Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2005-10-25

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:04 wilx has quit IRC (Read error: 104 (Connection reset by peer))
00:04 wilx` has joined #perl6
00:13 wilx` has quit IRC (Read error: 104 (Connection reset by peer))
00:13 wilx has joined #perl6
00:25 wilx` has joined #perl6
00:31 nz8EhV6 has joined #perl6
00:34 eric256_ has joined #perl6
00:37 wilx has quit IRC (Read error: 110 (Connection timed out))
00:43 wilx has joined #perl6
00:45 oylenshpeegul has joined #perl6
00:48 Blicero has quit IRC (Remote closed the connection)
00:52 Cryptic_K has joined #perl6
00:57 oylenshpeegul has left
00:57 crenz has quit IRC ("Computer goes to sleep!")
00:57 oylenshpeegul has joined #perl6
00:59 wilx` has quit IRC (Read error: 110 (Connection timed out))
01:00 nz8EhV6 has quit IRC (Client Quit)
01:05 sparrowsteeth has joined #perl6
01:09 wilx has quit IRC (Read error: 104 (Connection reset by peer))
01:09 wilx has joined #perl6
01:10 sparrowsteeth has quit IRC ("Chatzilla 0.9.68.5.1 [Firefox 1.0.6/20050717]")
01:14 wilx` has joined #perl6
01:14 oylenshpeegul has left
01:17 sili_ has joined #perl6
01:18 joepurl has joined #perl6
01:23 wilx` has quit IRC (Read error: 104 (Connection reset by peer))
01:24 wilx` has joined #perl6
01:29 wilx has quit IRC (Read error: 110 (Connection timed out))
01:32 * eric256_ whistles
01:32 eric256_ is now known as eric256__
01:36 runinshine has joined #perl6
01:37 runinshine has quit IRC (Client Quit)
01:38 wilx has joined #perl6
01:53 wilx` has quit IRC (Connection timed out)
02:05 eric256__ is now known as eric256
02:10 eric256 wow...lively in here. ;)
02:13 eric256 anyone know a way to make a loop output colums in HTML::Template? like three even colums of all the items in the loop?
02:29 wilx has quit IRC (Read error: 104 (Connection reset by peer))
02:29 wilx has joined #perl6
02:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
02:30 lisppaste3 has joined #perl6
02:44 wilx has quit IRC (Read error: 104 (Connection reset by peer))
02:44 wilx` has joined #perl6
03:02 rafl_ has quit IRC (Read error: 60 (Operation timed out))
03:04 eric256 http://feather.perl6.nl/~eric256/t_index/  for anyone who is interested... added templating and sub pages for each directory of tests. ;)  if only i knew someone with the HTML skills to make it pretty....hint hint. ;)
03:20 rafl has joined #perl6
03:47 buu ?eval my $a; say $a++ for 0..100;
03:47 evalbot_7731 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 undef
03:47 buu ?eval my $a="a"; say $a++ for 0..100;
03:47 evalbot_7731 a b c d e f g h i j k l m n o p q r s t u v w x y z aa ab ac ad ae af ag ah ai aj ak al am an ao ap aq ar as at au av aw ax ay az ba bb bc bd be bf bg bh bi bj bk bl bm bn bo bp bq br bs bt bu bv bw bx by bz ca cb cc cd ce cf cg ch ci cj ck cl cm cn co cp cq cr cs ct cu cv cw undef
03:48 buu Wtf?
03:48 buu ?eval my $a="cw"; say $a-- for 0..100;
03:48 evalbot_7731 cw -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 undef
03:48 buu I demand a magical decrement.
03:48 eric256 you probably wont get it. ;)
03:48 eric256 demand though you may
03:48 buu Why not?!
03:48 eric256 i don't remember the specific reason...maybe you will
03:49 buu Someone lied to me and said that magical increment could get to 0
03:49 eric256 it would actualy be pretty easy to add in the haskell code if you want
03:49 buu When it obviously can't.
03:49 eric256 ?eval my $x = '-10'; say($x++) for 0..20;
03:49 evalbot_7731 -10 -11 -12 -13 -14 -15 -16 -17 -18 -19 -20 -21 -22 -23 -24 -25 -26 -27 -28 -29 -30 undef
03:49 eric256 lol. thats not very good. lol
03:49 buu What does that have to do with anything?
03:49 buu Ok granted it's weird
03:50 buu But it's completely different
03:50 eric256 no if it had gone up like it should. then it would have gotten to 0
03:50 eric256 well "0"
03:50 buu Yes, but that's not a magical increment
03:50 eric256 ?eval my $x = '-'; say($x++) for 0..20;
03:50 evalbot_7731 - . / 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 undef
03:50 eric256 yes it was
03:50 buu ick
03:50 buu No it's not
03:50 buu incrementing numbers is standard ++ behaviour!
03:51 eric256 yes it is. i promise. if its a string at all then its magic
03:51 buu Since when has perl cared if your number is surrounded by quotes?
03:51 eric256 and -10 ++ usualy equals 11?
03:51 eric256 always
03:51 eric256 its a string, ++ on a string doesn't force it to be a number
03:51 buu -10 ++ should be 9
03:51 buu er
03:51 buu -9
03:51 buu Like god intended.
03:52 eric256 ?eval my $x = '-10'; for 0..10 { $x += 1; say $x; }
03:52 evalbot_7731 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 undef
03:52 buu and by god "b"-- should be a
03:52 eric256 non magic
03:52 eric256 ?eval my $x = '-10'; for 0..10 { $x++; say $x; }
03:52 evalbot_7731 -11 -12 -13 -14 -15 -16 -17 -18 -19 -20 -21 undef
03:52 eric256 magic
03:52 eric256 ?eval my $x = '-10 hello'; for 0..10 { $x++; say $x; }
03:52 evalbot_7731 -10 hellp -10 hellq -10 hellr -10 hells -10 hellt -10 hellu -10 hellv -10 hellw -10 hellx -10 helly -10 hellz undef
03:52 buu That's awful.
03:52 eric256 lol. not realy.
03:52 eric256 same things perl5 did
03:53 buu I don't care what perl5 did!
03:53 eric256 hehe
03:53 eric256 either way you can't say you want ++ to be magic, unless the thing inside the string might be a number
03:53 buu I don't see why it's so difficult to define a magic -- given the above constraints
03:53 eric256 its not difficult..where did you read that?
03:53 buu Here
03:53 eric256 i even just mentioned it would be quite easy to add to pugs
03:54 buu So what's the problem?
03:54 eric256 agreement, standard, like i said i can't remember the last reason why not, but i think its mostly just because no one realy needs it
03:54 Blicero has joined #perl6
03:56 eric256 i would say, if you want it, implement it in pugs.. pretty straight forward to do
03:58 wolverian buu, what's "a"--?
03:58 eric256 ?eval 'z'++
03:58 evalbot_7731 Error: Can't modify constant item: VStr "z"
03:58 eric256 ?eval my $x ='z'; $x++
03:58 evalbot_7731 "z"
03:59 eric256 ?eval my $x ='z'; $x++; $x
03:59 evalbot_7731 \"aa"
03:59 eric256 that was the reason. thanks wolverian
03:59 PerlJam Why is that reason good enough?
03:59 PerlJam Why not just do like perl5 does when you try to increment ";"  (for instance)
04:00 eric256 didn't say it was ;) just thats the last reason i heard. hehe. what would it be? ''?
04:00 Amnesiac has quit IRC ("Leaving")
04:00 PerlJam eric256: sure, why not.
04:00 PerlJam Same thing for $a = "0000000";   $a--
04:01 eric256 PerlJam: i think you would get many different answers.  besides i never said we shouldn't. lol people keep trying to put me on that side. i keep saying implement it in pugs. ;)
04:01 eric256 ?eval my $x =''; $x++; $x
04:01 evalbot_7731 \"1"
04:01 eric256 ?eval my $x ='1'; $x++; $x
04:01 evalbot_7731 \"2"
04:01 eric256 ?eval my $x ='0'; $x++; $x
04:01 evalbot_7731 \"1"
04:02 PerlJam eric256: indeed.  Talk is cheap.  I'm just entertaining buu's idea for a minute or two before I lose interest :)
04:02 eric256 interesting. ;) anyway i think the reason is just that ++ is pretty DWIM without special rules. -- could do many things at boundaries
04:15 buu hrm
04:15 buubot has joined #perl6
04:15 buu eval: $x=''; $x++; $x
04:15 buubot buu: Return: 1
04:15 buu eval: $x='z'; $x++; $x
04:15 buubot buu: Return: aa
04:15 buu eval: $x=';'; $x++; $x
04:15 buubot buu: Return: 1
04:15 buu heh
04:16 buu eval: $x='-'; $x++; $x
04:16 buubot buu: Return: 1
04:16 buu ?eval my $x='-'; $x++; $x
04:16 evalbot_7731 \"."
04:16 buu Teh weird.
04:16 eric256 ?eval:p5 my $x = 5; $x++; $x
04:16 evalbot_7731 Can't locate Scriptalicious.pm in @INC (@INC contains: /home/fibonaci/devel/pugs-base/perl5/PIL-Run/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/.. /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Value/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Container/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel/lib /etc/perl /usr/local/lib/perl/5.8.7
04:16 buubot eric256: Error: syntax error at (eval 109) line 1, near "p5 my "  
04:17 buu er
04:17 buu whups
04:17 eric256 yea
04:17 buu Didn't realize your 'eval' involved a colon
04:18 Cryptic_K has quit IRC (Read error: 104 (Connection reset by peer))
04:18 eric256 my eval?
04:19 eric256 your suppose to be able to eval with different backends.. could be usefull to eval p5 code for comparison too
04:19 buu ?eval:p5
04:19 evalbot_7731 Can't locate Scriptalicious.pm in @INC (@INC contains: /home/fibonaci/devel/pugs-base/perl5/PIL-Run/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/.. /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Value/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Container/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel/lib /etc/perl /usr/local/lib/perl/5.8.7
04:19 buubot buu: Return: p5
04:19 buu Causing my bot to pick it up
04:19 eric256 lol
04:19 eric256 gotcha
04:19 Khisanth you didn't anchor it?
04:19 buu I suppose I could change it
04:19 buu Apparently not
04:19 eric256 so i could just say eval:5
04:19 buubot eric256: Return: 5
04:20 eric256 might be considered a feature. i'm assuming buubot is a p5 bot?
04:20 buu yes
04:20 buu buubot: reload
04:20 buubot buu: Reloading!
04:20 buu ?eval: 42
04:20 evalbot_7731 Error:  unexpected ":" expecting program
04:20 buu eval: 42;
04:20 buubot buu: Return: 42
04:20 buu Anchored =]
04:22 Khisanth eval: kill KILL => 'buu';
04:22 buu =[
04:23 buu You made it die.
04:23 eric256 eval: duh
04:23 buubot eric256: Return: duh
04:23 buu Actualy, how does tha twork, I thought kill took a process id?
04:23 eric256 ?eval: 5 + 6
04:23 evalbot_7731 Error:  unexpected ":" expecting program
04:23 buu heh
04:23 buu Make up your mind!
04:23 eric256 ?eval kill $.
04:23 evalbot_7731 Error:  unexpected "." expecting "::"
04:23 eric256 ?eval kill me
04:23 evalbot_7731 Error: No compatible subroutine found: "&me"
04:23 buu ?eval kill $$
04:23 evalbot_7731 Error:  unexpected "$" expecting "::"
04:23 buu ?eval kill $?PID
04:23 evalbot_7731 Error: Undeclared variable: "$?PID"
04:23 * buu gives up
04:24 SamB ?eval kill $
04:24 evalbot_7731 Error:  unexpected end of input expecting "::"
04:24 eric256 eval: kill $$
04:24 buubot eric256: Return: 0
04:24 buu Two args.
04:24 buu eval: kill 9,$$;
04:24 buu Notice the lack of response.
04:24 eric256 kill your own bot.
04:24 eric256 soo mean
04:25 buu I know =[
04:25 coral oof
04:25 buu eval: "Help help I'm being opressed"
04:25 buubot buu: Return: Help help I'm being opressed
04:25 eric256 you might not want to have him on here. to easy to kill your system with a system command
04:25 buu That's what you think
04:26 eric256 what? did you regex for a system in the command or something? or does p5 have a safe mode i don't know about?
04:26 coral Safe.pm exists
04:26 buu I'm just cool like that =]
04:26 eric256 guess you could have used a Safe compartment
04:26 Khisanth eric256: trying to kill a line number? :P
04:26 buu I don't use Safe though
04:26 buu It didn't look safe =]
04:27 * eric256 fears for buu's computer
04:27 buu heh
04:27 eric256 eval:`start calc`
04:27 buubot eric256: Return:  
04:27 eric256 lol
04:27 buu rofls
04:27 coral heh
04:27 buu It's been beaten on by #perls on two networks now
04:27 coral eric256++
04:27 buu And the worst anyone managed to do was make it flood a channel =[
04:27 buu Which I fixed
04:28 eric256 built your own safe processings for it?
04:28 buu Yeah
04:28 eric256 based on? regex?
04:28 buu Naw
04:28 buu ulimits mostly
04:28 eric256 ahh
04:28 coral eval: use threads;
04:28 buubot coral: Error: Can't locate threads.pm in @INC (@INC contains: /etc/perl /usr/local/lib/perl/5.8.7 /usr/local/share/perl/5.8.7 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr
04:28 eric256 okay. thought it was win box,
04:28 coral rats
04:28 buu eric256: Er.. no
04:28 * buu shudders
04:28 eric256 now you know why i worried for your saftey
04:28 eric256 lol
04:28 buu What on earth made you think it was windows?
04:29 buu eval: $^O
04:29 buubot buu: Return: linux
04:29 Khisanth buu: your name conjures up visions of broken windows!
04:29 buu Khisanth: Hurray!
04:30 justatheory has quit IRC ()
04:30 sili_ uhh
04:30 sili_ buu is a windows whore. we all know that
04:31 buu =[
04:31 evalbot_7731 has quit IRC (Remote closed the connection)
04:31 buu http://perlmonks.org/?node_id=302287
04:31 buu DEAR MOTHER OF GOD
04:31 * eric256 is amused that evalbot seems to be quicker than svn bot right now
04:31 evalbot_7732 has joined #perl6
04:33 Blicero has quit IRC (Read error: 104 (Connection reset by peer))
04:33 svnbot6 r7732 | eric256++ | Added templates to the tests cataloge script, and some more output. test it at http://feather.perl6.nl/~eric256/t_index/
04:34 coral eval: use CPAN; CPAN->Shell(install('threads'));
04:34 buubot coral: Error: Can't locate CPAN.pm in @INC (@INC contains: /etc/perl /usr/local/lib/perl/5.8.7 /usr/local/share/perl/5.8.7 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr/lo
04:34 coral rats
04:34 Khisanth eval: use Socket;
04:34 coral eval: use LWP;
04:34 buubot Khisanth: Return:  
04:34 buubot coral: Error: Can't locate LWP.pm in @INC (@INC contains: /etc/perl /usr/local/lib/perl/5.8.7 /usr/local/share/perl/5.8.7 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr/loc
04:34 coral aha
04:34 Khisanth !
04:34 coral Khisanth++
04:34 buu eval: socket(foo) or die $!;
04:34 buubot buu: Error: Not enough arguments for socket at (eval 109) line 1, near "foo) "  
04:34 * buu shrugs
04:35 coral eval: use IO::File;
04:35 buubot coral: Return:  
04:35 mrborisguy eval: `create hole into system through shell`
04:35 buubot mrborisguy: Return:  
04:35 buu Socket and such are already loaded
04:35 buu You can't do anything remotely useful with them any more
04:35 buu So I didn't bother to disable using them
04:35 Blicero has joined #perl6
04:36 eric256 if someone with some html/css skillz wants to hack on the templates for the test cataloge, that would be great. ;)
04:36 coral eval: grep { /Socket/ } keys %INC;
04:36 buubot coral: Return: 6
04:37 coral eval: delete @INC{grep{/^Socket(?::|$)/}};
04:37 buubot coral: Error: syntax error at (eval 109) line 1, at EOF  
04:37 dduncan about labels ...
04:37 coral eval: delete @INC{grep{/^Socket(?::|$)/} keys %INC};
04:37 buubot coral: Return:  
04:37 coral eval: use Socket;
04:37 buu Buubot is also listening in #buubot and in private message for your destructive pleasures.
04:37 buubot coral: Return:  
04:37 dduncan in perl 6, do they look like the perl 5 ones, eg "foo:", or something else?
04:37 coral ok, i'll stop
04:37 * buu shrugs
04:37 dduncan I just want to put in something that'll compile, even if it doesn't run yet
04:37 buu Feel free to break it
04:38 PerlJam dduncan: yes, labels look the same.
04:38 buu But it keeps getting banned when people spam it in channel =[
04:38 dduncan the current pugs will compile "next FOO" and "last FOO", but not the "FOO:"
04:40 Khisanth buu: that sort of thing gets spammed here already!
04:40 * eric256 wonders why anyone ever wants or needs labels. ;)
04:41 Khisanth eric256: how else are you going to use goto?
04:41 Khisanth line numbers?!
04:41 dduncan how about in nested loops?
04:41 buu Khisanth: Well, yeah
04:41 dduncan when you want to use 'next' or 'last' and need to specify which loop you're controlling
04:41 buu Khisanth: I wasn't all that worried, I just thought I'd go ahead and point at the alternative methods.
04:41 eric256 why would i use goto?
04:41 eric256 in nested loops i use flags
04:42 dduncan I don't use goto
04:42 coral eval: goto: goto goto;
04:42 buubot coral: Error: goto must have label at (eval 109) line 1.  
04:42 eric256 and i've seldom even needed flags
04:42 coral eval: GOTO: goto GOTO;
04:43 eric256 has left
04:45 pasteling "dduncan" at 24.108.164.7 pasted "example method using labels" (59 lines, 2.5K) at http://sial.org/pbot/13906
04:45 dduncan eric256, here's where I want to use labels ... if anyone has an alternate suggestion to do the job which doesn't involve merging the 2 loops into one, speak up
04:51 buu coral: For my own edification, what do you think just happened with my bot when you ran that code?
04:51 coral eval/resource limit trap, possibly ignored it altogether, ulimit resource cap, none of the above
04:52 buu Heh, yeah, resource limit
04:52 buu I keep meaning to make it say something when the child gets killed off
04:54 bburton has joined #perl6
04:55 sili_ buu: you know what's funny
04:56 buu JESUS
04:56 sili_ buu: the fuckers on topcoder.com were bitching about how hard it would be to safely run external perl, and i basically told them to do what you did, but they kept dribbling
05:00 bburton has quit IRC ("Client Exiting")
05:13 bburton has joined #perl6
05:29 Cryptic_K has joined #perl6
05:29 r0nny has joined #perl6
05:33 scook0 has joined #perl6
05:35 elmex has joined #perl6
05:38 elmex has quit IRC (Client Quit)
05:49 mrborisguy has quit IRC (Read error: 110 (Connection timed out))
05:54 K_ has joined #perl6
05:54 geoffb Juerd, OT ping
06:03 Cryptic_K has quit IRC (Read error: 104 (Connection reset by peer))
06:08 GeJ has joined #perl6
06:14 pdcawley has joined #perl6
06:20 r0nny yo
06:20 r0nny does someone know, how to create a class behaves like hash syntax ?
06:20 dduncan question: if a BUILD submethod is put in a class, and it doesn't explicitly set a new object's attributes, then will perl 6 automatically set them after BUILD ends ...
06:21 dduncan my situation is this ...
06:21 r0nny dduncan: afaik default values are undef/0
06:22 dduncan I want to have default BUILD semantics, but that I can run a submethod around the same time that takes the same arguments and is just meant to test the arguments for correctness, throwing an exception if not
06:22 dduncan is there another special submethod I can define to do this, or does BUILD have to do that?
06:22 r0nny hmm
06:22 r0nny ask teh gurus
06:22 r0nny im a perl6 beginner, too
06:23 dduncan I did read something about traits or something that can be given to attribute definitions, one of whose names is CHECK, but I don't know if that's the route I want to take
06:24 dduncan I want to have something that is called once for the new object and can see all the initializer args given to it
06:24 r0nny hmm
06:24 saorge_ has quit IRC ("Leaving")
06:27 r0nny yay
06:27 r0nny my Deephash is Hash works
06:27 r0nny eh s/my/class/
06:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
06:30 lisppaste3 has joined #perl6
06:32 Juerd geoffb: OT pong
06:33 geoffb Juerd, I haven't backlogged, so you may have seen this already, but the dovecot problem turned out to have a simple workaround; just ln -s /usr/lib/libz.so.1 /usr/lib/libz.so
06:33 geoffb (I thought of you while I was fixing this on my own mail server.)
06:33 r0nny is method postfix:<.{}> ($key) correct for creating a new hash access operator ?
06:43 Juerd geoffb: Thanks
06:43 geoffb Juerd, np
06:44 Juerd r0nny: Probably so, but it feels wrong. Methods don't include the "." usually.
06:44 luqui has joined #perl6
06:44 Juerd r0nny: method postfix:<{ }> (*@keys) { ... }
06:44 Juerd (Certainly two arguments in <>)
06:45 luqui Juerd, I think perl 6 fears was a good idea
06:45 Juerd luqui: Thank you
06:45 pdcawley has quit IRC ()
06:45 luqui autrijus, ping
06:45 Juerd luqui: It'd be great if more people would write responses, but I may be a little impatient :)
06:46 luqui I think it's nice because it gives people who need to vent a place to do it, and feel like they've been heard without having to jam it down people's throats
06:47 Juerd Some people who originally complained have seen the list, and decided they were being silly
06:47 dduncan r0nny, I was looking at A12 again and it seems that BUILD will actually do what I want, when you have $.foo or $:foo in the signiture of BUILD, but that it does the default assignments from new() args prior to the BUILD call, rather than after ...
06:47 Juerd This by itself took away some fears. Wonderful, I think.
06:47 luqui :-)
06:47 Juerd It's probably partly because some fears are in contradiction
06:47 geoffb Yes, Juerd++ for that
06:47 dduncan if this is actually the preferred behaviour, then I may talk to Damian about his Class::Std doing it the same way (currently, it seems to do the default stuff after BUILD is called)
06:48 Juerd Obviously: "It will be slow" and "It will not be released" contradict
06:49 Juerd And: "Perl 6 will not look like Perl" versus "It will still look like line noise"
06:49 luqui right
06:56 luqui has quit IRC ("Leaving")
06:59 dduncan question ... when referring to a named object's private attribute, which syntax is better: "$obj.:foo" or "$obj:foo"? ... Pugs seems to parse both, while when referring to the topic object it only accepts "$:foo" but not "$.:foo"
07:00 iblechbot has joined #perl6
07:01 Juerd dduncan: It cannot be referred to by object, IIRC
07:01 Juerd dduncan: It's $foo, without :, and it has no accessor method. This is fairly new stuff though, I don't know if Pugs already reflects it.
07:02 dduncan I'll clarify that the code in question is within the class whose attributes I'm talking about ...
07:03 dduncan so privates are visible there
07:03 dduncan also, the named objects are ones that belong to the same class as the code in question, or it trusts the code in question
07:04 dduncan an issue here is that I see an inconsistency with what formats are parseable
07:04 dduncan "$:foo" will parse, "$.:foo" will not, ...
07:05 dduncan "$obj:foo" will parse, so will "$obj.:foo"
07:06 dduncan so part of my question is whether "$.:foo" *should* parse or whether its non-parseability is correct behaviour
07:08 Juerd $:foo variables are gone
07:08 dduncan I just want to be able to either use both "$.:foo" and "$obj.:foo" or not be able to use either of them
07:08 Juerd If you want externally visible variables, I don't think they can still be private
07:09 dduncan these aren't externally visible
07:09 dduncan I'm referring to allowed syntaxes for a classes own code to see its own private attrs
07:10 dduncan is the dot optional, or should it always be used, or should it never be used, when a class references its own privates?
07:11 dduncan I'm thinking that it should be possible to always use the dot, even if one can leave it out at times when doing so isn't ambiguous
07:12 K_ has quit IRC ("Leaving")
07:20 larsen has joined #perl6
07:21 Juerd dduncan: When it is used, accessor methods are created.
07:22 dduncan I think I'll defer this discussion until later ... since it may end up being a non-issue ... thanks anyway
07:25 dada has joined #perl6
07:30 sili_ has quit IRC ("Lost terminal")
07:34 svnbot6 r7733 | juerd++ | ,=response
07:35 wolverian ,=? :)
07:35 Juerd push ;)
07:35 Juerd This said, I want all of +=, ~=, ++, etcetera, to work on $_ by defaut.
07:35 wolverian I've always wanted that.
07:35 Juerd We should be able to write ++; instead of ++$_;
07:36 wolverian is that postfix or prefix?
07:36 Juerd Prefix
07:36 wolverian why?
07:36 Juerd Heads.
07:36 wolverian :)
07:36 Juerd I have very little influence on the coin :)
07:36 Juerd Hm, no, postfix makes more sense
07:36 Juerd Because += would also get $_ on the LEFT
07:37 Juerd I think that's way more useful than having it on the right, at least :)
07:37 wolverian btw, am I right in reading that /foo/ returns a match object?
07:37 wolverian er
07:37 wolverian s,match,regex,
07:37 Juerd Depends on its context.
07:37 wolverian right.
07:37 wolverian when does it match automatically against $_?
07:37 Juerd When used in boolean context, or with ~~, it automatically matches.
07:37 wolverian good. thanks. :)
07:37 wolverian more rationalisation.
07:38 Juerd Do note that you may need m// in case of: ($foo) = /(...)/
07:38 Juerd Or you just get a rx// in $foo
07:38 evalbot_7732 has quit IRC (Remote closed the connection)
07:38 wolverian yeah, thought about that too
07:39 wolverian I guess you can always do my ($foo) = .match(/(...)/); # :)
07:39 evalbot_7733 has joined #perl6
07:39 Juerd $foo = m/(...)/
07:40 wolverian yeah.
07:44 Juerd p6l'ed implicit $_ on mutation
07:45 wolverian you have a knack of suggesting things that you either love or hate.
07:45 wolverian s:2nd/you/one must/
07:46 Juerd Hm?
07:46 Juerd How so?
07:46 Juerd And does that make the things good or bad? :)
07:46 wolverian I think usually your suggestions aren't exactly small - or if they are, they are excessively cute - so people will polarise strongly on the issue
07:47 wolverian Juerd, it doesn't have a goodness factor per se, besides the fact that it starts discussion.
07:47 wolverian (or badness factor.)
07:47 Juerd I like to think they're all cute :)
07:47 Juerd Cuteness is important, because very often, it matches a way of thinking.
07:47 wolverian I agree.
07:48 Juerd junctions and -> subs are cute in such a way
07:48 Juerd Because of their cuteness, they're instantly understood in their respective idiomatic contexts
07:48 wolverian I just wish -> was spelled lambda (as in the letter)
07:48 Juerd (boolean, and loops)
07:49 wolverian (I don't really. but it'd be even cuter.)
07:49 Juerd The arrow makes it cute
07:49 Juerd I think using -> outside of loops is in many cases bad style.
07:50 Juerd But "for @foos -> $foo { ... }" is so incredibly cute that a newbie will probably not even notice that -> is a regular prefix op.
07:51 wolverian :)
07:55 Juerd Damian wondered what I was drinking when I suggested using 1c as the ascii equiv. of <cent thingy>
07:55 Juerd (It was Amsterdam tap water)
07:55 Liz has joined #perl6
07:55 wolverian hehe.
07:55 Juerd He first tried to argue his way out of it :)
07:55 wolverian ¢ is kind of cute, although I'm scared of a new sigil.
07:55 Juerd But eventually had to admit it's possible without any clashing with existing syntax :P
07:56 Juerd It's needed
07:58 wolverian yes.
08:14 Juerd I'm off for a few hours
08:15 Liz see ya later!
08:15 Juerd So if you should feel like responding to some fears without the possibility of clashing with my potential edits, now is the time! ;)
08:15 Juerd afk
08:23 brother Where was the fear-doc?
08:24 brother ahh /docs/quickref/
09:07 buu The document OF FEAR
09:09 Blicero has quit IRC (Remote closed the connection)
09:09 Blicero has joined #perl6
09:26 iblechbot has quit IRC (Read error: 104 (Connection reset by peer))
09:34 Blicero has quit IRC (Remote closed the connection)
09:50 evalbot_7733 has quit IRC (Remote closed the connection)
09:50 evalbot_7734 has joined #perl6
09:53 svnbot6 r7734 | Darren_Duncan++ | /ext/Rosetta-Incubator : wrote most of the initial Locale::KeyedText module ... present is nearly all the code, and a third of the POD
09:57 joepurl has quit IRC (Read error: 110 (Connection timed out))
10:02 ruz has quit IRC (Read error: 110 (Connection timed out))
10:08 dduncan has quit IRC ()
10:14 r0nny_ has joined #perl6
10:26 r0nny has quit IRC (Read error: 110 (Connection timed out))
10:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
10:30 lisppaste3 has joined #perl6
10:46 rafl_ has joined #perl6
10:57 rafl has quit IRC (Read error: 110 (Connection timed out))
10:59 spinclad lisppaste3: paste?
11:07 lisppaste3 spinclad pasted "Installation report for 6.2.10-3: libghc6-pugs-dev failed" at http://paste.lisp.org/display/12848
11:09 spinclad rafl: install report for you
11:12 rafl_ spinclad: Fuck.. I don't get it. I installed it >10 times without problems. The version in the archive isn't installable, though. :-(
11:13 Jooon something I can try?
11:14 iblechbot has joined #perl6
11:19 Jooon ah. got the same problem
11:19 rafl_ Jooon: You can try building it from source.
11:19 rafl_ Jooon: My idea is that the 30installdirs patch wasn't applied.
11:27 Jooon where is libghc6-plugins-dev?
11:28 autrijus *nod*
11:28 rafl_ In NEW. It needs to be approved by FTP-Masters before it goes into the archive. You can simply remove it from Build-Depends. That should not change the situation. You simply won't have hs eval support.
11:28 autrijus greetings :)
11:28 rafl_ Hello autrijus
11:28 rafl_ is now known as rafl
11:28 Liz morning!
11:29 * autrijus is back to life!
11:30 Jooon autrijus: smileys :) http://flickr.com/photos/jooon/55661134/
11:31 autrijus Liz: hi :) got a moment?
11:31 autrijus Jooon: ooh, my face was a virtual smiley too
11:32 Liz has quit IRC ("User disconnected")
11:33 * autrijus starts massive backlogging
11:36 spinclad rafl: great regret to always bear bad news.
11:38 spinclad i'll look at it some more, something may pop out at me.
11:39 rafl spinclad: The problem is, that there are directories from the src tree listed in install-dirs: in Pugs.cabal.
11:39 rafl spinclad: 30installdirs.dpatch should prevent that.
11:45 Jooon never built pugs before. Pugs.AST.Internals seems to take a while
11:55 Liz has joined #perl6
11:57 SamB Jooon: it does that
11:57 SamB and it has an annoying habbit of changing, too!
11:57 spinclad rafl: listed in library-dirs:, rather
11:58 rafl spinclad: Of course.
11:59 autrijus fears++ # wonerful group therapy thing
11:59 autrijus Liz: ping?
11:59 Liz ping
11:59 autrijus Liz: have you /msg'ed me? freenode doesn't allow /msg for nonregistered nicks
11:59 Liz fighting with all sorts of irc clients
11:59 autrijus ahh.
11:59 autrijus Liz: I think after .dk I'll be in .nl again
12:00 autrijus possiblity of staying at your place?
12:00 Liz cool
12:00 Liz yep
12:00 Liz did u get my email?
12:00 autrijus aye
12:01 Liz sent on friday, 6:41pm
12:01 Liz shall I send again?
12:01 autrijus no, I got the mail
12:02 autrijus the plan is to arrive to .nl around this weekend
12:02 Liz ok
12:02 autrijus and stay there till around nov 8th
12:02 rafl spinclad, Jooon: Is it reproduceable for you?
12:02 Liz schiphol?
12:03 autrijus I don't know the best routing from copenhagen to where you are
12:03 Liz lemme check and get back to you by email, ok?
12:03 autrijus sure!
12:03 autrijus thanks again :)
12:03 Jooon rafl: I have been building for over 20 minutes :)
12:03 Jooon rafl: but it looks like the patch is applied
12:03 autrijus Liz: oh also, I'll tell amazon to ship some books to your place. that's fine too?
12:04 Liz sure, no pb
12:04 autrijus cool!
12:05 Liz easyjet is out
12:06 lisppaste3 has quit IRC (Remote closed the connection)
12:06 theorbtwo Hey, Liz!  How are you?  I was just missing you the other day over at perlmonks.
12:07 Liz the other year you mean?
12:07 Liz ;-)
12:07 theorbtwo No, somebody mentioned you on perlmonks, and I thought "hmm, she was fun... I wonder why she stopped coming here".
12:08 Liz work
12:08 Liz ;-)
12:08 Liz paid work, that is...
12:10 autrijus bbiab...
12:11 spinclad rafl: i can tell that the 30 patch was applied; haven
12:12 spinclad 't tried a build yet; will do so this afternoon.
12:12 robkinyon autrijus - got a second?
12:12 robkinyon crud
12:12 rafl autrijus: Do you have an idea why cabal seems to add a library-dirs line to .installed-pkg-config? It isn't in Pugs.cabal.
12:15 autrijus robkinyon: mm?
12:15 autrijus rafl: no, no idea at all. what if we specify an empty one?
12:16 rafl autrijus: Do you know if the cabal version of ghc changed between 6.4 and 6.4.1? I thin 6.4 didn't do that.
12:17 rafl autrijus: I'll try specifying an empty one.
12:17 kolibrie has joined #perl6
12:18 autrijus no, it's the same cabal.
12:18 autrijus but maybe ghcbugs on 6.4.0
12:18 autrijus not sure
12:20 rafl autrijus: Specifying an empty one: Setup: Line 22: Unknown field 'library-dirs'
12:20 autrijus :/
12:20 autrijus library-dir perhaps?
12:20 autrijus in old cabal
12:20 autrijus not sure at all
12:22 lisppaste3 has joined #perl6
12:22 rafl library-dirs: seems to contain everything in extra-lib-dirs + $PREFIX/lib/Pugs-6.2.10
12:30 joepurl has joined #perl6
12:44 joepurl has quit IRC ("Lost terminal")
12:45 joepurl has joined #perl6
12:48 xinming has joined #perl6
12:49 scook0 has quit IRC ("Leaving")
12:56 Liz autrijus: see mail of 2:56pm
13:01 * clkao hugs autrijus
13:03 SamB has quit IRC (Connection timed out)
13:07 SamB has joined #perl6
13:07 Jooon rafl: same error with self built package
13:08 rafl Jooon: Currently trying if I can work that around with a dirty hack:
13:08 rafl grep -v library-dirs: .installed-pkg-config > $(CURDIR)/debian/libghc6-pugs-dev/usr/lib/haskell-packages/ghc6/lib/Pugs-6.2.10/installed-pkg-config
13:08 rafl instead of cp .installed-pkg-config $(CURDIR)/debian/libghc6-pugs-dev/usr/lib/haskell-packages/ghc6/lib/Pugs-6.2.10/installed-pkg-config
13:10 Qiang has joined #perl6
13:16 elmex has joined #perl6
13:24 ken has joined #perl6
13:24 joepurl has quit IRC (Read error: 110 (Connection timed out))
13:27 ken has left
13:31 joepurl has joined #perl6
13:51 eric256 has joined #perl6
13:51 eric256 has quit IRC (Read error: 104 (Connection reset by peer))
13:51 eric256 has joined #perl6
13:53 eric256 morning
13:55 hlen_ has joined #perl6
13:59 kolibrie eric256: morning
13:59 eric256 yea someone is alive ;)
13:59 kolibrie a little bit
14:08 hlen has quit IRC (Read error: 110 (Connection timed out))
14:09 xern has quit IRC ("Lost terminal")
14:11 xinming has quit IRC (Remote closed the connection)
14:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
14:30 lisppaste3 has joined #perl6
14:33 sapper has joined #perl6
14:56 eric256 has quit IRC (Read error: 110 (Connection timed out))
14:57 Blicero has joined #perl6
15:01 Blicero has quit IRC (Remote closed the connection)
15:06 Blicero has joined #perl6
15:06 saorge has joined #perl6
15:06 saorge has quit IRC (Read error: 104 (Connection reset by peer))
15:06 saorge_ has joined #perl6
15:06 saorge_ has quit IRC (Read error: 104 (Connection reset by peer))
15:06 saorge has joined #perl6
15:11 Blicero has quit IRC (Remote closed the connection)
15:13 justatheory has joined #perl6
15:31 dudley_ has joined #perl6
15:31 Blicero has joined #perl6
15:42 wilx` is now known as wilx
15:44 stevan autrijus: ping
15:51 rafl Can someone please check the pugs_6.2.10-4 packages at perlcabal.org/~rafl/debian?
15:52 rafl Espcially the libghc6-pugs-dev package.
15:54 mrborisguy has joined #perl6
15:56 rafl gaal, Jooon, spinclad?
15:57 eric256 has joined #perl6
16:02 eric256 autrijus: ping
16:02 svnbot6 r7735 | eric256++ | Updated documentation linkage in more tests.
16:02 svnbot6 r7736 | eric256++ | Updated test cataloge script to include the number of links in each test next to its entry in the index.
16:04 autrijus stevan: pong
16:04 autrijus eric256: pong
16:05 stevan hey autrijus :)
16:05 stevan mostly I just wanted to bug you regarding the Object Space stuff
16:05 evalbot_7734 has quit IRC (Remote closed the connection)
16:05 eric256 hey. would it be possible to reidrect tests.pugscode.org to http://feather.perl6.nl/~eric256/t_index/
16:06 evalbot_7736 has joined #perl6
16:06 eric256 i'm working on better templates but its functional now. and will soo be being autoupdated
16:07 mrborisguy good work, eric256
16:07 autrijus stevan: you are using that term, so I assume you've read through the PyPy literature? :)
16:07 stevan autrijus: some, but not all
16:07 autrijus eric256: excellent. will do that now
16:08 autrijus stevan: cool. I'm still halfway thru pickaxe
16:08 stevan autrijus: mostly I liked the idea of a seperate "part" of the compiler to deal with objects
16:08 autrijus "part"?
16:08 * stevan should really read that one again
16:09 stevan well, it seemed to me that they were seperating the flow-control, syntax, etc part from the object part
16:09 eric256 thanks.  mrborisguy ...care to work on the templates some?  or add links to documentation ;)
16:09 autrijus stevan: yes, and I think it makes ~total sense
16:09 stevan is that total in string context, or an approximate total
16:09 mrborisguy eric256: I can certainly help when I come by some time... what do you want me to do?
16:09 autrijus python doesn't allow type annotations
16:09 autrijus approximate
16:09 stevan yes
16:09 autrijus so their type analysis is limited to flow analysis
16:09 autrijus but we need another type engine
16:10 autrijus so that makes 3 parts
16:10 stevan ok
16:10 stevan what will the responsibilties of the type engine be?
16:10 autrijus it's done in 2 phases
16:10 autrijus phase 1 is heuristics
16:10 eric256 well the templates are in util/catalog_tmp/ ... they need some work to be less functional and more pleasing.  or just ad  L<S02/Heading> links in test files ;)
16:10 stevan just identification? or implementation
16:10 autrijus look at all annotations
16:10 autrijus and propagate them upward and downward
16:11 SamB boy, that sounds kinda familiar...
16:11 stevan downward?
16:11 autrijus phase 2 is standard typechecking / inference
16:11 * eric256 wonders if similar linking should be done in pugs source files.....
16:11 autrijus stevan: if you annotate an entire expression, "downward" means propagating things into subexpressions
16:11 stevan ok
16:11 autrijus canonical example being a annotated function and the expression in it
16:12 mrborisguy eric256: alright.  I'll do that.
16:12 stevan so once checking and inferencing is done, it will hand things off to the object space?
16:12 eric256 mrborisguy: which one? lol. thanks btw
16:12 autrijus stevan: yes
16:12 mrborisguy eric256: uh, the easy one ;)
16:13 stevan so that phase is really about squeezing the code for available info, so that by the time the object space gets it, there is very little left which is ambigious
16:13 mrborisguy eric256: another thing to consider would be taking note of all of the :todo items, and making another page that lists all of those
16:13 stevan correct? or am I not understanding
16:13 mrborisguy it might help lurkers find something to do that isn't implimented
16:14 xern has joined #perl6
16:14 eric256 yea. i was going to look at linking to the tests.yml file and putting that info next to each test file too. starting small and easy ;)
16:14 stevan eric256++
16:14 stevan eric256: are you writing this in perl6? or p5?
16:14 * eric256 anxiously awaits mrborisguy next commit. ;)
16:15 eric256 it was a script in the util directory that just needed some tlc, its in p5 currently
16:15 autrijus stevan: yup, ambiguity is either warning, rejection, or infereence
16:15 stevan autrijus: cool
16:16 stevan autrijus: so, I would like to get to work on the object space, using fglock's stuff as a basis, any thoughts?
16:16 Juerd Ambiguity doesn't exist!  </naive>
16:16 Juerd Oh, the world would be SO boring.
16:16 autrijus stevan: I think that'd be excellent.
16:17 autrijus and we can update the long overdue S12.5 :)
16:17 mrborisguy eric256: catalog_tests.pl ?
16:17 eric256 yep
16:17 stevan does what I wrote to p6c sound sane?
16:17 autrijus yes.
16:17 stevan autrijus: yes, S12.5 needs updating, but there is too much hand waving until I work out some of these object space details
16:17 eric256 stevan: no. ;)  j/k but most your stuff sounds insane to us mortals. ;)
16:17 autrijus eric256: tests. is live
16:17 mrborisguy alright... well don't wait too anxiously for my next commit... time to eat, then class in 20 minutes, then study for a big math test tomorrow.
16:17 autrijus stevan: yeah, exactly my thought
16:18 mrborisguy but I'll take a look at it some other time
16:18 eric256 autrijus sweet! thanks
16:18 stevan eric256: sorry hard to type in this straightjacket,.. but thanks :)
16:18 GeJ has quit IRC ("NetBSD: used on TabletPCs during the build of the Pyramids in 4000BC")
16:19 * eric256 gives up on mrborisguy for the moment and looks for other prey. ;)
16:20 mrborisguy :D
16:22 stevan autrijus: sounds good, I will probably make some more drawings before I start writing any code too
16:22 stevan I will try and make em flashy, so you can sell more t-shirts :)
16:23 * stevan wonders where putter has been recently
16:23 autrijus      styou++
16:23 autrijus gr, bad slow terminal
16:23 autrijus network going away in 30mins :/
16:23 autrijus next signon will be from .dk
16:23 stevan enjoy the trip :)
16:24 autrijus fortunately I'll stay at .nl from this weekend on for 10 days
16:24 autrijus then I can massively sync / update :)
16:24 Juerd autrijus: Where are you staying?
16:24 autrijus Juerd: liz's
16:24 autrijus in .dk will be defekt's
16:25 * Juerd wonders if this results in sane threads in a week ;)
16:25 autrijus eh? :)
16:25 Juerd Liz: By the way... Any chance of a lizperl-like thing for Perl 6? :)
16:25 autrijus btw, the "fears" stuff is absolutely wonderful.
16:25 Juerd autrijus: Thanks
16:29 * eric256 discovers thats his SEA was 2k commits behind...that would explain some missing parts ;)
16:31 Juerd Where are recent SEAs?
16:31 eric256 http://svn.perl.org/perl6
16:32 Juerd Ah
16:32 Juerd Nice to know
16:32 eric256 i did svk co .... but that doesn't get the updates? or did i just choose the wrong option somewhere? ;)
16:33 Juerd It doesn't seem to be more recent than those at http://dev.perl.org/perl6/doc/design/syn/
16:33 wolverian eric256, svk pull
16:35 wilx svk up -s
16:35 eric256 juerd: then i could be wrong  no idea realy...when i asked yesterday thats the one i got.. i could ask on p6l
16:36 wilx When you have things checked out already.
16:36 eric256 yea i'm doing a svk pull now. just didn't know it wasn't automatic on co
16:41 mtve has quit IRC (Read error: 110 (Connection timed out))
16:42 r0nny_ is now known as r0nny
16:55 sapper has quit IRC ("Leaving")
17:09 mtve has joined #perl6
17:13 eric256 hehe. its almost up to date ;)
17:20 Liz Juerd: with re to lizperl: you've got to be kidding...
17:24 svnbot6 r7737 | iblech++ | docs/quickref/fears:
17:24 svnbot6 r7737 | iblech++ | * Added a comment (should we $perl6_will_be_too_slow ~~ s/Perl 6/perl6/?)
17:24 svnbot6 r7737 | iblech++ | * "I'll have to *manually* separately compile my Perl 6 programs"
17:24 r0nny yo
17:26 r0nny is there a howto for creating new hash classes ?
17:27 evalbot_7736 has quit IRC (Remote closed the connection)
17:27 evalbot_7737 has joined #perl6
17:28 stevan r0nny: hey
17:28 stevan r0nny: do you mean a class which does(Hash)?
17:28 stevan so you can access it like as hash
17:29 stevan and treat it like a hash
17:30 Juerd Liz: I said lizperl-*like* :)
17:30 Juerd Liz: Preferably much simpler internally :)
17:31 r0nny stevan: oh - so that was my essos ;P
17:32 r0nny stevan: btw - is there a way to use methods of the rhole ?
17:32 r0nny since i need to overwrite them for my needs
17:32 stevan r0nny: I am not sure that functionality is implemented yet
17:32 stevan actually I am pretty sure it is not
17:32 r0nny :/
17:33 stevan that really requires the metamodel,.. which is not here yet
17:33 eric256 you can add methods that work on hashes though
17:33 stevan you can use multi-subs
17:33 stevan or even just subs with typed parameters
17:33 eric256 stevan's going to have that metamodel done and implemented in a week though, then we can play with it ;)
17:33 stevan eric256: it is already done,.. just not in Haskell ;)
17:33 eric256 hince *implemented*
17:34 stevan s/implemented/implemented in pugs/
17:34 eric256 well it aint any use in pugs unless its in there now is it!! ;)
17:35 stevan actually you can use the p5 version, and there is a hacked in p6 class parser too
17:35 stevan so its pretty close,..
17:35 stevan closer than nothing :)
17:35 eric256 yea. i've just been chomping at the bit for all the OO stuff. its the best part of perl6...and we keep waiting and waiting. ;) is the wait gonna be worth it?
17:35 stevan eric256: I think so
17:36 eric256 cool
17:36 stevan eric256: give the metamodel prototype a try, it can do quite a bit
17:36 eric256 is there any timeline on it yet?
17:36 stevan and its faster than pugs right now ;)
17:36 eric256 i don't wanna program in p5...thats old school
17:37 Juerd eric256: Time is circular anyway...
17:37 stevan eric256: to do the metamodel right, we need to have the object space, to have the objectspace, we need to rip the guts out of pugs, and in some ways start from scrath
17:37 stevan s/scrath/scratch/
17:37 eric256 yea /me feard that was your answer
17:37 eric256 ;)
17:38 stevan but autrijus had a plan to shoehorn the mm into it for the 6.28 release
17:38 stevan so it may not be all that long
17:38 eric256 i think it was juerd who was pointing out that there are some real issues with references in the current pugs that may require a large scale rewrite ;)
17:39 * eric256 just hopes a large scale rewrite wont stall pugs ;)
17:39 stevan anymore than it currently is?
17:39 * stevan is hoping that autrijus's 10 days in .nl will be very productive
17:39 Juerd eric256: Not likely.
17:40 stevan eric256: good things come to those who wait ;P
17:40 eric256 juerd...not likely that it was you, or that it wil stall?
17:40 Juerd I't likely that I said things were wrong
17:40 Juerd But not that I said anything requires large scale rewrites.
17:40 eric256 you said something about a new backend that autrijus was already working on /me might have misunderstood the implications
17:40 Juerd If I do hint about implementation time, it's usually that something ought to be trivial :)
17:41 Juerd eric256: I have no idea what autrijus is currently working on, so that wasn't me.
17:41 * eric256 needs memory pills ;)
17:41 Juerd mcflurry rolo ++
17:43 eric256 anyone know where .pick is defined in teh SAE?
17:43 Khisanth grep .pick *.pod? :)
17:44 eric256 no hits...which surprised me. :)  ... guess pugs just decided to implement that on its onw
17:45 * Juerd wonders what eric256 wants to know
17:45 Juerd pick's very straight forward
17:45 Juerd When used on any collection, it gives you a random element.
17:46 Juerd Hash, array, disjunction, conjunction, ...
17:46 svnbot6 r7738 | dakkar++ | fears: corrected a reference to Perl 6 where Perl 5 was intended, and added a pityful answer to the 'obfu/golf' fear
17:46 eric256 juerd...where to link to in the documentation
17:46 eric256 remember my new quest to increase cross references ;)
17:46 Juerd Ahh
17:47 Juerd I don't think lists of methods are officially defined yet
17:47 Juerd You may find something in pugs/docs/quickref/data
17:47 Juerd But that's inofficial
17:47 theorbtwo It'd be wonderful if all over the place there were cross-references between implementation, test, and documentation, and if one of the three were missing, to consider it a bug in it's own right.
17:47 evalbot_7737 has quit IRC (Remote closed the connection)
17:48 theorbtwo ...but getting there is hard.
17:48 Juerd theorbtwo: We need numbers, like some projects use bugnumbers
17:48 Juerd In a debian changelog, #238729 is all one needs to refer
17:48 Juerd The same notation is used in many places
17:48 evalbot_7738 has joined #perl6
17:48 PerlJam rt
17:48 Juerd However, referring to a subdocument is hard.
17:49 Juerd So perhaps it is time for a  spec cards wiki
17:49 eric256 theorbtwo yea i was thinking of linking in the haskell source too
17:49 PerlJam perl6 has a queue in rt.perl.org and svn gives a convienent way to reference any bit of it.
17:49 theorbtwo A long time ago, we started introducing L<document /regex/> links in the tests.
17:49 Juerd With lots of tiny subdocuments, named or numbered, that are categorized
17:49 eric256 theorbtwo thats what i'm currently working on
17:49 theorbtwo Great, eric.
17:49 eric256 theorbtwo: http://tests.pugscode.org ;)
17:50 eric256 makes it easy to find tests that need linked to documents, you can even surf the documents and see where tests are need (superscript T's mark the cross reference)
17:50 hexmode has joined #perl6
17:50 * theorbtwo remembers writing that code.
17:51 Juerd Wow, this way of cross referencing must be tedious
17:51 theorbtwo I really wanted tests to mark out extents in the file, but that'd make it more tedious.
17:51 theorbtwo OTOH, it'd also make it so that you could view the documentation with red spans where there weren't any tests.
17:51 eric256 a reged start and finsih theorbtwo?
17:52 theorbtwo Yeah, Juerd, I'd rather be able to say S3 § 3.4.8.7, or something, but I didn't consider changing Larry's writing style to be within my scope.
17:52 eric256 that would be interesting
17:54 Juerd I find it surprising that a spec can be maintainable in the form of only human language
17:54 * eric256 wonders what definition of maitained you are using ;)
17:54 Juerd theorbtwo: That still makes inserting sections hard
17:54 theorbtwo Yeah, Juerd, I know.
17:55 Juerd theorbtwo: The ultimate solution is to use *unstructured* pieces of spec, that by inclusion form larger documents
17:55 eric256 Juerd yes. i'm guessing commit access to those isn't easy to get? ;)
17:55 stevan geoffb: (see http://use.perl.org/~brian_d_foy/journal/)
17:55 eric256 then someone could mark sections in them with bugnums and link to that from tests. lots easier
17:56 geoffb stevan, what am I looking for?
17:56 stevan an article to write :)
17:56 stevan you are the official pugs scribe after all ;)
17:56 geoffb stevan, lol
17:57 theorbtwo What would be best, I think, possibly, is a wiki that uses nodeids.
17:57 Juerd theorbtwo: That's more or less what I suggested :)
17:57 theorbtwo ...so sections could be retitled without links breaking.
17:57 geoffb stevan, well, hmm.  Depends on when the deadline is, and whether it pays at all.  Right now I've got a lot of stuff on my plate (witness the fact that I haven't even been reading #perl6 for about half a week)
17:58 theorbtwo Juerd: We are in agreement, then.  Yey.
17:58 Juerd theorbtwo: But, I don't think we can get Larry to join in on such an effort
17:58 stevan geoffb: oh well, just a thought
17:58 Juerd theorbtwo: And I don't think Larry would allow others to help
17:58 geoffb stevan, and a good one.   Much appreciated.
17:58 Juerd Which, certainly combined, is a problem
17:58 theorbtwo Larry-- on this.
17:59 * geoffb needs to kill this $contract dead
17:59 Juerd It'd be great to have a wiki, with simply per node:
17:59 Juerd 1. title
17:59 Juerd 2. short description
17:59 Juerd 3. syntax
17:59 Juerd 4. semantics
17:59 eric256 yea ..../me is sad... guess i'll continue the regex linking for now
17:59 Juerd 5. tests
17:59 Juerd Then build test files from the spec sheets
17:59 theorbtwo Not even if we build some sort of red-letter bible system in, where the words of larry are in red, and the cabal is in purple?
17:59 Juerd And link in nodes to form synopses
17:59 theorbtwo That'd be wonderful.
18:00 * eric256 nominates juerd and theorbtwo as perl6 librarians and goes to lunch
18:00 Juerd All you need is a wiki without formatting syntax :)
18:00 theorbtwo (But not really likely to form good prose.)
18:00 Juerd (because plain text DOES rule.)
18:01 Khisanth Juerd: sounds a bit like the perl.com article on how the TT book was written :)
18:01 Juerd theorbtwo: It doesn't have to.
18:01 Juerd theorbtwo: Articles and books are for that.
18:01 Juerd Prose doesn't have to be complete. In fact, when complete, it overwhelms.
18:01 Juerd Specs do have to be complete.
18:01 Juerd Let's talk about something else
18:03 theorbtwo Juerd: I can't help but think some sort of coverage system on docs is highly important at this phase.
18:04 theorbtwo My #1 complaint about p6 is that it's poorly documented, and I think one part of a good solution is to make many mailing list messages directly into documentation.
18:06 theorbtwo 1: Emails already get written.
18:06 theorbtwo 2: Don't try to redocument all over the place, reuse existing wording instead.
18:07 theorbtwo 3: A single email is hopefully small enough to be readable as the answer to a simple question.
18:07 mrborisguy eric256: when I do svk pull, I don't get your template folder
18:08 larsen has quit IRC ("later")
18:08 mrborisguy is there something you need to do to make sure that gets into SVN, or is there something special that I need to do to pull it?
18:21 luqui has joined #perl6
18:29 Juerd theorbtwo: Documentation or specs?
18:29 Juerd theorbtwo: I think that at this point, specs are far more important.
18:29 theorbtwo Good point... I tend to use the terms interchangeably, but they really aren't...
18:30 Juerd theorbtwo: Mail is threaded and there are a lot of discussions. The summarizers don't even manage to summarize all of what is said, so we shouldn't expect outsiders to be able to follow anything of it.
18:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
18:30 theorbtwo The summeries used to do more summerizing, and less linking.
18:30 lisppaste3 has joined #perl6
18:36 Juerd Yes, but then the lists got more activity, and more fanned out discussion
18:37 Juerd There is very little philosophy left in today's language design discussions
18:37 Juerd It's all purity versus cuteness, I sometimes think
18:38 Qiang reading FEAR: Perl 6 is hard to learn!  
18:38 Qiang Perl has always been hard to learn. It pays off, though.
18:38 Qiang that's not true for everyone :)
18:39 Juerd The entire document assumes average programmers
18:39 Qiang I thought Perl has always been easy to learn.
18:39 Juerd But if you think you can refine it and make it better, by all means, please do so
18:39 mrborisguy how do I upload a folder with 4 new files to SVN?
18:39 mrborisguy (I've only been doing 'svk pull's recently, and I forget the syntax)
18:40 Juerd mrborisguy: man svk
18:40 Juerd svk --help
18:41 mrborisguy Juerd: thanks.
18:41 Qiang that sentence probably can be addressed bit flexible. though i am not sure how. me not good at english ;0
18:43 Juerd Qiang: "for the average programmer"
18:43 Juerd Qiang: But I prefer not to include this.
18:49 evalbot_7738 has quit IRC (Remote closed the connection)
18:49 svnbot6 r7739 | mrborisguy++ | Added template files for the test cataloger
18:50 evalbot_7739 has joined #perl6
19:03 eric256 mrborisguy - sorry forgot to add them to the repo...done now
19:06 eric256 svn check
19:06 eric256 ack  mrborisguy added his own templates?  /me is now realy confused
19:07 Maddingue has quit IRC (Read error: 104 (Connection reset by peer))
19:08 Maddingue has joined #perl6
19:08 luqui has quit IRC ("This computer has gone to sleep")
19:12 eric256 ahh mrborisguy must have got them out of my feather folder mrborisguy++ for determination ;)
19:12 xinming has joined #perl6
19:14 Khisanth Qiang: easy to learn, hard to master!
19:15 Juerd The "new sigil" thread is now the new largest thread that I didn't start
19:15 Juerd A few more posts and it's the largest thread in my entire p6l archive :)
19:25 r0nny yo
19:26 dada has quit IRC ("several species of small furry animals gathered together in a cave and grooving with a pict")
19:26 PerlJam In a way it's a shame that the new sigil isn't #.   #T looks nicer to me for some reason.
19:26 r0nny any date for "stuff like "class foo does hash" works if i want to change the way its operators work ?
19:26 * Khisanth comments out PerlJam's comments
19:29 eric256 r0nny you don't need that to change the way operators work...
19:29 Juerd PerlJam: I quite like # for comments.
19:29 eric256 well you do if you want to special case it i suppose, so that only certain hashes are treated differently
19:29 Juerd I think not picking ^ is a mistake.
19:30 r0nny eric256: can u gove me a hint ?
19:30 eric256 i think we should use /-# for comments ;)  maybe throw in a < just to be safe too. ;)
19:30 PerlJam Juerd: yeah, ^ already has that place-holder connotation.
19:30 evalbot_7739 has quit IRC (Remote closed the connection)
19:30 eric256 sub infix:<+> (Hash $a, $b) { ... };
19:30 Juerd PerlJam: And it is visually distinctive, easy to type, non unicode, and not already over used.
19:30 svnbot6 r7740 | juerd++ | re perl6 vs Perl 6
19:30 * eric256 likes ^ too... so macro ^ as "class " and be happy
19:30 evalbot_7740 has joined #perl6
19:30 Juerd I think "class" as a sigil is a huge loss
19:31 PerlJam Juerd: so reiterate that on p6l and see if it makes a dent
19:31 r0nny i want to translate stuff like %hash{this<sep>is<sep>a<sep>deep<sep>hash} to %hash{this}{is}{a}{deep}{hash}
19:31 eric256 i think it would be best if these don't end up with a sigil but become bare words.  so that my T $x; isn't cluttered with sigils
19:31 Juerd I can't rebute the unicode fear this way.
19:31 Juerd Good thing I did that before this was decided.
19:31 eric256 r0nny...ohh. hmmm. i wonder if you can
19:31 xinming hmm, May I know why the sigil exists?
19:32 xinming :-S
19:32 Juerd eric256: Possibly other syntax, like `` ;)
19:32 Juerd `T`
19:32 Juerd A sigil isn't a bad idea, I think
19:32 PerlJam eric256: how do you know whether or not to bind the type or look for an already existing type then?
19:32 eric256 sub circumfix:<{}> (Hash $x,  $y) ... but i can't remember if circumfix is right or even working yet
19:32 Juerd But anything \w+ cannot be a *sigil*
19:32 eric256 easiest way is to look for tests or examples that override
19:33 eric256 PerlJam:....what?
19:33 Juerd Also, "class" provides false symmetry
19:33 Juerd it also needs % to have a "hash" equivalent
19:33 eric256 yea i'm not a fan of class in this instance
19:33 Qiang Khisanth: yeah i agree with you.
19:33 r0nny eric256: will it work, if i create a class wich does hash, to use it to decide ?
19:33 Juerd And & was used as an example, but you can't use "sub" where you can use "&", so it's a bad example
19:34 eric256 i don't think does Hash works at all yet.
19:34 r0nny so i can use sub circumfix:<{}> (FooHash $x,  $y)?
19:34 PerlJam my T $x;  that suggests that $x is of type T.  my class T $x;  that suggests that T is some as yet unknown type but we're going to call it T for the duration of this scope.
19:34 eric256 r0nny. easiest way is to try it and see
19:34 Juerd PerlJam: For that matter, why not "T"? :)
19:34 Juerd my "T" $x
19:34 PerlJam heh
19:35 Juerd hm, think about this while I drive home
19:35 eric256 PerlJam.. i'd just disagree with that, but yea barewords would have an issue with name collissions
19:35 Juerd Let me know
19:35 PerlJam scare-quotes to the rescue!
19:35 Juerd afk
19:35 Khisanth the T reminds me of C++ :)
19:35 dudley_ has quit IRC ()
19:35 r0nny hmm
19:35 * r0nny tries
19:35 PerlJam Khisanth: yeah, where parameterized types use bracketing constructs, so why not  my <T> $x;  ?
19:36 integral why do you need type sigils if you can declare parameterised functions?
19:36 r0nny btw - how to use the operator of the initial hash ??
19:36 integral sub foo<Bar> (Bar $baz) { ... } !
19:36 PerlJam integral: I have no idea.
19:36 eric256 r0nny yea i think you just can't do that right now
19:36 eric256 but you could do
19:36 PerlJam integral: I'm taking on faith that there's some corner case that isn't covered.
19:37 eric256 %hash.processkey("asdf asdf asdf asd fasd f"); and have your processkey sub do the right thing
19:37 integral PerlJam: hmm, well I haven't seen it mentioned at all, and to me that's not necessarily a sign that anyone on p6l has thought about it outside of their tiny little part of the world
19:37 r0nny eric256: so it will be not-nice :/
19:37 r0nny hmm
19:37 r0nny will have many FIXME's
19:37 hexmode has quit IRC (Remote closed the connection)
19:37 eric256 that is the state of pugs/perl6 right now yes
19:38 PerlJam integral: so suggest this to p6l.
19:38 integral sure
19:39 PerlJam in fact, I thought at one time we'd talked about using my [T] $x to do the job of the class sigil
19:39 integral hmm, to pick a post at random to reply to, or a new thread
19:39 integral I thought perl used [] for parametric classes anyway, not <>?
19:39 PerlJam integral: yes. (this week anyway)
19:39 eric256 r0nny you could always override + for hashes... %x + "asdf asd fas dfas df" sets all those keys.
19:39 eric256 sub infix:<+> (Hash %x, $key) { warn $key }
19:40 integral ok, I'll modify my suggestion to that so that hopefully people don't want time griping about that
19:40 r0nny eric256: i think + is the wrong thing for this
19:40 * eric256 takes it back. don't do that because thats overriding+ for everything, even though its signature is just Hash.. thats annoying
19:41 eric256 r0nny, i think you have the wrong idea.  most of the *right* things, or at least lots of them, are either broken or unimplemented. that means most solutions are just get-it-done solutions, not get-it-done-all-nice-and-pretty
19:41 r0nny :/
19:41 eric256 r0nny you can always make up any operator  you want.   sub infix:<expand>
19:41 eric256 %h expand "asdf asd fasdf ";
19:42 r0nny i will go wit sub processkey and sub deepexists
19:42 r0nny well both will be multi subs
19:42 eric256 they don't need to be muti
19:42 eric256 the only time you need multi is if you are going to have a sub with the same name but a different signature
19:43 r0nny both will be overloaded
19:43 eric256 overloading has nothing to do with it
19:43 r0nny damn - its the wrong word
19:43 * eric256 might be thinking of overloading differently.  either way... two subs with the same name but different signatures need multi
19:44 r0nny eric256: im a c++ programmer
19:44 dduncan has joined #perl6
19:44 r0nny so im used to use overload with functions/methods, too
19:44 r0nny well - as word ;P
19:44 PerlJam r0nny: wow.  Don't admit that in public.  Keep your dirty secrets ... well, secret.
19:44 eric256 lol
19:44 r0nny c++ isnt a dirty secret
19:44 r0nny actually i think its a nice language
19:44 eric256 r0nny...of course it is
19:45 eric256 ack... lol
19:45 PerlJam r0nny: how long have you programmed in C++
19:45 PerlJam ?
19:45 r0nny but the std defining people make it suck more and more :/
19:45 PerlJam r0nny: and how many other languages have you use?
19:45 r0nny PerlJam: years
19:45 PerlJam s/use/used/
19:45 * eric256 used c++ for many years, then found perl...not needing a library to do hashes saves tons of time ;)
19:46 r0nny atm u use perl5 and sh besides of c++
19:46 PerlJam r0nny: well, that explains it.
19:46 eric256 lol
19:46 r0nny c++ has raw speed in some cases where perl just cant compare
19:46 PerlJam r0nny: I think you're thinking of C  ;)
19:47 r0nny PerlJam: no - c++
19:47 r0nny well its not as fast as pure c, but still very fast
19:47 r0nny and lots more of syntactic shugar
19:48 PerlJam r0nny: well, when you've only used one tool, of course that tool is the best tool in the world to you.  Get out and try more tools.
19:48 SamB syntactic nails and glass!
19:48 r0nny oh ;P
19:48 r0nny PerlJam: atm im trying lisp and haskell
19:48 PerlJam r0nny++ excellent choices
19:48 r0nny and perl6 ;P (or at least the parts are implemented)
19:49 PerlJam And if you don't learn to hate lisp, then you're doing something wrong! ;-)
19:49 integral scheme's not that hateful...
19:49 r0nny hehe - i hate it allready - one of my functions eded like this : "))))))))))))))))))))"
19:49 * PerlJam mumbles something about oatmeal and finger nail clippings.
19:50 integral r0nny: your function's too big :-)  factorise it more
19:50 stevan r0nny: then your function was too big
19:50 khaladan has quit IRC (Connection timed out)
19:50 r0nny it was made to eat the mind of peoples reading it
19:50 stevan :D
19:50 r0nny cant archive that wit sewwt small ones
19:51 r0nny sweet ;P
19:51 r0nny lisp has some cool features since its a very dynamic language
19:51 xern has quit IRC ("Lost terminal")
19:51 r0nny but i think perl6 will be a better choice ;P
19:52 PerlJam integral: scheme has a certain beauty to it.  Like that the entire spec and mathematical treatment can be expressed in under 60 pages.
19:52 r0nny nice
19:52 xinming hmm, I wonder, why Larry made the sigil "C" please?
19:52 r0nny hmm
19:52 r0nny sigil ?
19:53 PerlJam r0nny: perl6 is meant for humans, lisp is meant for machines (or weird humans)
19:53 PerlJam ;-)
19:53 PerlJam xinming: for "class"
19:53 r0nny i know some lisp fans
19:53 r0nny they are weird
19:53 integral one nice thing with haskell is that GHC is very much a compiler in the same format as a trad. unix C compiler
19:54 xinming In my understanding, It just do `my $o; try { class $class_name $o } ; ... ` ;
19:54 xinming right?
19:54 xern has joined #perl6
19:54 xinming PerlJam: hmm, I know, I've read the mailing-list. But not understand. :(
19:55 xinming where the "C|"T is becoming something like $class_name in the above example.
19:55 r0nny ??
19:56 xinming I mean the new sigil.
19:56 xinming some suggest C|
19:57 xinming I don't know how to type this. :-P
19:57 xinming `AltGr + C` not works
19:58 r0nny damn :/
19:58 PerlJam xinming: heh, I'm trying to figure out how to type it too.  In vi it would be ^KCt, but that doesn't work too well in irssi
19:58 r0nny using recursive methods in deep hashes is a disaster
19:58 integral In Theory, everyone can type Unicode, transport Unicode safely, and get perl6 released
19:59 justatheory integral: Only because I've had to learn a lot about character encodings lately.
19:59 r0nny is there a way to tell it all sub-hashes are the same class ?
19:59 integral heh.  hello :)
19:59 justatheory Unicode integral to the work I do, now. ;-)
19:59 SamB integral: In Theory, Haskell supports Unicode sources...
19:59 xinming But the fact is, even in Windows, the utf-8 isn't fully supported by default( IIRC )
20:00 justatheory SamB: No haskell in me, please.
20:00 ingy seen autrijus
20:00 jabbot ingy: autrijus was seen 3 hours 34 minutes 34 seconds ago
20:00 PerlJam xinming: anyway, my ¢T $x; says that $x is some type T that's not yet known, as soon as it's known, everywhere in that scope where it said T, is now whatever type was bound to T.
20:00 integral SamB: hmm, I guess real perl6 uses will just do what haskell people have done, and ignore it?
20:01 SamB integral: probably
20:01 SamB because they likely don't have those keys
20:01 integral PerlJam: hmm, so it's not like: my $x; my ::T := ref $x (or $x.class, or something)
20:01 PerlJam xinming: so ...  my Int $i = 5;  sub foo (¢T $x) { say $x }  foo($i)   # Since $i is an Int, so T becomes an Int
20:02 integral so anyone who does people unicode to c6pan will be allowed to, but everyone will mutter about incompatibilities
20:02 integral *post
20:02 PerlJam integral: I'm not sure what it's *really* supposed to be like.  This is just what I think given the discussion so far.
20:02 integral hmm.  Has someone posted a goal they want to use this to achieve?
20:02 integral (like why are non-linear type patterns useful without a unifying type checker?)
20:03 eric256 actualy they are pretty easy on windows.  oddly it seems to do okay with those, i havn't found a progam i can't type them in.  ¥ ¢
20:03 SamB ¢?
20:03 PerlJam integral: another good question (for you to ask p6l  :_
20:03 PerlJam er :)
20:03 SamB eric256: that sounds rather DOS-like!
20:04 eric256 it amuzed me when some of the arguments agianst it where becuase corp environments don't update, if they don't update then they arn't going to have perl6 anyway so it doesn't realy matter
20:04 xinming PerlJam: hmm, I think I understand, thanks.
20:04 integral and: sub foo ($x --> C|foo $y) { $y = C|foo.new } # hmm, so context has to include *full* type info?
20:05 xinming PerlJam: Why not use multi?
20:05 integral hmm, I guess multi is one place they would have an effect, since that effectively does typecheck
20:05 xinming IIRC, there is a type /Any/
20:06 xinming by the way, How do you type this character?
20:06 PerlJam xinming: yeah, but that won't quite do the same thing.  
20:07 PerlJam xinming: I typed it in vi and then cut+paste it here :)
20:07 r0nny damn
20:07 r0nny i got a problem with multi subs
20:07 * PerlJam is secretly hoping that @Larry will reply to his last post with "no,no,no, it works like *this*. and that useful *here* because ..."
20:08 eric256 perlbot nopaste
20:08 perlbot Paste your code here and #<channel> will be able to view it: http://sial.org/pbot/<channel>
20:08 eric256 put your stuff there r0nny....multi subs are a bit picky right now. ;) or maybei should say they arn't picky enough yet
20:10 xinming PerlJam: hmm, Could you please tell where the differences between multi and "C|" ? I really confused, hmm, if there is a type of any, then, multi can handle the case you showed above.
20:10 pasteling "r0nny" at 84.184.177.66 pasted "broke a multi-method - dont ask how" (74 lines, 1.6K) at http://sial.org/pbot/13927
20:11 r0nny oh - its nams processkey
20:15 r0nny btw - how to access a member var ?
20:16 eric256 with the * i've had trouble with multi
20:16 eric256 access member var?  has Str $.sep;   then my $x = new DeepHash; $x.sep = ":";
20:17 eric256 you probably want to start with a single multi and then add on, to see which one breaks it and when.
20:18 eric256 are these suppose to be methods? or subs?
20:18 r0nny methods
20:19 r0nny allready fixed the keyword
20:19 r0nny but it doesnt like them
20:19 eric256 if they are methods, then change the top two to multi method instad of sub, and drop the bottum one out of the class becuase it isn't a class method
20:19 eric256 you can't just access $self with no definition.
20:19 eric256 multi method processkey($self: Str $key) {
20:22 r0nny it breaks on  the first call to the first method on the code below the class decl
20:23 eric256 r0nny....i meant, after adding which multi does it break.  and break needs to be defined. does it give you an error? run the wrong mulit? etc.
20:24 eric256 besides that your first multi is going to be called when you are down to one element, it tries to split that element and then call itself, so you will get an infinite loop
20:25 r0nny eric256: it breaks becasue it cant find a fit
20:25 eric256 i havn't ever seen that error before.
20:26 r0nny *** No compatible subroutine found: "&processkey"
20:29 r0nny -_-
20:29 eric256 i would assume that does Hash isn't working at all
20:29 eric256 not that i see why you need it at all
20:29 r0nny well - time to go for multi-subs
20:30 eric256 yea. thats kinds what we've said since you started. ;)
20:30 eric256 or make info a full object
20:30 eric256 instead of a special hash
20:31 r0nny can i take a hash in the ctor ?
20:31 r0nny hmm and how to make a ctor
20:32 r0nny well - op = will do
20:34 shift8 has joined #perl6
20:37 eric256 btw the way you changed deepexists makes it fail
20:39 shift8 (slightly OT) anyone have experience w/ Text::Template?
20:41 shift8 has quit IRC ("Leaving")
20:44 r0nny eric256: i know
20:44 r0nny eric256: weill becone a method now ;P
20:49 shift8 has joined #perl6
20:49 eric256 still will be broken. you changed the order of the checks and it was important.  if you deepexist with only one element it always returns true even if that element doesn't exist
20:53 r0nny really ?
20:53 r0nny need to repair then
20:53 r0nny i will geuss on my own
20:53 eric256 yea because you check +@a first. but it will be empty for single element. ;)
20:54 r0nny hmm
20:55 r0nny if i make a method allways will return a bool, will it automatically convert the return value ?
20:56 eric256 no
20:56 r0nny how to convert to bool best ?
20:58 Qiang has quit IRC ("Leaving")
21:01 kolibrie has quit IRC ("leaving")
21:04 r0nny darn
21:05 r0nny class Foo { ... }; wont work :/
21:06 larsen has joined #perl6
21:08 xinming has quit IRC (Remote closed the connection)
21:27 putter has joined #perl6
21:29 putter stevan: ping
21:37 oylenshpeegul has joined #perl6
21:50 oylenshpeegul has left
21:55 iblechbot has quit IRC (Connection timed out)
21:56 Amnesiac has joined #perl6
22:02 nappy754 has joined #perl6
22:06 Steve_p has quit IRC ("OpenBsd 3.8 here I come")
22:07 elmex has quit IRC ("leaving")
22:07 Juerd I hate these "does he know that I know he knows that I know <foo>" questions I ask myself sometimes.
22:07 Liz reminds me of that Dutch poet who wrote:
22:08 Liz I wish I were two doggies, then I could play together outside...  ;-)
22:08 Juerd I like that one.
22:09 Juerd I do wonder why you associate that with the qusetion of knowing whether someone knows you know they know you know something...
22:10 Juerd s/qusetion/question/
22:10 Liz You were asking yourself...
22:11 Juerd Oh
22:11 lisppaste3 putter pasted "stevan: re object space" at http://paste.lisp.org/display/12866
22:11 Juerd You never consult yourself for advice? :)
22:11 Liz nope, I know what I need to know already  ;-)
22:11 Juerd Hmm
22:11 Juerd Interesting :)
22:11 Juerd And potentially very useful.
22:11 Juerd Could you teach me? :)
22:12 Liz It comes with age...
22:12 Liz ;-)
22:12 r0nny has quit IRC ("sleep")
22:13 Juerd Physical age?
22:13 Liz hmmm...  maybe I should say "experience" ?
22:13 Juerd I don't know :)
22:13 Juerd Were you asking yourself, by the way? :)
22:14 Juerd It should be obvious that I have no idea... ;)
22:14 Liz Well, actually you're right, I do consult myself occasionally
22:15 robkinyon has quit IRC (Remote closed the connection)
22:15 Liz I just don't consult myself on whether or not I should be consulting myself
22:15 Juerd Perhaps you should
22:15 Liz ;-)
22:15 Juerd Though the recursion really bites after a while.
22:16 Liz Indeed...  thermal overload
22:16 Juerd Only if you're hot headed :P
22:16 Juerd I try to use air cooling, into one ear, out the other
22:17 Juerd This has some unpleasant side effects though
22:17 Liz is there a wire between your ears?
22:18 luqui has joined #perl6
22:18 Juerd No
22:18 luqui Yes!
22:18 Juerd Just emptiness, to let the air flow along, and a radiator.
22:19 Liz wow!
22:19 Juerd luqui: You will not guess what we are talking about :)
22:19 Liz This introspection business is really great!
22:19 Juerd Liz: And you don't even need the little light!
22:21 Liz perhaps not...
22:21 Liz anyways...  enough of this banter  ;)
22:21 Juerd Banter?
22:21 Liz ;-)
22:22 Juerd This happens every time I try to discuss the more important world issues.
22:23 Liz That's experience...  ;-)
22:23 Juerd Ahhh
22:23 Juerd It must be, because I didn't ask myself what it was.
22:23 Liz I'm going to get some shuteye... its going to be a long to 2morrow...
22:23 Juerd Good to have learned something today. Thanks, Liz
22:23 Juerd slaap lekker
22:23 Liz you're welcome...  truste!
22:24 Juerd I'm also going to bed
22:24 Juerd Bye
22:24 Liz has quit IRC ("User disconnected")
22:25 nappy754 has quit IRC (Remote closed the connection)
22:30 lisppaste3 has quit IRC ("Want lisppaste3 in your channel? Email chandler AT unmutual.info.")
22:30 lisppaste3 has joined #perl6
22:38 nappy754 has joined #perl6
22:40 putter what was the pastebot before lisp.org?  (I'm hitting length limits...)
22:41 putter ah well.
22:41 integral http://sial.org/pbot/perl6
22:42 putter ah, thanks!
22:42 integral changes topic to: 6.2.10 released! http://xrl.us/hxnb | geoffb's column (/. ed): http://xrl.us/hxhk | pugscode.org | pugs.kwiki.org | paste: http://paste.lisp.org/new/perl6 or http://sial.org/pbot/perl6 | www.geeksunite.net
22:43 putter integral++
22:43 pasteling "putter" at 66.30.119.55 pasted "stevan, fglock: some random overhaul code" (3463 lines, 110K) at http://sial.org/pbot/13931
22:44 buu ?eval my $ x = 1;
22:44 evalbot_7740 Error:  unexpected " " expecting "^", "*", "?", ".", ":" or qualified identifier
22:44 buu NOOOOOOOOOOO.
22:45 theorbtwo buu: Whaa?
22:45 putter collective fyi. sial.org is accepting pastes more than 3x the size of lisp.org.  (Whether such a paste is a good idea is a separate issue...;)
22:45 theorbtwo The $ is part of the name.
22:45 buu theorbtwo: So?
22:45 buu I'm sad they don't allow it any more =[
22:46 theorbtwo I think it's a good thing that it's not allowed.
22:46 buu =[
22:46 buu But it's cool.
22:46 Khisanth putter: at the very least that is useful :)
22:46 theorbtwo The variable is named $x, not x with some sort of strange thingy in front of it.
22:47 buu theorbtwo: This is just another way that p5 is cooler than p6.
22:47 integral perl6 devalues the primary sigil way too much
22:47 theorbtwo Huh, integral?  It values it much more (but mostly not in ways that are user-visible, at least for normal users).
22:48 integral no, it makes it little more than a part of the variable name
22:48 buu integral: So I'm not the only one who hates %hash<ele> ?
22:48 theorbtwo The glob is gone; $x and @x are unrelated varables that happen to have similar names.
22:48 buu ?eval my $x; $x<a>
22:48 evalbot_7740 \undef
22:48 buu ?eval my $x; $x<a>=32;
22:48 evalbot_7740 \32
22:48 putter Khisanth: sial accepting large pastes?
22:48 buu ?eval my %x; %x<a>=32;
22:48 evalbot_7740 \32
22:48 buu What's the difference?
22:48 Khisanth putter: yes
22:48 buu What's named with a dollar sign and one isn't?
22:48 putter agreed
22:49 integral interpolation/context is the answer I got once to buu's question
22:49 buu Context?
22:49 buu And can I have my decrement operator now?
22:49 integral $a = ...; %a = ...; the yadas have different context.
22:49 theorbtwo $x-- ?  Sure, never changed.
22:49 buu integral: That makes a certain amount of sense
22:49 buu ?eval my $x = "b"; $x--; $x;
22:49 evalbot_7740 \-1.0
22:49 buu theorbtwo: It lacks goodness.
22:50 theorbtwo Oh.
22:50 theorbtwo Bring it up on p6l?
22:50 buu I suppose.
22:50 theorbtwo AFAIK, magic inc and dec have never been mentioned in the [AES]\d+es.
22:51 saorge has quit IRC (Read error: 110 (Connection timed out))
22:51 Eimi has quit IRC ("Leaving")
22:51 luqui buu, the sigils don't mean too much anymore, but they do mean something
22:51 saorge has joined #perl6
22:51 luqui I've been looking for a good meaning for them, and haven't found one
22:52 fglock__ putter: I started reading the new code - looks quite good (will finish later)
22:52 luqui so basically, @a flattens itself in list context and $a doesn't.  That's pretty much the only difference
22:52 buu Hmph.
22:52 integral don't we also still have @ as a prefix operator too?
22:52 luqui still?
22:52 luqui oh, like @$x
22:53 integral yep
22:53 luqui to my dismay, I think so
22:53 buu ?eval my $x; $x[1]; @$x
22:53 evalbot_7740 [undef, undef]
22:53 integral if it's really just denoting context, we should just have one x,  and have $x and @x differ purely on context, but be the same array
22:53 buu I agree.
22:53 buu Or something.
22:54 fglock__ has left "Fui embora"
22:54 integral golly, and I was griping about CL's splitting of func and var namespaces earlier too :-/
22:54 shift8 ?eval my $test = "no way!"; print $test;
22:54 evalbot_7740 no way!bool::true
22:55 shift8 (sorry all - great feature!)
22:55 buu ?eval:p5 %x
22:55 evalbot_7740 Can't locate Scriptalicious.pm in @INC (@INC contains: /home/fibonaci/devel/pugs-base/perl5/PIL-Run/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/.. /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Value/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-Container/lib /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel /home/fibonaci/devel/pugs-base/perl5/PIL-Run/../Perl6-MetaModel/lib /etc/perl /usr/local/lib/perl/5.8.7
22:55 integral o_O
22:56 shift8 m()m
22:56 integral &
22:56 theorbtwo ?eval eval :p5 "die 'foo'"
22:56 evalbot_7740 undef
22:57 buu eval: print "?eval say 1"
22:58 buubot buu: ?eval say 1Return: 1
22:58 buu hrm
22:58 buu Shux.
22:58 shift8 integral; &?  nothing more.  just an indication of repect.
22:59 buu shift8: invalid bash syntax.
23:00 Khisanth buu: does buubot have evalbot on ignore?
23:00 buu No =]
23:00 buu ?eval say "eval: 1 "
23:00 evalbot_7740 eval: 1  bool::true
23:00 buubot evalbot_7740: Error: syntax error at (eval 109) line 2, near "1  bool::true  "  
23:00 buu heh
23:00 * buu ponders
23:01 dduncan has left
23:01 mrborisguy infinite loop from buubot and evalbot?
23:01 mrborisguy that'd bef un
23:01 mrborisguy s/bef un/be fun
23:01 buu Naw, buubot can't trigger evalbot
23:01 dduncan has joined #perl6
23:02 theorbtwo I think that'd be less then fun, unless buu stopped buubot fairly quickly.
23:02 nappy754 has quit IRC (Remote closed the connection)
23:02 buu At least, I can't think of anyway to make it trigger
23:02 buu Since it automatically prepends the person who triggered it to it's response
23:02 eric256 has left
23:02 mrborisguy that's true, since buubot puts "the name:" in front anyway
23:02 luqui ?eval say "?eval say 'hi';"
23:03 evalbot_7740 ?eval say 'hi'; bool::true
23:03 luqui hmm, it doesn't listen to itself?
23:03 Khisanth it has no reason to :)
23:03 buu buubot: reload
23:03 buubot buu: Reloading!
23:03 buu ?eval say "eval: 1"
23:03 evalbot_7740 eval: 1 bool::true
23:03 buu eval: 1
23:03 buubot buu: Return: 1
23:05 Blicero has quit IRC (Remote closed the connection)
23:05 luqui eval: my %foo;  %foo<hello> = 'world';  %foo.kv
23:05 buubot luqui: Error: syntax error at (eval 109) line 1, near "hello>"  
23:05 luqui oh, buubot is not perl 6
23:06 luqui eval: dump
23:07 putter fglock__:  hi.  sorry for the delay (phone call).  there was something else I was going to mention... eh, can't remember.
23:07 buu luqui: You killed it =[
23:08 luqui :-(
23:08 buu I know.
23:08 buu Fortunately it respawns.
23:08 buu eval: "I'm not dead yet!"
23:08 buubot buu: Return: I'm not dead yet!
23:08 luqui heh
23:08 theorbtwo evalbot_7740: `cwd`
23:08 theorbtwo irssi--
23:09 buu Irssi? How is that irssi's fault?
23:09 luqui eval: `pwd`
23:09 buubot luqui: Return:  
23:09 putter fglock__: my fuzzy recollection is the main changes to P6::V/C code were - cleaned up method inheritance (methods which do nothing but call nextmethod, or which should be interited from others, can be deleted);
23:09 theorbtwo buu: irssi tends to autocomplete things I didn't want it to when I begin a line with \w+[:,]
23:10 buu um
23:10 buu No it doesn't.
23:10 theorbtwo evalbot_7740: `cd`
23:10 theorbtwo Yes, it does; what I typed was "eval: `cd`"
23:10 buu um
23:10 buu I've never ever seen irssi do that
23:10 buu Are you sure you aren't on bitchx?
23:11 Khisanth theorbtwo: perhaps you have a certain script loaded
23:11 Jooon /set completion_auto off
23:11 theorbtwo eval: 1
23:11 buubot theorbtwo: Return: 1
23:11 theorbtwo Thanks, Joon -- just a moment too slow, though
23:12 theorbtwo eval: `cd`
23:12 buubot theorbtwo: Return:  
23:12 theorbtwo eval: $ENV{USER}
23:12 buubot theorbtwo: Return:  
23:12 theorbtwo Hm, perhaps it is better locked-down then I'd thought.
23:12 theorbtwo buu: Yes, I am quite certian I am not on bitchx.
23:13 putter unpacked the monster "aUTOLOAD implements a half dozen methods" cruft; switched to a more p6-like mm sugar; let's see...
23:16 luqui ?eval ()
23:16 evalbot_7740 pugs: Prelude.tail: empty list
23:16 luqui woah
23:16 luqui :q
23:18 putter tightened up some of the code (either to remove what seemed to be development history artifacts, do some simple clearer-and-faster changes, or just reduce code size so I could more easily see what was going on);  switch to using runtime/mm for object creation and isa()/does(); ... maybe that's it.
23:19 luqui ?eval []
23:19 evalbot_7740 pugs: Prelude.tail: empty list
23:19 luqui ?eval my $x = [];  say "good";  $x
23:19 evalbot_7740 good pugs: Prelude.tail: empty list
23:21 shift8 quick question:  what is now considered "The Right Thing To Use" for templating ?
23:24 Juerd Template Toolkit!
23:24 shift8 although, like everyone i presume, i've implemented a "mini language" one, i dislike that route.  Text::Template is not meeting my needs.  is there any other suggestions?  (i know this is #perl6, but it's my best bet on freenode ;)
23:24 Juerd Templates without mini language?
23:24 Juerd Then you end up with pure perl.
23:24 Juerd Which is doable, but hard to teach newbies
23:24 shift8 yes... like Text::Template
23:25 Juerd afk  # .sleep(:really(bool::true))
23:26 putter The next step was to clean up and simplify the Containers and their helpers and Code (I was finding it hard, for example, to diagnose Code+Array bugs).  When encountering a particular flavor of bugs, the right response is to house clean, rather than chase them.  My impression is P6::V/C has reached a point of "throw one away".  Ie, "now we now what we are doing, and the things we depend on have matured, so now it can all be done much mo
23:26 putter re simply".
23:28 putter fglock__: ah well.  that's what comes to mind at the moment.  let me know if you have any questions/comments/whatever.  cheers.
23:28 putter &
23:28 putter has quit IRC ("Leaving")
23:31 larsen has quit IRC ("later")
23:34 rafl Jooon, spinclad, gaal: ping
23:36 putter has joined #perl6
23:36 putter s/The next step was to clean up/The next step was to have been to clean up/
23:43 putter Btw, the big picture was I took the pilrun p5 and p6 prelude files, and all the Perl6::Value and Perl6::Container files, and poured them into two big combined files, in the hope they could be distilled down into one.  with regexps to rip it all apart again at runtime.  basically, it all needed overhaul, but much of the changes would be non-local, so it seemed easiest to localize all the code.  basically trying to get from a succesful sp
23:43 putter ike (we have a p5 backend that actually does stuff), to something which can be a real p6 platform.

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo