»ö« | perl6-projects.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by sjohnson on 21 August 2009.
00:07 kent\n left, kent\n joined 00:12 pyrimidine joined 00:16 cotto left 00:17 cotto joined 00:20 pyrimidine left 00:25 pyrimidine joined 00:35 carlin joined 00:47 nErVe left 01:13 DakeDesu joined
pyrimidine jnthn: around? 01:14
01:18 KatrinaTheLamia left 01:29 pyrimidine left 01:30 pyrimidine joined 01:35 meppl left 01:38 bpetering joined
pyrimidine can we compose multiple traits into attributes yet with rakudo? 01:40
something like 'has $.foo is rw is aliased<bar>'
this popped up from looking at jonathan's example from his last post (maybe in response to takadonet's question a couple of days ago here) 01:45
01:45 bpetering is now known as missingthepoint
pyrimidine irclog.perlgeek.de/perl6/2009-08-18#i_1403903 01:45
01:47 KatrinaTheLamia joined
missingthepoint morning #perl'ers :) 01:56
02:02 pyrimidine left, DakeDesu left 02:03 lucs left, pyrimidine joined 02:05 lucs joined
pyrimidine missingthepoint: morning 02:05
missingthepoint morning. :) 02:07
02:11 beggars joined 02:13 Whiteknight left
carlin how do I throw an exception? 02:26
rakudo: sub a { return Exception; }; try { say a; CATCH { say 'foo'; } }; 02:30
p6eval rakudo 0d4fe0: ( no output )
carlin ^ seg faults
but I think that's wrong anyway
missingthepoint carlin: i think you just 'die' 02:37
rakudo: { die 'exception. :*(' CATCH { say "caught exception: $!" }} 02:38
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "CATCH { sa"␤in Main (src/gen_setting.pm:3390)␤»
missingthepoint oh hmm.
rakudo: do { die 'exception. :*(' CATCH { say "caught exception: $!" }}
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "CATCH { sa"␤in Main (src/gen_setting.pm:3390)␤»
02:38 nbrown left 02:40 nbrown joined, xomas joined
missingthepoint oh my goodness. i need caffeine. :O 02:42
rakudo: do { die 'exception. :*('; CATCH { say "caught exception: $!" }}
p6eval rakudo 0d4fe0: OUTPUT«caught exception: exception. :*(␤exception. :*(␤in Main (/tmp/5wqVgzZrOz:2)␤»
02:42 xomas_ left
missingthepoint that's probably more confusing (or just foolish) than helpful. :*( 02:44
FWIW, i was wondering the exact same thing the other day. 02:45
02:45 pyrimidine left
missingthepoint you might be right, but it might not be implemented in Rakudo. 02:52
but how would we know if it's not implemented? :/ 02:58
03:01 azawawi joined
azawawi moritz_: ping 03:01
carlin rakudo: sub a { die 'test'; }; try { say a; CATCH { say 'foo'; } }; 03:02
p6eval rakudo 0d4fe0: OUTPUT«foo␤»
azawawi moritz_: Could you please add #padre @ irc.perl.org to irclog.perlgeek.de/ ?
carlin missingthepoint++ # thanks, die is right :)
rakudo: say Exception; 03:04
p6eval rakudo 0d4fe0: ( no output )
carlin Hm, I wonder if that's known
03:05 pyrimidine joined
pyrimidine jnthn: nm, 'has Str $.description is rw is aliased<desc>;' works partially (desc is ro, description is rw). Just need to work out how to create the mutator for rw aliases. 03:07
03:11 azawawi left 03:12 pyrimidine left 03:17 Limbic_Region left, tak11 joined 03:21 justatheory joined 03:33 braceta left
s1n is it still not possible to set environment variables (%*ENV) in rakudo yet? 03:36
03:45 pyrimidine joined
pyrimidine rakudo: %*ENV.keys.join(',').say 03:48
p6eval rakudo 0d4fe0: OUTPUT«PERL6LIB,TERM,SHELL,USER,LD_LIBRARY_PATH,LS_COLORS,PATH,MAIL,PWD,EDITOR,LANG,HISTCONTROL,HOME,SHLVL,LOGNAME,LESSOPEN,LESSCLOSE,_␤»
pyrimidine s1n: just testing on p6eval and locally, doesn't look as if you can set %*ENV yet 03:52
rakudo: %*ENV<THIS_NEVER_EXISTS> = 123; say %*ENV<THIS_NEVER_EXISTS> 03:53
p6eval rakudo 0d4fe0: OUTPUT«␤»
carlin rakudo: %*ENV = { 'a' => 'b' }; say %*ENV; 03:54
p6eval rakudo 0d4fe0: OUTPUT«a b␤»
missingthepoint is rakudo.org being updated? 03:55
(i can't load it.) 03:56
03:58 macae joined 04:02 macae left
pyrimidine missingthepoint: rakudo.org is working for me 04:10
04:12 tak11 left
missingthepoint pyrimidine: oh noes, teh DNS iz broked! :*( 04:13
04:14 tak11 joined
pyrimidine missingthepoint: THE DEE-ENN-ESS ES BORKEN, RUN FOR TEH HILLZ 04:16
missingthepoint actually, it's just "broked"... 'www.rakudo.org' works, but 'rakudo.org' doesn't. that's not so bad... unless the first page google results for 'rakudo' link to is 'rakudo.org' 04:22
04:30 rjh left
pyrimidine missingthepoint: rakudo.org is working for me as well. maybe someone is watching? 04:30
pyrimidine queueing the spooky music...
04:36 wayland76 left, macae joined 04:37 tak_ joined, tak11 left 04:38 wayland76 joined, pyrimidine left
carlin rakudo: say $?LINE; 04:42
p6eval rakudo 0d4fe0: OUTPUT«Symbol '$?LINE' not predeclared in <anonymous> (/tmp/hJIrcUzgwq:2)␤in Main (src/gen_setting.pm:3390)␤»
carlin :(
04:45 wayland joined 04:50 nihiliad left 05:02 szabgab joined 05:06 wayland76 left 05:07 molaf joined, tak_ is now known as tak11
missingthepoint translate.google.com says someone is watching. ;) 05:08
05:13 wayland left 05:15 nsh left 05:19 sbp left 05:20 sri_kraih_ joined 05:22 frederico joined 05:25 wayland__ joined, meteorjay left 05:30 phenny left
missingthepoint rakudo: use LWP::Simple; my $l = LWP::Simple.new; say ?($l.get('rakudo.org') ~~ /'Get Rakudo'/) 05:30
p6eval rakudo 0d4fe0: OUTPUT«Can't find ./LWP/Simple in @*INC␤in Main (src/gen_setting.pm:445)␤»
05:31 SmokeMachine left 05:32 wayland76 joined 05:34 SmokeMachine joined 05:37 unitxt left, sri_kraih left, sri_kraih_ is now known as sri_kraih 05:38 unitxt joined 05:40 unitxt left 05:42 SmokeMachine left 05:45 orafu left, orafu joined 05:47 justatheory left 05:49 macae left 05:55 wayland__ left 06:04 wayland76 left 06:05 wayland76 joined, zloyrusskiy joined 06:13 tak11 left 06:15 tak11 joined 06:28 xomas left 06:33 zloyrusskiy left, cono_ joined 06:35 DHGE joined 06:36 zloyrusskiy joined 06:41 sbp joined, nsh joined 06:43 tak11 left 06:45 tak11 joined 06:50 agentzh joined, SmokeMachine joined, zloyrusskiy left 06:58 zloyrusskiy joined 07:17 zloyrusskiy left 07:22 beggars left 07:39 Eevee joined 07:42 Chillance joined 07:53 fredrecsky joined 08:00 frederico left 08:01 jaffa8 joined
jaffa8 hi 08:01
where can I see what tests have been passed?
moritz_ jaffa8: download rakudo, run 'make spectest' and see 08:02
jaffa8 ok 08:03
where can I download a pugs binary?
moritz_ I don't know of any up-to-date binaries 08:04
jaffa8 Can I see the pugs test results somewhere? 08:05
moritz_ it used to be smoke.pugscode.org. No idea if that still works. 08:06
jaffa8 it does not seem to. 08:10
carlin it doesn't, there is perlcabal.org/smoke.html but it's rather out of date
jaffa8 I guess rakudo inherited the test of pugs 08:13
Why was rakudo started? 08:16
What was not pugs continued? 08:17
08:23 adhoc_ is now known as adhoc
moritz_ well, pugs is written in haskell, and most people in the perl community don't speak that all too well 08:24
and it's not something you learn on a whim
carlin rakudo: say Exception; 08:26
p6eval rakudo 0d4fe0: ( no output )
carlin is that known?
moritz_ carlin: no; please report. 08:27
carlin filed bug 08:32
08:33 Jedai joined 08:38 Eevee left 08:44 duke_leto joined 08:52 Eevee joined 08:53 rblasch joined
duke_leto I’ve been wondering: A year (or more) ago, I was really hot for perl 6, because it was going to have all those nifty UTF-8 operators like ¥ and ← and ⇒ and stuff… I’ve tried some of them and rakudo never liked it :-( 08:56
Are they going to come or has that idea been dropped?
mberends duke_leto: the ability to define your own Unicode operators has now come :-) 09:03
09:16 Front_slash joined
duke_leto mberends: :-) yay! That’s a step forwards ;-) 09:24
Because I’d really love to read code like if $a ≤ $b {…} 09:25
(and write it) 09:31
09:36 M_o_C joined 09:40 DHGE left 09:41 masak joined
masak o/ 09:41
mberends o/ masak 09:42
rakudo: sub infix:<≤>($a,$b){$a<=$b}; if 4 ≤ 5 { "lessorequal".say} # see duke_leto 09:43
p6eval rakudo 0d4fe0: OUTPUT«lessorequal␤»
masak mberends: oh hai. whenever it's convenient for you, I'd like to pair up around proto, which has a number of pending refactors/fixes.
mberends: oh, and by the way, still enjoying your condiments on my bread. mberends++
mberends heh, well apart from unpredictable distractions, now would be good :) 09:44
masak I'm fairly available too. ok, here's the outline: 09:45
(1) the sources-as-classes refactor, (2) the independent-priority-graph refactor, (3) adapting proto for an installed Rakudo. 09:46
the two former I mentioned in a nice restaurant in Lisbon. the latter is perhaps the more urgent. anyone of these is up for grabs.
mberends I'm up for (3) first. 09:48
masak good. we should probably discuss that one a bit. 09:49
mberends (1) should be fairly simple, (2) depends on more abstraction redesign first
masak proto has never needed 'sudo' privs before. now that it does, I think that entails some other reconsideration.
maybe it's time to try that "central lib/" idea, for example...
the simplest thing that could possibly work with a central lib/ directory would be to have a directory somewhere, such as /usr/share/perl6-proto/lib (just a suggestion) 09:51
and then "install" the PIR files there -- or, if there is no make step, the .pm files.
slightly large rewrite, but it means our PERL6LIB worries are over, once and for all. 09:52
it also means people have to play nice(r) with module names.
mberends agreed. the dirname is a bikeshed topic, but we should try to avoid future upheavals 09:53
later today japhb++ planned to discuss similar issues for Parrot modules 09:54
masak I threw in '-proto' in the dirname because I still consider proto to be a very unofficial, under-the-radar phenomenon. thus, it shouldn't take an "official" path like /usr/share/perl6/lib 09:55
mberends masak++: yes, that's fine, if people agree with the proto defaults down the line, we can remove 6 characters from the dirname easily 09:56
masak sure, but remember that this solution _cannot_ be the final solution. 09:57
it doesn't do authorities and version.
s/version/versions/
mberends that's what the database-y alternative plans to handle
masak aye. 09:58
but does that mean that up in the filesystem, one can only have one module of a given name installed? 09:59
mberends that's "the simplest thing that could possibly work" ;)
masak it is indeed. 10:00
let's do that now, too.
mberends putting :ver and :auth into the filesystem is a bad idea imho 10:01
masak hm, changing the installation path is a pretty large change to proto.
M_o_C Is it intended that the parrot sources must be located in <parrot path>/src? And does "sources", i. e. the parrot source directory as specified in the makefile, refer to the svn co of trunk or is this something else?
masak I was thinking first about upgrading the config file and making it backwards-compatible, but that feels like a lot of work, to be honest.
instead, I'll just let proto self-deprecate if it notices it has been upgraded but has an old config file. 10:02
I'll start with that commit.
mberends most other installers (including CPAN) do everything apart from the final install step without su privileges. proto should also work in stages, if possible. That's a considerable change. 10:03
masak I can see the need for that. how would it work in practice? invoking proto several times? 10:04
I think I would deplore that slightly. 10:05
10:05 jauaor left
jaffa8 Does anyone know about cabal install Pugs? 10:06
How should it work?
mberends masak: unfortunately, that's probably the common solution. OTOH, shelling to a 'sudo' might prompt for a password.
masak mberends: I'd be more comfortable with. "the module and all its dependencies have been built and tested. we're now ready to install" -- user gets put into a shell either inside or outside of proto, and is told what to write. 10:08
something in the order of 'sudo ./proto finish-install' 10:09
speaking of './proto', it's probably time to put proto in the path, too.
mberends M_o_C: yes, although Parrot experts can override numerous settings. The Rakudo Configure.pl finds out about your Parrot from <parrot>/bin/parrot_config 10:11
masak: shell inside proto if possible, that way proto has a bit more control 10:12
masak good.
mberends masak: putting proto in the path will hinder its main design objective, which was to become obsolete ;) 10:13
masak perhaps a full-blown shell is overkill, though. perhaps it should be "do you want to install this module globally? (will require sudo password) [Y/n]" 10:14
mberends likes that
masak mberends: true, and then the config file would have to go somewhere more global as well. so I'll stand down on that idea.
mberends are we ever going to port proto to Windows? 10:15
jnthn++ was reading create_new_project recently and snickering 10:16
masak aye. I saw that.
things in proto are made on an as-needed basis. often the needs are mine.
mberends the simplest thing... is Unix-y right now 10:17
masak the Windows compat is no different -- if someone comes along and wants it badly enough, it'll happen.
M_o_C Thanks, mberends.
masak jaffa8: it's been a long time since I used 'cabal install Pugs', but I got it to work then. "how should it work?" is difficult to answer sensibly... it should install Pugs. it did for me, after some persuasion. 10:18
mberends masak: if we can count on $HOME existing, we can base proto's lib there 10:19
10:19 xinming_ joined
masak instead of globally, you mean? yes, perhaps that's prederable. 10:19
jaffa8 masak, i am trying on windows
I get an error
masak jaffa8: oh, that might make things less... straightforward.
jaffa8 missing dependency on foreign librarz 10:20
masak jaffa8: that's pretty clear, at least.
mberends jaffa8: which library name?
masak jaffa8: your quest, should you choose to accept it, is to find and install this library.
jaffa8 perl5/p5embed.h
I set the include path
mberends jaffa8: that's from the Perl 5 source code 10:21
jaffa8 masak, I started that "quest
I found the files in the pug distribution
I set the path
cabal install Pugs --extra-lib-dirs=D:\standalone\pugs\src
but it has no effect 10:22
jnthn dobry den :-)
mberends jaffa8: have you tried also installing Strawberry Perl 5?
jaffa8 what for? 10:23
mberends jnthn: dobor dan :)
jaffa8: I think Strawberry Perl includes the file you are lacking 10:24
masak jnthn: добрый день :)
10:25 Su-Shee joined
Su-Shee heyho. 10:25
jaffa8 jnthn, can you tell me how to install pugs?
mberends, I have the header file 10:26
mberends jaffa8: cool! one step closer...
jnthn jaffa8: No. 10:28
jaffa8 How come? You must have done many times
jnthn jaffa8: Not for (literally) years.
jaffa8 jnthn, do you have a version of pugs? 10:29
jnthn jaffa8: And even then I didn't install it; I built from source.
no, I use the evalbot here in channel if I want to check something
jaffa8 Does that use pugs?
mberends jaffa8: dunno if this helps, perl6.cz/wiki/Pugs_on_MS_Windows (it's old) 10:30
jnthn jaffa8: It can
pugs: say 42
p6eval pugs: OUTPUT«42␤»
jnthn jaffa8: Normally I'm only interested in the output of Pugs in certain specific cases. 10:31
jaffa8 pugs:while ( $u ~~ m:p/a/ ) {}
pugs:my $u;( $u ~~ m:p/a/ ) {}
it is not working
pugs: y $u;( $u ~~ m:p/a/ ) {} 10:32
p6eval pugs: OUTPUT«*** ␤ Unexpected ";("␤ expecting "::"␤ Variable "$u" requires predeclaration or explicit package name␤ at /tmp/xEydusvUpt line 1, column 5␤»
jaffa8 pugs: m $u;( $u ~~ m:p/a/ ) {}
p6eval pugs: OUTPUT«*** ␤ Unexpected "u"␤ expecting term postfix or operator␤ at /tmp/6j2JJ1gdAz line 1, column 9␤»
jaffa8 pugs: my u; if ($u ~~ m:p/a/ ) {}
p6eval pugs: OUTPUT«*** ␤ Unexpected ";"␤ expecting word character, ":" or "("␤ at /tmp/dc0uhEgh98 line 1, column 5␤»
jaffa8 pugs: my $u; if ($u ~~ m:p/a/ ) {}
p6eval pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/evalenv/pugs/perl5/Pugs-Compiler-Rule/lib';␤ unshift @INC, '/home/evalenv/pugs/third-party/Parse-Yapp/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime…
jaffa8 jnthn, Is this the most recent pugs? 10:33
mberends jaffa8: unless you have a good reason, rather use Rakudo
jaffa8 I have a good reason.
mberends ok􏿽x9E
jnthn jaffa8: Probably...remember Pugs hasn't really seen active development for a couple of years. 10:34
jaffa8 Is that supposed to be an error message?
pugs: my $u; if ($u =~ m:p/a/ ) {} 10:35
p6eval pugs: OUTPUT«*** ␤ Unexpected "m"␤ There is no =~ operator in Perl 6 -- did you mean ~~ (match) or ~= (concat-assign)?␤ at /tmp/pT9kJTEbZn line 1, column 18␤»
jaffa8 pugs: my $u; if ($u ~~ /a/ ) {}
p6eval pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/evalenv/pugs/perl5/Pugs-Compiler-Rule/lib';␤ unshift @INC, '/home/evalenv/pugs/third-party/Parse-Yapp/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime…
jaffa8 pugs: my $u; if ($u ~~ m/a/ ) {} 10:36
p6eval pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/evalenv/pugs/perl5/Pugs-Compiler-Rule/lib';␤ unshift @INC, '/home/evalenv/pugs/third-party/Parse-Yapp/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime…
jnthn jaffa8: I think it's just something broken in the regex handling.
mberends pugs: my $u = "yay"; if ($u ~~ m:p5/a/ ) { "matched".say }
p6eval pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/evalenv/pugs/perl5/Pugs-Compiler-Rule/lib';␤ unshift @INC, '/home/evalenv/pugs/third-party/Parse-Yapp/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime…
carlin jaffa8: why do you need to use pugs instead of rakudo? 10:37
jaffa8 rakudo did not handle the perl6 statement I used
this-->>> my $u; if ($u =~ m:p/a/ ) {} 10:38
masak jaffa8: it's ~~
jaffa8 even then
mberends rakudo: my $u = "yay"; if ($u ~~ m/a/ ) { "matched".say }
masak and it might work if you put the :p inside the //
p6eval rakudo 0d4fe0: OUTPUT«matched␤»
jaffa8 rakudo: my $u; if ($u =~ m:p/a/ ) {}
p6eval rakudo 0d4fe0: OUTPUT«Colons cannot be used as delimiters in quoting constructs at line 2, near ":p/a/ ) {}"␤in Main (src/gen_setting.pm:3390)␤»
reqamst Is AUTOLOAD implemented currently? 10:39
masak jaffa8: you know, it occurs to me that it might be more effective to ask about such things here, rather than try a whole different implementation. :)
mberends the :p is the problem
jaffa8 rakudo: my $u; if ($u =~ /:p a/ ) {}
p6eval rakudo 0d4fe0: ( no output )
masak see? :)
jaffa8 Can some try this: my $u="􏿽xE4 a a"; while ($u =~ /:p a/ ) {print "got "} 10:40
M_o_C rakudo: my $u; if ($u =~ :p/ a/ ) {}
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near "/ ) {}"␤in Main (src/gen_setting.pm:3390)␤»
jaffa8 Can some try this: my $u="􏿽xE4 a a"; while ($u ~~ /:p a/ ) {print "got "}
rakudo: my $u="􏿽xE4 a a"; while ($u ~~ /:p a/ ) {print "got "} 10:41
p6eval rakudo 0d4fe0: OUTPUT«got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got
..got go…
jaffa8 rakudo: my $u="􏿽xE4 a a"; while ($u ~~ m:g /:p a/ ) {print "got "}
p6eval rakudo 0d4fe0: OUTPUT«Colons cannot be used as delimiters in quoting constructs at line 2, near ":g /:p a/ "␤in Main (src/gen_setting.pm:3390)␤»
masak jaffa8: I'm sorry, you have a strange byte in your $u string. 10:42
neither my client nor the IRC logs can see what it is.
jaffa8 How would you parse a string for matches repeatedly?
rakudo: my $u="a a a"; while ($u ~~ m:g /:p a/ ) {print "got "}
p6eval rakudo 0d4fe0: OUTPUT«Colons cannot be used as delimiters in quoting constructs at line 2, near ":g /:p a/ "␤in Main (src/gen_setting.pm:3390)␤»
masak jaffa8: I think I'd just use .comb
jaffa8 what is .comb? 10:43
masak rakudo: say "foo foooo foooo fo".comb(/fo+/).perl
p6eval rakudo 0d4fe0: OUTPUT«["foo", "foooo", "foooo", "fo"]␤»
masak that's .comb
S32/Str contains the details. 10:44
jaffa8 rakudo: my $u="a a a"; while ($u ~~ m:g /:g :p a/ ) {print "got "}
p6eval rakudo 0d4fe0: OUTPUT«Colons cannot be used as delimiters in quoting constructs at line 2, near ":g /:g :p "␤in Main (src/gen_setting.pm:3390)␤»
jaffa8 rakudo: my $u="a a a"; while ($u ~~ /:g :p a/ ) {print "got "}
p6eval rakudo 0d4fe0: OUTPUT«got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got got
..got go…
jaffa8 masak,ok
it is quite a weird name
masak jaffa8: my approach has the advantage of actually working. :)
M_o_C Is there a difference between perlcabal.org/syn/ and feather.perl6.nl/syn/?
jaffa8 it does not remind me of anything
masak jaffa8: no, 'comb' is a verb meaning 'pick out the things you're interested in'. 10:45
it's the figure-ground opposite of 'split'.
jaffa8 what is that .perl at the end? 10:46
10:47 molaf left
mberends jaffa8: it's a built in Data::Dumper 10:47
carlin M_o_C: No, feather.perl6.nl/ and perlcabal.org/ look like they're just different domains pointed to the same server
jnthn -> bbq, bakc later
M_o_C Ok, thanks 10:48
jaffa8 ok
rakudo: say "foo foooo foooo fo".comb(/(f)o+/).perl 10:50
p6eval rakudo 0d4fe0: OUTPUT«[Match.new(␤ # WARNING: this is not working perl code␤ # and for debugging purposes only␤ ast => "foo",␤ Str => "foo",␤ from => 0,␤ to => 3,␤ positional => [␤ Match.new(␤ ast => "f",␤ Str => "f",␤ from => 0,␤ to => 1,␤ ),␤ ],␤), Match.new(␤ # WARNING: this
..is n…
10:51 Whiteknight joined
mberends jaffa8: don't worry about the warning, .perl for pattern matches is under construction 10:51
jaffa8 I wonder if I use brackets, then how do I get the value of those? 10:52
in comb?
mberends rakudo: "foo foooo foooo fo" ~~ /(f)o+/; say $/[0] 10:54
p6eval rakudo 0d4fe0: OUTPUT«f␤» 10:55
jaffa8 rakudo: "foo foooo foooo fo" ~~ /(f)o+/; say $/[1] 10:56
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
mberends rakudo: "foo foooo foooo fo" ~~ /(fo+)o/; say $/[0]; say $/[1]
p6eval rakudo 0d4fe0: OUTPUT«fo␤Use of uninitialized value␤␤»
reqamst rakudo: my num $number 10:59
p6eval rakudo 0d4fe0: OUTPUT«Malformed declaration at line 2, near "num $numbe"␤in Main (src/gen_setting.pm:3390)␤»
reqamst Why I can't use native types? Is it a bug?
carlin rakudo: my Num $number; 11:00
p6eval rakudo 0d4fe0: ( no output )
reqamst yes, there is no problem with immutable types
mberends reqamst: native types are not yet implemented 11:01
reqamst mberends: I see, thanks. 11:02
gdzie ten TBH? 11:04
oops, wrong channel
masak Polish, right? 11:06
reqamst right 11:07
masak mberends: I just pushed a deprecation message for proto. could you run it and see that it works?
(at present, if you follow the instructions, you'll just get the deprecation message again. but that's next on the agenda.)
11:09 rafl joined
masak & # hunting for food 11:09
11:13 azawawi joined 11:14 eiro_ joined
azawawi moritz_: Thanks a lot for adding it. That really help with searching IRC logs. cool :) 11:14
moritz_: s/help/helps 11:15
11:16 azawawi left, Infinoid_ joined, wayland_ joined, Front_slash left, wayland76 left, gfldex left, patmat left, mberends left, Maddingue left, eiro left, Infinoid left, gabiruh left, Aisling left, awwaiid left, Aisling joined 11:17 mberends joined, Front_slash joined, wayland76 joined, gfldex joined, gabiruh joined, Maddingue joined, awwaiid joined, patmat joined, irc.freenode.net sets mode: +o mberends 11:22 meppl joined 11:23 wayland76 left
M_o_C Win32 Problem: I have a SVN CO in .\Parrot\src and Parrot is installed in .\Parrot. Rakudo git-clone resides in .\rakudo-git-clone. When I try to mingw32-make Rakudo I get the following error: 11:24
lisppaste3 MoC pasted "Win32 make fails" at paste.lisp.org/display/85891 11:26
11:29 leedo_ joined, leedo left 11:46 masak left, masak joined 11:47 DarkWolf84 joined
DarkWolf84 hi all 11:49
:)
masak DarkWolf84: o/
DarkWolf84 I have strange problem with installed rakudo and unicode 11:50
masak let's hear it.
DarkWolf84 when i try to print some unicode chars it writes a jubberish 11:52
moritz_ M_o_C: did you do a 'make install-dev' in parrot?
DarkWolf84 like
no
11:53 hanekomu joined
masak rakudo: say 'ʀɐƙʊðɵ' 11:53
p6eval rakudo 0d4fe0: OUTPUT«ʀɐƙʊðɵ␤»
moritz_ DarkWolf84: rakudo does input and output in UTF-8. If you use a diferent character encoding on your console you'll only see rubbish
masak works here.
DarkWolf84 yeah
rakudo: say "􏿽xFA􏿽xC4􏿽xD2􏿽xC1􏿽xD7􏿽xC5􏿽xCA"
p6eval rakudo 0d4fe0: OUTPUT«�������␤»
masak DarkWolf84: what you entered there was not UTF-8. 11:54
M_o_C Ah, there's the problem, I only did make install. Thanks moritz_
DarkWolf84 strange
masak DarkWolf84: it might have been some latin-* encoding, perhaps.
DarkWolf84 why is that
masak DarkWolf84: here on IRC, it's probably because of the way your IRC client is set up. 11:55
for the Rakudo execution, it's probably the way your shell or your filesystem is set up. 11:56
DarkWolf84 i have to check the web logs
masak s/filesystem/editor/
DarkWolf84 my client should use utf8
11:57 phenny joined
masak phenny: welcome back. 11:59
phenny masak: 22 Aug 19:32Z <s1n> tell masak i noticed there was no longer a Web.pm. am i mistaken or is that an ironic name for the project now
masak: 22 Aug 19:36Z <s1n> tell masak LolDispatch doesn't compile, rakudo thinks it's missing an =end, might be bug worthy
masak: 22 Aug 23:33Z <s1n> tell masak i can't seem to clone november, it keeps timing out... any idea what's up with that?
masak phenny: tell s1n not an ironic name. but I wouldn't say that the use case is very clear either right now. Web.pm is a set of modules at different levels. maybe it'll make sense to load them through 'use Web;' at the top, maybe not. 12:00
phenny masak: I'll pass that on when s1n is around.
12:01 hanekomu left, DarkWolf84 left, DarkWolf84 joined
DarkWolf84 rakudo: say "Здрасти" 12:02
p6eval rakudo 0d4fe0: OUTPUT«Здрасти␤»
DarkWolf84 works
masak aye, much better. 12:03
DarkWolf84 that was "hi" in Bulgarian :)
masak ah, pretty similar to Russian, then. 12:04
DarkWolf84 yeah
masak is there a singular and a plural form, as in Russian?
phenny: tell s1n cloning from git://github.com/viklund/november.git works fine here.
phenny masak: I'll pass that on when s1n is around.
DarkWolf84 yeah
most of the plural forms in Russian end with и 12:05
masak да. 12:06
DarkWolf84 in Bulgarian too
masak like пироги. 12:07
DarkWolf84 да
is the evalbot here installed or just from source dir 12:10
i tried perl6 in the rakudo source dir and utf8 works 12:11
but installed version is not
masak that's pretty interesting. 12:12
I could imagine that the evalbot runs Rakudo in the current directory. 12:13
DarkWolf84 maybe that is related to icu linking 12:14
moritz_ masak: it does not 12:15
masak: it installs rakudo
masak I sit corrected.
moritz_ rakudo: say %*VM<config><prefix>
p6eval rakudo 0d4fe0: OUTPUT«/home/evalenv/p2␤» 12:16
DarkWolf84 ok I'll make it again 12:18
is something needed for the parrot lib directory 12:23
moritz_ disc space
or what do you mean?
DarkWolf84 like ld path 12:24
or something like that
moritz_ if libicu is installed normally on your system everything should be fine.
DarkWolf84 ok
moritz_ tried to visualize object creation and initialization in Perl 6: moritz.faui2k3.org/files/build.svg 12:25
masak moritz_: nice. 12:31
moritz_ masak: thanks
moritz_ is quite fond of SVG these days...
masak too
mberends moritz_++ 12:33
phenny mberends: 22 Aug 16:11Z <masak> tell mberends TimToady would like you to rewrite create-new-project in TECO for greater platform independence...
moritz_ TECO? 12:34
12:34 duke_leto left
mberends :) pull the other one, it haz bells on 12:34
masak moritz_: en.wikipedia.org/wiki/Text_Editor_and_Corrector
mberends "TECO command sequence more closely resembles transmission line noise" sounds familiar ;) 12:37
masak $TECO is readonly 12:40
DarkWolf84 oh I got it it is the interactive shell
oneliners write corect utf8 and the normal programs 12:41
masak DarkWolf84: I can confirm that. 12:42
moritz_ too 12:43
DarkWolf84: would you please submit a bug report to [email@hidden.address]
mberends masak: proto's deprecation notice is a good start. It would be nice to save the list of existing projects somewhere, because once we move to a different base dir we cannot expect to find the old projects any more. 12:45
masak nod. 12:46
mberends perhaps just keep the old basedir in the proto config file 12:47
jaffa8 hi, the rakudo does not give good error messages
moritz_ jaffa8: in what cases?
masak another thought I had: if we put proto itself through it installation process, it would be antisocial to keep the module names Installer and Ecosystem. they would have to be renamed to App::Proto::Installer and Perl6::Ecosystem, or the like.
jaffa8 for bad regular expression
it does not give a line number
moritz_ rakudo: /,/ 12:48
p6eval rakudo 0d4fe0: OUTPUT«Syntax error at line 2, near "/,/"␤in Main (src/gen_setting.pm:3390)␤»
moritz_ well, in this case it gives a line number.
masak mberends: yes, keeping the old basedir is fine, but I think I plan to remove all PERL6LIB handling for those...
jaffa8 rakudo: while ( $filenamelist =~ /\G(.+?),|\G(.+)/g ) {}
DarkWolf84 masak, I'll try but I've never pasted bug before
p6eval rakudo 0d4fe0: OUTPUT«perl6regex parse error: Alphanumeric metacharacters are reserved at offset 38, found '('␤in Main (src/gen_setting.pm:3390)␤»
moritz_ oh.
jaffa8 here you are
masak DarkWolf84: you meant moritz_. :) it's not that hard.
DarkWolf84 I'm not shure what to do
ok
moritz_ jaffa8: care to submit a bug report?
masak DarkWolf84: paste enough information to reproduce the problem. 12:49
DarkWolf84 ok
jaffa8 where is it?
moritz_ jaffa8: just write an email to [email@hidden.address] includiing this example and the complaint that it gives no line number.
DarkWolf84 ok
I'll do it
masak DarkWolf84++ 12:51
jaffa8 I have done it
mberends masak: yes, the old basedir must not exist in PERL6LIB. Would you be happy to ask the user for permission to wipe the old config.proto and automatically create a new one? 12:53
masak jaffa8++
moritz_ do you really have enough users alreay to care about migration/backward compatibility? 12:54
masak mberends: that's even better. we'll go with that.
mberends @karma jaffa8
lambdabot jaffa8 has a karma of 1
masak moritz_: (1) yes, (2) what we're doing is the next simplest thing from not caring at all.
moritz_ masak: ok 12:55
masak: do you happen to know how many users you have, roughly?
masak if the user upgrades, she will need a new version of the config file anyhow.
mberends moritz_: all of Austria for starters
masak moritz_: that's hard to tell. we don't phone home or anything.
moritz_ mberends: oh, I forgot :)
masak moritz_: but I'm sometimes pleasantly surprised at seeing projects created by mberends' script. 12:56
so people seem to tend to use it.
& # commuting
12:56 masak left
mberends the main reason proto doesn't phone home is the hardware and bandwidth it would cost us 12:57
moritz_ ;-)
DarkWolf84 oops 13:00
i've pasted the bug too 13:01
13:04 jaffa8 left 13:05 xinming_ is now known as xinming 13:09 jaffa8 joined
jaffa8 I have this in my code 13:09
print <<eof;
What can I use in Perl 6? 13:10
13:22 masak joined
colomon jaffa8: that's not implemented in Rakudo yet. 13:25
mberends jaffa8: print " 13:26
jaffa8: hello
jaffa8: "; # multiline string
pmichaud there's also: 13:28
print qq{
hello
};
jaffa8 ok 13:29
mberends hi ^
pmichaud (good morning, #perl6)
mberends oops hi pmichaud
jaffa8 rakudo: $b=qq<wewewewe>;
p6eval rakudo 0d4fe0: OUTPUT«Symbol '$b' not predeclared in <anonymous> (/tmp/Cx4nd5be0U:2)␤in Main (src/gen_setting.pm:3390)␤»
jaffa8 rakudo: my $b=qq<wewewewe>; 13:30
p6eval rakudo 0d4fe0: ( no output )
jaffa8 where can I get and std parser?
pmichaud if you want to build the std parser for yourself, it's in the pugs repository 13:31
if you just want something online, use "std: " on this channel
std: my $b=qq<wewewewe>;
p6eval std 28050: OUTPUT«ok 00:04 39m␤»
pmichaud see also perl6-projects.org for links to the std parser 13:32
s1n morning all
phenny s1n: 12:00Z <masak> tell s1n not an ironic name. but I wouldn't say that the use case is very clear either right now. Web.pm is a set of modules at different levels. maybe it'll make sense to load them through 'use Web;' at the top, maybe not.
s1n: 12:04Z <masak> tell s1n cloning from git://github.com/viklund/november.git works fine here.
masak s1n: the 'Web.pm' brand is important, but I don't know yet whether it'll be part of use statements. it's first and foremost a social project, an incubator for web projects that we need. I'm pretty sure that the MVC framework will not be called 'Web'. that's why I removed those stubbed module files. 13:35
mornin', pmichaud. 13:36
s1n masak: it's gathered attention as Web.pm, not Web, you've built a brand around a non-existant module...
masak: i understand where you're coming from, but newcomers wishing to use it may not
masak s1n: that's a valid point. 13:37
s1n: if there's a simple mechanism by which Web.pm can pull in those other 'endorsed' modules, I think that'd be a good use for it.
s1n pmichaud: do hll_globals have any mechanism for callback when being assigned? i want to perform an action in C/PIR when one is set
masak: sure, it doesn't sound like an easy fix, but you should look into it, i was rather surprised to see there was no Web.pm, maybe rebuild the brand or come up with a sort of component loading module named Web.pm... 13:40
masak s1n: it could be an easy fix. I'm sure it'll come back before week 20.
mberends pmichaud: after using rakudo/tools/test_summary.pl rather a lot, there seems little reason for both 'test' and 'plan' counts per script, so I've posted a deprecation notice on line 132. Do you sometimes need the separate figures or can one be dropped, and if so, when?
s1n masak: okay, it's your project, i just wanted to give you an outsider's perspective 13:41
masak s1n: much appreciated. thanks.
pmichaud mberends: looking 13:42
mberends pmichaud: the polite timing would be after the September release 13:43
btw, the planless testing discussed recently would wreck our spectest statistics :-( 13:44
moritz_ mberends: I know, which is why I use it only sparingly 13:45
pmichaud mberends: all that I really need/use personally at the moment are the statistics at the end
mberends ok, I reformatted them a little as well for you :) 13:46
13:46 braceta joined
masak I've come to really like planless testing. but I see the argument for not using it in the Perl 6 spectests. 13:47
13:49 icwiener joined
pmichaud mberends: I'll run my daily tests with the revised test-summary.pl script and can give better feedback then 13:50
(running them now, will take an hour or two)
mberends the 'test' column can be dropped because (1) it's the sum of several other columns (2) it should match 'plan' (3) it doesn't occur in the subtotals at the end
pmichaud I think I'm fine with dropping the 'test' column, yes.
we can go ahead and do that now, and if it's a problem for someone we'll put it back
(or come up with some other answer)
mberends pmichaud++ 13:51
13:52 simcop2387 left
moritz_ as long as the format of spectest-progress.csv doesn't change I'm fine with that 13:53
mberends moritz_: yes, that's staying the same
pmichaud afk # one last trip to the airport 13:58
jaffa8 rakudo: ">${filename}.dbg" 14:00
p6eval rakudo 0d4fe0: OUTPUT«Can't use $ as non-variable in interpolated string at line 2, near "${filename"␤in Main (src/gen_setting.pm:3390)␤»
jaffa8 std: ">${filename}.dbg"
p6eval std 28050: ( no output ) 14:01
jaffa8 how would you do this in rakudo: "${fn}"?
masak rakudoÖ sub foo { "OH HAI" }; say "like this: {foo}" 14:02
rakudo: sub foo { "OH HAI" }; say "like this: {foo}"
p6eval rakudo 0d4fe0: OUTPUT«like this: OH HAI␤»
jaffa8 I mean a variable 14:03
how would you do this in rakudo: "${thisisavar}thisisnotavariablename"?
masak rakudo: my $variable = "OH HAI"; say "{$variable}likethis" 14:04
p6eval rakudo 0d4fe0: OUTPUT«OH HAIlikethis␤»
masak rakudo: class Austria { BEGIN { say self } }
p6eval rakudo 0d4fe0: OUTPUT«Lexical 'self' not found␤»
masak pmichaud: do you know if this has been ticketed?
14:10 nihiliad joined
jaffa8 std: if ( $l ~~ /^(.+?);(.+?);(.+?);(.+?)(?=\s)/ ) {} 14:12
p6eval std 28050: OUTPUT«===SORRY!===␤Semicolon must be quoted at /tmp/0GlqyoJKBH line 1:␤------> if ( $l ~~ /^(.+?);⏏(.+?);(.+?);(.+?)(?=\s)/ ) {}␤Other potential difficulties:␤ Variable $l is not predeclared at /tmp/0GlqyoJKBH line 1:␤------> if ( $l⏏
..…
masak jaffa8: so, the rule is: if it's an alphanumeric, you don't need to escape of quote it. if it's not, you do. (if you mean it literally and not as an operator) 14:17
jaffa8 ok, $1 is $0 no 14:18
w
masak aye. 14:19
ok, my latest Rakudo doesn't build.
p6opaque.c:275: error: 'struct PMC' has no member named 'pmc_ext' 14:20
no activity today 'sides mberends++' patch, so it's probably something Parrot did. 14:21
jaffa8 How would you do something like this? 14:23
push @{ $table{ lc "$1.$2" } }, lc $3;
moritz_ perlgeek.de/blog-en/perl-6/object-c...ation.html there you go, corrections very welcome
jaffa8: %table{ lc "$0.$1" }.push: lc $2;
rakudo: say %*VM<config><revision> 14:24
p6eval rakudo 0d4fe0: OUTPUT«40556␤»
jaffa8 rakudo: %table{ lc "$0.$1" }.push: lc $2;
p6eval rakudo 0d4fe0: OUTPUT«Symbol '%table' not predeclared in <anonymous> (/tmp/d85m6y719M:2)␤in Main (src/gen_setting.pm:3390)␤»
masak rakudo: my %table = "foo.bar" => []; "foo bar baz" ~~ / (\w+) \s* (\w+) \s* (\w+) /; %table{ lc "$0.$1" }.push: lc $2; say %table.perl
p6eval rakudo 0d4fe0: OUTPUT«{"foo.bar" => ["baz"]}␤»
jaffa8 moritz_, Did you use : 14:25
rakudo: %table{ lc "$0.$1" }.push(lc $2);
p6eval rakudo 0d4fe0: OUTPUT«Symbol '%table' not predeclared in <anonymous> (/tmp/VkUhvxwpad:2)␤in Main (src/gen_setting.pm:3390)␤»
moritz_ jaffa8: you have to declare variables with 'my' before you can use them.
like masak++ did
jaffa8 that I understand
What about >>:<<?
moritz_ that's a syntax for a method call 14:26
$foo.method: a, b, c
masak 'invocant colon'
moritz_ is the same as $foo.method(a, b, c)
jaffa8 you mean this is alternative syntax
moritz_ yes
masak artificial sweetener. 14:27
jaffa8 funny
moritz_ convenient.
14:27 Psyche^ joined
masak unnecessary. 14:27
moritz_ like perl ;-)
jaffa8 faster
masak error-prone.
moritz_ don't we all love listops? ;-) 14:28
masak no.
jaffa8 Why is it error-prone?
carlin masak: rakudo builds alright for me after git pull and svn up'ing parrot
masak jaffa8: I guess it's a matter of taste, but I find the parentheses clearer.
carlin: that's interesting.
moritz_ tries
14:29 jan_ left, icwiener left
masak carlin: I did realclean on Parrot, and install-dev. then I did make clean on Rakudo and tried to build it. did you do the same? 14:30
moritz_ p6opaque.c: In function ‘Parrot_P6opaque_clone’:
p6opaque.c:275: error: ‘PMC’ has no member named ‘pmc_ext’
that's what I get with latest rakudo+parrot
jaffa8 􏿽xCDs it possible to change the character set of the source?
What if I do not want utf8? 14:31
masak moritz_: yup, same as mine. 14:32
jaffa8: Rakudo is a bit weak on encodings at present. it's one of the things being worked on. contributions are welcome. 14:33
moritz_ jaffa8: implementations should auto-detect character encodings on their BOM if present. In other cases you have to pass an option to the compiler 14:34
something like perl6 --encoding=UTF-16 script.pl
don't know what the exact syntax is, but it's not yet supported
japhb urgle. The problem with having kids on a school day schedule is that they stay on it on the weekends, even when you need more sleep .... 14:35
moritz_ ouch. 14:36
japhb: I think it was you who asked about the various object initialization methods in Perl 6. I wrote something up here: perlgeek.de/blog-en/perl-6/object-c...ation.html
carlin after make realclean on parrot it still wfm ... so I must be doing something different 14:37
or wrong
mberends moritz_++: nice article. spello: initiliaztions
jaffa8 what is this in Perl 6: $l = <dic>? 14:38
japhb moritz_, Ah, thank you. I'll look at that in a bit, as soon as I start some rebuilds.
mberends jaffa8: <a b c> is the new qw{a b c}
jaffa8 Is it now readline? 14:39
moritz_ japhb: thanks
jaffa8 Is qw throuwn out?
moritz_ it's spelled <...> now 14:40
mberends jaffa8: it's quote words. yes, afair qw{} is gone.
moritz_ of or q:w{...}
which can be shortened to qw{...} I think
std: qw{a b}
p6eval std 28050: OUTPUT«ok 00:03 37m␤»
moritz_ yes
mberends timtowtdi 14:41
jaffa8 std: print qw{a b}
p6eval std 28050: OUTPUT«ok 00:02 37m␤»
DarkWolf84 I made a executable with pbc_to_exe but it ends with "Class '[ 'parrot' ; 'Perl6MultiSub' ]' not found"
moritz_ but notice that you can't use qw() with round parenthesis, that's a sub call
japhb So what's the canonical Parrot / Rakudo build sequence now that the "builds against installed Parrot" thing happened? I'm guessing 'cd <parrot_build_dir>; perl Configure.pl; make install; cd <rakudo_build_dir>; perl Configure.pl; make install'?
moritz_ japhb: you're missing 'make install-dev' for parrot
japhb moritz_, instead of or in addition to 'make install' for parrot?
moritz_ japhb: in addition
mberends japhb: and Rakudo's Configure.pl needs --parrot-config= 14:42
japhb moritz_, ah, so install-dev adds additional files needed for building things?
mberends japhb: right
moritz_ japhb: yes
mberends: unless it's installed in rakudo/parrot_install/, which is the default invocation with --gen-parrot 14:43
japhb mberends, --parrot-config= what? Directory of parrot install, or path to config binary?
mberends the binary
14:43 Patterner left, Psyche^ is now known as Patterner
japhb mberends, OK, got it. 14:43
jaffa8 is pbc to exe reliable?
japhb
.oO( Why would anyone install parrot except to build stuff on it? )
 14:44
moritz_ DarkWolf84: did you install rakudo? ('make install')
japhb goes to look at install-dev target
mberends jaffa8: for making perl6.exe, definitely. for other programs ymmv.
DarkWolf84 yeah 14:45
jaffa8 ymmv?
japhb Your Mileage May Vary
DarkWolf84 but it instals in ./parrot_install
mberends jaffa8: your success will depend on many other things :) 14:46
jaffa8 SUch as?
luck?
14:46 c9s___ is now known as c9s
jaffa8 What are the known problems? 14:46
mberends jaffa8: there are other ways to make .pbc files, and depending on what goes into them, the result may either work or fail 14:47
jaffa8 that sound pretty cryptic to me. 14:48
What are other ways?
Hex editor
?
I guess everybody uses parrot.exe to make pbc fies.
14:49 DanielC joined
mberends jaffa8: there are lots of examples of .pir source programs that your could compile to .pbc and maybe to .exe 14:50
in the parrot source tree
jaffa8 How can one deploy files? 14:51
IS there program for that?
mberends ouch, we're struggling to make that still
proto (being discussed today) is an early attempt, there are other projects being worked on. 14:52
japhb proto had discussions earlier today? What time frame?
I'm working on the parrot module ecosystem today, in fact.
mberends irclog.perlgeek.de/perl6/2009-08-23#i_1420618 14:53
jaffa8 how many files are needed for a simple deployment?
E.g. hello world
I guess parrot exe and lib are needed. 14:54
japhb mberends, excellent, thank you
mberends japhb: I like your plan so far
japhb prepares to eat proto's brains
mberends, you mean trac.parrot.org/parrot/wiki/ModuleEcosystem ? 14:55
mberends japhb++: yes
japhb mberends, ah, thank you!
mberends jaffa8: 14:56
jaffa8: yes, and the perl6 executable
jaffa8 Has anyone seen this: Pugs can compile Perl 6 programs into Perl 5 source code 15:02
209.85.129.132/search?q=cache:-mnYw...&gl=hu
mberends japhb: this project may overlap a small part or yours, for the Rakudo case only: gitorious.org/parrot-module-lib/mai...er/ROADMAP 15:06
*of yours
15:06 Front_slash left
japhb mberends, reading .... 15:07
mberends, ooh, yeah, all of that looks useful. 15:09
japhb adds parrot-module-lib to the @eat-these-brains
carlin freshly cloned rakudo, I ran Configure --gen-parrot, make clean, cd to parrot, make realclean, svn up, Configure, make install-dev, cd to rakudo, Configure, make install. Did I miss anything? 15:10
masak carlin: no, it all looks good. I'm running Rakudo-in-Parrot, though. but that shouldn't make a difference. 15:20
carlin builds fine :s 15:21
15:23 molaf joined
jnthn oh hai 15:27
masak yayitsjnthn
jnthn japhb: I'm around for the rest of my $day
yayitsmasak!
masak: What's cooking today?
15:27 justatheory joined
japhb jnthn, Great! How long will that last? :-) 15:27
(timezones are one thing, sleeping schedules another ...)
jnthn japhb: Well, I got up quite early-ish today, but I reckon there's a good 6-7 hours yet. :-) 15:28
15:29 simcop2387 joined
japhb jnthn, :-) 15:29
jnthn masak: Is the build fail with latest Parrot rather than recommended one? 15:31
Su-Shee is poll and select already implemented? 15:32
masak jnthn: yes. some parrot after 40719 breaks Rakudo.
carlin: ah, and that explains why it runs for you. of course... :/ 15:33
jnthn Parrot is really good at breaking the Rakudo build these days. 15:34
In fact, it's really excelled at it since the last release.
masak it doesn't bother me much, as long as there's not a recent feature/fix I really need.
jnthn masak: Sure, but really best is to build the Parrot that Rakudo says it needs. 15:35
masak jnthn: I'm sure it is. I just like it this way. :)
I'm usually among the first to spot this type of error.
jnthn Aye. 15:38
mberends Su-Shee: no select: irclog.perlgeek.de/perl6/2009-08-10#i_1379225
Su-Shee just found another interesting posting on the subject... 15:39
perl-yarg.blogspot.com/2009/08/work...arrot.html
ah. which is the same fellow. :)
15:44 rindolf joined
rindolf Hi all. 15:44
Su-Shee: here?
Su-Shee yes. 15:45
rindolf Su-Shee: hi.
Su-Shee: what's up?
Su-Shee: BTW, why aren't you present on #perl no longer?
Su-Shee rindolf: it's a little too much alphamale-ing sometimes, so I take a break. ;) 15:46
rindolf Su-Shee: ah.
Su-Shee rindolf: just tested a teenytiny blog engine based on mojo, very nice. 15:47
DarkWolf84 what i need to use executable generated from pbc_to_exe 15:48
jaffa8 std:/^-h|--help$/ 15:50
std: /^-h|--help$/ 15:51
p6eval std 28050: OUTPUT«===SORRY!===␤Unable to parse regex; couldn't find final '/' at /tmp/ZbKfXvUEHo line 1:␤------> /^⏏-h|--help$/␤ expecting any of:␤ infix stopper␤ quantifier␤ regex atom␤ regex_infix␤ standard stopper␤ terminator␤
..ws␤FAILED 00:02 37m␤»
jaffa8 std: /^\-h|\-\-help$/
p6eval std 28050: OUTPUT«ok 00:02 37m␤»
sri_kraih really hopes rakudo star will have async io 15:52
rindolf Su-Shee: ah. 15:53
Su-Shee: does it have comments?
Su-Shee: Altreus and I are writing our own blog engine. Based on Cat, though.
Su-Shee rindolf: I know, I'm very much waiting for it. ;)
rindolf Su-Shee: it will be ready faster if you contribute<tm>. 15:54
sri_kraih rindolf: comments are so 2008...
in 2009 everybody just tweets
rindolf sri_kraih: you mean now people post responses as separate blog entries on their blogs?
Su-Shee rindolf: I focus on gui stuff. I simply can't to everything :/
ah, THOSE comments :)) I thought comments in the source :)) 15:55
jnthn sri_kraih: Was perl-yarg.blogspot.com/2009/08/work...arrot.html your post?
sri_kraih jnthn: nope
rindolf Su-Shee: yes, blog comments.
jnthn sri_kraih: Ah, OK.
sri_kraih but i can relate to it 15:56
jnthn sri_kraih: It's a good post, that's all. :-)
rindolf sri_kraih: I'm using plurk.com which is a micro-blogging service that has micro-comments.
sri_kraih would love to port Mojo to Rakudo...
15:56 Chillance left, missingthepoint left
japhb rindolf, 'micro-comments'? What are those, 14 chars max? 15:56
rindolf japhb: no. 140
japhb "Ogg agree." 15:57
rindolf, just teasing. :-)
rindolf japhb: ah.
15:57 Confield joined
Su-Shee japhb: "OMG!" and "SRY :)" fits in less. 15:57
jnthn sri_kraih: Seeing people saying "we could really use this to do X concrete application" is good to know. :-)
masak here's a post that audreyt++ just made: pugs.blogs.com/audrey/2009/08/why-such-me.html
sri_kraih i think as soon as rakudo does async io i can have port a full http 1.1 server over
rindolf Su-Shee: there was a "Noise to Signal" strip about 1-char micro-blogging.
15:57 FCO joined
sri_kraih *-have 15:58
jnthn sri_kraih: Since it helps give some focus.
15:58 Confield left
masak sri_kraih: let me know when you embark on such a project. I might want to be at least tangentially involved. 15:58
15:59 Confield joined
rindolf masak: had audrey been sick? 15:59
masak rindolf: yes. hepatitis.
and burnout.
rindolf masak: ah.
15:59 rblasch_ joined
sri_kraih masak: sure 15:59
rindolf masak: yes, I wondered where she has been. 16:00
16:00 molaf left
japhb rindolf, the burnout as I understand it contributed mightily to the hepatitis. 16:01
rindolf japhb: ah.
jaffa8 Who burnt out?
rindolf jaffa8: audreyt.
jaffa8 When?
masak now that I think about it, the vacuum after audreyt++'s disappearance from this channel, and from Pugs development, hit me just as hard as the _why disappearance seems to hit the Ruby community.
jaffa8: around 2007.
jaffa8 What has he been doing since then? 16:02
moritz_ she, actually
jaffa8 that is 2,5 years
masak jaffa8: recovering, I imagine.
moritz_ working on other things
not as exposed as before
japhb masak, yeah, losing audreyt++ was painful. Kept asking after her, but noone who knew was telling. 16:03
rindolf en.wikipedia.org/wiki/Hepatitis - wow, sounds nasty.
jaffa8 She was employed
was not she?
masak japhb: the news trickled down over the years after that. still, there's a clear hole left after her. 16:04
rindolf he.wikipedia.org/wiki/%D7%93%D7%9C%...7%91%D7%93
masak jaffa8: yes.
japhb masak, NODNOD
moritz_ www.perlmonks.org/?node_id=790671 perl 5.10.1 released! YaY!
jaffa8 Why are you showing Hebrew webpages? 16:05
japhb YES!
M_o_C Huh, released on the weekend :o
16:05 SmokeMachine left
rindolf jaffa8: because Hebrew is my native tongue. 16:05
jaffa8 ok 16:06
rindolf jaffa8: I didn't know what Hepatitis was, but I was familiar with the Hebrew term.
jaffa8 ok
rindolf jaffa8: there's a link to the English version.
jaffa8 Are you in Israel?
rindolf jaffa8: yes, I live in Tel Aviv.
jaffa8 must be fun 16:07
16:08 rblasch left
M_o_C Did someone apply Whiteknights patch to master yet? 16:08
jaffa8 Since when have you lived there?
rindolf jaffa8: since I was 5 years old.
jaffa8: I was born here. 16:09
But my parents moved to the states when I was 1 y.o.
jaffa8 Are you Russian?
rindolf jaffa8: no.
jaffa8: I'm Jewish.
M_o_C nvm, apparently they still test it.
rindolf jaffa8: and Israeli born.
japhb chuckles at lines 99-101 of gitorious.org/parrot-module-lib/mai...ster/HOWTO
rindolf jaffa8: I know many Russians though. 16:10
jaffa8 DId you parent come from Russia?
rindolf jaffa8: no.
jaffa8: my mother was born in Iraq.
16:10 rblasch__ joined
rindolf jaffa8: and my father is also Israeli born, to Polish and Lithuanian immigrants. 16:10
jaffa8 Is not that unusual?
rindolf jaffa8: not all Israelis are Russian.
jaffa8: or their parents came from Russia.
jaffa8 that is clear from me 16:11
I heard there were so many Russians there, one can hear Russian speaking everywhere
on the beach.
rindolf There are many Moroccans, Romanians, Tunisians, Syrians, Yemenites.
jaffa8: hearing Russian spoken is not uncommon. 16:12
We've received many immigrants from the former USSR.
jaffa8 Do you practice your religion? 16:13
rindolf jaffa8: no, I'm not religious. 16:14
jaffa8: I'm a Jew by peopleship - not by faith.
jaffa8: I do however, not eat leavened bread on Passover.
But I use electricity on Sabbath, etc. 16:15
And eat non-kosher foods.
jaffa8 Why do you keep one rule and not another?
japhb He likes matzoh? 16:16
:-)
rindolf jaffa8: well, I consider the Passover matzos thing an ethnic custom.
Which I like to observe.
M_o_C Hm, so some jews also refuse the use of electricity on Sabbath?
Su-Shee jaffa8: for the same reasons on fridays in germany many people eat fish and no meat - "custom derived from relegion" :) 16:17
japhb M_o_C, there is a *lot* that is forbidden on the Sabbath, for a strict observer.
rindolf I think that God, if he exists, cares little for old testament ethics.
jaffa8 probably
rindolf M_o_C: yes, some Jews don't use electricity on Sabbath.
jaffa8 or he may be in another galaxy
rindolf jaffa8: heh.
jaffa8 leading another group of people
japhb obStargateReference 16:18
M_o_C japhb: Well, I'm not really that well informed in regards of the Tora...
rindolf Su-Shee: is this something related to Christianity?
16:18 rblasch_ left
rindolf M_o_C: according to the interpretation of the Torah, it was forbidden to light fire. 16:18
But you could still keep a fire burning.
M_o_C: light fire in Sabbath I mean. 16:19
M_o_C: and Jews concluded that Electricity was fire-like.
cognominal rindolf, since the old testament, jewish have been in the entertainment business meaning writing and now producing fictions
M_o_C Yep, I interpreted it that way. Thanks for explaining.
Su-Shee rindolf: yes, a "fasting on friday" thing.
rindolf cognominal: :-)
16:19 adam-pwgsc joined
rindolf cognominal: "God is my favourite fictional character." -- Homer Simpson. 16:20
Su-Shee rindolf: and fish is/was considered no meat and therefore fasting.
rindolf Su-Shee: ah.
Su-Shee: what about vegetables?
cognominal rindolf, I suppose that if you are not restricted to kasher, you can eat camels too.
I don't remember the exact rule about 2 toed animals... 16:21
rindolf cognominal: Camels may be kosher.
cognominal: not sure.
KatrinaTheLamia rindolf, why is it that Homer Simpson gets all the most intelligent lines... despite being a rather dumb characters
rindolf cognominal: they should either make a hoof print or recycle their food.
Su-Shee rindolf: also allowed, I think. several german dishes go back to this "no meat" thing and have meat not visually but hidden included in the dish. ;)
rindolf KatrinaTheLamia: I don't know.
Su-Shee: ah. 16:22
KatrinaTheLamia: it's been a while since I watched the Simpsons.
But I used to like it.
jaffa8 what is minus in regular expressions? 16:23
rindolf jaffa8: in p6?
jaffa8 yes
rindolf why does my spell checker highlights Jaffa? It's a very old city for crying out loud. 16:24
masak rakudo: say "foo" ~~ /<-[bar]>/
p6eval rakudo 0d4fe0: OUTPUT«f␤»
rindolf masak: negation?
masak jaffa8: it's a negated character class.
rindolf: it's a negated character class. 16:25
rindolf masak: ah.
masak: but not negation in general?
masak rindolf: what would that mean inside a regex?
rindolf masak: like match anything except "masak"?
japhb mberends, am I correct in reading gitorious.org/parrot-module-lib/mai...c/json.pir that the runtime/parrot/library/JSON.pir does to-json only, and not from-json, by design?
rindolf masak: a lot of people ask us about it. 16:26
masak :)
rindolf: the closest two things I can think about are the !~~ operator and <!after ...> negative lookahead.
rindolf masak: perl.net.au/wiki/Freenode_Sharp_Per...ression.3F
masak: I'm not saying implementing it would be a good idea.
masak rindolf: ISTR regexes can have !~~ inside of them. 16:27
japhb Why doesn't 'perl6 --version' include either a release number or a git rev?
rindolf masak: ah.
masak but I don't know if it's implemented in Rakudo/PGE yet.
mberends japhb: lines 13-21 should do from-json
masak japhb: you're right, it definitely should, at least the release name.
japhb: care to submit a Rakudobug? 16:28
japhb masak, I can, but it will have to wait. I'm trying to be head down right now.
masak ok. later works.
japhb (Yesterday was my birthday, but I had to go to someone else's party. So for my birthday I'm getting a day to do nothing but hack on Parrot, and dangit I am making the *most* of it!) 16:29
masak :D
Su-Shee got herself a new job for her birthday last week. ;) 16:30
japhb Su-Shee, awesome
jaffa8 What kind of job? 16:31
16:31 DarkWolf84 left
Su-Shee japhb: yeah, only the gift wrapping was a little difficult.. ;) 16:31
jaffa8: coding perl. :)
dalek kudo: e2b5e8f | moritz++ | (2 files):
re-enable building on latest parrot. Also bump PARROT_REVISION
rindolf Su-Shee: mazal tov!
Su-Shee: for your birthday I mean. 16:32
Su-Shee rindolf: that's for birthdays as well?
rindolf Su-Shee: yes.
masak Su-Shee: grattis! # sv 'congratulations'
rindolf masak: slovekian? Slovenian?
Su-Shee rindolf: ah, ok. :) 16:33
masak Swedish.
rindolf masak: oh.
:-S
masak: gratis means costless in English.
Must be confusing.
masak in Swedish too.
Su-Shee rindolf: in german as well.
masak it's not confusing, the vowels are different.
jnthn
.oO( gratis is accepted as an English word now? )
Su-Shee masak: ttis makes the i short and tis is with long i?
masak Su-Shee: no, the 'a' before the 't' or 'tt'. 16:34
Su-Shee ah. ok.
jnthn: wasn't everything latin? ;)
masak Su-Shee: and 'short'/'long' actually changes the kind of vowel, not just the length.
japhb jnthn, yes, quite 16:35
masak the 'a' in 'gratis' sounds like the one in 'father'. the 'a' in 'grattis' sounds like the one in 'pattern'.
japhb heck, half of English is Latin by way of French anyway ....
masak indeed.
rindolf japhb: some of it is directly from Latin. 16:36
jaffa8 Rakudo is not very fast
japhb rindolf, sure
jaffa8 it ran for 45 seconds for me
Su-Shee rindolf: what isn't in indoeuropean languages.. ;))
rindolf But it's all lies. Half of English is from Yiddish..
masak japhb: it hasn't been optimized much.
rindolf Su-Shee: also in Persian and Indian languages? 16:37
japhb rindolf, the swearing?
;-)
rindolf Su-Shee: they are IE too.
Su-Shee: but I imagine they are less influenced than Latin than European IE languages.
cognominal rindolf, you got some chutzpah to pretend that 16:38
rindolf got stuck at work today.
cognominal: :-)
Su-Shee
.oO(chuzpe..) :)
rindolf chutzpah is an Hebrew word in origin. 16:39
16:39 zloyrusskiy joined
rindolf I think it appears in the Bible. 16:39
Su-Shee rindolf: I know, we use it in german, but it's writte chuzpe here. 16:40
16:40 jaffa8 left
rindolf I once heard a speech given in Anglo-Saxon. 16:40
cognominal it is unkown in French
rindolf Su-Shee: how is it pronounced?
KatrinaTheLamia English? There is no language known as English. English describes a style of syntax for matching words together. The words themselves come almost entirely from other languages.
Su-Shee rindolf: oh yeah... how do I write that.. :) 16:41
rindolf cognominal: en.wiktionary.org/wiki/chutzpah
huf KatrinaTheLamia: that's true for all languages as far as i know
japhb KatrinaTheLamia, that makes it no different than any other natural language ....
huf japhb++
rindolf japhb++
en.wiktionary.org/wiki/Chuzpe#German 16:42
No pronunciation unfortunately.
No wonder you prefer it over Dreistigkeit
Su-Shee rindolf: the chu a little like in church and then with a sharp sz, short p and the e is like hm... 16:43
moritz_ prefers Dreistigkeit, actually
rindolf moritz_: ah.
KatrinaTheLamia japhb, huf yeah... I've heard all sorts of wild and crazy definitions for what constitutes something as a "natural" language. I dunno, most of what I read makes me just want to learn and use sanskrit instead >.>
rindolf moritz_: you have some chutzpe to prefer it.
moritz_ rindolf: no, freedom ;-)
the 'e' is not pronounced at all, I think
rindolf moritz_: I was joking . :-) 16:44
masak rindolf: my Google-fu cannot help me answer the question whether chutzpah is a word from the Bible or not. :/
rindolf We pronounce Hutzpah with the stress on the "ah".
As opposed to the Yiddish Hooooooooooootzpah
Su-Shee moritz_: my grandmother used it and she pronounced it like tschuszpe.
japhb mberends, let me see if I understand ... compilers/json/JSON.pir first does a PGE match against the JSON grammar, then uses JSON::PIR to convert the match to PIR source, then uses PIRC to compile it to bytecode, then tailcalls the bytecode to produce the thawed data structure?
rindolf Su-Shee: is verboten pronounced as fferboten? 16:45
Su-Shee rindolf: yes, in that case.
mberends japhb: checking.. 16:46
lisppaste3 MoC pasted "Rakudo Makefile" at paste.lisp.org/display/85906
moritz_ www.nntp.perl.org/group/perl.module...67299.html # my mail to [email@hidden.address]
Su-Shee rindolf: if a word derives from greek/latin and starts with a "v" it's more like "w"
rindolf en.wiktionary.org/wiki/verboten - what is its etymology?
Su-Shee: ah.
Su-Shee: "more like"?
Su-Shee rindolf: pronounced more like a w.
rindolf [OE. forbeden, AS. forbeódan; pref. for- + beódan to bid; akin to D. verbieden, G. verbieten, Icel., fyrirbj&omac;&edh;a, forbo&edh;a, Sw. förbjuda, Dan. forbyde. See Bid, v. t.] 16:47
Su-Shee rindolf: verboten comes from "gebote" (10 commentments). if you _have_ to do something, it's a "gebot". if you must NOT do something it's a "verbot"
rindolf Seems like its Germanic in origin. 16:48
machaut.uchicago.edu/?resource=Webs...use1913=on
cognominal I think american english is not frozen by an academy like french.
mberends moritz_++ for Perl 6 modules on CPAN email suggestion
Su-Shee rindolf: yeah, looks like it. the important part is the "bot" in the middle. ;) 16:49
cognominal French evolves thx to semi-illiterate people like sons of arabic immigration that don't give a shit about the Académie Française.
Su-Shee cognominal: but the french have extremely cool dictionaries thanks to the academy. :)
cognominal I don't think you have it on line like the very good Littré 16:50
There is even a free Littré app on the iPhone.
rindolf cognominal: there's fr.wiktionary.org 16:51
It's bigger than en.wiktionary.org
cognominal en.wikipedia.org/wiki/Dictionnaire_...e_(Littré)
Su-Shee cognominal: yeah, but I'm loughing every day about the "petit robert" and its 9cm "petitness" in my shelf. ;)
rindolf Su-Shee: heh, heh.
Su-Shee: lough is a lake, BTW.
Like Loch.
My spell-checker highlights it. 16:52
It didn't use to.
moritz_ except that most English can't pronounce "loch" ;-)
rindolf Maybe the en_GB one is incopmlete.
lokh, no?
moritz_ no
rindolf Like Genghis Khan.
cognominal for an up to date online dictionnary (but not free) there is Antidote RX done by people from Quebec lacks etymology though
moritz_ that's what most say, and that's what's wrong
rindolf Or Khalif.
moritz_: ah.
moritz_ rindolf: it's pronounced with soft ch like you find in German 16:53
Su-Shee rindolf: I _think_ our ch is like an arabian h.
cognominal loughing?
rindolf Su-Shee: which one? a Kh or a 3? Muhhamad or Khalif/Khan? 16:54
Su-Shee: two different sounds.
cognominal I don't think the Larousse or the Petit Robert are available on computers in any useable form. Last time I checked one used to have the CD/DVD in the reader to use it. 16:55
Su-Shee rindolf: the muhammad h.
masak mberends: I got distracted by some ideas I had for a file-backended mini-ORM, as a precursor to an early MVC for Web.pm.
cognominal come on. Why not dongles why they are at it.
mberends japhb: by golly you're right, the JSON data is translated to PIR which is then compiled in line 70 and executed in line 71
masak mberends: I'd be happy to pick up the proto restructuring at some later time.
Su-Shee moritz_: let's hope, someone answers.. 16:58
mberends masak: yes distracted randomly here too. getting ready for a pretty week ahead too.
masak mberends: aye. same here.
mberends *pretty busy
17:01 hanekomu joined
japhb mberends, OK, so I'm not entirely insane. 17:01
Su-Shee moritz_: why don't you write a nice perl 6 programming book? 17:02
japhb mberends, are the JSON grammar and emitter fully compliant? Or are there things in JSON still not handled correctly?
moritz_ Su-Shee: I'm really thinking about it.
Su-Shee moritz_: I suggest "contemporary object oriented programming with perl 6" 17:03
mberends japhb: it is a fairly whacky way to convert a data file to a data structure. when your only tool is a hammer, every problem looks like a nail ;)
japhb mberends, :-)
moritz_ Su-Shee: "Perl 6" implies "object oriented" already ;-) 17:04
japhb mberends, question still applies though ... can I assume that the current JSON code is just "correct", and not worry about edge cases? 17:05
Su-Shee moritz_: yeah, but I meant something focussing on architecture and roles and all this stuff contrary to something like "the handbook" or "perl 6 for computer linguists" (which would be probably later be called "the grammar for grammars" or something like this. ;)
moritz_ Su-Shee: problem is, I don't understand most of Perl 6's OO :) 17:06
Su-Shee moritz_: well if YOU don't - what will the rest of us do? ;)
mberends japhb: there is meant to be a 'make test' to give some idea of correctness. I just passed all 78 tests here :-) 17:07
Su-Shee I feel bad every day because I don't use all the cool features.
japhb mberends, all right, I'll call that infrastructure for now then, thx
moritz_ Su-Shee: jnthn, ruoso and TimToady probably understand it 17:08
Su-Shee oh, that many? ;)
mberends japhb: agreed
moritz_ Su-Shee: presumably pmichaud too
Su-Shee moritz_: "gee, they need a second car soon.." ;) 17:09
rindolf They tell that after Millikan received his Nobel Prize, someone at the news conference asked him "Is it true that only three people understand Einstein's General Theory of Relativity". So he replied: "Really? Who's the third?" 17:11
en.wikipedia.org/wiki/Robert_Millikan - him. 17:12
My spell-checker also highlights Millikan.
dalek kudo: e781e94 | mberends++ | tools/test_summary.pl:
tools/test_summary.pl: remove the almost redundant 'test' report column
 17:17
kudo: c4c67da | mberends++ | :
Merge branch 'master' of [email@hidden.address]
17:20 masak left
japhb ARG, github failing 17:21
sigh
Tene github is down?
japhb Seems to be coming back slowly. 17:22
It was internally timing out tree browsing. (You got the fail octopus)
Tene considers harassing/trolling his friend who works at engineyard.
Su-Shee "his up until now but not tommorow anymore" friend... ;) 17:23
-m+r 17:24
carlin rakudo: method foo() {}; multi sub foo {}; 17:26
p6eval rakudo 0d4fe0: ( no output )
carlin anyone know if that is known?
japhb Tene, feel free to troll. github is sleeping again. :-) 17:27
Tene already did: 11:25 <Tene> OMG GITHUB IS LIKE SLOW TODAY U WORK AT ENYA CANT U FIX IT?????
japhb heh 17:28
carlin (oh, it's a seg fault not a no-output)
japhb Frack. OK, of the git-hosting services, any of them that can be claimed to be stable? (That's a serious question -- no point in switching to SF or gitorious or what have you only to see same fail level) 17:30
17:30 rblasch__ left
Tene "running your own" has always been pretty stable for me. 17:31
17:31 DanielC left
japhb Seriously github? Failing on front page? 17:31
Tene except for that time when I was traveling for work and the flatmate pulled out the ethernet cable.
japhb Tene, ouch
Tene, I'm looking for the social stuff to work as well, not just pure repo hosting 17:32
Tene we could always have mirrors on several of the popular hosting sites.
japhb True
17:33 nbrown left
mikehh rakudo (c4c67da) builds on parrot r40742 - make test / make spectest (up to 28050) PASS - Ubuntu 9.04 amd64 (g++) 17:38
17:39 nbrown joined 17:48 jauaor joined
carlin rakudo: multi sub infix:<~> {}; sub infix:<~> {}; say "a"~"b".perl; 17:57
p6eval rakudo 0d4fe0: OUTPUT«␤»
carlin rakudo: multi sub infix:<~> {}; sub infix:<~> {}; say "a"~"b";
p6eval rakudo 0d4fe0: OUTPUT«␤» 17:58
carlin rakudo: multi sub foo {}; sub foo {}; foo
p6eval rakudo 0d4fe0: ( no output )
jnthn std: multi sub foo {}; sub foo {}; foo 17:59
p6eval std 28050: OUTPUT«===SORRY!===␤Illegal redeclaration of package routine foo at /tmp/sjUNJHro9E line 1:␤------> multi sub foo {}; sub foo⏏ {}; foo␤ expecting new name to be defined␤FAILED 00:02 37m␤»
jnthn std++ # we'll steal that check at some point :-)
carlin atm it's a seg fault 18:00
moritz_ doesn't segfault here.
carlin > multi sub foo {}; sub foo {}; foo 18:01
Segmentation fault
Hmm
lambdabot <no location info>: Empty record update
moritz_ carlin: it might be different on the REPL 18:02
even on the REPL it doesn't segfault here.
18:07 zloyrusskiy left
carlin fresh rakudo build with recommended parrot revision still does it 18:10
moritz_ that's bad.
jnthn If it's a memory corrutpion issue, it's completely feasible that it would manifest itself differently on different platforms. 18:11
carlin: If you know how to use GDB, then a backtrace would be interesting to see.
moritz_ no problems with -G and that code
now running with gcdebug 18:12
18:16 adam-pwgsc left, mberends left
carlin jnthn: I have gdb how I do I create a backtrace? 18:22
jnthn gdb perl6
r
moritz_ run $yourscript.pl
jnthn enter the code
moritz_ watch it die
then 18:23
bt
jnthn then bt
moritz_
.oO( concurrency gone wrong ;-)
jnthn and moritz_ star in the GDB duet!
18:25 davef joined 18:26 davef left
carlin jnthn: gist.github.com/173374 18:27
moritz_ oh, that way it also segfaults here 18:28
I tried with the multi first
carlin Ah 18:29
So I'm not insane
rakudo: multi sub foo {}; sub foo {}; foo; # doesn't die 18:32
p6eval rakudo 0d4fe0: ( no output )
carlin rakudo: multi sub foo {}; sub foo {}; foo # does
p6eval rakudo 0d4fe0: ( no output )
18:33 nbrown_ joined
moritz_ p6eval is bad in showing segfaults or exceeded resource limits 18:33
(patches welcome)
carlin: could you please write a bug report (including backtrace)? 18:34
18:35 nbrown__ joined
moritz_ Su-Shee: just wanted to tell you: I like your idea of a perl6-projects.org with sub pages; if you come up with a neat design for the sub pages, I'll contribute content 18:35
Su-Shee moritz_: is in the making, I'm a little slow behind due to the new job.. 18:36
moritz_ Su-Shee: don't need to rush. I'd be fine with having it in a sub folder for a while 18:37
Su-Shee moritz_: I realized on masak's comment on doing work twice that there isn't a perl 6 cms.. ;) 18:38
moritz_ ;-)
Su-Shee: well, SSI exists and is the simplest thing to avoid duplicate work
Su-Shee: we have a virtual host on apache2, and I can configure it to allow anything we need 18:39
Su-Shee SSI like server side includes or...did I miss something yet again?
moritz_ server side includes, yes
Su-Shee *pew* :) 18:40
jnthn carlin: Ouch.
carlin can someone run this twice from the perl6 cli:
jnthn carlin: Thanks for backtrace
carlin rakudo: multi sub infix:<~> {}; sub infix:<~> {}; say "a"~"b";
p6eval rakudo 0d4fe0: OUTPUT«␤»
carlin only segfaults on the second execution here 18:41
jnthn oh
it actually looks not so bad...
appears we deref a null pointer
18:42 hanekomu left
moritz_ that sounds much more debuggable than a classical memory corruption 18:42
jnthn moritz_: indeed 18:43
carlin bug report sent 18:45
moritz_ thanks, carlin++
jnthn can probably patch that at some point 18:46
carlin Good luck :) I better go, night all 18:48
18:48 carlin left 18:49 nbrown left 18:52 nbrown_ left 19:09 nbrown__ left 19:14 Aisling left 19:15 nbrown joined 19:17 rblasch joined 19:19 justatheory left
Tene jnthn: My friend was already woken up about the github thing, apparently. From what he tells me, the app/db aren't sharded at all, so if one person does a big query or downloads a big repo, it sucks for everyone else using the service. 19:22
japhb ouch
Tene and then he's irritated that they then claim it's a problem with EnYa. 19:23
japhb I can imagine that.
How does one compile a program written purely in NQP? (I'm crawling all over the directory tree piecing it together, but if someone knows off hand ...) 19:27
Tene parrot compilers/nqp/nqp.pbc
--target=pir will generate PIR 19:28
are either of those what you're asking for?
jnthn Tene: Ouch. 19:29
Tene eh? 19:30
jnthn Tene: The github thing.
japhb Tene: from an installed parrot? Wouldn't that have to be parrot <parrot_install_dir>/lib/<ver>-devel/library/nqp/nqp.pbc --target=pir <sourcefile>.nqp?
jnthn (Was afk...only just noticed it)
Tene jnthn: I got confused by s/jnthn/japhb/ and thought he was repeating himself.
jnthn lol 19:31
Tene japhb: from installed, it's $WHATEVER/languages/nqp/nqp.pbc
jnthn japhb: I think you're right on the installed Parrot case, yes.
Tene nqp is a language, not a library.
jnthn Ah, but that.
Tene also, I really need to add inter-lang support to nqp, so you can eval and load_library and such. 19:32
jnthn That'd be quite cool.
japhb Tene, oh yes, please
Tene oh, that would be useful?
japhb That would probably prove more than just useful.
Tene I just meant from a standpoint of completeness.
What would you want to use it for?
japhb Does NQP do the MAIN magic? 19:33
jnthn japhb: I believe not.
japhb Tene, I'm writing Parrot Plumage in NQP ... and JSON parsing is implemented as a language.
jnthn, OK, so does it just start running the file at the top of the mainline code? 19:34
Tene, to be clear: I didn't implement the JSON parser that way -- that's the way it *already* exists.
.oO( Ah, duh, NQP is based on PCT, so supports --target that way ... )
 19:36
Tene japhb: that's how I'd do the JSON stuff, fwiw.
IMO, the following is at least a very nice way to deal with such formats: my $json = read_whatever(); my $content = eval($json, :lang<JSON>); 19:37
japhb Tene, right. Which is why I want HLL interop in NQP. ;-) 19:38
Tene japhb: now, what I originally was talking about, was adding uspport to NQP for other languages to load libraries written in it.
but... I can certainly do what you want.
I'll try to get it done in the next day.
leaving for seattle tonight, though.
so... travel can get in the way.
japhb Both will be useful in the long run ... since people will want to use the brains of Parrot Plumage as modules, I'm sure. 19:39
But loading the JSON language from NQP is my priority. Thank you for putting it on your list. :-)
19:45 rfordinal3643 left
Tene japhb: go ahead and assume that it'll be there the next time I have a chance to spend 30 minutes working on it. 19:45
japhb :-) 19:46
Tene AFK shower/lunch/laundry/packing/travel
japhb good luck
19:57 rblasch left
moritz_ when rakudo supports pluggable grammars I'll modify the quoting constructs not to take a single quote between n and t as a quote terminator 19:58
so that you can write say 'don't do that'
japhb heh
20:05 pyrimidine joined
japhb Oh, very nice. I'm only just started writing tests to see what NQP provides already and what it doesn't, and have been pleasantly surprised so far. Good job, guys. 20:12
pugs_svn r28051 | moritz++ | [t/spec] test for spec change r27054 20:14
jnthn japhb: I've written a fair amount of it and while it's fairly minimal in terms of feture set, it seems to have got a good balance of small yet enough to not feel annoying. 20:18
20:21 sri_kraih left, Su-Shee left, sri_kraih_ joined
japhb jnthn, indeed 20:24
jnthn pmichaud++ for that 20:25
japhb afk for bday lunch ... bbl, same bat channel
pugs_svn r28052 | moritz++ | [t/spec/TODO] remove two items that I believe are covered now 20:26
r28053 | moritz++ | [t/spec] test that "for" statement modifier temporizes $_ on the thunk
20:27 sri_kraih_ left 20:28 sri_kraih joined, nbrown_ joined 20:38 masak joined, Aisling joined
masak brunorc.wordpress.com/2009/08/23/wh...ting-bugs/ # nice YAPC::EU TimToady keynote summary post 20:39
moritz_ btw r28051 contains pretty scary code 20:41
that second test I added...
moritz_ shudders 20:42
one-pass parsing is not for the faint of hearted.
masak that's what I keep telling you. :)
sometimes I kinda wish the alternative weren't worse. 20:43
moritz_ aye
jnthn lolitsmasak :-)
masak: I might if you're really unlucky turn your create new project script into Perl. 20:44
;-)
masak: Because manually creating the project shell based upon what it does will take me a similar amount of time once I've done it a few times. :-)
masak jnthn: cool!
I went offline to bake bread. but I also accidentally wrote an ORM persistence module.
jnthn Accidents happen. 20:45
We'll forgive you.
BTW does it have a cool name yet?
masak aye. 20:46
I think I will consider that week 14 of Web.pm and blog about it.
it will take some time to retrace my steps.
pugs_svn r28054 | moritz++ | [t/spec/TODO] rephrase moving intructions
r28054 | moritz++ |
r28054 | moritz++ | I consider the moving of tests from t/ to t/spec/ mostly done. What's left is
r28054 | moritz++ | some tests for run() of which I'm not sure how portable they are, some rather
r28054 | moritz++ | obscure tests for which I haven't been able to find out if they are still
r28054 | moritz++ | conforming to the spec, some regression tests and some disabled tests.
jnthn moritz++: w00t! 20:47
masak jnthn: I'm using custom traits on attributes!
jnthn: yes. it's called Viper.
and I don't expect it to last too long.
it's just bootsmacking, er bootshackling.
whatever.
20:47 nbrown left, braceta left
jnthn masak: Oh noes! How many bugs did you find in the custom traits on attributes implementation? 20:47
masak jnthn: none. 20:48
jnthn: but I do have a question or two.
moritz_ masak: try harder, then ;-)
jnthn masak: No bugs in fresh code? 20:49
masak moritz_: I was baking bread!
jnthn checks masak's temperature...
masak jnthn: maybe you hit the Ballmer peak with that traits implementation... :P
moritz_: btw, that second test is a great obfuscation technique. 20:50
jnthn masak: Oooh, plausible explanation!
reqamst Masak was baking bread, Su-Shee was backing bread... Huh, bread6 channel?
jnthn ...wait, did I check it in during evening or afternoon?
masak oh, and I think 'whitespace' is actually an uncountable.
:) 20:51
jnthn /join #bread6
masak reqamst: baking bread is simply a very nice thing to do.
it brings a certain sense of control over the chaotic and impersonal forces that rule our universe. 20:52
reqamst masak: I must try. :)
masak you really should.
reqamst Yes! Baking bread must be like writing program. Implementing algorithm 20:53
masak reqamst: actually, there are similarities. but baking bread is slightly more fysical. you're building concrete structures instead of abstract ones. 20:55
reqamst: if you want some tips, we're discussing them on #bread6 right now. :)
reqamst masak: Oh, I see. 20:56
jnthn ...the worst thing is that we actually *are*.
masak no, the worst thing is that I don't have any garlic butter right now.
20:56 braceta joined, jauaor left 20:57 jauaor joined
colomon No, the worst thing is that now you've got me wanting fresh baked bread, and the only thing I know how to bake takes about 20 hours from start to finish... 21:04
moritz_ fast-forward-bread? ;-)
masak
.oO( ▶▶ bread )
 21:06
21:12 jauaor left 21:13 takadonet joined
takadonet hey all 21:13
masak takadonet: o/
moritz_ wonders if masak always raises the same hand, or if we sometimes see him from the and front and sometimes from the back 21:14
s1n masak: oddly enough, i still can't clone november, though i can clone other things on github fine
masak s1n: that is strange. I had no trouble cloning it. 21:15
s1n masak: yes, i have been cloning other github projects all weekend 21:16
moritz_ s1n: do you use http:// or git://? 21:17
s1n masak: git clone git://guthub.com/viklund/november.git
masak s1n: the git:// one. 21:19
s1n masak: yes, i know :)
masak oh, sorry. :) not reading nicks correctly. 21:20
takadonet Is www.rakudo.org/documentation broken for anyone else? 21:26
moritz_ takadonet: all rakudo.org pages are broken for me as soons as I allow it to set a cookie :( 21:27
jnthn moritz_: Same.
moritz_: That's why I haven't posted my last couple of things on there.
takadonet who runs rakudo.org? 21:28
moritz_ takadonet: alester aka petdance and pmichaud
21:29 justatheory joined 21:30 sri_kraih_ joined 21:31 rhr joined, sri_kraih left, sri_kraih_ is now known as sri_kraih 21:32 rhr_ left 21:39 pyrimidine left 21:43 kidd_ joined
japhb is back ... Back to the Hack 21:44
What is the qx'' equivalent for NQP? 21:45
Or even system() for that matter?
moritz_ system is spelled run() in Perl 6. Don't know if NQP supports either. 21:46
jnthn NQP doesn't really have built-ins.
Erm, built-in subs.
It fully supportings defining and calling subs. 21:47
But I don't know that it includes anyway.
(I'm quite sure it doesn't.)
moritz_ is ruby's { |foo| ... } the same as Perl 6's -> $foo { ... } ? 21:48
jnthn moritz_: That was my understanding, but I've only really skimmed Ruby.
moritz_ jnthn: thanks 21:49
rakudo: clas A { has $.b is rw }; my @a = map { A.new }, ^3; @a».b = 4; say @a[1].b 21:50
p6eval rakudo 0d4fe0: OUTPUT«Attempt to define attribute $!b outside of class, role, or grammar at line 2, near "}; my @a ="␤in Main (src/gen_setting.pm:1789)␤» 21:51
jnthn japhb: Maybe best bet for now is to use inline PIR to do whatever Rakudo's run does, and stick it inside a sub run(...) { ... }
japhb jnthn, I was thinking that myself.
moritz_ rakudo: clas A { has $.b is rw }; my @a = map { A.new }, ^3; @a»b = 4; say @a[1].b
p6eval rakudo 0d4fe0: OUTPUT«Attempt to define attribute $!b outside of class, role, or grammar at line 2, near "}; my @a ="␤in Main (src/gen_setting.pm:1789)␤»
moritz_ rakudo: clas A { has $.b is rw }; my @a = map { A.new }, ^3; @a.»b = 4; say @a[1].b
jnthn ...wtf?
p6eval rakudo 0d4fe0: OUTPUT«Attempt to define attribute $!b outside of class, role, or grammar at line 2, near "}; my @a ="␤in Main (src/gen_setting.pm:1789)␤»
moritz_ I mis-spelled "class"
jnthn rakudo: clas A { has $.b is rw }
p6eval rakudo 0d4fe0: OUTPUT«Attempt to define attribute $!b outside of class, role, or grammar at line 2, near "}"␤in Main (src/gen_setting.pm:1789)␤»
jnthn moritz_: I know, but...it parses!
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a».b = 4; say @a[1].b
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
jnthn std: clas A { has $.b is rw }
p6eval std 28054: OUTPUT«Undeclared name:␤ A used at 1 ␤Undeclared routine:␤ clas used at 1 ␤ok 00:02 38m␤» 21:52
jnthn oh
so does STD
heh, now I want to know what it parses as. :-)
moritz_ rakudo: clas A { has $.b is rw }; my @a = map { A.new }, ^3; @a.»b = 1..3; say @a[1].b
p6eval rakudo 0d4fe0: OUTPUT«Attempt to define attribute $!b outside of class, role, or grammar at line 2, near "}; my @a ="␤in Main (src/gen_setting.pm:1789)␤»
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a.»b = 1..3; say @a[1].b
p6eval rakudo 0d4fe0: OUTPUT«Statement not terminated properly at line 2, near ".\x{bb}b = 1..3"␤in Main (src/gen_setting.pm:3460)␤»
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a».b = 1..3; say @a[1].b
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a».b = 1..3; say @a[0].b 21:53
p6eval rakudo 0d4fe0: OUTPUT«Use of uninitialized value␤␤»
moritz_ so it's just igoring the assignment?
jnthn moritz_: Hmm 21:55
I'm not sure what it's doing tbh.
rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; say (@>>.b).perl
p6eval rakudo 0d4fe0: OUTPUT«say requires an argument at line 2, near " (@>>.b).p"␤in Main (src/gen_setting.pm:2550)␤»
jnthn rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; say (@a>>.b).perl 21:56
p6eval rakudo 0d4fe0: OUTPUT«[undef, undef, undef]␤»
jnthn rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; say (@a>>.b).PARROT 21:57
p6eval rakudo 0d4fe0: OUTPUT«Perl6Array␤»
jnthn Hmm, I wonder if that wants to be List. 21:58
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a[1].b = 4; say (@a>>.b).PARROT
p6eval rakudo 0d4fe0: OUTPUT«Perl6Array␤»
moritz_ rakudo: class A { has $.b is rw }; my @a = map { A.new }, ^3; @a[1].b = 4; say (@a>>.b).perl
p6eval rakudo 0d4fe0: OUTPUT«[undef, 4, undef]␤»
22:14 pyrimidine joined
nothingmuch where is junction eigenstate extraction best documented? 22:14
masak nothingmuch: S32/Containers 22:15
nothingmuch and the best permalink for that is the pugscode svn repo?
22:15 cotto left
masak 'permalink' in the sense that you want to link to the current version? 22:16
nothingmuch yeah
nothingmuch is writing a blog post comparing McCarthy's amb to junctions for backtracking 22:17
masak yes, probably the SVN repo then.
perlcabal.org/syn/S32/Containers.html keeps being regenerated.
nothingmuch basically you can't really say if ( $x >= $y ) { } as if it wasa an assertion because you need to collapse the junction yourself
masak nothingmuch: sounds interesting.
nothingmuch $x and $y aren't mutated by that
nothingmuch is not really sure what to recocomend 22:18
maybe there'sa better construct for this?
22:19 justatheory left, rindolf left
masak nothingmuch: not built into Perl 6, methinks. 22:19
nothingmuch what's the most idiomatic way to construct junctions for $x and $y that contain the subset of values that pass that expression 22:20
?
masak uhm, using all(), perhaps?
or any()... 22:21
nothingmuch ?
masak I guess you're asking how to extract the values in a good way.
nothingmuch scsys.co.uk:8001/32892 22:22
that's the canonical amb example
masak since you want to loop over both sets in all combinations, you might want to do 'for $x.eigenstates X $y.eigenstates { ... }'
nothingmuch: ah. so you're actually changing $x and $y as you go.
nothingmuch if my $x = any(1, 2, 3); my $y = any(1, 2, 3); i don't see a way to keep only the valid values with a single expression
jnthn nothingmuch: Ah, you're basically added extra conditions that stack up? 22:23
nothingmuch well, not exactly, $x and $y aren't junctions in this example
amb is an implicit backtracking point
the first and third assertions are simple 22:24
that's just $x = $x >= 2;
jnthn Right.
The second one is a bit more interesting.
nothingmuch exactly =)
japhb It looks like Q:PIR{} syntax was partially implemented in NQP, but it wasn't finished. Is there a reason for that?
jnthn japhb: No, that's odd. 22:25
nothingmuch: I think the issue is that with junctions, you end up computing various permuations of the values in $x + $y, that gives you a new junction, then you see if any of those values == 5 22:26
But the link to the junctions $x and $y is lost at that point.
nothingmuch yeah
any takers for an idiomatic way to write that code with junctions? 22:27
22:27 braceta left
nothingmuch it doesn't need to be similar, it should just end up with some association of $x to 3 and $y to 2 in the end 22:27
jnthn Hmm. 22:28
Off hand, I can't think of one. 22:29
It feels like a related problem to the one junctions solve, but not quite the same one.
nothingmuch yeah
22:29 frederico_br joined
masak nod. 22:30
22:30 pyrimidine left
masak the question has been asked a few times on p6l. 22:30
all people seem to agree on is that junctions do not easily bend to solve such problems.
nothingmuch probably by me as well at some point =)
jnthn I think it was December/January-ish something like this last came up on p6l. 22:32
I think the outcome was, Junctions are designed to deal with a certain set of problems, and are more a linguistic device than a mathematical one, and that things that do Junction-alike stuff that doesn't really fit what Junctions do probably wants to be a different data structure. 22:35
masak coincidentally, my journey with junctions has followed the same broad arc. 22:36
I feel very confident with using them in conditions.
nothingmuch <p>That said, although junctions are not truly capable of creating backtracking logic that doesn't mean they are more or less useful than <tt>amb</tt>, their applictions are simply different.They are most useful for collecting permuted results is what really interests you, sort of like <tt>amb-collect</tt>, but when overloading the data (and not the control flow) is really what you want.</p>
that's how I summarized it
actually let me post the whole draft 22:37
jnthn nothingmuch: That sounds like a sensible conclusion.
nothingmuch i was just hoping to provide an additional example that explains how this would be done more idiomatically
but that's a flawed idea
jnthn You've thought about it more than I have, of course, but it matches my feeling on this.
nothingmuch the example doesn't want a permutation, it wants two scalar eigenstates
jnthn I think it'd be possible to implement such a data structure in Perl 6 without too much effort. 22:38
Though I suspect it's probably library rather than core language.
It is a cool construct.
nothingmuch nothingmuch.woobling.org/tmp/amb.html' 22:39
i think the second paragraph is a little bit inaccurate 22:40
nothingmuch will reword
masak nothingmuch: suddenly I want 'amb' (or the tools to make it) in Perl 6. 22:41
nothingmuch you have it 22:42
look at the test
i implemented it in perl 5
22:42 lollan joined
nothingmuch (caveat tester, it actually segfaults in its urrent form due to a bug in unwinding subroutine invocation PERL_CONTEXT frames, but if you fiddle it a bit the test passes with Continuation::Delimited) 22:42
jnthn nothingmuch++ # cool write-up :-)
masak aye, nothingmuch++
jnthn Thanks for pointing it out. And now you've mentioned hwo it works, it's even more interesting. 22:43
nothingmuch github.com/nothingmuch/continuation...t/04_amb.t
comments and everything ;-)
the p6 impl would be much more elegant
this is what it needs to pass right now: scsys.co.uk:8001/32885 22:44
and this is the output with tracing: scsys.co.uk:8001/32884
masak lastofthecarelessmen.blogspot.com/2...ystem.html # someone blogs about proto :) 22:45
use.perl.org/~masak/journal/39513 # lolibloggedtoo
japhb I love how going to the pub is a todo item 22:47
jnthn It's one of those todo items that doesn't need to get taken off the todo list after it's done.
22:48 ihrd joined
jnthn masakpost! :-D 22:48
japhb jnthn, is a leading '!' the Rakudo way of saying "internal subroutine"?
22:48 fredrecsky left
japhb As in: .sub '!qx' 22:49
I'm guessing it's by analogy with private class attributes
jnthn japhb: Yeah, it's a convention that we use. 22:50
masak std: class A { method !private() {} }
p6eval std 28054: OUTPUT«ok 00:02 38m␤»
masak japhb: not just internal.
jnthn Taken from that concention. :-)
japhb masak, oh?
jnthn masak: I think japhb was talking about subs rather than methods.
masak: For subs it's just a convention.
masak oh, right. indeed.
japhb OK, right, we're clear
nothingmuch is dev.perl.org/perl6/doc/design/exe/E..._Junctions still the most perma link?
jnthn Whereas method !private() { } is actually Perl 6 syntax. :-)
masak: In your post, what are you using the trait for? 22:52
masak jnthn: nothing, yet.
jnthn: but it will provide various convinience methods eventually. 22:53
right now it's just a stub.
jnthn ah, ok
nothingmuch: It doesn't immediately look out of date.
my Floor::Wax|Dessert::Topping $shimmer; is
22:53 cotto joined
jnthn S09 is The Spec that is maintained. 22:54
Junctions section in svn.pugscode.org/pugs/docs/Perl6/Sp...9-data.pod
I forget where it's HTML-ized...
s1n cabal
jnthn ah, yes 22:55
perlcabal.org/syn/S09.html#Junctions
s1n perlcabal.org/syn/S09.html#Junctions
lol
nothingmuch i was considering using that but it's not good introductory material I think
jnthn rakudo: say ('great minds think alike', 'all fools are the same').pick
nothingmuch it's hard to see what you can do with them
p6eval rakudo 0d4fe0: OUTPUT«great minds think alike␤» 22:56
jnthn nothingmuch: Aye. The lesser evil may be your original suggestion.
nothingmuch till you grok the whole thing that is
jnthn I can't think of a better one off hand.
nothingmuch exe it is then =)
ok, now to see if I can make it stop segfaulting
^_^
s1n nothingmuch: the spec is somewhere between spec and chatter
nothingmuch s1n: i want a link in a blog post where someone who doesn't grok junctions can get an idea of what they are about 22:57
s1n nothingmuch: aside from a few minor things, that first link you posted is more n00b-friendly 22:58
jnthn nothingmuch: Happy debugging. 22:59
s1n nothingmuch: junctions aren't rocket science though, despite their popularity, the n00b can think of them as sets with boolean eval (few other nice things, like pick)
jnthn Don't let TimToady see you equating junctions and sets. ;-) 23:00
nothingmuch s1n: the exegesis's advantage is that it provides a clear example of the linguistic advantage
whereas the synopsis does not
s1n jnthn: just wait until people _abuse_ junctions and make them sets
nothingmuch been there, done that, that's why we have the Set type (or at least had it back in 2005-2006 when I was last up to date ;-)
s1n nothingmuch: if you need real examples, i suggest the tests
23:01 nbrown joined
jnthn s1n: Yes, the problem is that said people will try and get bitten. 23:01
s1n nothingmuch: more often than not, i give up with the spec and just read the tests
nothingmuch why not the exegesis?
s1n jnthn: that's what you think :)
nothingmuch it's got clear examples and explanatory text
not just examples that work
jnthn nothingmuch: The general issue with the exegesis is that they aren't maintained relative to the spec. 23:02
s1n nothingmuch: because the tests are far more complete :)
jnthn nothingmuch: In thise case however, the two haven't really diverged.
So it's a fine thing to link to still.
s1n the cabal embeds smartlinks
jnthn grabs a beer to read masak's blog post with
nothingmuch yeah, i'm not trying to detail how they should really wokr, just to get a quik handle on them if you haven't been exposed
s1n: again, not the point of the blog post 23:03
s1n nothingmuch: i know, just telling you how i grok things :)
nothingmuch: btw, nice talk on haskell at yapc :)
nothingmuch i'd link to Perl6::Junction, it as a great synopsis, but it doesn't actually support eigenstate extraction 23:04
and Q::SP's synopsis isn't that great either
thanks =)
bad delivery though, I thought I was going to faint (saw black spots and everything) 23:05
s1n nothingmuch: no it was fine, you seemed kinda nervous, but content is king :) 23:06
nothingmuch s/nervous/sick/;
jnthn I imaigne that's kinda distracting when trying to give a talk.
nothingmuch very very sick
i thought i was going to cancel
jnthn nothingmuch: Hope you're feeling much better now.
nothingmuch yeah, it was just airplane flu
jnthn eww 23:07
nothingmuch it only ruined pittsburgh
the other legs of the trip were successful
most importantly: gallery.woobling.org/cathedral_peak, gallery.woobling.org/after_six
also got to go mountain biking in santa cruz
and the big NP in Maine 23:08
i forget its name
Acadia
usaw many friends
anyway, fun times
jnthn Wow, looks like awesome fun. 23:09
nothingmuch: Planning to do YAPC::Asia? 23:10
23:10 wayland_ left
jnthn masak: "Note that we haven't even bothered to define the trait we're using. Wonder when that'll bite us?" - in the future, it may have to be "sooner than it does now" 23:11
masak jnthn: glad to hear that. 23:12
23:12 wayland_ joined
masak jnthn: I was actually surprised that I got away with sticking a bareword into the code like that. 23:12
jnthn: also, I think you'll have to explain the reasoning behind the parameter list to trait_mod:<is>. I did a double-take on that one. 23:13
jnthn Text::CSV.parse-file($filename, :output($type)); # :output($type) is _so_ cool 23:14
nothingmuch jnthn: yp
also climbing in .tw and Myogi in Japan: i-cjw.com/blog/2009/06/28/where-eagles-darent/
jnthn If it does what I think it does...
masak jnthn: I know! I hope to blog about it soon. :)
jnthn: it does.
jnthn: still mulling over how to combine it with typed attributes, though. 23:15
that was one of the problems I ran into this evening.
jnthn Mozes introspect the types of the attributes
erm
You can ...
masak true.
maybe that's the solution.
but even then, how do I convert from a Str to... something arbitrary?
there'd have to be an adapter function sent along. 23:16
23:16 nbrown_ left
jnthn masak: Rely on coercion, perhaps. 23:18
$type($thingy-you-read)
masak jnthn: aye, that might work.
jnthn (yes, nyi, but maybe you can talk some Rakudo dev into it ;-)) 23:19
masak :)
jnthn masak: I'd avoid defining your own coercion mechanism without good reason.
masak will think a bit more about it
I have a feeling people will like :output(MyType). so it's worth it to make it work in many of the common cases. typed attributes is definitely a common case. 23:20
jnthn *nod* 23:21
23:21 M_o_C left
jnthn masak++ # awesome post 23:21
One of my favorites so far.
masak thanks. :)
jnthn I'm really excited to see the things people are starting to create in Perl 6. 23:22
masak lately, I've come to focus on showing bare-metal thinking rather than high-level discussion in the Web.pm posts. I think that constitutes an improvement.
jnthn masak: It can be nice to see the big picture once in a while, but yes, I'm enjoying it. 23:24
jnthn hasn't actually done a high-level post for ages.
I think I may do one in late Nov / early Dec.
masak I think we'll be building up to the big picture as we approach week 20.
jnthn Which marks the 2 year point.
masak wow. 2 years. 23:25
jnthn (2 years of my involvement in Rakudo, that is.)
Yes, it scares me too.
I really didn't know what I was getting myself into.
masak :) 23:27
I'm glad I got to blogging week 14 today. I think this marks the first Web.pm week that was actually a week long. :P 23:30
23:30 nbrown left
jnthn It's OK, I have a liberal definition of "week" for my grant too. 23:31
23:31 nbrown joined
jnthn Well, "month" 23:31
:-)
I actually did get my current one done closer to on time than the previous one. :-) 23:32
masak I'm not sure how to count with Web.pm.
technically, I'm doing ihrd's weeks now, and my part is finished.
jnthn Well, to me the important thing is: there's visible progress in the right direction. 23:33
masak nod.
jnthn And Web.pm certainly feels that way to me.
masak to me too. 23:34
there's just so _much_ to do! :/
and by that, I don't necessarily mean the parts of Web.pm that fall within the grant.
those parts are pretty much wrapped up already.
jnthn knows the feeling of so _much_ to do. 23:35
23:35 wayland_ is now known as wayland76
wayland76 does too, with his 6 or so perl 6 projects :) 23:36
jnthn tries to plan what on earth to do this week. 23:38
masak washes up after the baking session
wayland76 is glad jnthn is on earth this week
wayland76 wonders whether jnthn is planning to do role {...} this week :)
jnthn Yeah, hacking on the moon last week was a real drag.
AWFUL ping times.
wayland76 ...and the added danger of becoming a ..... lunatic :) 23:39
jnthn wayland76: erm, role Foo { ... }?
wayland76 jnthn: Yes :)
jnthn wayland76: Ouch.
wayland76 I expect it's not ready to be done yet, but I figured it couldn't hurt to ask :) 23:40
jnthn wayland76: I'll keep it in mind. Maybe.
Hmm. Ideally I'd do a Rakudo day tomorrow. In reality, I probably need to clear up a couple of non-Perl 6 bits. 23:41
wayland76 Btw, is it supposed to tell me what file an error is in, as well as the line #?
jnthn If it knows it.
Well, if it can know it. If you think it should be telling you, it probably should be.
File rakudobug if you've got a way to produce a missing one where it should not be missing. 23:42
wayland76 ok, because I had a project that was blocking on that -- I'll revisit it, and see if I can narrow down the bug, and submit a bug :)
jnthn OK, thanks. :-) 23:43
masak wow. it's very late. 23:44
I'd better start trying to sleep now, so that I'll succeed before it's morning... 23:45
see y'all tomorrow!
23:46 masak left
nothingmuch $x is to addition what factorization is to product 23:47
what's the word I'm looking for?
japhb nothingmuch, what constraints do you want on the things you're breaking the sum into? 23:48
(factorization has a clear constraint: the factors must be integers) 23:49
nothingmuch <p>The expression <tt>1|2|3 + 1|2|3 == 5</tt> that at least one of the permuted sums is equal to <tt>5</tt>. Although it can't tell you which ____ amount to 5, the ability to do <tt>1|2|3 + 1|2|3</tt> is useful in its own right.</p>
japhb summands?
(which is actually a match for 'factors' not 'factorization') 23:50
jnthn *nod*
summands fits there.
nothingmuch yes, thanks =)
yeah i rephrased it more since I asked ;-)
jnthn I can't think of something to put in the $x though.
nothingmuch neither
wayland76 Or operands is the general term, if you prefer that 23:51
nothingmuch but if we refactor that we don't need to =)
23:52 jauaor joined
japhb jnthn, what is the 'root_new' op? 23:53
japhb reading src/builtins/globals.pir
jnthn japhb: Allows you to create an instane of something in a different HLL. 23:54
Or potentially in a different HLL.
Normally new is relative to the current HLL.
japhb jnthn, OK, gotcha. 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!