»ö« | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moritz_ on 25 June 2010.
00:02 Psyche^ joined 00:03 Patterner left, Psyche^ is now known as Patterner 00:04 masonkramer left 00:09 Boyo joined, Boyo is now known as Boyo17 00:10 meppl left 00:11 risou left
dalek odel: 78103c0 | ++ | java/runtime/ (7 files):
[java] expand the dependency circle with Context and ThreadContext
 00:11
01:07 hsb left 01:11 isBEKaml joined 01:30 drbean joined 01:39 sekimura left 01:46 justatheory left 01:47 masonkramer joined, justatheory joined 01:48 jaldhar joined 02:14 masonkramer left 02:15 Boyo17 left 02:40 Holy_Cow joined, Holy_Cow left 02:56 justatheory left 03:06 whiteknight left 03:07 sftp_ left 03:33 tylercurtis joined 03:44 molaf joined 03:45 envi^home joined 03:52 _twitch joined 04:07 zorgnax left, synth joined 04:17 hercynium left 04:19 molaf left 04:35 plobsing left
tylercurtis What context is the last statement in a compilation unit in? Sink? 04:42
isBEKaml .u 􏿽xB8 04:46
phenny UnicodeDecodeError: 'utf8' codec can't decode byte 0xb8 in position 0: unexpected code byte (file "/usr/lib/python2.5/encodings/utf_8.py", line 16, in decode)
isBEKaml pytohn? :)
Not bad.
04:51 hans__ joined
hans__ hi I am new to perl 6 I'll need some help 04:54
tylercurtis hans__: What do you need help with?
hans__ some tests because I'm new to irc, too 05:00
05:00 patspam left
hans__ how can I add 2 numbers and print to the screen? 05:01
\msg p6eva rakudo say 37 + 5 05:05
\msg p6eval rakudo say 37 +5 05:06
isBEKaml hans__: you need a forward slash. Like so, "/msg p6eval rakudo: say 37 + 5"
hans__ thank you
05:07 azert0x left
isBEKaml np and welcome to perl6! hang out here more often. :) 05:07
05:11 drbean left 05:16 cognominal left
hans__ how do you parse options in a perl6 program? (e.g. getopt in perl5) 05:23
tylercurtis hans__: perlcabal.org/syn/S06.html#Declarin...subroutine
hans__: perlgeek.de/en/article/5-to-6#post_14 is a more tutorial-like article about MAIN subs. 05:26
05:28 cognominal joined 05:30 fod left 05:31 plobsing joined
isBEKaml perl6advent.wordpress.com/2009/12/2...d-grammar/ 05:38
the link to Rakudo's Grammar.pm in the page is broken. It points to ng. 05:39
I think it should point to master always.
05:49 Guest23195 joined 05:50 plobsing left
sorear if you want to get fancier, you can also use the @*ARGS array 05:55
ng is master, no? 05:57
isBEKaml sorear: I thought so, when I clicked on the link. github threw me a 404. 05:58
Do anyone of you have a link to get the text of day24 in p6advent2009 calendar? firefox here is hung over that I'm building rakudo. :( 06:00
Or is available anywhere in pugs repo?
mberends isBEKaml: svn.pugscode.org/pugs/t/spec/integration/ 06:01
isBEKaml mberends: that's the test. I wanted "Text". 06:02
mberends: I'm viewing it over lynx, but that's pretty uncomfortable. :/
mberends oops, misread (just woken up). I'll try. 06:03
06:04 thepler left
isBEKaml mberends: :) thanks. 06:07
mberends isBEKaml: nopaste.snit.ch/23180 06:08
isBEKaml mberends: oh, you still haven't had your coffee? Day24. :D 06:09
mberends oh :D
isBEKaml mberends: well, never mind. I'm about to finish it. :D
hans__ thank you for the tip, bye 06:11
06:11 hans__ left
mberends isBEKaml: no, gotta get it right now... ;) (Booting the numb skull) 06:11
isBEKaml mberends++ # MOAR COPHEEE 06:12
mberends isBEKaml: nopaste.snit.ch/23181 : Perl 6 Advent Calendar 24th December 2009 - The Perl 6 standard grammar - pmichaud 06:14
ok, I feel awake now :)
isBEKaml mberends: :) 06:18
lue how can I make a range/series count by, say, every other number? (i.e. have 0..10 count 0,2,4,6,8,10) 06:24
isBEKaml rakudo: (0,2 ... *).munch(6).say 06:28
p6eval rakudo a20e62: OUTPUT«0246810␤»
isBEKaml lue: ^^
lue rakudo: say (0,2 ... 10) 06:29
p6eval rakudo a20e62: OUTPUT«0246810␤»
lue thank you. [the spec said that 0 ... *+2, 10 would work. It doesn't, so I was a bit confused]
isBEKaml pfft, YAGNI moment. But never mind. :) 06:30
TimToady 0, *+2 ... 10 is what the spec says 06:31
06:34 jhuni joined
lue S03:3380 states otherwise [which is where I found information]. 06:35
06:39 molaf joined
pugssvn r32108 | lwall++ | [S03] fix fossil found by lue++ 06:41
TimToady the section that is about the series operator is more authoritative
06:41 molaf left
lue rakudo: my $a = "hello there"; my @b = $a.comb[$_] for 0..3; say @b 06:42
p6eval rakudo a20e62: OUTPUT«l␤»
lue now why am I suddenly unable to do this? [I'm pretty sure I have it right]
TimToady you donh't 06:43
you're reiniting @b each time
lue aah. 06:44
rakudo: my $a = "hello there"; my @b; @b = $a.comb[$_] for 0..3; say @b
p6eval rakudo a20e62: OUTPUT«l␤»
TimToady the assignment operator clobbers @b
it's not the same as a push 06:45
tylercurtis rakudo: my $a = "hello there"; my @b = ($a.comb[$_] for 0..3); say @b
p6eval rakudo a20e62: OUTPUT«␤»
tylercurtis rakudo: my $a = "hello there"; my @b ,= $a.comb[$_] for 0..3; say @b
p6eval rakudo a20e62: OUTPUT«hell␤»
TimToady what you want is a 0..3 subscript to slice it
lue rakudo: my $a = "hello!"; my @b = $a.comb[0, 2 ... $a.chars]; say @b 06:47
p6eval rakudo a20e62: OUTPUT«hloAny()␤»
06:48 shade\ left
lue rakudo: my $a = "hello!"; my @b = $a.comb[0, 2 ... ^$a.chars]; say @b 06:48
p6eval rakudo a20e62: OUTPUT«hello!␤»
lue rakudo: my $a = "hello!"; my @b = $a.comb[0, 2 ...^ $a.chars]; say @b
p6eval rakudo a20e62: OUTPUT«hello!␤»
TimToady rakudo: my $a = "hello!"; say $a.comb[0, 2 ... *] 06:49
p6eval rakudo a20e62: OUTPUT«(timeout)»
TimToady heh
lue rakudo: my $a = "hello!"; say $a.comb[$_] for 0, 2 ... * 06:50
p6eval rakudo a20e62: OUTPUT«(timeout)()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤Any()␤»
TimToady rakudo: my $a = "hello!" say $a.comb[0,2...$a.chars-1] 06:52
p6eval rakudo a20e62: OUTPUT«===SORRY!===␤Confused at line 22, near "my $a = \"h"␤»
TimToady rakudo: my $a = "hello!"; say $a.comb[0,2...$a.chars-1]
p6eval rakudo a20e62: OUTPUT«hlo␤»
TimToady that will have to do for now, till subscripts trim infinite lists correctly
or till ...^ implemented 06:53
lue rakudo: my $a = "hello"; say $a.comb[0,2...$a.chars-1] 06:54
p6eval rakudo a20e62: OUTPUT«hlo␤»
06:55 shade\ joined 06:58 shade\ left
lue rakudo: my $a = "hello!"; say $a.comb[1,3...$a.chars-1] 06:58
p6eval rakudo a20e62: OUTPUT«el!␤»
lue rakudo: my $a = "hello"; say $a.comb[1,3...$a.chars-1] 06:59
p6eval rakudo a20e62: OUTPUT«el␤»
07:02 isBEKaml left, shade\ joined 07:03 isBEKaml joined
lue is there an easy way to turn every element in an array (say [1,2,3]) into a string? 07:05
moritz_ ~<<[1, 2, 3]
.map ~*
TimToady .map *.=Str to do it in place 07:06
lue Ah, I forgot that ~ strigifies.
TimToady well, with a colon
lue [IIUC, I'm trying to implement the Luhn test for Rosetta Code]
isBEKaml rakudo: (0...^ 6).perl.say 07:08
p6eval rakudo a20e62: OUTPUT«((0), [1, 2, 3, 4, 5])␤»
isBEKaml rakudo: (0 ... ^6).perl.say
p6eval rakudo a20e62: OUTPUT«((0), [1, 2, 3, 4, 5])␤»
TimToady doesn't know that yet
std: 0 [...^] 6
p6eval std 32108: OUTPUT«===SORRY!===␤Unable to parse bracketed infix at /tmp/C2mxzvJEsY line 1:␤------> 0 ⏏[...^] 6␤Couldn't find final ']'; gave up at /tmp/C2mxzvJEsY line 1:␤------> 0 [...⏏^] 6␤Parse failed␤FAILED 00:01 115m␤»
isBEKaml Why is '0' separated from the rest? 07:09
TimToady curious
isBEKaml and, I'm curious about the difference between '0 ... ^6' and '0 ...^ 6'. Is there any?
TimToady no 07:10
not yet
not till it's implemented
07:10 _twitch left
isBEKaml :0 07:10
TimToady rakudo: say 0 ... 0,1,2,3,4,5
p6eval rakudo a20e62: OUTPUT«012345␤»
isBEKaml what would be the difference then?
TimToady rakudo: say 0 ... (0,1,2,3,4,5) 07:11
p6eval rakudo a20e62: OUTPUT«012345␤»
TimToady ...^ would do what you want, and ... ^ wouldn't
rakudo: say 0 ... [0,1,2,3,4,5]
p6eval rakudo a20e62: OUTPUT«012345␤»
TimToady rakudo: say 0 ... 0..5
p6eval rakudo a20e62: OUTPUT«012345␤» 07:12
07:12 drbean joined 07:13 gfx joined
isBEKaml I think I get it. Also, the output here is somewhat confusing - a20e62: OUTPUT􏿽xAB((0), [1, 2, 3, 4, 5])NL􏿽xBB 07:13
TimToady kinda looks like someone tried to implement ...^ and failed
but that wouldn't explain 0 ... ^6 07:14
isBEKaml it's the same in both (with and without space)
TimToady rakudo: say 0 ... (^6)
p6eval rakudo a20e62: OUTPUT«012345␤»
TimToady beats me
isBEKaml I think .perl is the culprit. 07:15
rakudo: say (0 ... (^6)).perl;
p6eval rakudo a20e62: OUTPUT«((0), [1, 2, 3, 4, 5])␤»
isBEKaml :|
lue rakudo: say (0 ...^ 6).Str; say (0 ... ^6).Str;
p6eval rakudo a20e62: OUTPUT«0 1 2 3 4 5␤0 1 2 3 4 5␤» 07:16
moritz_ (grammar.pm link fixed on the advent calendar) 07:26
07:38 zulon joined
dalek odel: 451f57c | ++ | java/runtime/ (3 files):
[java/runtime/Makefile] update the circular dependency diagram, uncomment most parts of Context
 07:42
isBEKaml java runtime? why don't I see the author here than just the '++' :) 07:43
sorear it's mberends, but dalek fails as usual on it 07:44
mberends isBEKaml: my git settings are slightly br0ken on this project
isBEKaml mberends! awesome. :)
mberends other projects work fine from the same computer
I like the ascii art in the Makefile :) 07:45
moritz_ mberends: usually it's fine to have username and email in the global git config, and no user/email in the per-repo config at all
sorear mberends: you are familiar with the way dalek gets nicknames, right?
sorear achieves Enlightenment Level 1 with the nqp-rx parser engine 07:46
07:47 zulon left
mberends moritz_, sorear: I'm totally ignorant about dalek's workings. I have only global git settings and they work with other projects (rakudo etc). Also 6model commits displayed fine initially but dalek began dropping my ID after a while. The 6model git log still records my name the same as before. I can't be bothered much to look into it right now, in a few days I'll be at home and will try pushing from a clone on another computer. 07:51
moritz_ mberends: I also noticed that git log showed your name correctly. (which is why I wondered why you attributed it your git config, not dalek) 07:53
isBEKaml mberends: I think you'll have to look at user.name and user.email listings in git config --list
mberends isBEKaml: just checked, and it's only global user.name and user.email. :-( 07:54
isBEKaml mberends: I think if your commits show up properly int he logs, it's dalek that's the problem not your config settings. 07:55
moritz_ curl github.com/api/v2/yaml/commits/list...del/master
shows that the 'login' field is empty
which I don't understand
mberends: ah, you use a different email address than that registered with github 07:56
(github vs. mberends) @ autoexec.demon.nl 07:57
maybe that causes github not to recognize your commits as coming from the 'mberends' github account
isBEKaml moritz_: I'm curious. I can see mberend's commits in the git repo logs I just cloned. 07:58
moritz_: even the email you refer to is seen to the same in the git logs and the curl output. 07:59
or am I missing something?
moritz_ isBEKaml: but it's not the one you see on github.com/mberends/ 08:00
isBEKaml: and since git allows you to push other people's commits, there has to be a way to map the commits back to a nickname
mberends moritz_: that makes a lot of sense. I should rename my global user.email to github@autoexec, it's part of my spam management strategy.
moritz_ s/nickname/github login/
I've also asked Infinoid to make the author info fallback to the name if login is empty 08:01
isBEKaml moritz_: ah, makes sense. gravatars and email mapping. 08:02
mberends: is there any reason you have put xml like entries in a string of single line comments? 08:07
mberends: I kind of find that distracting when looking over the java code. :)
08:08 zulon joined
mberends isBEKaml: I usually prefer a mx line length of 72 chars (for POD etc) but this was trying to deviate as little as possible from the C# original. So I blame jnthn++ ;) 08:09
08:09 wamba joined
isBEKaml mberends: if you're trying to document the methods that are there in the java code, you're probably better off with javadoc comments unless you want to generate pods from them. :) 08:10
mberends: also, you can use /* ... */ multiline comments. /// is rather distracting to code-reading. 08:11
mberends isBEKaml: so I agree with you, I don't like the long lines either. I do plan to improve the comments. It's nice to be motivated by knowing people like you are reading them :-) 08:12
sorear aha 08:14
using github's data is _so much smarter_ 08:15
back before the move to github api, dalek got usernames by scraping CREDITS
parrot's CREDITS.
no matter what project you were on
isBEKaml mberends: I was only curious when I was cloning the repo. But now I'm interested! =) 08:16
tylercurtis Good night, #perl6.
mberends isBEKaml: jnthn++ has been above average in commenting his C# code, but to me it is still not explanatory enough. After I've hacked some more code to 1) compile, and 2) run? I'll definitely add more docs. 08:17
isBEKaml mberends: I cannot speak for C#, but I sure can help around with java.
mberends isBEKaml: cool! Patches Welcome (tm) as always :) 08:18
isBEKaml mberends: ah, looking over C# code, I see that you have tried to retain much of what is there. :)
08:19 tylercurtis left
mberends isBEKaml: especially when I don't understand (yet) what it all does. 08:19
08:19 Mowah joined
isBEKaml mberends: I don't have the full picture either. More than once, I was wondering "What is he trying to do?" 08:20
mberends isBEKaml: learning more is my main reason working on this. Oh, and collecting karma points. 08:22
isBEKaml mberends++ I will try going over it sometime. If I do "get" something, you'll definitely see me pushing patches! :) 08:23
s/pushing/\1 git pull/ 08:24
mberends isBEKaml: an interesting "challenge" is guessing what is in some of the files that jnthn has not committed yet, but uses on the dotnet side ;)
isBEKaml mberends: not committed yet? 08:25
I don't understand.
mberends: Ah, mind reading and anticipation! :D 08:26
mberends isBEKaml: in the dotnet/ tree there are references to classes for which no source files have been pushed.
mberends must show jnthn the 'git status' command some time 08:27
isBEKaml mberends: your makefile is pretty weird. you have added every class as a target in it. 08:35
mberends: why not just compile them all together? or simpler, use ant?
mberends isBEKaml: I'm trying to keep things portable for Windows nmake, which is not as clever as GNU make. 08:36
isBEKaml mberends: No, that is not what I meant. javac has some switches that enable you to compile all classes. 08:37
mberends: ant works fine on windows too.
dalek odel: 3aa2fd7 | mberends++ | java/runtime/Rakudo/Metamodel/KnowHOW/KnowHOWBootstrapper.java:
[KnowHOWBootstrapper.java] a mostly commented out version, mainly for karma points
 08:38
mberends isBEKaml: I'll look into it, unless you beat me to it. But if ant is an extra dependency to install on Windows, I'm reluctant to add that to the beginner's requirements list. 08:39
isBEKaml mberends: every java user would know about ant i think. Besides, I don't think you're going to allow the end user to build classes themselves, right? YOu're only going to provide the compiled binaries. So it's just a build dependency. 08:41
mberends: s:1st/user/developer/ 08:42
mberends isBEKaml: yes Ant is well known. The beginner developer may or may not have it, and so far I do not mind maintaining a few extra Makefile lines to lower the barrier to getting started. I hope jnthn will soon be trying to compile the java code, and I'd rather not hassle him with needing to install Ant if he doesn't have it yet. 08:46
moritz_, sorear: \o/ I got my karma points, thanks for the help! 08:49
isBEKaml mberends: hmmm, forked the repo. I expect some miracle will happen soon! :) 08:50
isBEKaml hopes for a miracle opening....
mberends isBEKaml: anticip....ation! # sorry, Rocky Horror
afk & # England -> Slovenia -> Netherlands 08:54
moritz_ Slovenia doesn't look like the shortest route to me, if you want to travel from England to the Netherlands :-) 08:55
isBEKaml mberends: :) 09:03
mberends the route is more -Ofun that -Ofast 09:07
sorear does anyone else want to share the karma from niecza?
;)
mberends sorear: yes, please. Do you charge in tuits? 09:08
sorear tuits and inspiration 09:09
rakudo: say "foo" ~~ /<from>/ 09:11
p6eval rakudo a20e62: OUTPUT«Method '!cursor_names' not found for invocant of class 'Integer'␤ in <anon> at line 22:/tmp/MurOlDi3PK␤ in 'Cool::match' at line 2404:CORE.setting␤ in 'Regex::ACCEPTS' at line 5812:CORE.setting␤ in main program body at line 22:/tmp/MurOlDi3PK␤»
sorear rakudo: say "foo" ~~ /<next>/
p6eval rakudo a20e62: OUTPUT«␤»
09:12 zulon left
isBEKaml tuits,what do they mean? 09:14
tuits seems to be the currency here. :) 09:15
sorear puns.
moritz_ sometimes you just don't get aroundtoit
that's when you need a round tuit
isBEKaml punnderrrific!
09:19 M_o_C joined, M_o_C left, M_o_C joined 09:23 drbean left
patch hasn't been updated since the star 2010.07 download: rakudo.org/how-to-get-rakudo 09:23
09:26 mberends left
moritz_ updated, thanks 09:29
phenny: tell pmichaud when you write a release guide for *, be sure to mention updating rakudo.org/how-to-get-rakudo
phenny moritz_: I'll pass that on when pmichaud is around.
09:38 gfx left 09:42 meppl joined 09:43 risou_ joined 09:46 dalek left, dalek joined, ChanServ sets mode: +v dalek 09:50 risou_ left
dalek kudo: a3d4be0 | moritz++ | src/core/Range.pm:
delete trailing blanks in Range.pm
 09:57
09:59 risou_ joined
TiMBuS is .so still part of the perl 6 spec? 10:00
there's already .Bool and prefix:<?>
moritz_ rakudo: say 3.so
p6eval rakudo a20e62: OUTPUT«1␤»
TiMBuS or is .Bool not the proper way to do it 10:03
moritz_ would be fine with removing .s
.so
but since it's spec, I've implemented it 10:04
TiMBuS hmm
rakudo: say (my @a).so
p6eval rakudo a20e62: OUTPUT«0␤»
TiMBuS oh right, thats different to what i was thinking 10:05
rakudo: say (my @a).defined
p6eval rakudo a20e62: OUTPUT«1␤»
TiMBuS also, is there a reason behind .Bool and .Str being Ucfirst as opposed to .list and .hash? 10:07
10:10 risou_ left 10:12 risou_ joined
moritz_ I guess it's because .list and .hash refer to syntactic forms 10:13
@() -> .list, %() -> .hash
not sure if that's a good reason though
TiMBuS thats what I was thinking
10:13 risou_ left 10:16 hans__ joined
moritz_ phenny: tell ash_ that I made FORBID_PIR persist between REPL lines 10:22
phenny moritz_: I'll pass that on when ash_ is around.
10:24 hans__ left
moritz_ rakudo: my $1 10:24
p6eval rakudo a20e62: ( no output )
dalek kudo: a821f58 | moritz++ | src/Perl6/Actions.pm:
change FORBID_PIR from contextual to global; this makes it persist through multiple REPL lines
 10:27
10:30 bbkr left
sorear TiMBuS: in S13 or S14, .(type name) is how Perl 6 does explicit coercions 10:35
right now it's just a convention but we plan to tie it into my Int $x and similar... eventually... somehow...
Bool and Str are type names 10:36
TiMBuS as is Hash and List
moritz_ it's used by the sub f ($x as Int) { ... } syntax
sorear %() does not return a Hash, @() does not return a List
TiMBuS er.
moritz_ what does it return? Associative and Positional? 10:37
sorear in the former case, I think so
but that's not central to the purpose of either operator 10:38
e.g. Match ~~ Associative, but %($/) is emphatically not the same as $/ 10:39
moritz_ STD.pm6 line 2376: <[\@\%\&]> # is there a good reason for those backslasshes?
sorear (if it was, why would the %() special case exist)
moritz_ sorear: good point
TiMBuS well calling the method forms .hash and .list was probably incorrect then
sorear TiMBuS: no, they're contextualizers, look at S02 again
moritz_ how would you name them? 10:40
sorear moritz_: perl 5 character classes are double quotish interpolating
moritz_ but STD.pm6 is suposed to be Perl 6, no? 10:41
TiMBuS if .hash doesnt make a Hash, it shouldnt be .hash? im not sure what I would call it
sorear TiMBuS: .hash is the contextualizer for hash context. calling it anything else would be wrong
moritz_: I am not going to touch the viv character class translator 10:42
moritz_ so it's a hack. Ok
TiMBuS .hashify
hashualize
sorear hashualize context is questionable grammar 10:43
moritz_ gesundheit!
TiMBuS well fine we can add hasualise as well :p
moritz_ out
10:49 masonkramer joined 10:51 zulon joined, bbkr joined 10:57 masonkramer left 11:01 whiteknight joined 11:04 M_o_C left 11:10 [Coke] left 11:11 [Coke] joined 11:17 Italian_Plumber joined 11:26 [Coke] left 11:27 [Coke] joined 11:32 azert0x joined 11:40 shade\ left 11:41 shade\ joined 11:43 isBEKaml_ joined 11:46 zulon left 12:04 stepnem left 12:06 stepnem joined, drbean joined
colomon loliblogged! justrakudoit.wordpress.com/2010/08/...moization/ 12:08
12:17 shade\ left 12:25 snearch joined
gfldex colomon: i feel well informed now 12:29
12:29 shade\ joined 12:30 shade\ left, shade\ joined 12:31 shade\ left, sftp joined
dalek kudo: 4ca5220 | moritz++ | src/Perl6/Grammar.pm:
forbid my $1
 12:32
12:33 shade\ joined 12:34 amkrankruleuen joined
amkrankruleuen Hi. 12:35
12:36 risou_ joined 12:48 timbunce joined 12:50 jhuni left 12:51 Ross joined 13:09 orafu left, orafu joined 13:12 Ross left 13:13 [Coke] left 13:20 Ross joined 13:26 drbean left 13:35 timbunce left 13:47 zulon joined 13:49 amkrankruleuen left 13:51 janosik joined 13:53 janosik left 14:09 zby_home_ joined 14:17 wamba left 14:19 ruoso joined 14:20 M_o_C joined 14:25 Ross left 14:27 M_o_C left 14:28 M_o_C joined, M_o_C left 14:29 M_o_C joined 14:31 plobsing joined 14:39 M_o_C left 14:55 zulon left 14:58 risou_ left 15:01 isBEKaml_ left 15:03 [Coke] joined 15:04 justatheory joined 15:06 molaf joined, M_o_C joined 15:07 M_o_C left 15:11 M_o_C joined 15:14 lanny joined
lanny rakudo: my $a = [0..300]; $a[^300].join(' ').say 15:15
p6eval rakudo 4ca522: OUTPUT«src/string/api.c:3341: failed assertion 'pos <= res->strstart + Buffer_buflen(res) + 1'␤Backtrace - Obtained 13 stack frames (max trace depth is 32).␤/home/p6eval//p2/lib/libparrot.so.2.7.0 [0x2b2f00dec1a3]␤/home/p6eval//p2/lib/libparrot.so.2.7.0(Parrot_confess+0x87)
..[0x2b2f00dec2d…
15:18 grondilu joined
lanny That came in on the parrot upgrade in commit 03a9388. I've submitted a rakudobug but hoping it can get looked at so it doesn't get into R*. 15:18
Great work in perl6 and rakudo, all. /me goes back to lurking. 15:19
15:19 lanny left
moritz_ I'll try to bisect, and see which parrot revision caused it 15:21
I'm afraid it already went into the 2010.08 R* release 15:23
15:25 grondilu left 15:33 Mowah left
moritz_ star: my $a = [0..300]; $a[^300].join(' ').say 15:37
p6eval star 2010.07: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
..96 9…
moritz_ on rakudo 7b0031 I get some garbage out 15:38
scalar scalar scalar ... scalar named named named named named named named named named named repr_P6opaque named $!llsig named 105 106 107 ... 194 195 1`�_}$!llsig2`�_}!llsig 2`�_}subtype_realtype`�_}Mu 213 2`�_}named7 2`�_}self21 2`�_}repr_P6opaque7 2`�_}!llsig 2`�_}subtype_realtype`�_}�2 243 2`�_}$!llsig2`�_}named1 2`�_}P6opaque`�_}$!flat 2`�_}@!rest 2`�_}$OUT67 2`�_}!llsig 2`�_}subtype_realtype`�_}Mu 281 2`�_}Mu 285 2`
wtf? 15:39
15:39 Mowah joined 15:44 M_o_C left 15:45 M_o_C joined 15:46 M_o_C left, M_o_C joined, zulon joined 15:50 amiri joined 15:52 stepnem left, stepnem joined 16:04 patspam joined 16:10 amiri left 16:11 amiri joined 16:14 snearch left
isBEKaml moritz_: my local rakudo build at a20e62d shows the right output and doesn't spew garbage. 16:15
moritz_: masak++'s any-list was the last commit there. You might want to bisect it from there. 16:16
TimToady phenny: tell @tylercurtis depends on the compunit: eval returns a value, main doesn't
phenny TimToady: I'll pass that on when @tylercurtis is around.
16:17 justatheory left
moritz_ isBEKaml: it's a bit hard to bisect when I don't get the same output as the ticket suggests 16:18
isBEKaml: it means that something is broken locally
(probably on my machine)
which is not a good precondition for bisecting
isBEKaml moritz_: oh 16:19
moritz_: I was thinking since I could get the right output, this brokenness was something recent, but then yours suggests that it goes way back.
moritz_ I get the garbage locally even on newest rakudo, and no segfault 16:22
isBEKaml moritz_: what's the ticket id anyway? 16:23
moritz_ isBEKaml: I haven't actually looked in the ticket, just saw the report from lanny++ in the backlog 16:24
TimToady I got garbage like that some days ago, and mentioned it here 16:26
here: irclog.perlgeek.de/perl6/2010-08-24#i_2739992 16:29
I'm running 2010.08-8-gd1015f0, and still makes the garbage. will try a pull 16:31
16:37 tylercurtis joined 16:38 pmurias joined
TimToady It still malfs, and what's more, it's non-deterministic from run to run 16:39
rakudo: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say ~hailstone(27);
p6eval rakudo 4ca522: OUTPUT«src/string/api.c:3341: failed assertion 'pos <= res->strstart + Buffer_buflen(res) + 1'␤Backtrace - Obtained 24 stack frames (max trace depth is 32).␤/home/p6eval//p2/lib/libparrot.so.2.7.0 [0x2b72dd2081a3]␤/home/p6eval//p2/lib/libparrot.so.2.7.0(Parrot_confess+0x87)
..[0x2b72dd2082d…
TimToady rakudo: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say ~hailstone(27); 16:40
isBEKaml uh, yeah. that's what I got here too.
p6eval rakudo 4ca522: OUTPUT«src/string/api.c:3341: failed assertion 'pos <= res->strstart + Buffer_buflen(res) + 1'␤Backtrace - Obtained 24 stack frames (max trace depth is 32).␤/home/p6eval//p2/lib/libparrot.so.2.7.0 [0x2b445e9901a3]␤/home/p6eval//p2/lib/libparrot.so.2.7.0(Parrot_confess+0x87)
..[0x2b445e9902d…
TimToady that's not what I get at home
is the server 64-bit?
more to the point, is parrot/rakudo compiled 64 bit? 16:41
isBEKaml star: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say ~hailstone(27); 16:43
p6eval star 2010.07: OUTPUT«27 82 41 124 62 31 94 47 142 71 214 107 322 161 484 242 121 364 182 91 274 137 412 206 103 310 155 466 233 700 350 175 526 263 790 395 1186 593 1780 890 445 1336 668 334 167 502 251 754 377 1132 566 283 850 425 1276 638 319 958 479 1438 719 2158 1079 3238 1619 4858 2429 7288 16:44
..3644 1…
moritz_ server is 64bit
and parrot too
isBEKaml it's 32 bit here and I get those stack frames.
TimToady I'm comiled 32 16:47
*p
though I'm on a 64-bit arch 16:48
isBEKaml rakudo: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say hailstone(27); 16:50
p6eval rakudo 4ca522: 16:51
..OUTPUT«278241124623194471427121410732216148424212136418291274137412206103310155466233700350175526263790395118659317808904451336668334167502251754377113256628385042512766383199584791438719215810793238161948582429728836441822911273413674102205161543077923246162308115457717328664331300650325…
TimToady works on the server
but maybe it's "Hey, I'm 64-bits, I don't need no stinkin' GC."
16:51 stepnem left
TimToady or maybe it's just the indeterminacy being differently indeterminate... 16:53
16:53 stepnem joined
TimToady rakudo: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say ~hailstone(77031); 16:57
p6eval rakudo 4ca522: OUTPUT«(timeout)»
TimToady rakudo: sub hailstone($n) { $n, -> $n { $n %% 2 ?? $n / 2 !! $n * 3 + 1 } ... 1; }; say ~hailstone(77031);
p6eval rakudo 4ca522: OUTPUT«(timeout)» 16:58
TimToady that number makes a very pretty pattern of garbage at home...
(it's the longest sequence under n == 100000) 16:59
17:03 M_o_C left
TimToady are the words in the garbage any kind of clue? It seems to generate fruns from the set: $!llsig named %.15g repr_P6opaque named self scalar 17:04
17:04 envi^home left
TimToady BTW, someone should go through all the rosettacodes and make sure all appropriate entries are added to the test suite 17:07
that would be a good task for someone who wants to contribute but doesn't know how
moritz_ TimToady: the names seem to be related to rakudo dynops or their usage 17:08
TimToady the rosettas tend to put stress on the design of everything
tylercurtis sorear: looking for help on Niecza? 17:09
17:10 grondilu joined
TimToady oh bleh, twitter brain damage, I sent the phenny message to @tylercurtis 17:11
I am officially twitterpated.
grondilu Hi, I like the idea of a Grammar object. Has anyone ever written for the Mediawiki syntax ?
(written one)
TimToady haven't heard of one, but this is the right place to ask 17:12
tylercurtis TimToady: I saw it in the backlog. I hadn't thought about evals. Thanks.
moritz_ grondilu: I don't know one either; but would be nice to have one :-)
grondilu I could try it, but I'm not an expert and I don't want to do it if someone already did.
TimToady not even for the learning experience? :) 17:13
grondilu yes, sure
moritz_ nobody starts out as an expert :-)
if you want some advice, start with *very* simple grammar, improve it step by step, and test it after each step
TimToady and start by only writing tokens, not rules 17:14
grondilu ok
TimToady people get into trouble with implicit whitespace matching
moritz_ since the debugging facilities for grammars are somewhat limited, that's the only way not to get stuck with frustration
grondilu so far I have the perl6 book : github.com/downloads/perl6/book/2010.08.a4.pdf. Will this doc be enough for me to start ? 17:15
moritz_ yes; if you know some Perl 5, perlgeek.de/en/article/5-to-6 might be good additional information 17:16
TimToady knowledge of how to grep the synopses is also quite helpful
grondilu "synopses" ? 17:17
TimToady the specs
moritz_ and don't hesitate to come back here and ask questions
TimToady perlcabal.org/syn/
Juerd grondilu: It's probably impossible to do everything mediawiki does with a clean grammar, given the way it's parsed by the php code. 17:19
TimToady that's what dirty grammars are for :) 17:20
we have those too...
.oO(STD)
grondilu well if I could do 90%, that will be fine.
17:20 zulon left
grondilu I'll just verbatim the rest. 17:20
Juerd grondilu: In any case, I'd love to see a mediawiki grammar emerge 17:21
And then wiki software. And then... :)
TimToady ...profit
moritz_ I don't know much about mediawiki syntax; but for some formats it's also a good idea to split up on empty lines first, and then parse the chunks separately
Juerd moritz_: Not so for MW, IIRC 17:22
grondilu that wouldn't work for ''' and '' (bold and italic format) for instance, I'm afraid. Since several paragraphs can be enclosed in a '' or ''' pair. 17:23
Juerd Paragraphs are separated by empty lines, but not everything is a paragraph
TimToady grammars can certainly heed/ignore line boundaries at need 17:24
grondilu so I guess we can't use empy lines as separator for anything else than paragraphs. 17:25
moritz_ Juerd, grondilu: that's why I warned that I don't know mediawiki syntax enough :-)
17:26 molaf left
TimToady Perl 6 grammars are not restricted to context-free. 17:27
grondilu A good mediawiki example would be the Wikipedia:SandBox page I guess. I'll make my experiments on that. en.wikipedia.org/wiki/Wikipedia:SandBox
17:31 colomon left 17:32 grondilu left
TimToady rakudo: my $a = [<0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 17:37
p6eval rakudo 4ca522: OUTPUT«===SORRY!===␤Confused at line 22, near "my $a = [<"␤»
TimToady truncated, I imagine
but that doesn't mess up here, only use of .. or ...
17:37 plobsing left
TimToady btw the end of that was: 296 297 298 299 300>]; $a[@($a)].join(' ').say 17:39
so not using range in the subscript either
moritz_ rakudo: say True xor True 17:41
p6eval rakudo 4ca522: OUTPUT«1␤»
17:41 pochi left
moritz_ rakudo: say True xor False 17:41
p6eval rakudo 4ca522: OUTPUT«1␤»
moritz_ huh?
rakudo: say True ?^ False
p6eval rakudo 4ca522: OUTPUT«1␤»
moritz_ rakudo: say True !^ False
p6eval rakudo 4ca522: OUTPUT«0␤»
TimToady rakudo: say (True xor True) 17:42
p6eval rakudo 4ca522: OUTPUT«␤»
TimToady marriage rotting your brain already? :P
moritz_ seemslike
17:43 zulon joined
moritz_ (though $wife says "cuddling should prevent it") 17:43
TimToady lovely word, "should" 17:44
moritz_ hugme: hug me 17:45
hugme hugs moritz_
TimToady makes no commitment to whether it actually does or not...
hugme: hug yourself
hugme hugs yourself
TimToady
.oO(Babel 17)
 17:46
moritz_ speaking of brain rot... my installed rakudo segfaults a simple test script that loads a grammar from a .pm file 17:47
17:48 pmurias left
TimToady hides behind a lollipop 17:48
17:48 mmmpork left 17:57 Mowah left 17:59 justatheory joined 18:01 justatheory left
moritz_ huh 18:03
"use Test;" segfaults. 18:04
(on latest parrot)
TimToady that seems...suboptimal... 18:05
18:18 IllvilJa left
moritz_ indeed 18:19
18:23 [Coke] left
moritz_ works on HEAD~2 18:23
18:27 c9s joined 18:38 pochi joined 18:40 zulon left 19:01 tadzik joined
tadzik oh hai 19:01
19:10 [Coke] joined
TimToady rosettacode.org/wiki/Combinations#Perl_6 very much resembles the Haskell solution 19:18
19:28 IllvilJa joined 19:29 colomon joined 19:33 wamba joined
diakopter . 19:50
19:56 Packetknife joined, Packetknife left, Packetknife joined
TimToady hmm, seems like one could solve spiral and zigzag arrays with a grammar 19:57
diakopter "solve"? 19:58
TimToady like turtle graphics with backtracking
rosettacode.org/wiki/Spiral_matrix 19:59
rosettacode.org/wiki/Zig-zag_matrix
20:00 Packetknife left
TimToady course, that probably means it can be done with signature matching as well 20:01
diakopter the Python functional solution could be golfed, I think
dalek kudo: d9aa575 | moritz++ | src/Perl6/Grammar.pm:
reset $*IN_DECL for regex bodies
TimToady yes, well, there are various approaches to either problem 20:02
but a turtle's point of view is rather close to a human's
diakopter thinks about a grammar way 20:04
TimToady so such a solution could be very readable, I suspect, assuming a good set of turtle primitives for moving around a matrix
diakopter o wait 20:05
diakopter thinks about a flattening way
TimToady basically, a turtle can keep egocentric directions, not just geographic 20:06
perigrin Part of the Me Generation.
TimToady and you can then turn the entire coordinate system merely by turning the turgle
fsdo turgle
perigrin rummages around and finds a bag of 't's for TimToady so he won't have to replace them with g's anymore. 20:07
diakopter turtles get squished all the time
perigrin er 'g's
diakopter: that's because they stack so well.
diakopter :) esp the ones toward the bottom, yeah
TimToady see also "Whorf Revisited", www.nytimes.com/2010/08/29/magazine...nted=print 20:08
perigrin yeah you can't fool me there ... there is no bottom ... it's turtles all the way down.
well and a few turgles for when you need another cuppa
20:08 patrickas joined
patrickas o/ 20:09
TimToady don't think another cuppa will help cure my neuronal woes when it comes to turgles
perigrin TimToady: out of *that much* T(ea)?
ouch.
diakopter tea is easier to gurgle than coffee 20:10
20:10 Ross joined
TimToady not really, coffee has a velar, and tea only has a dental 20:10
patrickas TimToady: Got a minute for a series spec question ?
TimToady well, alveolar 20:11
diakopter hah
TimToady patrickas: well, yeah, though I'm overdue for lunch...
if you don't mind munching sounds
patrickas What should this produce ? (10,11,12,3,4,5,6,7...4) : Nil because rhs on the wrong side or (10,11,12,3,4) because lhs should get "processed" anyway ?
diakopter thinks of "coffee" from Smash Bros
TimToady I can argue that one both way, alas 20:12
patrickas but I can't implement it both ways :-(
TimToady the question is how far back the filtration of the end value goes 20:13
patrickas the whole return Nil for rhs o nthe wrong side feels weird to me 20:14
but I can see where it would make some sense
but there are so many weird edge cases ... I had to give up on trying to reason it out on my own :-) 20:15
TimToady I would prefer (10,11,12,3,4) myself
20:15 stepnem left, stepnem joined
patrickas Well ... your preferences are my orders :-) 20:16
TimToady the retroactive bit only works on intuited functions, anyway, I think, so it shouldn't go back further than the last 3 values, since those may be "tacked on" to a different sequence
patrickas ok so I use the last three values to determin if on the wrong side 20:18
but I also make sure that the rhs is not between any two items of the lhs
does that seem reasonable enough ? 20:19
diakopter identifies the odd vs even distinction in the spiral matrix
patrickas or do you mean i drop the whole check when the lhs has more than three items? 20:20
20:20 timbunce joined
TimToady "turn ($x++ %% 0 ?? left !! right)" 20:20
%% 2 rather
no, just apply to the last three values 20:21
patrickas so in that case we get Nil because looking at the last three values ( 5,6,7 ) 4 is on the wrong side 20:22
TimToady the question is whether we've already returned anything in front of that irretrievably
^10, 5,6,7 ... 4 20:23
patrickas yes I think that case should produce 0,1,2,3,4 20:24
and the previous one 10,11,12,3,4
but this means we are checking all items of the lhs to decide if 4 is on the wrong side .... otherwise result of the whole series op is just Nil 20:26
TimToady why should that chop all the way back to the 4? what if they really want 0..9 followed by 5,6,7...4
patrickas precedence ? 20:27
i my head ... has the lowest precedence
TimToady well, higher than listops and such, but yes, pretty low
I suppose they can always parenthesize
which is the argument on the other side 20:28
patrickas so you want (^10) , ( 5,6,7 ... 4) but you are asking for ( ^10 , 5,6,7 ) ... 4
the major reason that is bothering me for not stipping at the first 4 20:29
is things like the prime number generator series
patrickas trying to get it right 20:30
diakopter TimToady: well, it's n is odd/even, not x or y on the grid 20:32
20:32 M_o_C joined
diakopter for me 20:32
TimToady yes, that's why I incremented it myself 20:33
diakopter hum 20:34
TimToady re they NYT article, I speak an egocentric language, but my brain is nonetheless tracking directionality continually like a geographic language speaker 20:36
it would not take me long to get used to speaking in a geographic language
"turn north, the look east"
"there's a spider crawling on your west arm" 20:37
patrickas I was wrong about the prime number thingy ... it does not have more than three items on the lhs
20:39 Mowah joined
patrickas I am inclined to check all of the lhs for on-the-wrong-side-ness but I have no strong feelings about it wither ways 20:39
20:39 Ross left
TimToady well, how do you know which side is wrong for 1, *+2 ... $x where $x == 0 20:39
or $x == 2 20:40
patrickas we do not do the check for "code" generated series
TimToady have to
@array[0, *+2 ... $max]
patrickas only intuited ones
TimToady mmm 20:41
patrickas I dn't know I never thought about it ... I thought the spec was clear on this part
TimToady it seems hard to argue that 0,2 ... $max is different from 0, *+2 ... $max
esp when used as a subscript 20:42
patrickas I thought the argument was whenever you have Code all bests are off ... since you could have something generating numbers .. somethign depending on io ....
s/bests/bets/ 20:43
s/something generating numbers/something generating random numbers/
moritz_ TimToady++ # sharing that nyt link
TimToady you mean like, say, hailstones. :)
patrickas moritz_ \o/ congratsssssssss 20:44
TimToady which, however, still would work with the inequality test
but point taken
moritz_ thanks patrickas 20:45
TimToady but yeah, all my memories are tagged with the direction I was facing 20:46
patrickas I wanted to say MABROUUUKKKKK but I thought it would sound to weird in here!
TimToady if I have any visual memory, that is
patrickas It just feels like I'd need to solve the halting problem first ... then I could check if the rhs in on the wrong side of the Code part... not sure if that's really the case though
20:46 shade\ left
TimToady we can certainly limit the inequality to intuited cases for now, and see how it goes 20:48
patrickas ok.. 20:49
20:49 Ross joined
patrickas and for the original question? intuited series with more than 3 items on the lhs ? 20:50
20:50 shade\ joined
TimToady the last three have to be literal, but the ones before that don't, so I'm still inclined to only treat the last three as part of the controlled list 20:52
1 .. 10, 9, 8 ... 1
20:53 ruoso left
moritz_ rakudo: my $a = '.*'; say 'abc' ~~ / $a / 20:53
TimToady clearly wants to be mean 1,2,3,4,5,6,7,8,9,10,9,8,7,6,5,4,3,2,1 somehow
p6eval rakudo 4ca522: OUTPUT«␤»
moritz_ rakudo: my $a = '.*'; say '.*' ~~ / $a /
p6eval rakudo 4ca522: OUTPUT«.*␤»
patrickas parentisize!
TimToady I suspect we might want to require the series intuiter to operator at compile time, not run time 20:54
patrickas besides
1 .. 10, 9, 8 ... 1 will mean that in all cases
TimToady so the above just gets compiled to 1 .. 10, 9, 8, *-1 ... 1
patrickas At least regarding the "is on the wrong side issue" 20:55
TimToady okay, ... 5
patrickas because in that case the rhs is clearly not on the wrong side
TimToady 1 .. 10, 9, 8, *-1 ... 5
patrickas I see your point ... but that's another issue. 20:56
TimToady well, has to handle ... 5.5 too
moritz_ is scared by the complextiy of the series operator
patrickas no I mean
sorear good * #perl6
patrickas my question was about the "is on the wrong side" check
20:57 nimiezko joined
patrickas none of your last few examples are ambiguous 20:57
the answer is NO in all cases
TimToady I'd still argue it belongs only on the part of the series that is intuited
20:57 masak joined
moritz_ good morning sorear 20:57
masak ahoy, #perl6!
20:57 Ross left
TimToady it applies to the ... 5case above 20:58
20:58 Ross joined
TimToady whether you get the 1,2,3 out 20:58
moritz_ has an interesting parsing question, but hesitates to hijack the current discussion 20:59
TimToady directionality is determined from the final values 10 9 8, not the 1
patrickas Yes ... this is another issue that needs to be discussed too
TimToady also, part of the reason it probably must be syntactic is that the 10 in 1..10, 9, 8 wants to particpate differently in the range and in the intuition 21:00
patrickas yes but the dilemma in your examples is should we return 1,2,3,4,5 OR 1,2,3,4,5,6,7,8,9,10,9,8,7,6,5
TimToady the latter
patrickas In noen of the cases should Nil be returned 21:01
TimToady directionality by the last items, not the first
patrickas that's what I means it is anothe rissue
TimToady I think I said that already
patrickas thee isn't even a question if we should return Nil or something else .... we definetely want to return something else in all of your last examples, but there is an argument about that something else. 21:02
TimToady Nil is just a limiting case of what I've already said 21:03
patrickas Oh ok
TimToady but I'd really prefer that intuited inequality not veto anything before the last 3 elements 21:04
where that can be examined in the parser
@values, 10, 9, 8 ... 5 has guaranteed semantics regardless of what's in @values 21:05
patrickas ok but that means a whole lots of tests need to be changed
TimToady it always makes @values, followed by 10,9,8 ... 5
patrickas ok if no one else has a problem with that, I'll about implementing it that way. 21:07
21:07 ruoso joined
patrickas to me it just feels weird er that 1,2,3,4,5,6 ... $x 21:08
would behave inone way when x is 4,5 or 6 and a different way when $x is 1,2 or 3
TimToady lemme think about it smore
patrickas OK 21:09
moritz_ sorry for now giving you a chance to speak! 21:10
s/now/not/ :-(
moritz_ np, it's a bit wordy anyway
I'm writing a mail to p6u right now
masak yay! someone using p6u :) 21:11
21:11 tylercurtis left
masak moritz_++ 21:11
I'm a Perl 6 user, and I hardly ever use it. I tend to use a certain channel on freenode.
has anyone else gotten this error message lately? 'postcircumfix:<Nil()> not defined for type Array()' 21:12
I find it weird and disconcerting.
TimToady maybe that's because it *is* weird and disconcerting
moritz_ rakudo: say Nil ~~ Parcel
p6eval rakudo d9aa57: OUTPUT«1␤»
moritz_ rakudo: say Nil ~~ Positional 21:13
p6eval rakudo d9aa57: OUTPUT«1␤»
tadzik github.com/masak/web/blob/master/li...atch.pm#L7 -- where are awesome things like this one documented?
masak I'll try to find the cause of it. it occurred in Yapsi when I was youing with making closures first-class.
tadzik: hardly anywhere at all, sadly :(
tadzik: if you go through my blog posts about Web.pm, you'll find a lot of clues. but maybe not about that one, that's Tene++'s code. 21:14
tadzik then it needs a blag post :) 21:15
masak I couldn't agree more. that module is pretty awesome.
maybe it'd even run, too. under alpha, at least.
moritz_ I think it did 21:16
tadzik duh. I feel like fixing it after exams just to witness the awesomeness
masak moritz_: oh, it definitely did, once.
tadzik: that would be, for lack of a better word, awesome. 21:17
tadzik masak: I need to pass my *$*&$! linear algebra first
Tene tadzik: LolDispatch was just a short proof-of-concept.
masak ooh! linear algebra! why didn't you *say* so? :D
tadzik masak: hum?
masak likes it
tadzik oh, you can help me then :) 21:18
masak I certainly can.
Tene tadzik: The traits aren't being done right and such. It was just intended to explore how traits were implemented in rakudo at the time.
tadzik I didn't have and problems today, but I will certainly have in the next few days. Got xmpp masak?
Tene: I see
masak tadzik: I have GChat.
21:18 Ross left
masak er, GTalk. 21:18
tadzik masak: you mean this thing from Gmail? That's xmpp 21:19
masak that's why I brought it up :)
tadzik mind sharing?
masak sounds like something a Vulcan would do... :P
Tene tadzik: I don't recall why LolDispatch got committed. It wasn't intended for wide public use.
If it does end up useful or inspiring work on something else, that's great. 21:21
21:21 wamba left 21:22 Ross joined
Tene anyway, it should work fine if you fix the syntax for the trait definitions, and maybe other minor changes. iirc, I was relying on intenrals that have changed significantly by now. 21:23
masak but the intent of the module should still very much work. 21:24
once all those changes are made, I mean.
Tene That's right.
I did something like that with my POE-alike toy
github.com/tene/Doten/blob/master/dt.pl 21:25
still not done the right way there either, though 21:26
I also experimented with "is memoized" or "is cached" or something, but I couldn't manage to replace or wrap the sub from the trait handler. 21:27
moritz_ sent is mail, and now goes to bed 21:28
masak can haz it o/ 21:29
interesting question. 21:31
gfldex I wonder what a guugu yimithirr would do aboard the ISS. North is flipping every 45 minutes.
nice article :)
szbalint Hm, so "run" returns the return code of a command that was executed. How do I get back the output the executed command wrote to STDOUT? 21:32
TimToady no, north stays the same
give or take a bit of inclination 21:33
but I can testify that travel in a vehicle is a problem for geographic brains
usually I just peg one end of the vehicle as notional north, and leave it there until I have to forcibly reorient on landing 21:34
reorienting can be quite difficult when you're brain is hardwired to track the direction you think you're facing 21:35
it's a bit like asking a synasthete to see different colors around letters and numbers than they're used to...
I don't have much control of my brain circuitry on that level 21:36
diakopter get a rootkit; they're free :P
TimToady often I think I've reoriented, and my brain suddenly snaps back to the old orientation 21:37
gfldex My head-north of my hometown is off by about 15' . Proly because I got my head before I got a map. :) 21:38
TimToady and the sun angles really screwed me up in south america, because my brain thinks it knows where the ecliptic is 21:39
I have to be very careful the first time I visit somewhere
I spent the entire second half of my childhood with the sun rising in the north and setting in the south 21:40
because I have Bremerton WA 90° off in my head
diakopter TimToady: a spiral matrix one-liner will be difficult. 21:41
TimToady my parents old house was 180° off on the inside, and there was nothing to be done about it
not hard if you have an appropriate turtle 21:42
21:42 Ross left
TimToady it's just lay an egg, move forward || right 21:42
assuming forward fails outside the matrix or where there's already an egg 21:43
masak my room is 180° off on the inside. it's the half-twist of the stairs to the second floor that never registered somehow. 21:45
diakopter egg already right ?? move forward !! turn right (don't need to worry about the edges, just go until eggs run out)
er, turn left....
heh
tadzik rakudo: say 'alive'; #= testing 21:49
p6eval rakudo d9aa57: OUTPUT«===SORRY!===␤Whitespace character is not allowed as a delimiter at line 22, near " testing"␤»
tadzik what kind of thing is this?
masak rakudo: say 'alive'; #=[ testing ] 21:50
p6eval rakudo d9aa57: OUTPUT«alive␤»
masak tadzik: I don't like it, but it's spec.
TimToady std: say 'alive'; #= testing
masak tadzik: my plan is to change the spec.
p6eval std 32108: OUTPUT«ok 00:01 115m␤»
masak hm. or not spec.
:)
tadzik duh 21:51
TimToady looks like LTM failure to me
tadzik . o O ( masak submits a rakudobug )
masak I think it's just failure, period.
masak submits rakudobug
diakopter ouch
masak tadzik:
tadzik++
tadzik Druid++
. o O ( class masak does Druid)
szbalint Is it possible to run an external program from Rakudo and capture it's output, without resorting to temporary files? 21:55
tadzik qx//
or qxx//, if you want interpolation
szbalint ty :) 21:56
masak qqx// 21:57
(to parallel qq[])
tadzik or right
I miss backticks though :)
masak tadzik: backticks now serve a higher purpose... 21:58
...nothing. :)
tadzik (: 21:59
oh, not really :)
rakudo: say 'alive'; #`[ fasd ] say 'and well';
p6eval rakudo d9aa57: OUTPUT«alive␤and well␤»
tadzik well, that's a backtick
masak bleh.
tadzik I liked this #=[] notatio better though 22:00
masak I think the design decision behind that one was "intentionally uncommon".
tadzik rakudo: #=[ fasd ] say 'test';
p6eval rakudo d9aa57: OUTPUT«test␤»
masak double bleh.
tadzik hmm
that's actually fun
reads like: Comment equals fasd
masak for a very long time, the response to delimited comments was: "there's line-ending comments, and there's Pod. deal with it." 22:01
then @Larry succumbed, and the result was the ugly syntax we have today.
tadzik is it that ugly really? 22:02
masak oh, believe me.
tadzik There is plenty of #=[] in Druid, and it looks not that bad for me
masak :(
tadzik hugme: hug masak 22:03
hugme hugs masak
masak feels better now
Tene What's wrong with using =begin comment/=end comment ?
tadzik verbosity
imho
hmmmm 22:04
rakudo: say 'alive'; "test"; say 'and well';
p6eval rakudo d9aa57: OUTPUT«alive␤and well␤»
tadzik oh funny
rakudo: say 'alive'; """I can haz pythonz?"""; say 'and well'; 22:05
p6eval rakudo d9aa57: OUTPUT«===SORRY!===␤Confused at line 22, near "\"\"\"I can h"␤»
masak we've had delimited comments all along, and we didn't know it. they're called "literal strings" :P
tadzik I wonder if they're even ignored by Rakudo on this stage
nope :) 22:06
masak tadzik++ # rosettacode.org/wiki/Combinations#Perl_6 22:09
tadzik ...me? 22:10
TimToady++ # passing along
masak sorry, that was today's tab fail for me. :)
tadzik well, I don't mind a nice, friendly karma :) 22:11
22:12 Mowah left
szbalint finally! It's 00:19, but I've completed my first real Perl 6 thing: paste.scsys.co.uk/49497 22:21
*00:21
22:21 timbunce left
tadzik congratulations :) 22:22
and goodnight guys o/
22:22 tadzik left 22:24 zby_home_ left, zby_home_ joined, justatheory joined 22:27 timbunce joined 22:28 nimiezko left 22:31 timbunce left 22:33 masonkramer joined 22:36 M_o_C left 22:41 masonkramer left
masak "Perl 6 -- there's some madness to the method after all" 22:47
would that work as the title for a talk?
22:48 Guest23195 left 22:57 masonkramer joined, masonkramer left 23:00 risou_ joined 23:02 jhuni joined
szbalint masak: yes! 23:03
masak I might use that for Paris, then. 23:04
23:16 macroron joined
lue ohai o/ 23:21
23:22 zby_home_ left
masak lue: はい！ 23:25
lue
.oO(quick! Where is my transliterator?)
 23:26
masak just said 'hai' :) 23:28
lue I could guess that. [Although I'm not familiar with my kana yet] 23:29
masak you need pulactice! 23:30
:) 23:31
sorear purakuteisu? 23:32
masak right; that. 23:33
lue
.oO(I have yet to find a decent ogg recording that I can play of the japanese 'u' and 'r')
 23:34
23:34 patrickas left
masak rolling 'r', 'u' like in most of the world. 23:34
sorear lue: there's no substitute for a flesh teacher if you want to speak correctly 23:35
lue Of course, I wouldn't expect to learn perfect pronunciation any other way. 23:36
masak it's pronounced as it sounds. 23:37
Japanese is easy that way. :)
lue I need to know how it sounds though! :) .oO(If I just want to hear it, I should find the next time a Japanese game show comes on air) 23:38
23:40 masak left, shade\ left 23:43 shade\ joined, shade\ left 23:44 awwaiid joined
sorear lue: japonic u, when it's pronounced at all, sounds exactly like english short u in e.g. rune 23:47
the trick is that Japanese orthography does not admit terminal consonants or paired consonants 23:48
23:49 shade\ joined
TimToady but has more variation than english, since in a five vowel system it isn't competing with, say "oo" 23:49
as in "took"
sometimes get a bit ü ish too
sorear school would become sukūl; the first u is inserted just to avoid /sk/, which is unwritable 23:50
TimToady since there's still a culture left over from samurai days of not moving your lips much while talking
sorear erm, sukūru 23:51
TimToady from a standpoint of syllable length, it's not so much dropped there as turns into a double long ss
sskul
ss ku and ru are all basically the same length
japanese is very syllable timed, like spanish but unlike english 23:52
which is stress timed
sorear as for r, the japanese only have one consonant between l and r 23:53
TimToady and we have four, m, n, p, and q. :D
sorear this has the very useful consequence that the ear of a native Japanese /cannot hear/ the difference 23:54
if you say lu, or ru, or anything in between, they will hear HIRIGANA SYLLABLE LU
so... don't worry about it
(it cuts both ways. the most painful part of Japanese I for me was learning to distinguish o from ō) 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!