Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2008-01-16

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 pugs_svn r19501 | lwall++ | Changed #?perl6: to #?rakudo:
00:06 cosimo joined #perl6
00:06 cosimo left #perl6
00:07 cosimo joined #perl6
00:09 cosimo left #perl6
00:16 justatheory joined #perl6
00:17 IllvilJa joined #perl6
00:34 namit left #perl6
00:34 thoughtpolice joined #perl6
00:35 Blasi joined #perl6
00:37 jiing joined #perl6
00:43 crashmatrix joined #perl6
00:48 Auzon joined #perl6
01:01 cognominal_ joined #perl6
01:05 Tene So, any way to execute a closure in the current context instead of the context it was created in?
01:06 Tene I want to be able to pass a closure to a function and then run that closure, except with some setup and teardown code inserted around it.
01:06 cnhackTNT joined #perl6
01:06 Tene If there was a way to insert extra code at arbitrary points in a closure, that would be okay too.
01:07 Tene I guess I could do it with macros, too.
01:07 skids joined #perl6
01:15 Lorn joined #perl6
01:20 wknight-away joined #perl6
01:23 wknight8111 joined #perl6
02:32 lisppaste3 joined #perl6
02:38 pugs_svn r19502 | lwall++ | [fudge] first whack at test preprocessor
02:47 kanru joined #perl6
02:59 [particle1 joined #perl6
03:12 stevan_ joined #perl6
03:14 aindilis joined #perl6
03:24 alester joined #perl6
03:29 [particle] joined #perl6
03:49 [particle1 joined #perl6
04:02 pugs_svn r19503 | agentz++ | [smartlinks.pl] prints usage info when GetOptions fails
04:02 [particle2 joined #perl6
04:12 penk joined #perl6
04:14 [particle2 joined #perl6
04:26 Auzon joined #perl6
04:35 thoughtp1lice joined #perl6
04:40 pugs_svn r19504 | agentz++ | [smartlinks.pl] fixed a bug when --test-res is set while test file info is not available (aka the newline prbolem) reported by Cosimo++
04:53 [particle] joined #perl6
04:57 lyokato_ joined #perl6
05:29 masak joined #perl6
05:34 alc joined #perl6
06:41 pen1 joined #perl6
06:49 Schwern joined #perl6
06:52 Schwern joined #perl6
07:16 pugs_svn r19505 | lwall++ | [fudge] bug removal
07:22 jjore joined #perl6
07:22 pugs_svn r19506 | lwall++ | Changing todo's to "fudge normal form"
07:39 bsb left #perl6
07:39 pugs_svn r19507 | lwall++ | more "fudge normal form"
07:40 pugs_svn r19508 | lwall++ | [fudge] allow indentation of #?foo: line
07:51 penk joined #perl6
07:57 pugs_svn r19509 | cosimo++ | Perl script to set svn props, for those poor souls that work on Win32.
07:57 pugs_svn r19509 | cosimo++ | Bonus: it added the svn props to itself.
07:59 pugs_svn r19510 | cosimo++ | [t/spec] Added svn props where not set
08:11 jisom joined #perl6
08:22 iblechbot joined #perl6
08:23 [particle1 joined #perl6
08:28 Aankhen`` joined #perl6
08:30 devogon joined #perl6
08:43 kane_ joined #perl6
08:45 cosimo joined #perl6
08:59 moritz_ so is perl6 on parrot officially renamed to rakudo?
10:04 wknight-away joined #perl6
10:08 Alias_ rakudo?
10:09 ruoso joined #perl6
10:10 masak rakudo!
10:10 masak sounds japanese
10:17 * ruoso rofl with the last xkcd strip
10:21 luqui joined #perl6
10:29 pugs_svn r19511 | Darren_Duncan++ | ext/Muldis-DB/ : reverted last SeeAlso change (r19354) as ::Interface will take the simpler AST input directly, no X wrapper needed
10:38 chris2 joined #perl6
11:00 barney joined #perl6
11:17 ebassi joined #perl6
11:20 jferrero joined #perl6
11:45 cmarcelo joined #perl6
11:50 polettix joined #perl6
11:54 marmic joined #perl6
12:13 DarkWolf84 joined #perl6
12:17 Alias_ joined #perl6
12:32 Lorn joined #perl6
12:53 rindolf joined #perl6
12:54 rindolf Hi all.
12:54 wknight-away joined #perl6
12:55 moritz_ hi rindolf
12:56 rindolf Hi moritz_
12:56 rindolf moritz_: what's up?
13:00 moritz_ rindolf: the ususal stuff - my studies, perlmonks, hacking my websites etc.
13:01 moritz_ rindolf: what about you?
13:01 rindolf moritz_: well, I started working in a new job (for a one-month trial) on Sunday.
13:01 rindolf moritz_: and yesterday I wrote some code.
13:01 moritz_ rindolf: what kind of job?
13:02 rindolf moritz_: C++ programmer.
13:02 rindolf moritz_: on Linux.
13:02 rindolf :-)
13:02 pugs_svn r19512 | ruoso++ | [smop] the ___POINTER___ macro replaces the object's RI to a NOOP RI so it can be put as a parameter inside a stack node, instead of just returning a proxy object.
13:04 [particle] joined #perl6
13:04 rindolf moritz_: oh! And on Saturday I wrote my first PIR program.
13:04 moritz_ both sound cool
13:05 Alias_ I have seen the future, and it is inline PIR
13:08 moritz_ and mod_parrot ;)
13:08 Alias_ Quite probably
13:08 Alias_ What struck me was chromatic's speed benchmarks
13:08 Alias_ It's not quite as fast as haskell, but it's damned impressive
13:09 penk joined #perl6
13:09 Alias_ So it would seem to me that embedding not-quote-as-fast-as-c-but-way-faster-than-perl code would be advantagous in extreme
13:09 avar where are those benchmarks?
13:10 Alias_ recent use.perl or perl.com post, one of the two
13:10 moritz_ especially if it doesn't cause segmentation faults when you do something wrong ;)
13:10 Alias_ indeed :)
13:10 ebassi avar: http://use.perl.org/~chroma​tic/journal/35333?from=rss
13:10 lambdabot Title: Journal of chromatic (983)
13:11 avar Alias_: you can have stand-alone perl6
13:11 Alias_ avar: I know
13:11 Alias_ Although not QUITE standalone right?
13:11 pasteling "avar" at 208.78.101.240 pasted "ldd -r perl6" (15 lines, 781B) at http://sial.org/pbot/29872
13:12 Alias_ You still need to keep it inside a parrot checkout or something?
13:12 Alias_ Because of make problems?
13:12 avar since mod_parrot works I expect linking it to libparrot should too
13:12 Alias_ libgmp built in? Neato
13:12 Alias_ So large number support from the start
13:13 avar ./perl6 -e 'say 9**9**9'
13:13 avar 1.96627e+77
13:13 avar sh-3.1$ ./perl6 -e 'say 9**9**9**9'
13:13 avar inf
13:13 Alias_ nice
13:14 Alias_ I've been thinking a bit over the last few days
13:14 Alias_ Is anyone building parrot/perl6 with MinGW?
13:14 avar they're waiting for you
13:14 Alias_ :)
13:14 Alias_ Well, kinda :)
13:15 Alias_ I'm not one to deal with bugs in the C, I just take stuff that works and package it up
13:15 Alias_ BUT, I might be able to add parrot/perl6 support to Perl::Dist
13:15 Alias_ "Vanilla Perl 6"
13:24 [particle1 joined #perl6
13:26 meppuru joined #perl6
13:27 rindolf Hi Alias_
13:28 rindolf Alias_: there was something I've been meaning to ask you, but I don't recall what it was.
13:28 moritz_ rindolf: then you'll have to ask a generic question
13:28 moritz_ like "?"
13:28 moritz_ ;)
13:29 rindolf Alias_: when are you going to restore http://ali.as/devel/cvsmonitor/ ?
13:31 njbartlett_ joined #perl6
13:32 Alias_ rindolf: As it just so happens I'm in the middle of recovering it right now
13:32 Juerd_ How does >> work in regexes, given that > also ends <foo ...>?
13:33 moritz_ what's >> supposed to be?
13:33 Juerd_ Right word boundary
13:34 * Juerd_ doesn't like its spelling because it's much different from >> in Perl 6 itself, but mostly wonders how it can be parsed alright
13:34 moritz_ disambiguation by requiring whitespaces?
13:34 moritz_ that's ugly, I know
13:35 Juerd_ Ugh
13:35 rindolf Alias_: oh cool.
13:35 rindolf Alias_++
13:35 Alias_ rindolf: In getting the Strawberry websites working, I discovered how to link svn to dreamhost sites properly
13:36 Alias_ I have the ali.as site in some random condition in version control, just didn't have a good way to deploy it
13:36 Alias_ Setting up the cron now in fact
13:36 Alias_ (dreahost recently adding cron support was the final piece I needed)
13:42 Alias_ rindolf: You actually use CVS Monitor?
13:42 Alias_ I thought most people had pretty much moved on from CVS now
13:43 rindolf Alias_: no, but it's a broken link in the dmoz.org CVS category.
13:43 Alias_ ah :)
13:43 Alias_ Which you look after?
13:44 rindolf Alias_: I'm an editor of the "Configuration Management/Tools" directory, which contains it.
13:44 Alias_ ah
13:44 Alias_ You'd probably be interested in one of my current work projects then
13:45 Alias_ I've added schema version control to Oracle
13:52 iblechbot joined #perl6
14:10 bphillip1 Alias_: sounds interesting, available to the public?
14:10 Alias_ bphillip1: It's too company-specific atm
14:11 Alias_ bphillip1: But once I'm no longer employed by the company (and I've got all the kinks worked out) I definitely want to do a CPAN variant
14:11 Alias_ We're still hitting problems with it not being quite paranoid enough
14:11 Alias_ And it doesn't support branch tracking properly
14:11 Alias_ It does, however, seriously kick ass
14:12 Alias_ Inspired a bit by some of the Ruby stuff, but improved
14:12 Alias_ (so as to be more appropriate for mere mortals that can't code)
14:13 bphillip1 Alias_: do you tend to reserve the right to release what you develop at a company as open source stuff at a later date?
14:13 Alias_ Sometimes
14:13 Alias_ Or at least, I try to negotiate the ability to open source stuff that doesn't have a business value to the company
14:13 bphillip1 where I work, we sign something about everything we create on company time/equipment being proprietary to The Company
14:14 Alias_ Well, contracting helps for one
14:14 bphillip1 true
14:14 Alias_ Also, I've done it enough times that I can demonstrate all the additional value I bring to the company from previous iterations
14:14 Alias_ The giant stinking pile of CPAN modules etc
14:15 Alias_ And failing that, I sucker them on the "It means I'll keep supporting the code even after you stop paying me" argument
14:15 bphillip1 nice :-)
14:15 Alias_ Of course, I'd never propose to do it with anything that generates income
14:15 Alias_ I have one system in particular that I'd dearly love to CPANify
14:15 Alias_ For screen scraping banks websites
14:15 Alias_ But it's a critical advantage for my client in his market
14:16 moritz_ something like a kicking ass WWW::Mechanize?
14:16 moritz_ erm, ass kicking
14:16 moritz_ ykwim
14:16 Alias_ More like an ass-kicking Finance::Bank
14:17 moritz_ ok
14:17 Alias_ Awesome code, very very robust
14:17 Alias_ But utterly sekrit
14:17 Alias_ And so it stays private
14:18 Alias_ But really, 50% of most code everyone writes has little business value
14:18 rindolf Alias_: regarding Ada - besides the Ada-ish VHDL there's also Verilog which is more Cish
14:18 [particle] joined #perl6
14:18 Alias_ rindolf: Ya, but I get the impression that VHDL is what the big boys all use
14:19 moritz_ it is indeed
14:19 Alias_ And now I understand why
14:19 rindolf Alias_: my previous workplace used Verilog.
14:19 rindolf Alias_: and so did my friend's previous workplace.
14:19 Alias_ Were they massive chip companies?
14:19 rindolf Alias_: no.
14:19 rindolf Alias_: they were small companies.
14:20 Alias_ Wel there you go :)
14:20 rindolf I did work a bit with VHDL in the Technion.
14:20 Alias_ Anyways, I need to sleep now
14:20 rindolf Alias_: good night.
14:20 Alias_ If this trip of mine comes off, I'll wanted to spend some quality time on the Perl 6 / parrot toolchain anyways
14:21 Alias_ So hopefully all the niggly compiler issues are worked out by then
14:22 Alias_ rindolf: Won't get cvsmonitor pages working today
14:22 Alias_ But I've got cron up, so it should just be a matter of pushing the pages over from old version control
14:22 Alias_ night all
14:32 skids joined #perl6
14:34 jferrero joined #perl6
14:43 jhorwitz joined #perl6
14:53 ebassi joined #perl6
14:59 rdice joined #perl6
15:25 gbacon joined #perl6
15:29 jferrero joined #perl6
15:31 kanru joined #perl6
15:31 rindolf joined #perl6
15:45 araujo joined #perl6
16:00 ispy_ joined #perl6
16:09 pugs_svn r19513 | ruoso++ | [smop] the ___POINTER___ macro is actually POINTLESS... :) The Free lowlevel call should ignore the refcount anyway and clean the reference to the object in the current stack.
16:12 exp_evalbot joined #perl6
16:12 pugs_svn r19514 | moritz++ | [evalbot] perl6 on parrot is now called "Rakudo"
16:13 exp_evalbot joined #perl6
16:13 moritz_ rakudo: say "hi there, with new name"
16:13 exp_evalbot OUTPUT[hi there, with new name␤]
16:15 pro21ms4 joined #perl6
16:22 [particle] moritz_++
16:31 moritz_ pmichaud++
16:32 * ispy_ wonders if he can ++ himself?
16:33 moritz_ ispy_: you can, but it won't do you any good ;)
16:33 * ispy_ kicks the dirt...
16:40 [particle] ispy_++ for being a karma whore
16:41 [particle] who's your daddy?
16:41 ispy_ you are :(
16:41 [particle] :D
16:41 ispy_ haha...
16:44 gbacon integral: you clown
16:44 gbacon a total of three messages and you ban me?
16:44 diakopter moritz_: how about having exp_evalbot prepend the implementation name before 'OUTPUT'
16:46 moritz_ diakopter: why? ususally it's quite clear from context which on is to blame ;)
16:47 pugs_svn r19515 | ruoso++ | [smop] For sanity reasons, I defined q:sm0p which is a quoting type that is a "stack definition language" that translates one statement to one stack node without creating a new stack.
16:47 diakopter usually, yeah. I guess I'm a stickler for perspicuity.
16:47 diakopter but, you're right.
16:47 diakopter so it's probably not worth the extra noise
16:48 ruoso I just love Perl 6... which other language would allow me to create a new syntax inside it in such a clear way...
16:49 moritz_ diakopter: if you can't map eval command to output from context, you'd have to prepend the output lines with the evaluated command as well
16:49 moritz_ ruoso: perhaps lisp?
16:49 moritz_ I never actually used lisp, so I'm not sure ;)
16:49 jferrero joined #perl6
16:50 pasteling "ruoso" at 213.13.127.198 pasted "If you think stack manipulation was nice in SMOP, take a look in the q:sm0p quoting type, in the example CPS implementation for DESTROYALL..." (18 lines, 639B) at http://sial.org/pbot/29875
16:51 diakopter moritz: or, prepend with requestorname: IMPL|STDOUT
16:51 ruoso moritz_, see the paste above
16:53 moritz_ diakopter: I think I'll do that if it turns out that the current approach isn't working in the wild
16:54 araujo joined #perl6
16:54 [particle] perl6: say(captialize('hElLo, WoRlD'));
16:54 exp_evalbot kp6: OUTPUT[no method 'APPLY' in Class 'Undef'␤ at compiled/perl5-kp6-mp6/lib/Ki​ndaPerl6/Runtime/Perl5/MOP.pm line 345␤   KindaPerl6::Runtime::Perl5::MO​P::__ANON__('HASH(0x824bda4)', 'APPLY', 'HASH(0x8e4afe0)') called at compiled/perl5-kp6-mp6/lib/Ki​ndaPerl6/Runtime/Perl5/MOP.pm line 169␤
16:54 exp_evalbot ..main::DISPATCH('HASH(0x824bda4)', 'APPLY', 'HASH(0x8e4afe0)') ca...
16:54 exp_evalbot ..pugs: OUTPUT[*** No such subroutine: "&captialize"␤    at /tmp/G24xX4oUq0 line 1, column 5-31␤]
16:54 exp_evalbot ..p6: No output (you need to produce output to STDOUT)
16:55 [particle] where's rakudo in there?
16:57 pugs_svn r19516 | moritz++ | [evalbot] one more s/p6/rakudo/, [particle]++
16:58 pugs_svn joined #perl6
17:04 [particle] would folks here be interested in commit messages from rakudo, and is it easy enough to create a bot that reports only commits to parrot's languages/perl6 dir?
17:07 * diakopter fires up svnbotl for parrot/languages/perl6 in #perl6
17:07 moritz_ [particle]: yes
17:09 [particle] diakopter: it'd be nice to have that bot on irc.perl.org#rakudo too
17:09 [particle] ...and some other bots, if i can figure out how to get them there
17:09 moritz_ [particle]: do you want logging in that channel?
17:09 gbacon anyone have ops on #Perl?
17:10 [particle] sure, you can log #rakudo
17:13 moritz_ let's see if my bot works on irc.perl.org
17:14 moritz_ that server tends to produce connection timeouts
17:15 rakudo_svn joined #perl6
17:15 diakopter here goes nothing
17:16 diakopter c'mon, someone with commit access to rakudo... do your thing
17:20 manfred joined #perl6
17:20 pugs_svn r19517 | lwall++ | [fudge] recognize is_* ok_* for user-defined tester subs
17:54 ruoso TimToady, I've came up with an alternative way to create a new set of stack nodes in smop, if you have some free time, could you please take a look at it? http://sial.org/pbot/29875
17:54 lambdabot Title: Paste #29875 from "ruoso" at 213.13.127.198
17:55 pmurias joined #perl6
17:55 pmurias moritz_: lisp dosn't have syntax itself
17:55 ruoso heh
17:57 pmurias ruoso: tcl would let you create a new syntax
18:04 * ruoso having weird thoughts about using the stage 0 and 1 languages inside the C code, including a pre-processor before cpp that would translate the stage 0 and 1 languages to C code directly
18:07 pmurias ruoso: i think an important question is "Is Perl 6 a better low-level language then C?"
18:08 ruoso pmurias, not at all, but the thing is that the stage 0 language and the stage 1 language are supposed to be translattable to C
18:09 [particle] is stage 1 built on stage 0?
18:10 [particle] like, could you build up a chain of emitters?
18:11 ruoso [particle], I still didn't stress it, but I think so...
18:11 ruoso because the stage 0 language is a stack node builder
18:11 ruoso and the stage 1 is something that will be translated into a set of stack nodes
18:14 ruoso [particle], but I think it won't be a simple chain of emitters... because the stage1 has some boot C code before creating the actual nodes...
18:15 ruoso so, the stage 1 would be translated to some C code, than the stage0 lang and then some more C code...
18:19 skids joined #perl6
18:21 rindolf joined #perl6
18:36 braceta joined #perl6
18:38 skids Suppose I have a class with tons of constants some of which are needed to invoke methods, but maybe the constants are not so useful outside the class.  Has any sugar been proposed to allow a y.meth($foo, $bar) to look up $foo, $bar in y's class before the regular search?  Just wondrin.
18:40 pmurias skids: $foo,$bar are variables?
18:41 skids Well, could be.  For me I'm just thinking constants, but that could be useful too.
18:42 skids There's that class variable $^var, right?  But that would still need a Package:: in front of each one used.
18:43 pmurias skids: $^var is a placeholder variable
18:43 skids Hrm oh yeah what was that sigil then...
18:44 pmurias skids: i don't think there are class variables
18:44 pmurias i think it would be $ClassName::var
18:46 skids No there was something about non-per-instance methods IIRC, looking now.
18:47 pmurias they are methods on the metaobject instance method ^foo {}
18:47 skids $:, @:, %: is what I was referring to. "Class attributes"
18:51 skids I suppose y.meth(y.:foo,y.:bar,y.:blah ...) is better than y.meth($MyClass::foo, $MyClass::bar, $MyClass::blah ...) for short variable names.
18:57 pmurias S12:715
18:58 alester joined #perl6
19:02 cosimo joined #perl6
19:02 cosimo hi there!
19:02 lambdabot cosimo: You have 1 new message. '/msg lambdabot @messages' to read it.
19:03 moritz_ cosimo: that message is now superflous, since agentzh++ fixed smartlinks.pl ;)
19:03 cosimo moritz_: yes, I know. Thanks anyway
19:04 cosimo BTW, newbie q:
19:04 cosimo I don't remember my nick pwd, how can I "reset" it?
19:05 TimToady pugs passwd?
19:05 moritz_ cosimo: pugs svn?
19:05 pugs_svn r19518 | putter++ | Add ruby backend option to script/kp6.
19:05 [particle] irc nickserv password?
19:05 cosimo no, sorry. I'm talking about freenode irc.  particle: yea
19:05 TimToady as usual, the expert answer is "it depends" :)
19:06 cosimo :-)
19:06 skids pmurias: ahh, the ^ was in my synapses from S12:441
19:07 * skids relieved his synapses are functioning on some level
19:07 cosimo what time is it in the states, now??? Usually at this time in my TZ there's not many people
19:07 skids 2pm EST.
19:08 skids People too busy working off their mortgages, I guess.
19:08 gbacon and selling off gold too
19:09 cosimo TimToady: have you seen the add-svn-props.pl? Don't know if it's ok, but I thought it's more cross-platform than /bin/sh
19:09 skids speaking of which, meeting...
19:17 TimToady cosimo: no, haven't looked at it, but if it works, it works...
19:18 pugs_svn r19519 | putter++ | Emit/Ruby.pm now defines Emit::Ruby, not Emit::Perl5.
19:18 pugs_svn r19519 | putter++ | perl script/kp6 -Cruby -e 'say 3' still doesn't work:
19:18 pugs_svn r19519 | putter++ |   Can't load Emit::Ruby plugin: Can't locate KindaPerl6/Visitor/Emit/Ruby.pm
19:18 pugs_svn r19519 | putter++ |    in @INC
19:18 cosimo TimToady: yep, it does
19:18 cosimo :-)
19:31 pasteling "mncharity" at 76.24.28.215 pasted "perl Makefile.PL kp6_mp6 "Please inform the author." error message" (9 lines, 400B) at http://sial.org/pbot/29880
19:31 mncharity @tell fglock fyi: perl Makefile.PL kp6_mp6 "Please inform the author." error message: http://sial.org/pbot/29880
19:31 lambdabot Consider it noted.
19:32 pmurias @tell putter did you run make? ;)
19:32 lambdabot Consider it noted.
19:33 mncharity no.  :)
19:33 mncharity @tell fglock I didn't run make.  Oops.  Never mind.  Thanks.  pmurias++
19:33 lambdabot Consider it noted.
19:35 jferrero joined #perl6
19:35 pmurias mncharity: perl5v6 now has a primitive MOP
19:35 pmurias src/KindaPerl6/Runtime/Perl5V6/MOP.pm
19:36 mncharity looking...
19:37 mncharity (while make kp6_mp6 is running...  it's compiling more than Emit/Ruby.pm, which was unexpected)
19:39 mncharity re MOP.pm, interesting, thanks.  (I don't understand the oo bootstrap or mop well enough to comment, but I look forward to copying from it:)
19:40 * mncharity wonders why make kp6_mp6 is recompiling KindaPerl6/Grammar/*.pm :/
19:42 pmurias mncharity: did you use make forcerecompile?
19:42 pugs_svn r19520 | ruoso++ | [smop] finally I have the code for the first test... the only problem is that this test will only be able to run when I have a looooooot of code.
19:42 pmurias (svn sometimes screws up timestamps)
19:44 mncharity re forcerecompile, no.  and i don't immediately see a kp6_mp6 dependency on the makefiles.  curious.
19:45 mncharity hmm... timestamps..
19:46 pugs_svn r19521 | ruoso++ | [smop] a small note regarding the specific feature the test is trying to address..
19:47 mncharity oh, duh.  yes, the compiled/perl5-kp6-mp6/lib/KindaPerl6/Grammar/*.pm are from Jan 5, vs Jan 11 for the source.
19:49 mncharity and it was indeed an svn issue.  removing and reverting bumpted the *.pm to the 11th.
19:49 jisom joined #perl6
19:57 crashmatrix left #perl6
20:04 pugs_svn r19522 | putter++ | perl script/kp6 -Cruby -e 'say 3' works.
20:04 pugs_svn r19522 | putter++ | docs/ruby_backend.txt: added 'how to get Emit/Ruby.pm built' notes.  pmurias++
20:04 pugs_svn r19522 | putter++ | Emit/Ruby.pm: emit_perl5 methods renamed emit_ruby.
20:04 pugs_svn r19522 | putter++ | Bug: while -Cruby behaves as -Cperl5, -Bruby also behaves as -Cperl5.
20:14 pugs_svn r19523 | putter++ | ruby_backend.txt: change mentioned but missed in r19522.
20:25 _sprocket_ joined #perl6
20:25 pugs_svn r19524 | putter++ | script/kp6 -Bruby -e 'say 3' sort of works.
20:25 pugs_svn r19524 | putter++ | It doesn't run, because the emitted p5 is changed:
20:25 pugs_svn r19524 | putter++ | -        ::DISPATCH( $Code_say, 'APPLY', ::DISPATCH( $::Int, 'new', 3 ) );
20:25 pugs_svn r19524 | putter++ | +        ::DISPATCH( ( $GLOBAL::Code_say = $GLOBAL::Code_say || ::DISPATCH( $::Routine, "new", ) ), 'APPLY', ::DISPATCH( $::Int, 'new', 3 ) );
20:25 pugs_svn r19524 | putter++ | script/kp6: added missing -B/-C handling for "ruby".  I'm unsure why -C already
20:25 pugs_svn r19524 | putter++ |  seemed to be working.
20:25 pugs_svn r19524 | putter++ | Emit/Ruby.pm: modified an emitted header comment to make origin unambiguous.
20:29 Coke joined #perl6
20:29 * Coke yawns.
20:40 lorn joined #perl6
20:41 pugs_svn r19525 | putter++ | Milestone: script/kp6 understands a ruby backend exists.
20:41 pugs_svn r19525 | putter++ | Currently it's a copy of Perl5.pm.
20:41 pugs_svn r19525 | putter++ | It is emitting slightly broken p5 code, cause unknown.
21:08 pmurias mncharity: you didn't add all nessesary visitors to the processing chain
21:09 pmurias script/kp6 --do ExtractRuleBlock,Token,MetaClass,Global,Emit::Ruby ../../t/01-sanity/01-tap.t | perl -I compiled/perl5-kp6-mp6/lib/
21:09 pmurias works
21:15 pmurias script/kp6 -Cnoexistent defaults to -Cperl5
21:16 TJCRI joined #perl6
21:18 pmurias mncharity: why did you change your nick, i only made the connection that you are putter by looking at AUTHORS
21:21 Gregor__ joined #perl6
21:23 wknight-away joined #perl6
21:23 mmorrow joined #perl6
21:26 pugs_svn r19526 | cosimo++ | Test suite refactoring.
21:26 pugs_svn r19526 | cosimo++ | Moved builtins/control_flow/sleep.t under t/spec/S29-context
21:28 icwiener joined #perl6
21:29 pmurias mncharity: just asking out of curiosity ;)
21:31 Gregor__ left #perl6
21:32 pugs_svn r19527 | cosimo++ | Refactored keys/values tests for arrays and hashes (S29)
21:35 meteorjay joined #perl6
21:36 pugs_svn r19528 | lwall++ | [fudgeall] mung filenames to .fud and run fudge preproprocessor over /spec files only
21:50 jferrero joined #perl6
22:01 Limbic_Region joined #perl6
22:05 pugs_svn r19529 | putter++ | [kp6] ruby backend now actually runs ruby, and emits some ruby code.
22:05 pugs_svn r19529 | putter++ |   script/kp6 -Bruby -e 'say 3'  works, but little else.
22:05 pugs_svn r19529 | putter++ | And the calling convention is likely to change.
22:05 pugs_svn r19529 | putter++ | Ruby version >= 1.9.0 is currently required.
22:08 japhb joined #perl6
22:09 ispy_ joined #perl6
22:10 mncharity pmurias: re nick, people in perl land had short, creative nicks.  back when I didn't realize how much of my life would be spent on p6, I thought "ah, here's a chance to experiment with having a short creative nick!".  perhaps if it were shorter or more creative it would have had a more lasting appeal.  but didn't.
22:12 pmurias mncharity: thanks
22:12 moritz_ perl  script/kp6 -Bruby -e 'say 3'
22:12 moritz_ No such file or directory at script/kp6 line 528.
22:12 mncharity np :)
22:12 pmurias mncharity: did you make
22:13 [particle] mncharity: do you want an mncharity-shaped commit bit?
22:13 mncharity docs/ruby_backend.txt has the "get kp6 to compile Emit/Ruby.pm into the right place" receipe
22:14 pugs_svn r19530 | pmurias++ | [kp6] added missing compiled file
22:14 mncharity [particle]: hmm, I think I'll stick with putter for now.  traditional.  but thanks. :)
22:15 pmurias moritz_: do you have ruby?
22:15 mncharity pmurias: ah, good idea.  thanks.
22:15 moritz_ pmurias: aye
22:15 moritz_ ah, wait
22:15 moritz_ I have ruby1.9 but no 'ruby' symlink
22:16 mncharity I was wondering what to call it.  Is that a vote for "ruby1.9" ?
22:17 moritz_ in Debian Stable ruby is ruby1.8, if you want ruby1.9 you have to say that explicitly
22:17 moritz_ so for me ruby1.9 is more convenient ;)
22:17 mncharity changing script/kp6 ...
22:18 pmurias sleep&
22:18 pugs_svn r19531 | putter++ | [kp6] script/kp6 now uses "ruby1.9" instead of "ruby"
22:18 pugs_svn r19531 | putter++ | to get a >=1.9.0 version of ruby.
22:19 mncharity pmurias: g'night.  thanks for your help.
22:21 moritz_ perl  script/kp6 -Bruby -e 'say 3'
22:21 moritz_ -:5: syntax error, unexpected '('
22:21 moritz_ c_say.( 3)
22:21 moritz_ ^
22:22 mncharity hmm...
22:23 mncharity that's the error message I see when running a ruby1.8 against the -C output.
22:23 mncharity so much for giving a nice error message to people running 1.8 :/
22:23 mncharity did you svn up to r19531 ?
22:31 jjore joined #perl6
22:32 moritz_ yes
22:33 moritz_ and I deinstalled ruby 1.8
22:33 moritz_ so if it runs anything rubyish at all, it is 1.9
22:34 Auzon joined #perl6
22:34 moritz_ I have a libruby1.8 still, but I don't think that should interfere
22:36 mncharity (bbiab - call)
22:37 mncharity ruby1.9 --version  says 1.9.0?
22:37 moritz_ aye
22:37 mncharity mumph
22:38 mncharity such pain because 1.8 doesn't have real lambdas. :/  let's see...
22:39 moritz_ I can see if I can get a newer ruby version from backports.org
22:40 mncharity could you try .call( instead of .( ?
22:40 mncharity my guess is debian is providing a prerelease version of 1.9.0 (as the release version works).
22:40 moritz_ .c_say(2) is valid syntax
22:41 moritz_ ruby 1.9.0 (2006-06-08) [i486-linux]
22:41 mncharity c_say.call(2)
22:41 moritz_ how do I obtain all of the emitted ruby code?
22:41 mncharity yeah, 1.9.0 was released at end of year.  but the development builds from the preceeding year called themselves 1.9.0.  sigh.
22:42 [particle] now that's just silly.
22:42 mncharity released 25 Dec 2007  http://www.ruby-lang.org/en/
22:42 lambdabot Title: Ruby Programming Language
22:43 mncharity moritz_:  perl script/kp6 -Cruby -e 'whatever'
22:43 mncharity [particle]: yeah. :/
22:44 mncharity moritz_:  ruby1.9 -Icompiled/ruby-kp6-mp6 emitted.rb
22:45 moritz_ testing has 1.9.0+20071016-1
22:45 moritz_ not really new either
22:45 _sprocket_ left #perl6
22:46 cspencer joined #perl6
22:46 mncharity moritz_: actually, I think I'm going to leave the calling convention .( .  Running old 1.9 development snapshots opens up a world of random pain, so failing seems the better outcome than continuing.
22:50 moritz_ mncharity: good reasning ;)
22:53 [particle] it'd might be helpful to have back-end appropriate help text, something like :script/kp6 -Bruby -h
22:54 mncharity moritz_: if you haven't yet upgraded, could you wait a moment, and check out a new kp6 with warning message?
22:54 meppl good night
22:54 moritz_ mncharity: ok
22:56 pugs_svn r19532 | putter++ | [kp6] script/kp6: check for old versions of ruby (1.8, and prerelease 1.9).
22:57 mncharity moritz_: ok, should now politely complain.
22:57 moritz_ it does
22:57 moritz_ mncharity++
22:57 mncharity :)
22:58 mncharity re testing has 1.9.0+20071016-1, :(
23:00 mncharity I'll try to remove the 1.9 dependency once things are further along.  to use 1.8, one has to play gensym games, because 1.8 doesn't have a normally scoped lambda.
23:00 gbacon @pl \a b -> head a == head b
23:00 lambdabot (. head) . (==) . head
23:01 moritz_ mncharity: I don't think that's high priority
23:02 mncharity but it's hard to emit code, and use blocks for lexical scoping, where variables may be mis-bound.  eg,
23:03 mncharity x = 4; ->(x){x=3}; #x == 4 works, but -> is 1.9
23:03 mncharity x = 4; (proc {|x| x=3}).call(14); # x==3 :(
23:03 mncharity err,
23:04 mncharity x = 4; (->(x){x=3}).call(14); #x == 4
23:05 Jamtech joined #perl6
23:05 mncharity or something like that.  know recognized as a Bad Idea.  but... still have to live with the concequences for another year or so.
23:05 mncharity s/know/now/
23:06 * moritz_ drifts off to sleep
23:06 mncharity g'night!  thanks for trying it, and turning up bugs :)
23:07 jferrero joined #perl6
23:11 Jamtech_ joined #perl6
23:12 pugs_svn r19533 | putter++ | [kp6] script/kp6: fixed check for ruby 1.9.0 prerelease versions.
23:16 cspencer have the new #?perl6 test skipping commands been added into the p6 distributed with parrot yet?
23:17 Auzon joined #perl6
23:21 [particle] we're working on it.
23:21 [particle] #?perl6: todo(...) works
23:21 [particle] skip() doesn't
23:23 cspencer ah ok
23:23 pugs_svn r19534 | putter++ | [kp6] Emit/Ruby.pm: removed ruby version check from emitted code.
23:23 pugs_svn r19534 | putter++ | Older versions parsefail, so it wasn't running anyway.
23:23 cspencer is that in the current check svn revision?
23:23 pugs_svn r19535 | cosimo++ | Refactored `kv' tests under S29-array & S29-hash.
23:23 pugs_svn r19535 | cosimo++ | Fudge-ify :todo test markers.
23:23 cspencer er...checked in, rather
23:25 cosimo [particle]: isn't it  #?perl6: todo 'bla' ? without parens?
23:25 [particle] cosimo: now it is. but that doesn't work yet :)
23:25 [particle] things are changing very quickly
23:28 cosimo I have changed 2-3 times the tests in t/spec ... :-)
23:30 TimToady hopefully we're converging on a solution, but I haven't yet figured out how to shim fudgeall into Makefile.PL for pugs...
23:31 cosimo No problem, really.
23:31 TimToady well, I'd just like the fudging to be automatic where it's needed
23:31 alester joined #perl6
23:31 TimToady and I'm scared of hacking on Test::Harness directly :)
23:32 [particle] i'll be changing t/spec shortly if you don't beat me to it
23:32 [particle] s/perl6/rakudo/g
23:32 TimToady I beat you to it more than a day ago...
23:32 cosimo TimToady: I didn't understand the "fudge" thing until I read your emails twice :-)
23:33 TimToady it's a colloqial use of the term
23:33 TimToady fudging is a kind of cheathing
23:33 TimToady cheating
23:33 cosimo IIUC, it's a way to "hide" perl6 implementation differences from a test point of view. Isn't it?
23:33 TimToady as in Fudge Factor
23:34 Khisanth you mean I wasn't supposed to pour fudge on that ... hmmm
23:34 TimToady it's a way to get more information out of a partial success than just by blowing up halfway through, or not parsing at all
23:34 [particle] TimToady: C<#?perl6: skip_block('...');> becomes <$?rakudo: skip '...'>, correct?
23:34 TimToady yes
23:35 cosimo I tried to ack for skip_block recently (maybe a week ago) but nothing showed up...
23:35 TimToady except #? still not $?
23:35 [particle] oops, right, #?
23:36 TimToady and you can say #?rakudo: 3 skip '...' to skip the next 3 tests without using a block
23:36 TimToady putting the count in front makes it easier to keep it separate from the "reason" arguments
23:36 REPLeffect joined #perl6
23:37 TimToady (which are actually extensible, just being a string that is passed through)
23:41 pugs_svn r19536 | cosimo++ | Added svn props to test scripts
23:43 cosimo Is it normal that http://perlsix.org/svn/pugs/revision?rev=19536 , which is only svn props added, shows all rows removed and re-added?
23:43 lambdabot Title: revision: /pugs (Rev: 19536, via SVN::Web)
23:53 pugs_svn r19537 | cosimo++ | Refactored pairs tests into S29-arrays and S29-hash
23:54 TimToady I would guess it's diffing a file with CRLF with a file with just LF
23:54 TimToady ETOOMANYWITHS

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

Perl 6 | Reference Documentation | Rakudo