able to browse it
sjn rakudo: say "somestring".methods
p6eval rakudo 5e7b43: OUTPUT«Method 'methods' not found for invocant of class 'Str'␤ in main program body at line 22:/tmp/itjnngoeAb␤»
hudnix rakudo: class A{method a($a){}}; A.can('a')
TimToady use .^methods
p6eval rakudo 5e7b43: ( no output )
sjn rakudo: say "somestring".^methods
p6eval rakudo 5e7b43: 15:48
..OUTPUT«ACCEPTSperlpredsuccWHICHBoolStrencodeNumericRealIntRatNumabsexploglog10sqrtrootsto-radiansfrom-radiansfloorceilingroundtruncatesigncisunpolarchrrandsincostanseccoseccotansinhcoshtanhsechcosechcotanhasinacosatanatan2asecacosecacotanasinhacoshatanhasechacosechacotanhbytescapitalizech…
sjn right
TimToady then stringify somehow to get whitespace
sjn rakudo: say "somestring".^methods.perl
p6eval rakudo 5e7b43: OUTPUT«[{ ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ...
..}, { …
TimToady and use :local to omit Cool methods
but think Cool and Any should be omitted by default
sjn rakudo: say "somestring".^methods(:local).perl # ?
p6eval rakudo 5e7b43: OUTPUT«[{ ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }]␤»
sjn is confused 15:49
TimToady rakudo doesn't know how to stringify closures
*perlify 15:50
masak it gets the curlies right :) 15:51
sjn rakudo: say "somestring".^methods(:local).join(" ")
p6eval rakudo 5e7b43: OUTPUT«ACCEPTS perl pred succ WHICH Bool Str encode␤»
sjn rakudo: say round.^methods(:local).join(" ")
p6eval rakudo 5e7b43: OUTPUT«Not enough positional parameters passed; got 0 but expected between 1 and 2␤ in 'round' at line 1979:CORE.setting␤ in main program body at line 22:/tmp/vIM9aPYZMQ␤»
moritz_ TimToady: S11 says that require <EXPR>; takes the expression as file name (and not module name). Do you think it's still a good idea?
15:51 rrgau_ joined
TimToady no 15:51
sjn rakudo: say round(1,2).^methods(:local).join(" ") 15:52
p6eval rakudo 5e7b43: OUTPUT«perl pred succ WHICH Str Bridge Int Rat Num sign ACCEPTS Real Bool Complex reals isNaN abs exp ln sqrt roots floor ceiling truncate round cis unpolar rand sin asin cos acos tan atan sec asec cosec acosec cotan acotan sinh asinh cosh acosh tanh atanh sech asech cosech acosech
..cotanh…
moritz_ TimToady: that's good, because implementing it as a module name is easier for now :-)
15:52 nymacro left
TimToady that being said, a module name is unlikely to contain a '.' 15:52
flussence rakudo: my @x = "something".^methods(:local); say ~(@x Z=> @x.perl);
p6eval rakudo 5e7b43: OUTPUT«ACCEPTS [{ ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }]␤»
flussence ack. 15:53
15:53 LionMadeOfLions left
sjn would it be sensible to ignore signature checks when one is playing around with introspection? 15:54
TimToady sjn: .Str or ~ is a shorter way to do that join
sjn rakudo: say round.^methods(:local).Str
p6eval rakudo 5e7b43: OUTPUT«Not enough positional parameters passed; got 0 but expected between 1 and 2␤ in 'round' at line 1979:CORE.setting␤ in main program body at line 22:/tmp/u5HF9mqUfz␤»
masak sjn: why do you start by calling &round?
sjn: seems like that's not what you wnat.
PerlJam TimToady: Just reading the last few minutes of conversation, switching require's behavior based on the presence of a '.' sounds a little too magical to me right now.
masak s/wnat/want/
sjn masak: I'm using round as "the method I'd like to found out more about" 15:55
PerlJam masak: he only wants the round methods, not the square ones or those of other shapes
TimToady but you used it as a verb, not as anoun
masak sjn: I think you mean &round
sjn: without the sigil, it's a call.
sjn rakudo: say &round.^methods(:local).Str
p6eval rakudo 5e7b43: OUTPUT«candidates multi name␤»
flussence rakudo: my @x = "something".^methods(:local); say ~(@x.Str Z=> @x.perl);
p6eval rakudo 5e7b43: OUTPUT«ACCEPTS perl pred succ WHICH Bool Str encode [{ ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }, { ... }]␤»
flussence oh.
sjn right
flussence rakudo: my @x = "something".^methods(:local); say ~(@x».Str Z=> @x».perl);
p6eval rakudo 5e7b43: OUTPUT«get_attr_str() not implemented in class 'Sub'␤ in main program body at line 1␤» 15:56
PerlJam rakudo: say &round.signature.perl
p6eval rakudo 5e7b43: OUTPUT«get_attr_str() not implemented in class 'Perl6MultiSub'␤ in main program body at line 1␤»
flussence rakudo: my @x = "something".^methods(:local); say ~(@x».Str Z=> @x.perl);
p6eval rakudo 5e7b43: OUTPUT«get_attr_str() not implemented in class 'Sub'␤ in main program body at line 1␤»
flussence :/
15:56 masak left, masak joined
sjn seems just "playing around" isn't as easy as it could be :-\ 15:57
PerlJam If I were wishing for things, I'd wish that not so much parrot bleeds through to Perl 6.
masak sjn: we're very open to suggestion for how to make it easier.
15:58 kjeldahl left, rindolf joined
sjn well, my (rather naïve) stab at this would be to start looking for an easy way to list available methods, then a way to describe a method in any sensible way 15:59
with those two, I'd be able to explore and learn everything
PerlJam sjn: seems you've got the first.
moritz_ for the second, we need Pod and .WHY methods 16:00
sjn rakudo: say &round.^methods(:local).Str
p6eval rakudo 5e7b43: OUTPUT«candidates multi name␤»
sjn yeah, the first one is ok, although I'd live the defaults to be simpler :)
(if possible) 16:01
masak sjn: sounds like you could benefit from briefing through S32, as well.
PerlJam sjn: you'd like :local to be default?
sjn PerlJam: perhaps?
TimToady local should not be the default
PerlJam sjn: I'm wondering what other "defaults" you're referring to.
TimToady something in the middle should be the default
sjn middle is fine too
TimToady everything out to Cool or Any is a fine default, I think 16:02
sjn PerlJam: "defaults" as in "what I get when I don't add any extra arguments"
PerlJam "".^methods(:middle).Str
:)
masak I'm wary-ish about special-casing Cool and Any in that way.
TimToady in some sense, the real 'base' type is the first type derived from Cool or Any
anything outside that is "linguistic" methods 16:03
PerlJam .^methods(:base) ?
TimToady that is, methods that are defined across the language to be somewhat coercive
PerlJam .^methods(:up-level(2)) ?
sjn what's the ^ for again?
rokoteko what is the most important thing to know about perl6 type system?
PerlJam rokoteko: it's not done yet :)
rokoteko Haha :)
PerlJam (whatever "done" means) 16:04
TimToady rokoteko: the most important thing to realize is that it's there mostly to help with multiple dispatch
16:04 jaldhar left
sjn Here's my use case: Tell a kid "look, you can learn Perl 6 just by exploring the language in the REPL" 16:04
masak sjn: here it's for accessing the MOP.
ash_gti masak: btw, do you want to help write tutorials for try.rakudo.org? :P 16:05
rokoteko TimToady: well if you are answering, you could go with top 10 most important things as well.
masak sjn: but ^ in general has many meanings.
sjn: most are related to "up", though :)
16:05 cotto left
TimToady most of them mean "within the domain of" 16:05
sjn ...and then proceed to show 1 or 2 methods that are simple, non-magical, with sane defaults
ash_gti rakudo: say "foo".HOW.methods(:local)
p6eval rakudo 5e7b43: OUTPUT«too few positional arguments: 1 passed, 2 (or more) expected␤ in main program body at line 22:/tmp/QIsaL4ytTJ␤»
ash_gti oops
masak ash_gti: I'm a tad overbooked at present. but I'd be happy to review and give suggestions.
16:05 patspam1 joined, patspam left
sjn masak: MOP is a good enough answer for me, thanks :) 16:05
TimToady .HOW requires the first arg to be a repeat of "foo"
ash_gti rakudo: say ~"Foo".HOW.methods(Str, :local) 16:06
TimToady which is one of the reasons we have the .^ shortcut
p6eval rakudo 5e7b43: OUTPUT«ACCEPTS perl pred succ WHICH Bool Str encode␤»
16:06 meppl joined
ash_gti rakudo: say ~"Foo".^methods(:local); 16:06
p6eval rakudo 5e7b43: OUTPUT«ACCEPTS perl pred succ WHICH Bool Str encode␤»
PerlJam TimToady: I don't think I've ever understand that particular thing.
ash_gti sjn: those are two ways of doing the same thing
PerlJam er, s/understand/understood/
sjn ash_gti: ah, ok, thanks :) 16:07
rokoteko TimToady: is there any big 'reason' behind the type system as in haskell related to category theory? .. Not necessarily to mathmeatics of course, it would be foolish to expect that. :)
TimToady you're switching the method call to a different object, so the invocant doesn't tell the system the original object
rokoteko: on that level most of the coolness is just cargo culted. we talk about Mu as the top type, and Nil as the bottom type, but it's not really rigorous 16:08
sjn ash_gti: although the topic I'm trying to raise here is more of a pedagogical type than a implementation details-related one
ash_gti masak: if you have a moment, does my outline look okay github.com/moritz/try.rakudo.org/bl...s/index.js
rokoteko TimToady: ahh. Im not aware what bottom type means.
TimToady then Perl 6's type system is probably good enough for you :) 16:09
masak ash_gti: looks inspirational. best of luck. 16:10
ash_gti isn't multi-dispatch mostly a tool for LTM? 16:11
rokoteko No, I mean if Im planning of using this type system one day (in productio), Id be more comfortable doing it if I understand what it really does does.
16:11 Lorn joined 16:12 tomaw joined
ash_gti what part's confuse you? it supports traditional and prototype inheritance, it supports roles (traits from smalltalk, or categories from objc, or modules from ruby), its very flexible 16:12
rokoteko so the most important thing to know it's not done yet. ok Im getting PerlJam. :)
PerlJam++
16:13 snearch left, alester joined
moritz_ ash_gti: looking now 16:17
... at least if github responds at some point :-) 16:19
ash_gti its been slow for me too in the last few days 16:20
wonder whats up
moritz_ ash_gti: depending on how much you talk about regexes in Chapter 2, diving into Grammars with no exhaustive chapter on regexes might be confusing
ash_gti: apart from that, it looks pretty good
ash_gti i think i need to change chapter 1 16:21
i don't know if i should get into list ops early
16:21 wallberg joined
ash_gti i thought they a nice feature of perl6, but i think they could get confusing if your just getting your feet wet 16:21
moritz_ probably better to do it as part of chapter 3
ash_gti which is what chapter 1 should be about, i think
moritz_ maybe Chapter 0 - Teasers 16:22
with a few nice examples :-)
ash_gti 1, 1, * + * ... 21 :P 16:23
since calculating fib is really important for programmers
moritz_ rakudo: say (35, 25, *-* ... 0)[*-2] 16:25
p6eval rakudo 5e7b43: OUTPUT«(timeout)»
moritz_ rakudo: say (35, 25, *%* ... 0)[*-2]
p6eval rakudo 5e7b43: OUTPUT«5␤»
moritz_ can't remember his own inventions
GCD with serie^Wsequence operator 16:26
ash_gti does rakudo support tcp/udp yet? 16:27
moritz_ tcp yes
rokoteko ok. what does 'rakudo' mean in this context?
moritz_ rokoteko: it's the name of a Perl 6 compiler 16:28
rokoteko the modules provided within rakudo installation package?
so the perl6 *compiler* does support tcp but not udp?!?!
moritz_ currently, yes
rokoteko ok. I hear you.
what are the reasons?
moritz_ nobody implemented udp support yet 16:29
rokoteko to what?
moritz_ doesn't understand the question
16:30 cotto joined
masak heh. 'for () { ... }' doesn't work in Perl 5, but 'for (()) { ... }' does. I guess the outer parens are part of the form somehow. 16:30
rokoteko so what I have learnt this far is that perl 6 compiler doesnt support udp, but it supports tcp. I havent understood why?
moritz_ masak: they are
18:29 < moritz_> nobody implemented udp support yet
masak ok, makes sense. 16:31
rokoteko moritz_: Ill drop it for tonight. thank for all the effort you took on the subject. :)
moritz_ sees no reason for dropping. It just hasn't been done yet, that's all.
rokoteko I thinkg perl6 compiler vs some osi layer stack are orthogonal. 16:32
moritz_ rokoteko: it's just that some libraries are part of language
rokoteko: so they are included in the compiler
and nobody has written UDP libraries that rakudo could include and ship 16:33
rokoteko in which language should this udp support be implemented?
parrot? perl6? c? 16:34
moritz_ anything you can use in Rakudo: as a parrot PMC (mostly written in C), PIR, NQP, Perl 6
masak: funny detail: perl -wE 'for $a qw(foo bar baz) { say $a }'
rokoteko so in which language was the TCP stack implemented with?
moritz_ masak: works in Perl 5 without parens (but is deprecated)
masak moritz_: aye, MJD blogged about that once.
PerlJam rokoteko: IMHO, in general, the higher level you can get, the better (i.e. favor NQP over PIR or C if you can)
florz rokoteko: linux' tcp stack was implemented in C 16:35
moritz_ rokoteko: it's a Perl 6 wrapper around a parrot socket PMC, which uses the POSIX API under the hood
"it's wrappers all the way down"
ash_gti NCI means you can probably not have to drop into C to implement UDP, you might even be able to stay in NQP or perl6 16:36
rokoteko The overhead just blinds me.
moritz_ you get reusability in return
rokoteko I see many times?
moritz_ doesn't know what rokoteko sees :-) 16:37
rokoteko or in which colors!
anyhow. I just had some trouble understanding why isnt UDP "implemented in the compiler" .. I still dont completely get it, but I dont have to do that right now. :) So thanks everyone for explanation. 16:38
moritz_ rokoteko: just remember that not all parts of a compiler are concerned with transforming source code into bytecode 16:39
ash_gti at some point it becomes a runtime dependency, most socket communication is runtime not compile time
rokoteko ash_gti: yes. but I was told that "udp doesn't work" because it is "not implemented in the compiler" 16:41
moritz_ yes, and that's true.
rokoteko oh wait. I was supposed to let this go. :)
moritz_ the libraries are part of the compiler
there's no UDP library, so it's not implemented in the compiler
flussence there's always "use NativeCall;" 16:42
16:42 masak left
flussence then you can get someone else to do all the network stuff for you and use their libs instead :D 16:42
ash_gti someone could implement it outside the compiler though, i don't see any reason why it has to be implemented in the core lib, it will probably eventually make it into the core lib, but it doesn't seem to my anyway to be a required feature of the compiler
rokoteko please, let's talk about something more interesting. anyone mind explaining as much they can about my $x = @list; vs my @x = @list; vs my $x := @list; ... and vs my @x := @list ..?
ash_gti = is copy assignment 16:43
:= will bind 2 vars to point to the same thing
flussence $x = @list is like $x = [@list] in perl5, except you don't have to bother dereferencing it
ash_gti rakudo: my $a = 1; my $b = $a; $a = 2; say $b;
flussence or something like that
moritz_ flussence: more like $x = \@list in Perl 5
p6eval rakudo 5e7b43: OUTPUT«1␤»
moritz_ flussence: [...] copies
ash_gti rakudo: my $a = 1; my $b := $a; $a = 2; say $b; 16:44
p6eval rakudo 5e7b43: OUTPUT«2␤»
16:45 perlygatekeeper joined
rokoteko Hmmm. ok. that makes sence. I just once heard said that there are no references in perl6. 16:46
moritz_ that's kind of a marketing lie
flussence rakudo: say ~(*.name => *.perl for 'a'.^methods(:local))
dalek ecs: c7808c3 | TimToady++ | unknown:
introspection excludes Cool and Any by default

Methods at the level of Cool and Any are language-defined methods more than they are exclusive methods of a specific type, so should be excluded from the user's view by default.
p6eval rakudo 5e7b43: OUTPUT«===SORRY!===␤set_number_native() not implemented in class 'Num'␤»
moritz_ in Perl 6, nearly everything is a reference
16:46 wallberg left
rokoteko eval: my @arr = <a 1 b 2>; my %hash := @arr; say @arr.perl ~ " vs " ~ %hash.perl 16:48
PerlJam rokoteko: what it really means is that you don't have to think about references in Perl 6. They are largely (if not completely) transparent.
moritz_ %hash := @arr should be a type check failure
it's not yet in Rakudo, which is a known bug 16:49
rokoteko moritz_: where I could learn more about this?
link to synopsis etc is what Im looking for :) 16:50
flussence rakudo: say ('a'.^methods(:local).map: -> $a { $a.name => $a }).perl
p6eval rakudo 5e7b43: OUTPUT«("ACCEPTS" => { ... }, "perl" => { ... }, "pred" => { ... }, "succ" => { ... }, "WHICH" => { ... }, "Bool" => { ... }, "Str" => { ... }, "encode" => { ... })␤»
moritz_ perlgeek.de/en/article/5-to-6 has a short section about it
flussence there we go.
moritz_ and I think S03
rokoteko rakudo: my @arr = <a 1 b 2>; my %hash := @arr; say @arr.perl ~ " vs " ~ %hash.perl
p6eval rakudo 5e7b43: OUTPUT«["a", "1", "b", "2"] vs ["a", "1", "b", "2"]␤»
flussence not as short as I'd like however
rokoteko Ahh. Wrong bot.
moritz_ why wrong bot?
flussence std? 16:51
moritz_ std only does syntax checking
rokoteko moritz_: earlier I prefex it with "eval:" not "rakudo:" ?
flussence oh
p6eval
moritz_ right, that didn't do anything
ash_gti std: my $a = 5;
p6eval std : OUTPUT«ok 00:03 117m␤»
flussence wait,
you want "perl6:"
flussence still hasn't woken up today
moritz_ well, perl6: just runs both pugs and rakudo 16:52
ash_gti perl6: say 1;
moritz_ pugs: my @a = <1 2 3>; my %h = @a; say %h.perl
p6eval pugs, rakudo 5e7b43: OUTPUT«1␤»
pugs: OUTPUT«*** Odd number of elements found where hash expected: VList [VStr "1",VStr "2",VStr "3"]␤ at /tmp/PgZNtrBGGT line 1, column 18-28␤»
moritz_ pugs doesn't do type checking yet
s/yet// 16:53
rokoteko pugs: my @arr = <a 1 b 2>; my %hash := @arr; say @arr.perl ~ " vs " ~ %hash.perl
p6eval pugs: OUTPUT«["a", "1", "b", "2"] vs ["a", "1", "b", "2"]␤»
flussence rakudo: my @a = <a 1 b 2 quack>; my %h := @a; say @a.perl; say %h.perl;
p6eval rakudo 5e7b43: OUTPUT«["a", "1", "b", "2", "quack"]␤["a", "1", "b", "2", "quack"]␤» 16:54
flussence that looks wrong.
16:54 cdarroch joined, cdarroch left, cdarroch joined
moritz_ flussence: right; should be a type error, as mentioned above 16:54
flussence rakudo: my @a = <a 1 b 2 quack>; my %h := @a; say %h.WHAT
p6eval rakudo 5e7b43: OUTPUT«Array()␤»
flussence ouch. 16:55
ash_gti o.0
rokoteko rakudo: my @a = <a 1 b 2 quack>; my %h := @a; my %x := %h; say %x.WHAT 16:56
ash_gti std: my @a = <a 1 b 2 quack>; my %h := @a; say %h.WHAT
p6eval rakudo 5e7b43: OUTPUT«Array()␤»
std : OUTPUT«ok 00:01 118m␤»
dalek ast: ef86bef | moritz++ | S03-se (10 files):
[sequence] is the new series
16:56 dakkar left
rokoteko rakudo: my @a = <a 1 b 2 quack>; my %h := @a; my %x := %h; say %x.WHAT 16:57
p6eval rakudo 5e7b43: OUTPUT«Array()␤» 16:58
moritz_ TimToady: S11 seems to imply that 'require Foo'; stubs a 'Foo' lexpad entry into the current scope; std doesn't. Which one should I believe? 17:00
also 17:01
require "/home/non/Sense.pm" <common @horse>;
looks scarily like two arbitrary terms in a row
17:04 wamba joined
ash_gti TimToady: does the new spec change mean it skips Cool and Any or it stops at Cool or Any? 17:06
17:06 sjn left
moritz_ this whole 'require' thing looks like badly ported from Perl 5 17:06
17:09 sjn joined
dalek kudo: c94bfe1 | moritz++ | src/Perl6/Grammar.pm:
parse require
 17:09
kudo: b3efa7a | moritz++ | src/Perl6/Actions.pm:
implement argumentless form of require
kudo: 7b21bf1 | moritz++ | src/Perl6/ (2 files):
implement require EXPR;
kudo: 82ebb54 | moritz++ | t/spectest.data:
[t/spectest.data] track file name changes in roast
moritz_ require "string" <import list>; # evaluates "string" at run time, and <import list> at compile time 17:10
that's quite confusing
so, require needs a redesign, but I have no idea how the redisgn might look like 17:12
ideas welcome.
flussence does require have a verbose form, like module names?
moritz_ flussence: what do you mean?
flussence require Module:file<"string">:import<x y z>; or something like that
where I guess :file<> would override the default search path 17:13
TimToady moritz_: believe the spec on require; that particular form is nyi in std, apparently. or maybe nli...
moritz_ std: require Foo <$x>; $x 17:14
p6eval std : OUTPUT«===SORRY!===␤Variable $x is not predeclared at /tmp/CpiEgfgdUe line 1:␤------> require Foo <$x⏏>; $x␤Bogus term at /tmp/CpiEgfgdUe line 1:␤------> require Foo <$x>⏏; $x␤Parse failed␤FAILED 00:01 117m␤»
moritz_ right, NYI
TimToady or NLI
:)
moritz_ where L = ?
TimToady ash_gti: stops at Cool or Any
ash_gti okay, just curious 17:15
TimToady no longer implemented :/
moritz_ I thought "Not Likely to by Implemented" :-) 17:16
TimToady ah, getting aced out on LTM
moritz_ uhm, did STD try to interpolate $x into the <> quote? 17:17
TimToady shouldn't be getting aced out, though...
it parsed it as a listop rather than as a package_declarator, for some reason
moritz_ should I bother p6l about the require thing? 17:20
TimToady why? 17:21
tadzik wklej.org/id/392982/ -- what am I missing? . is in INC, it says
flussence
.oO( require Foo:in<../lib/ ./>:for<import1 import2>; )
ash_gti TimToady: with .^methods(:excl(Foo)); would that skip foo or stop at foo? Assuming your inheritance was Bar -> Foo -> Baz and you invoked .^methods on Baz
TimToady :excl doesn't take an argument
moritz_ TimToady: because I'm looking for ways to rethink or improve require(), and I don't have any good ideas 17:22
TimToady it's just short for "exclusive"
ash_gti oh
okay, sorry, misunderstood its purpose, i get it now
moritz_ wait... can't we just stub the lexicals with 'need'?
TimToady not if the module doesn't exist yet 17:23
require doesn't commit to that
"assume there will be a module named Foo by the time you run this, and define these variables that will come from it then"
tadzik or, what am I missing this time? wklej.org/id/392983/
moritz_ tadzik: you need 'our sub foo', and then call m1::foo 17:24
tadzik moritz_: doesn't 'is export' make foo visible? 17:25
moritz_ tadzik: no, it just marks it for exporting
tadzik: but since importing is done into the lexical pad, and the lexical pad is immutable at run time, you can't import at run time
tadzik ah 17:26
moritz_ which is why the second form require m1 <foo>; exists
which stubs the &foo entry in the lexpad at compile time
and then binds at runtime
(but that form is NYI)
tadzik so how do I call a function if I have a module name in string? Through eval? 17:27
I have to 'require $str; $str::foo()'?
moritz_ TimToady: my complaints with the current require specs are 1) the 'require "string" <importlist>' is two terms in a row (with no sufficient special-casing of the first term) and 2) that it has two arguments, the first evaluated at run time, the second at compile time. I think that's very confusing 17:28
tadzik: require $str <foo>; foo; # once it's implemented
tadzik oh, ok
TimToady moritz_: it's not a function; declarators cheat on TTIAR all over the place 17:29
moritz_ TimToady: that works for me, if we have precise rules of what is allowed as one of the terms 17:30
and S11 is very vague on that
TimToady we will as soon as I fix STD
moritz_ ok
moritz_ fears a fix that can't easily be ported to rakudo 17:31
TimToady it's the same syntax as use, more or less
moritz_ use doesn't allow strings as first term though 17:32
TimToady that's the less part :) 17:33
moritz_ right :-) 17:34
that's the one I worried about from the start
ash_gti hand wavey magic
17:35 patspam joined
TimToady shower & 17:35
17:36 envi^home left
moritz_ std: Foo::Bar 17:37
17:37 patspam1 left
p6eval std : OUTPUT«===SORRY!===␤Undeclared name:␤ 'Foo::Bar' used at line 1␤Check failed␤FAILED 00:02 114m␤» 17:37
17:37 wamba left 17:38 rrgau_ left 17:39 rrgau_ joined 17:41 stkowski joined 17:42 rrgau_` joined
dalek ast: e9774db | moritz++ | S11-modules/require.t:
basic tests for require
 17:42
17:44 rrgau_ left
moritz_ feels unusually productive 17:45
colomon moritz_++ 17:46
dalek kudo: 107cc16 | moritz++ | t/spectest.data:
run tests for require
kudo: 13c8a23 | moritz++ | docs/ChangeLog:
[docs] update Changelog with require and Bool stringification
colomon feels unusually unproductive 17:47
17:49 daxim left 17:55 perlygatekeeper left 18:00 perlygatekeeper joined 18:03 Italian_Plumber left 18:14 jest joined 18:15 jest left, jest joined
jest hi guys 18:15
18:16 mberends joined
jest I have problem building Using Perl 6 18:16
when running make I receive error:
pastie.org/1179671 18:18
tadzik oh, I had something similar
jest When I enter build/ and run 'makeindex UsingPerl6.a4.idx' and then 'make -I ../lib -f ../lib/Makefile' it produces PDF 18:19
with the most serious "error" message: Overfull \hbox (171.8775pt too wide) in paragraph at lines 3623--3636 18:20
(which is the last table in 11.2 much to wide) 18:21
does anyone feel competent to correct the Makefile's makeindex problem? 18:22
18:24 M_o_C joined
jest tadzik: did you find the reason? 18:26
tadzik jest: nope
moritz_ jest: please file a bug report at github.com/perl6/book/issues 18:28
18:30 ash_gti left 18:36 rlb3 joined 18:39 rlb3 left, [Coke] joined 18:40 tadzik left 18:41 tadzik joined
patspam std: 1..2.PARROT 18:45
p6eval std : OUTPUT«ok 00:01 116m␤»
patspam segfaults for me (after a while)
18:45 stkowski left
TimToady in rakudo? 18:53
rakudo: say 2.PARROT
p6eval rakudo 5e7b43: OUTPUT«Int␤»
TimToady rakudo: say 1..Int 18:54
p6eval rakudo 5e7b43: OUTPUT«␤»
jest moritz_: done, github.com/perl6/book/issues/issue/20
moritz_ jest++
18:56 jferrero joined
patspam TimToady: in rakudo star 18:57
18:57 timbunce left 18:58 stkowski joined
TimToady star: say 1..Int 19:02
p6eval star 2010.07: OUTPUT«␤»
TimToady note that it parses as 1 .. (2.PARROT) 19:03
19:03 timbunce joined 19:17 fod left
colomon std: (1..2).PARROT 19:21
p6eval std : OUTPUT«ok 00:01 117m␤» 19:22
colomon rakudo: say (1..2).PARROT
p6eval rakudo 13c8a2: OUTPUT«Range␤»
19:26 hanekomu_ joined 19:30 plainhao left
patspam 2.PARROT is ok. 1..(2.PARROT) segfaults 19:31
does this mean I get to submit my first rakudo bug? 19:32
colomon yes 19:34
patspam wooh!
colomon what platform are you on? 19:35
star: 1..(2.PARROT)
p6eval star 2010.07: ( no output )
19:36 ggoebel left
patspam Ubuntu 10.04 19:36
colomon I can't duplicate it here, but if 1..(2.PARROT) segfaults for you, that's definitely something that needs to be reported. 19:37
(at least, assuming other things on your build work -- can you do (1..2).PARROT?) 19:38
patspam yeah that one works 19:40
19:40 mfollett joined 19:42 pyrimidine joined
colomon do you know how to report a bug? 19:45
patspam following: rakudo.org/submit-a-bug-report
colomon patspam++ 19:46
thanks!
(I mean, thanks for reporting a bug and thereby helping.) 19:48
patspam pleasure!
gets more interesting.. if you run it as perl6 -e '1..2.PARROT' it does nothing (including not segfaulting). whereas if you put a "say" in there, it spews numbers onto the screen 19:49
for a while, until it segfaults
colomon oooo 19:53
rakudo: say 1..2.PARROT
p6eval rakudo 13c8a2: OUTPUT«␤»
19:54 timbunce left
colomon what version of Rakudo are you using? 19:54
patspam "version 2010.07-47-g9fd5eaa built on parrot 2.6.0 r48152" 19:55
tadzik star: say 1..2.PARROT
p6eval star 2010.07:
..OUTPUT«(timeout)101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130…
patspam there she blows!
19:55 wallberg joined 20:00 timbunce joined
patspam bug submitted as #78034 20:01
20:03 tadzik left
patspam btw, I found it while reading the Using Perl 6 book (in preparation for tomorrow's Rakudo Star Study Group via Perl Seminar NY) - so yay for the book 20:04
slavik omg omg omg 20:07
star: say 2 ** 64
p6eval star 2010.07: OUTPUT«1.84467440737096e+19␤»
slavik aww :(
still no bignum
flussence rakudo: say ('0b1' ~ '0' x 60 ~ '1000').Int === ('0b1' ~ '0' x 60 ~ '1001').Int 20:09
p6eval rakudo 13c8a2: OUTPUT«Bool::True␤»
slavik huh? 20:10
flussence oh... that adds up to 65
rakudo: say ('0b1' ~ '0' x 60 ~ '100').Int === ('0b1' ~ '0' x 60 ~ '101').Int
p6eval rakudo 13c8a2: OUTPUT«Bool::True␤»
flussence I think those are getting eaten by floating-point underflow somewhere... 20:11
colomon rakudo: say ('0b1' ~ '0' x 60 ~ '101').Int.WHAT 20:17
p6eval rakudo 13c8a2: OUTPUT«Int()␤»
colomon :\
moritz_ rakudo: say Inf == Inf 20:18
p6eval rakudo 13c8a2: OUTPUT«Bool::True␤»
20:18 Italian_Plumber joined
colomon rakudo: say ('0b1' ~ '0' x 60 ~ '101').Int 20:18
flussence they come out as -9223372036854775808, which is the same I get for really big numbers in perl5 & php
p6eval rakudo 13c8a2: OUTPUT«-9223372036854775808␤»
colomon rakudo: say ('0b1' ~ '0' x 60 ~ '101') 20:19
p6eval rakudo 13c8a2: OUTPUT«0b1000000000000000000000000000000000000000000000000000000000000101␤»
colomon rakudo: say +('0b1' ~ '0' x 60 ~ '101')
20:19 bluescreen left
p6eval rakudo 13c8a2: OUTPUT«9.22337203685478e+18␤» 20:19
colomon so this is really a story about converting Nums which are too big to Ints... 20:20
flussence rakudo: say 2**63 + 1 20:21
p6eval rakudo 13c8a2: OUTPUT«9.22337203685478e+18␤»
flussence rakudo: say (2**63 + 1).Int
p6eval rakudo 13c8a2: OUTPUT«-9223372036854775808␤»
flussence eww, chrome's v8 gets it completely wrong. 20:23
Math.pow(2, 63) == 9223372036854776000
mfollett Where would I find a list of all the things I can call on $obj.HOW at this point in time? 20:25
20:25 M_o_C left
moritz_ mfollett: src/metamodel/*HOW.pir 20:26
mfollett moritz_: Thanks!
20:31 Italian_Plumber left, bluescreen joined 20:33 stkowski left
mfollett Why do I have to provide an object to add_method when I'm adding a method to a class, e.g.: Dog.HOW.add_method( Dog.new(), 'bark', {say 'woof'}) 20:33
in action: 20:40
oh, maybe I can't fit that on one line with the declaration of the dog class. 20:41
rakudo: class Dog {}; Dog.HOW.add_method( 'bark', {say 'woof'})
p6eval rakudo 13c8a2: OUTPUT«too few positional arguments: 3 passed, 4 (or more) expected␤ in main program body at line 22:/tmp/EdbpJHhtI4␤» 20:42
mfollett class Dog {}; Dog.HOW.add_method( Dog.new(), 'bark', {say 'woof'})
oops
rakudo: class Dog {}; Dog.HOW.add_method( 'bark', {say 'woof'}); Dog.new.bark()
p6eval rakudo 13c8a2: OUTPUT«too few positional arguments: 3 passed, 4 (or more) expected␤ in main program body at line 22:/tmp/tyR4YoysU5␤»
mfollett rakudo: class Dog {}; Dog.HOW.add_method( Dog.new(), 'bark', {say 'woof'}); Dog.new.bark 20:43
p6eval rakudo 13c8a2: OUTPUT«woof␤»
PerlJam rakudo: class Dog {}; Dog.HOW.add_method( Dog(), 'bark', {say 'woof'}); Dog.new.bark 20:51
p6eval rakudo 13c8a2: OUTPUT«Could not find sub &Dog␤ in main program body at line 22:/tmp/2aJltlhEFf␤»
mfollett turns out you can do this: 20:52
rakudo: class Dog{}; Dog.new.HOW.add_method( 'Dog', 'bark', { say 'woof' }); Dog.new.bark 20:53
p6eval rakudo 13c8a2: OUTPUT«woof␤»
mfollett Why does it need that second parameter though? 20:54
20:54 rindolf left, plobsing joined
mfollett I guess ClassHOW does not know what class it would be operating on in that case? 20:57
21:07 mfollett left 21:10 amkrankruleuen left 21:16 kjeldahl_ left 21:17 jest left 21:18 mfollett joined 21:21 alext_ joined 21:22 patspam left
alext_ hi, can someone point me to the regex grammars tutorial or documentation ? 21:22
PerlJam alext_: There's a book in the works with a chapter on regex and another on grammars at github.com/perl6/book 21:24
21:25 bluescreen left
alext_ thanks! 21:25
PerlJam alext_: there's also more resources on perl6.org
21:28 Ross left
alext_ perl 6 tutorial on the tpf site in german only 21:29
21:36 pyrimidine left 21:37 patspam joined
dalek href="https://perl6.org:">perl6.org: a4d5b66 | perlpilot++ | source/documentation/index.html:
update the advent calendar

  * change the URL to point to the page that lists all days
  * Add the year in the link in preparation for the 2010 advent season		 21:42
21:47 ruoso left, Chillance left 21:58 timbunce left 21:59 Holy_Cow left 22:07 mikehh joined 22:11 mfollett left 22:21 icwiener left 22:28 thepler left 22:42 hanekomu_ left 22:43 ggoebel joined, wallberg left 22:50 Italian_Plumber joined
sorear good * #perl6 22:55
22:55 patspam1 joined 22:57 patspam left 22:59 Grrrr left
sorear PerlJam: You want to see less Parrot in your Perl 6? Join me ;) 23:00
diakopter star: say &say.WHO.HOW 23:13
p6eval star 2010.07: OUTPUT«Method 'HOW' not found for invocant of class 'Multi'␤ in main program body at line 22:/tmp/gov1vpH_l8␤»
23:13 patspam1 left 23:19 patspam joined
sorear How is Perl 6 doing on the 'Report' part of the backronym? I notice we don't have a one-liner for generating a nice table display of Pascal's Triangle 23:24
colomon rakudo: say ([1], -> @a { @a, 0 >>+<< 0, @a } ... *).munch(5).perl 23:33
p6eval rakudo 13c8a2: OUTPUT«([1], (1, 0, 1), (1, 0, 1, 0, 1, 0, 1), (1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1), (1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1))␤»
23:33 kid51 joined
sorear I don't call that "nice" 23:33
colomon rakudo: say ([1], -> @a { (@a, 0) >>+<< (0, @a) } ... *).munch(5).perl
p6eval rakudo 13c8a2: OUTPUT«([1], [1, 1], [1, 2, 1], [1, 3, 3, 1], [1, 4, 6, 4, 1])␤»
colomon I don't call it pascal's triangle, either
rakudo: say ([1], -> @a { (@a, 0) >>+<< (0, @a) } ... *).map({ $_ ~ "\n" }).munch(5).say 23:35
p6eval rakudo 13c8a2: OUTPUT«1␤1 1␤1 2 1␤1 3 3 1␤1 4 6 4 1␤␤Bool::True␤»
colomon rakudo: say ([1], -> @a { (@a, 0) >>+<< (0, @a) } ... *).map({ $_ ~ "\n" }).munch(5)
p6eval rakudo 13c8a2: OUTPUT«1␤1 1␤1 2 1␤1 3 3 1␤1 4 6 4 1␤␤»
colomon rakudo: say ([1], -> @a { (@a, 0) >>+<< (0, @a) } ... *).map({ $_ ~ "\n" }).munch(6)
p6eval rakudo 13c8a2: OUTPUT«1␤1 1␤1 2 1␤1 3 3 1␤1 4 6 4 1␤1 5 10 10 5 1␤␤»
colomon dunno if that counts as nice or not 23:36
sorear colomon: I'd like each number in a box of equal width, staggered so the center column is in a constant position, half of boxes in the triangle empty
colomon is there any language out there wherein that is a one liner? 23:37
sorear There should be!
TimToady you should really be using Z+, then you don't need the parens 23:40
sorear ponders a practical report language for the 21st century 23:41
(we have extraction down)
TimToady printing pascal's triangle is not "practical" to anyone but a mathematician :) 23:42
the specific problem of the triangle is that the cells have to be sized by later results
but if you abandon cells, we could pretty easily do centering of pre-rendered lines 23:45
23:46 ruoso joined
TimToady or we could just embed TeX... 23:47
sorear tex likes points 23:50
a lot
embedding nroff might work better
TimToady hard to find the center character of '20'
we need half-spaces :)
oh wait, we'll just use the double-wides 23:51
２０ 23:52
sorear hmm, niecza is going to have issues with my %x; in protopads 23:53
it'll be interesting to ensure that class Hash; is defined first
sorear thinks double-wide ASCII chars are silly
TimToady sure, but you'll note it lets you center a digit 23:54
pity it's probably in a different font face...
also a pity it doesn't actually give you generalized half-spacing :) 23:55
unless you count double-spacing as normal, of course 23:56

Powered by Cro and the Raku® Programming Language.

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

Thank you!