Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2009-02-02

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:29 jhuni joined #perl6
00:41 eternaleye joined #perl6
00:43 spx2 joined #perl6
00:48 spx2_ joined #perl6
01:01 |jedai| joined #perl6
01:02 spx2 joined #perl6
01:21 mj41 joined #perl6
01:27 |jedai| joined #perl6
01:27 spx2_ joined #perl6
01:37 Whiteknight rakudo: my $x = { say 5; }; $x();
01:37 p6eval rakudo 36267: OUTPUT«5␤»
01:37 Whiteknight rakudo: my $x = -> $a { say $a; }; $x("hello world");
01:37 p6eval rakudo 36267: OUTPUT«hello world␤»
01:38 Whiteknight always happy when my code examples not only parse, but actually produce the right output
01:40 shinobi-cl_ joined #perl6
01:59 c1sung joined #perl6
02:02 jfredett joined #perl6
02:02 shinobi-cl_ perl6: my %Example ; push %Example{"a"}, 1, 2, 3; say %Example{"a"};
02:02 p6eval pugs: OUTPUT«1 2 3␤»
02:02 p6eval ..rakudo 36267: OUTPUT«No applicable methods.␤␤current instr.: '_block14' pc 104 (EVAL_16:50)␤»
02:02 p6eval ..elf 25169: OUTPUT«␤»
02:03 alester joined #perl6
02:12 |jedai| joined #perl6
02:19 s1n hmm, does rakudo from github currently build for anyone else?
02:19 s1n i'm getting PARROT_FLOATVAL_INF_NEGATIVE and POSITIVE errors and don't see it defined in parrot anymore
02:28 aindilis joined #perl6
02:36 s1n n/m, outdated repo
02:38 spx2 joined #perl6
02:43 TheOrz joined #perl6
03:01 spx2 joined #perl6
03:22 spx2 joined #perl6
03:28 pugs_svn r25170 | diakopter++ | deleting cruft.
03:37 justatheory joined #perl6
03:38 justatheory joined #perl6
03:51 |jedai| joined #perl6
03:53 mberends joined #perl6
03:55 meppuru joined #perl6
04:15 Sepheebear joined #perl6
04:21 spx2_ joined #perl6
04:36 spx2 joined #perl6
04:48 alester joined #perl6
04:52 |jedai| joined #perl6
04:55 |jedai| joined #perl6
05:11 scrottie joined #perl6
05:16 Tene_ joined #perl6
05:26 s1n moritz_: i think you can close rt #59372, i can't seem to reproduce it anymore
05:27 justatheory joined #perl6
05:32 s1n moritz_: while you're at it, close rt #60510 too
05:43 spx2 joined #perl6
05:49 spx2_ joined #perl6
05:50 mberends is anyone with OS X lurking here right now?
05:56 jrockway joined #perl6
05:59 mberends on OSX, what does "perl6 -e 'say %*VM<config><osname>'" print? osx?
06:02 |jedai| joined #perl6
06:04 maerzhase joined #perl6
06:05 |jedai| joined #perl6
06:09 TheOrz joined #perl6
06:34 spx2 joined #perl6
06:52 moritz_ s1n: closed, thanks
06:56 |jedai| joined #perl6
06:57 Sepheebear joined #perl6
07:10 ayrnieu joined #perl6
07:14 maerzhase joined #perl6
07:27 eternaleye_ joined #perl6
07:33 iblechbot joined #perl6
07:35 |jedai| joined #perl6
07:41 eternaleye joined #perl6
08:03 timebomb joined #perl6
08:05 DemoFreak joined #perl6
08:07 ayrnieu joined #perl6
08:18 ashizawa joined #perl6
08:26 szabgab_ I saw in the docs the way to open directory is   S16-io.pod:  my $dir = IO::Dir::open('.');
08:27 szabgab_ has this been implemented in Rakudo ?
08:29 szabgab_ in other words, how can I list all the files in a directory ?
08:30 moritz_ I'd be surprised if you could
08:31 moritz_ run('ls > tempfile');
08:32 szabgab_ tx
08:38 szabgab_ regarding open() for files, is the normal behavior now to throw an exception if it cannot open the file?
08:39 moritz_ the specced behaviour is to fail(), which throws an exception only if 'use fatal;' is in scope
09:02 ejs joined #perl6
09:10 maerzhase joined #perl6
09:11 |jedai| joined #perl6
09:11 szabgab_ std: my $line = "bla"; if ($line ~~ /x/ ff $line ~~ /y/) {say "in" }
09:12 p6eval std 25170: OUTPUT«00:04 34m␤»
09:12 szabgab_ rakudo: my $line = "bla"; if ($line ~~ /x/ ff $line ~~ /y/) {say "in" }
09:12 p6eval rakudo 36276: OUTPUT«Statement not terminated properly at line 1, near "ff $line ~"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
09:14 maerzhase1 joined #perl6
09:16 maerzhase1 joined #perl6
09:37 |jedai| joined #perl6
09:43 maerzhase joined #perl6
09:46 meppl joined #perl6
10:12 masak joined #perl6
10:25 maerzhase joined #perl6
10:37 yves joined #perl6
10:39 mberends masak: hi
10:39 masak mberends: greetings
10:40 moritz_ hi there
10:40 masak all the people starting with 'm' are here.
10:40 masak where are the others?
10:41 mberends Hmmm
10:41 mberends what does your computer say for perl6 -e'say %*VM<config><osname>' ?
10:41 mberends is it osx ?
10:41 Matt-W Is it just that people beginning with M are more likely to be in Europe?
10:42 masak mberends: darwin
10:43 moritz_ masak: yesterday night TimToady tried to @tell you something, but ENOλBOT
10:43 masak mberends: are you slipping into the dangerous territory of OS detection? :)
10:43 masak moritz_: ok, will backlog.
10:43 mberends thanks, that goes into my Makefile.p6
10:43 masak mberends: just make sure the fallback is sensible. :)
10:44 Matt-W surely the default os is windows 95 :)
10:44 mberends the fallback is my local OS :)
10:45 masak lunch &
10:48 szabgab_ masak: when you are back from lunch, how should I reuse parts of November in some other application ?
10:48 szabgab_ right now I am writing a static web site so I'd like to reuse HTML::Template
10:49 szabgab_ for now I start by copy-ing the required files from the November source tree to my directory
10:49 xinming_ joined #perl6
10:51 masak szabgab_: not technically back from lunch, but short reply:
10:52 masak HTML::Template is being obsoleted in the long run. we're starting a Web project, whose objective is re-use. in the meantime, feel free to use whatever you want. any feedback to prevent code drift and increase collaboration is appreciated.
10:52 * masak recloaks
10:54 szabgab_ and why do you drop HTML::Template ?
10:54 moritz_ if "they" drop it, I'll adopt it
10:55 szabgab_ can modules already loaded from other directories (is @INC already in place ?) or only from . ?
10:57 moritz_ you can BEGIN { @*INC.push('dir') }
10:57 szabgab_ oh and just to catch up with the changes in last week, I see the parrot SVN repo moved, what about Rakudo, is it still in the languages/perl6 subdir of parrot ?
10:58 szabgab_ moritz_: oh, that's better than copying the files
10:58 szabgab_ I could not compile Dispatcher.pm
10:59 moritz_ szabgab_: it has moved to git (or so I think), but there hasn't been an announcement yet
11:00 Maghnus joined #perl6
11:03 |Jedai| joined #perl6
11:06 hercynium joined #perl6
11:15 maerzhase joined #perl6
11:19 masak szabgab_: we'll drop HTML::Template because November is XML-based, so a text-based templating system feels both dangerous and wasteful.
11:19 iblechbot joined #perl6
11:24 masak TimToady: re a class lying to itself: maybe I'm using 'handles' for the wrong thing then. I have a class hierarchy in which plenty of delegation is made to a central class, and I found it made things a lot shorter and more convenient to use 'handles'. however, only the class itself uses those delegated methods, and it even feels a bit odd to be exposing them. hence the question about 'handles' and private methods.
11:30 moritz_ if you use 'handles' excessively, can't you use roles + composition instead?
11:31 masak moritz_: no, because part of the reason I use it is so that I can have separation of concerns.
11:31 masak moritz_: https://github.com/masak/druid/
11:32 masak Druid::Game is in the middle, Druid::View and Druid::Player depend on it.
11:32 masak but both the latter are interested in things like board size and the position of pieces, information found in the former.
11:33 masak uh, I'm consciously blurring the object/class distinction in the above, as is often done.
11:35 Jedai joined #perl6
11:35 masak anyway, doing roles+composition would be counter to the idea that Druid::Game holds all the important information (model), while surrounding classes (views) query it for things.
11:43 szabgab_ is there an .ini or .conf file reader implemented in Perl 6 ?
11:44 masak szabgab_: not to my knowledge.
11:44 masak we usually take the eval(slurp) route.
11:44 Matt-W eww
11:44 masak suboptimal, but fast.
11:45 Matt-W that's not very safe
11:45 masak Matt-W: that's correct.
11:45 masak it's generally advised against.
11:45 Matt-W a handy stopgap until you implement something better though
11:45 szabgab_ but then you assume the file format to be perl 6 , not the "standard" [section]   and key = value lines
11:45 masak we figure that we're allowed to make such transgressions if we put up big warning signs and produce working code :)
11:45 Matt-W lol
11:45 Matt-W and replace it with a proper configuration file later?
11:45 masak aye.
11:46 Matt-W you're forgiven
11:46 masak szabgab_: right.
11:46 masak szabgab_: writing a grammar for that ought to be a pleasant task.
11:46 szabgab_ oh, so now I should learn how to that, :-),
11:46 szabgab_ lots of yaks to be shaven here
11:48 masak szabgab_: if you're here for an hour or so, I can walk you through it.
11:48 masak it should be really easy.
11:48 masak that's the nice thing about Perl 6 grammars.
11:49 masak rakudo: grammar Conf { regex TOP { <heading> }; regex heading { ^^ '[' <ident> ']' $$ }; regex ident { \w+ } }; say '[test]' ~~ /<Conf::TOP>/
11:49 p6eval rakudo 36277: OUTPUT«[test]␤»
11:50 masak szabgab_: there you go, just to get you started :)
11:50 masak now, go forth and spread the Perl 6 joy in the world.
11:51 szabgab_ masak: I want my code work first and start creating the web site
11:52 szabgab_ and I'd like to make it happen today
11:52 szabgab_ I can always improve tomorrow :-)
11:52 masak szabgab_: go with eval(slurp) so far, then.
11:52 masak or no conf file at all.
12:05 pugs_svn r25171 | masak++ | [S12] fixed minor typo in class name
12:05 pugs_svn r25172 | masak++ | [S16] added :async flag to getc
12:05 Matt-W masak: I spent a while the other day going mad trying to write a grammar
12:06 Matt-W masak: I may need to ask many questions about it at some point
12:06 masak Matt-W: I'd be glad to help.
12:06 masak Matt-W: writing grammars is supposed to be rewarding, not frustrating. you're doing something wrong. :P
12:07 Matt-W I know!
12:07 Matt-W I think it's just some basic understanding about whitespace handling and things like that
12:07 Matt-W if I can get my head around all that, I should be fine
12:07 masak aye.
12:07 Matt-W Then I shall just kick myself over using _something_ to denote emphasis instead of some sort of matching brackets
12:07 masak to be honest, I haven't grokked that either.
12:08 Matt-W lunch &
12:16 donaldh joined #perl6
12:21 jnthn hi from Bulgaria
12:21 |Jedai| joined #perl6
12:21 masak jnthn: say 'hi' back :)
12:24 jnthn :)
12:24 jnthn Flying home tomorrow.
12:24 jnthn And then it'll be back to Rakudo hacking. :-)
12:27 masak jnthn: it will be nice to have you back. when both you and pmichaud are absent, reality feels strangely hollow and unfulfilling.
12:28 masak plus, I don't know who to complain to. :P
12:29 jnthn RT? ;-)
12:30 masak jnthn: come on, I've done that to death. :P
12:31 masak I think I have 150 open bugs or some such reported by me.
12:31 masak new/open
12:31 jnthn That used to be our entire queue size! :-|
12:31 masak aye.
12:31 * jnthn will try to reduce the bug count using his Rakudo Days.
12:32 maerzhase1 joined #perl6
12:32 masak jnthn++
12:32 jnthn OK, so what's the git equivalent to svn up?
12:32 jnthn git pull?
12:33 masak jnthn: basically, yes.
12:33 masak seems I have only 114 new/open bugs. :)
12:33 masak jnthn: pull = fetch + merge
12:34 masak http://git.or.cz/gitwiki/GitFaq#head-47​40bcc3bab81a8b3f11258fdab5d7c853acf5a7
12:34 |Jedai| joined #perl6
12:36 maerzhase joined #perl6
12:38 maerzhase joined #perl6
12:42 maerzhase joined #perl6
12:43 maerzhase1 joined #perl6
12:47 maerzhase1 joined #perl6
12:48 maerzhase joined #perl6
12:57 _Jedai_ joined #perl6
13:07 TheOrz joined #perl6
13:12 mberends masak: down the slippery slope, what does darwin show for 'ps o args' ?
13:15 masak mberends: it shows something I don't understand. what is it I'm looking at?
13:17 mberends the 'o' option of ps does user defined columns of output. On GNU/Linux it gives 'COMMAND'
13:18 mberends it's a little used section of 'man ps'
13:18 masak I see.
13:18 masak well, it does work here.
13:18 masak but it shows 'ARGS' and then some empty lines.
13:19 mberends oh dear, empty lines not good
13:19 masak they seem to correspond to the command '-bash' in full ps.
13:20 mberends that's not the same as GNU. I'm trying to replace your Perl 5 scaffolding with Perl 6, and make it autodetect differences between your OS and mine.
13:21 mberends it looks like 'o' is a valid option for you, but what follows must differ.
13:22 masak aye.
13:23 mberends could you peruse 'man ps' for me and summarize the available arguments after the 'o' option? I would like to read the full command line, including its arguments. The purpose is to detect the paths used by parrot and perl6.pbc
13:24 |Jedai| joined #perl6
13:25 masak mberends: http://gist.github.com/56912
13:27 mberends brilliant, standby for my work in progress
13:29 mberends masak: http://gist.github.com/56913
13:30 ruoso joined #perl6
13:35 _Jedai_ joined #perl6
13:36 masak mberends: nice.
13:36 masak for those first debug statements, you might want to consider a 'given %*VM<config> { ... }' block
13:37 masak oh! you did that later, I see.
13:38 mberends yes, the first lines are deprecated, only retained for possible copy/pasting
13:38 mberends you have to start somewhere...
13:38 masak mberends: why do you need to find yourself in the ps listing?
13:39 mberends 13:23 : to get the possibly independent parrot and perl6.pbc directories
13:39 masak not sure I understand.
13:39 masak ps shows processes, not directories :)
13:40 masak just get people to set the damn $PARROT_DIR env variable. :)
13:40 mberends ah, but the right ps incantation will show the executable filenames and args. Works a treat on Linux.
13:40 masak $ perl6 -e 'say @*ARGS' 1 2 3
13:40 masak 123
13:41 masak wot's wrong with dat?
13:42 masak mberends: no, I confess to being utterly confused as to what information it is you are trying to extract from ps.
13:42 literal masak: nothing?
13:42 masak literal: exactly!
13:43 mberends this approach was: don't give people unnecessary work. bearing in mind that some people want to use 'perl6' and others 'parrot perl6.pbc', let them choose and then detect how the Makefile.p6 was invoked.
13:43 literal does it put spaces between the numbers if you do say ~@*ARGS ?
13:43 masak literal: yep.
13:43 literal cool
13:43 * jnthn wondes why they'd do different things...
13:44 masak mberends: I'd say detect as much as possible (via $PATH and $PARROT_DIR or whatever) but don't go overboard trying to please.
13:44 masak sometimes what the user needs is a good slap-around and a message telling them to set their environment variables.
13:44 mberends it already works for me, I was just trying to get it to work for you too.
13:45 mberends being a good neighbour an' all that
13:45 masak yes, but why do you have to mix 'ps' into it all?
13:45 masak I just don't get it.
13:45 masak I mean, given that ps apparently exists in a multitude of flavours.
13:46 masak oh well, I guess we just have different priorities. you seem to be willing to sacrifice platform independence for that little extra service.
13:50 |Jedai| joined #perl6
13:58 ruoso masak, I've just replied to your last commit message...
13:58 masak ruoso: the :async one?
13:58 ruoso yes
13:58 * masak reads
13:59 ruoso I agree with leon that it should be called non-blocking instead
13:59 ruoso but more importantly, I think being blocking-or-not is an attribute of the IO object,
13:59 ruoso not a parameter of that specific io operation
13:59 masak yes, I agree with you both about the name.
14:00 masak ruoso: my main objective was to make sure the functionality is in S16.
14:00 masak I'm not particularly attached to the name of the parameter or to its exact location.
14:00 ruoso so, role IO { has $.blocking; }
14:01 masak has Bool $.blocking;
14:02 masak anyway, I don't have the time to make elaborate changes right now. feel free to change it any way you please.
14:04 * ruoso out-of-time
14:05 masak I'm sure someone will take pity on the change sooner or later.
14:06 meppl joined #perl6
14:11 DemoFreak joined #perl6
14:15 |Jedai| joined #perl6
14:20 maerzhase joined #perl6
14:21 maerzhase joined #perl6
14:21 spx2 joined #perl6
14:24 szabgab_ std: my %h; %h<a><b> = 42;
14:24 p6eval std 25172: OUTPUT«00:02 33m␤»
14:24 szabgab_ rakudo: my %h; %h<a><b> = 42;
14:24 p6eval rakudo 36282: OUTPUT«Method 'postcircumfix:{ }' not found for invocant of class 'Failure'␤current instr.: 'postcircumfix:{ }' pc 3633 (src/classes/Associative.pir:77)␤»
14:24 szabgab_ rakudo: my %h{Str;Str}; %h<a><b> = 42;
14:25 p6eval rakudo 36282: OUTPUT«Statement not terminated properly at line 1, near "{Str;Str};"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
14:25 masak szabgab_: autovivification not implemented yet in Rakudo.
14:25 masak neither are chained-dimensional indices.
14:25 szabgab_ rakudo: my %h; %h<a> = {}; %h<a><b>= 42;
14:25 p6eval rakudo 36282: RESULT«42»
14:25 moritz_ or typed arrays/hashes
14:26 szabgab_ so is this the way to use multi dimensianl hashes ?
14:26 masak moritz_: not even typed arrays? I thought they were just broken.
14:26 masak szabgab_: aye.
14:27 szabgab_ rakudo: my %h; %h<a> = {}; %h{"a"}{"b"}= 42;
14:27 p6eval rakudo 36282: RESULT«42»
14:27 szabgab_ rakudo: my %h; %h<a> = {}; %h{"a"; "b"}= 42;
14:28 p6eval rakudo 36282: OUTPUT«Statement not terminated properly at line 1, near "{\"a\"; \"b\"}"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
14:28 szabgab_ is the {;} syntax only used for declaration or also for the indexing ?
14:28 masak szabgab_: no, really not implemented yet. :)
14:28 masak szabgab_: I think for both.
14:29 szabgab_ std: my %h; %h<a> = {}; %h{"a"; "b"}= 42;
14:29 p6eval std 25172: OUTPUT«00:04 33m␤»
14:29 maerzhase1 joined #perl6
14:32 spx2_ joined #perl6
14:37 pmurias joined #perl6
14:38 pmurias ruoso: hi
14:38 ruoso hi pmurias
14:40 spx2__ joined #perl6
14:43 pmurias ruoso: i'm thinking if i would be worth to make the $scope register special so we could have &?ROUTINE.lexical withought having to call .set_lexical
14:44 pmurias s/thinking/considering
14:45 ruoso pmurias, I'd argue to keep the set_lexical specially and make $scope special in m0ld code
14:46 spx2 joined #perl6
14:46 szabgab_ how do I compare two deep data structures if they contain the same data ?
14:47 szabgab_ shouldn't ~~ do the job ?
14:47 masak Test.pm's is_deeply does the job.
14:47 masak not sure if ~~ is the right tool for that.
14:48 szabgab_ hmm, I missed is_deeply when searched for it a few minutes ago, thanks
14:49 szabgab_ actually there should be some tool without involving Test.pm  as well
14:50 szabgab_ but for now I need it in a test anyway
14:51 ruoso szabgab_, in theory eqv should do it
14:51 Exodist joined #perl6
14:52 pmurias ruoso: the $scope register would be special by convention rather then syntactially, .set_lexical would also be kept
14:52 szabgab_ is_deeply(%data, {"names" => {"fname" => "foo", "lname" => "bar"}}, $file);
14:52 ruoso pmurias, I mean it would be treated by the compiler specially...
14:52 szabgab_ says they are not the same, even though if I %data.perl.say
14:52 szabgab_ it shows the same thing
14:53 ruoso pmurias, so you don't need to mess with the register count
14:53 masak szabgab_: please submit a rakudobug for that.
14:53 pmurias ruoso: it's already treated specially by the compiler
14:53 ruoso I mean the m0ld compiler, not mildew
14:54 szabgab_ I have BEGIN { @*INC.push("{%*ENV<PARROT_DIR>}/languages/perl6/") } in my test file
14:55 szabgab_ but is that the right version of rakudo?
14:56 masak the git project resides in languages/rakudo, no?
14:57 szabgab_ so has rakudo moved to git or is it still in the parrot svn?
14:57 masak yes.
14:57 masak :)
14:58 masak szabgab_: pmichaud did the move last week, and then he mysteriously disappeared.
14:59 pmurias ruoso: the m0ld compiler would only need support for a volatile declaration so that the register would have the appropriate number
14:59 pmurias my $scope is volatile(0); ?
15:00 ruoso I was thinking it would be easier if you just assumed that "my $scope;" meant that register would already contain the lexical scope of that mold
15:00 ruoso like if "my $scope;" actually meant "my $scope = $mold.lexical"
15:03 pmurias is volatile seems easier to implement and we currently depend on handwritten blocks not being optimised in a few places already
15:04 ruoso pmurias, if you say so, go for it ;)
15:04 ruoso it looks great to me ;)
15:05 |Jedai| joined #perl6
15:41 spx2_ joined #perl6
15:47 khisanth_ joined #perl6
15:54 justatheory joined #perl6
15:54 aindilis` joined #perl6
16:17 maerzhase joined #perl6
16:21 alester joined #perl6
16:26 stef_ joined #perl6
16:29 stef_ joined #perl6
16:30 nihiliad joined #perl6
16:31 alester Where's the pmichaud?
16:31 alester Oh Paaaatrick
16:33 Tene joined #perl6
16:34 icwiener joined #perl6
16:39 hercynium joined #perl6
16:39 |Jedai| joined #perl6
16:56 maerzhase joined #perl6
16:57 maerzhase joined #perl6
17:02 alc joined #perl6
17:04 |Jedai| joined #perl6
17:06 jferrero joined #perl6
17:13 ayrnieu_ joined #perl6
17:19 sahadev joined #perl6
17:19 sahadev hello, i am not svn-savvy, so can someone tell when this means in svn-speak?
17:19 sahadev >> svn update
17:19 sahadev svn: PROPFIND request failed on '/parrot/trunk'
17:19 sahadev svn: PROPFIND of '/parrot/trunk': 302 Found (https://svn.perl.org)
17:20 moritz_ sahadev: it means that you've missed the switch to the parrot.org svn server
17:20 moritz_ sahadev: see https://www.parrot.org/
17:20 sahadev ah. how do i fix that?
17:20 moritz_ there are some instructions on that page
17:20 sahadev ok. thanks.
17:22 sahadev moritz_: i recently came across your series of "Perl 5 to 6" lessons. thanks for them. very useful.
17:25 moritz_ sahadev: good to hear; you're welcome
17:46 dalek joined #perl6
17:49 |Jedai| joined #perl6
17:49 _Jedai_ joined #perl6
17:49 smtms joined #perl6
18:02 Psyche^ joined #perl6
18:15 _Jedai_ joined #perl6
18:26 dmpk2k joined #perl6
18:30 Caelum_ joined #perl6
18:30 smtms joined #perl6
18:32 cognominal joined #perl6
18:35 mtnviewmark joined #perl6
18:40 _Jedai_ joined #perl6
18:41 barney joined #perl6
18:44 preflex joined #perl6
18:56 icwiener joined #perl6
19:00 japhb joined #perl6
19:06 sri_kraih joined #perl6
19:09 maerzhase joined #perl6
19:10 kcwu joined #perl6
19:10 simcop2387 joined #perl6
19:10 pugs_svn joined #perl6
19:17 kcwu joined #perl6
19:17 simcop2387 joined #perl6
19:17 pugs_svn joined #perl6
19:20 timbunce joined #perl6
19:25 mberends timbunce: hi, do visit here regularly?
19:25 timbunce mberends: no, just when I run my irc client, which isn't very often at the moment
19:27 mberends ah. I can tell you that DBI is one of the most yearned after features not yet in Rakudo.
19:28 timbunce feel free to write one layered over Simon Cozens driver code
19:28 mberends if only I could...
19:29 rodi out of curiosity, why do people want DBI in rakudo at this stage?
19:29 rodi </delurk>
19:30 _Jedai_ joined #perl6
19:30 mberends the most complete application is a Wiki engine, http://november-wiki.org
19:31 mberends in addition, there is enough of Perl 6 working now that the data it manipulates is worth storing in something smarter than flat files.
19:33 mberends the November project is http://github.com/viklund/november/tree/master
19:34 meppl joined #perl6
19:37 rodi mberends: cool, thanks for the link.
19:39 maerzhase1 joined #perl6
19:39 mberends the other must-use link is http://irclog.perlgeek.de/perl6/today
19:41 rodi feels very recursive to look over there while I'm typing over here...
19:41 mberends rorrim a ekil | like a mirror
19:42 _timbunce joined #perl6
19:45 simcop2387 joined #perl6
19:51 cognominal joined #perl6
19:53 M_o_C joined #perl6
20:03 lisppaste3 joined #perl6
20:23 sahadev left #perl6
20:24 kcwu joined #perl6
20:24 pugs_svn joined #perl6
20:37 jan_ joined #perl6
20:37 c9s_ joined #perl6
20:54 _Jedai_ joined #perl6
21:09 maerzhase joined #perl6
21:18 ayrnieu_ joined #perl6
21:29 jferrero joined #perl6
21:29 Khisanth hmm
21:29 Khisanth perl6: 1..10:by(2)
21:30 p6eval pugs: OUTPUT«*** ␤    Unexpected ":by"␤    expecting "_", fraction, exponent, term postfix or operator␤    at /tmp/LmU4e4iw1T line 1, column 6␤»
21:30 p6eval ..elf 25172: OUTPUT«Parse error in: /tmp/G2wP4wnErK␤panic at line 1 column 11 (pos 11): No previous operator visible to adverbial pair ([#<Match:0x819d6bc @on_str="1..10:by(2)", @from=5, @to=11, @bool=true, @hash={:value=>#<Match:0x819d874 @on_str="1..10:by(2)", @from=6, @to=11, @bool=true,
21:30 p6eval ..@hash={:...
21:30 p6eval ..rakudo 36300: OUTPUT«Statement not terminated properly at line 1, near ":by(2)"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
21:31 moritz_ no implementation does adverbs on operators yet :/
21:31 rodi I can help with that, actually:
21:31 rodi (2,4,6,8,10)
21:31 rodi next?
21:32 moritz_ rodi: now do that for 2..*:by(2) ;-)
21:32 rodi moritz_: done.  it's lazy, though, so you won't get
21:32 rodi 'em til you need 'em.
21:32 rodi :-P
21:34 rodi crap, I can never remember if I'm supposed to be parsimonious or lazy there.
21:34 Khisanth moritz_: I just try that once every couple of months :P
21:34 moritz_ Khisanth: ;)
21:50 TimToady I can do that one.  (2,4,6,8,10...*)
21:51 avar my @fib = (2,3,5,7...*)
21:52 TimToady currently just (1,1...&[+])
21:52 moritz_ &[+] works?
21:52 moritz_ I'm surprised
21:52 TimToady specced
21:52 moritz_ is that new?
21:52 TimToady less than a week :)
21:52 moritz_ ie "yes"
21:53 TimToady but not new :)
21:53 moritz_ not .new ;-)
21:53 TimToady but given that &[+] is just short for &infix:<+> it shouldn't be hard to implement
21:56 TimToady 'course, some folks define fib() as (0,1...&[+])
22:00 TimToady so obviously the standard P6 definition should be (0|1, 1 ... &[+])
22:03 * moritz_ cries out in agony
22:04 araujo (1,1 ... &[+]) , means .... (1,1, 1+1, 2+1, 3+2 .... ) ?
22:04 moritz_ yes
22:05 moritz_ although... not sure
22:05 rob joined #perl6
22:05 TimToady same as (1,1 ... { $^a + $^b }) only without creating a new closure
22:05 araujo nice fib
22:05 moritz_ TimToady: doesn't [+] expect arbitrary many arguments?
22:05 pugs_svn r25173 | putter++ | [elfish/on_sbcl] A start at supporting multiple CL implementations - libraries are now organized into implementation specific copies.
22:05 pugs_svn r25173 | putter++ | [STD_red] More UTF fiddling.  Unbreaks on_sbcl
22:05 moritz_ so it would be (1, 1, 1+1, 1+1+2, ...)
22:06 TimToady no, specced as left associative
22:06 moritz_ ok
22:06 TimToady and perhaps the &[] form always picks a 2-ary form
22:07 pugs_svn r25174 | putter++ | [elfparse] rx_on_re - A start at cleaning up field accesses.
22:08 TimToady much as [~] $a, $b, $c implies $a ~ $b ~ $c despite being list associative
22:09 araujo so like a fold?
22:09 TimToady that's how we spell it in p6
22:10 araujo I see ...
22:10 araujo :)
22:11 * araujo likes it
22:12 araujo TimToady, is this true for any operator? .. I mean, [+], [-], .....
22:12 araujo ?
22:13 TimToady pugs: [-] 10,1,1,1
22:13 p6eval pugs: RESULT«7»
22:13 [particle]1 any infix operator
22:13 TimToady any simple infix
22:13 araujo right, good :)
22:13 TimToady doesn't do thunky stuff
22:13 TimToady and lvalues are unlikely to work well
22:14 TimToady but anything without side effects, sure
22:14 TimToady including user-define operators
22:14 araujo yeah, it is a very declarative feature
22:15 TimToady pugs: sub infix:<foo> ($a,$b) { "$a:$b" }; [foo] 1,2,3,4,5
22:15 p6eval pugs: RESULT«1»
22:15 TimToady pugs: sub infix:<foo> ($a,$b) { "$a:$b" }; say [foo] 1,2,3,4,5
22:15 p6eval pugs: OUTPUT«1␤»
22:15 TimToady hmm
22:15 TimToady well, seems buggy
22:15 araujo :(
22:15 araujo but it is cool indeed
22:16 araujo who is in charge of pugs these days?
22:16 [particle]1 yes.
22:16 TimToady no one
22:16 araujo mm..
22:16 * jnthn finally gets around to scribbling something about his hacking in his blogs.
22:17 TimToady officially, no one was ever in charge of pugs :)
22:17 araujo Long time I don't look at pugs code...
22:17 araujo wonder how much it has changed.... will take a look
22:17 TimToady almost nothing has changed...if you want to bring it more up to date, feel free
22:18 TimToady or write a STD backend that spits out Haskell...
22:18 araujo I can play a bit with it for sure
22:19 avar araujo: yay
22:20 araujo avar, :)
22:22 TimToady it's suffered a certain amount of "design rot" as the P6 specs have evolved
22:22 Exodist joined #perl6
22:23 mtnviewmark ah, so the fib example shows that the &[op] syntax is clearly being confused by people with [op] metaop
22:23 mtnviewmark perhaps &<op> would be better?  (and I know what yer gonna say.... :-)  )
22:24 mtnviewmark the comment moritz_: so it would be (1, 1, 1+1, 1+1+2, ...)
22:24 mtnviewmark implies that he thought it was passing the [+] operator, not the + operator
22:25 diakopter araujo: see the recent Pugs releases on hackage
22:26 diakopter araujo: I'm curious to know how Pugs has changed apart from svn.pugscode
22:27 araujo diakopter, that one is the latest 6.2.13.14?
22:27 araujo diakopter, do you use it?
22:28 Whiteknight joined #perl6
22:29 diakopter araujo: I have a local install of that release, yeah, that I was using to build smop
22:30 moritz_ I tried to install it on the server on which p6eval lives, but it fell victim to the OOM-killer
22:30 araujo diakopter, does it work latest ghc?, I heard it had some problems
22:30 araujo with*
22:31 araujo ok, I will try this
22:36 pugs_svn r25175 | lwall++ | [S03] remove frivolous sub [op] form of reduce declaration
22:36 pugs_svn r25175 | lwall++ | &[op] always refers to a binary infix even for list associative ops
22:37 TimToady mtnviewmark: can't really use &<op> because far more ops contain < and > than contain [ and ]
22:38 TimToady and I'm trying to change the meaning of [op] to mean infix, and the reduce form works that way merely because it's being used where a term is expected
22:39 TimToady and if that's what you expected me to say, I guess it doesn't hurt to repeat it for everyone else :)
22:40 araujo :)
22:40 TimToady now I'm trying to figure out what $a [\op] $b means :)
22:40 araujo haha
22:41 araujo diakopter, i can see pugs is available from cabal-install .....
22:42 TimToady not sure how close to HEAD that is; was audreyt doing that on a private branch?
22:43 TimToady or did that get merged?
22:43 moritz_ there were no pugs commits in the last couple of months
22:45 araujo well, the hackage tarball package seems a bit outdated
22:45 TimToady it was some months ago in any case
22:46 TimToady anyway, I don't know how much success you'll have compiling src/Pugs directly, but you'd need that in order to hack on it
22:47 araujo isn't audrey around here these days?
22:47 araujo TimToady, next to it...
22:47 pugs_svn r25176 | moritz++ | [t/TASKS] remove task for which the spec was just removed
22:47 phade joined #perl6
22:47 araujo looking at the cabal file, I think i won't have too much problems
22:47 araujo i am checking the svn repo though
22:48 diakopter araujo: how is the hackage tarball outdated? there's been a new release posted quite often the past few months
22:50 araujo Upload dateTue Dec 16 09:01:01 UTC 2008
22:50 araujo for latest tarball?
22:50 diakopter (that seems fairly recent to me)
22:51 diakopter relative to the last time someone committed something to Pugs in svn here
22:51 TimToady my impression is that audrey lurks, but doesn't want to promise (implictly or explicitly) anything that might exceed her current bandwidth, which seems completely sane to me :)
22:56 wolv joined #perl6
22:57 araujo diakopter, ok, just asked, because i got a 2006 package in my repo for the pugs tarball .. version 6.2.13
22:57 diakopter oh.... 2006?  wow.
22:57 araujo yeah :P
22:58 araujo but i guess it is just us (Gentoo) too outdated
22:58 araujo :|
22:58 araujo I am testing svn though ...
23:00 AzureStone_ joined #perl6
23:00 diakopter which ghc version do you have? I recommend the cabal/hackage release of Pugs 6.2.13.14 (you mentioned above), on ghc 6.10.1.
23:01 araujo svn: REPORT of '/pugs/!svn/vcc/default': Could not read response body: connection was closed by server (http://svn.pugscode.org)
23:01 araujo diakopter, i've got 6.8
23:01 moritz_ that's too old
23:02 araujo i need that version for some packages
23:03 araujo moritz_, will test first
23:05 phade left #perl6
23:05 moritz_ oh wait, the 6.6.* are too old, not sure about 6.10
23:11 lambdabot joined #perl6
23:11 diakopter wb lambdabot
23:13 TimToady @messages
23:13 lambdabot You don't have any new messages.
23:19 ruoso joined #perl6
23:22 araujo moritz_, yeah, 6.6 is the 'old' one
23:22 araujo 6.8 is quite good yet
23:22 araujo and 6.10 is latest
23:29 maerzhase left #perl6
23:29 maerzhase joined #perl6
23:37 araujo diakopter, which version do you use?
23:48 diakopter araujo: ghc 6.10.1
23:48 diakopter but, only to build smop a little ago
23:50 araujo oh i mean
23:50 araujo of pugs
23:50 diakopter the latest hackage/cabal release, 6.2.13.14
23:50 araujo diakopter, through caball install?
23:50 diakopter yeah
23:50 diakopter on windows, no less.
23:57 jhorwitz joined #perl6

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

Perl 6 | Reference Documentation | Rakudo