»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:08 ymasory joined
dalek ecza: 3859825 | sorear++ | lib/ (2 files):
Switch BigInteger implementation, reduce Rat literals
 00:09
00:25 donri left 00:41 tyatpi joined, gbacon_ joined 00:50 Eevee left 01:35 kst left 01:36 bacek left 01:39 Eevee joined 01:40 bacek joined 01:43 eternaleye left, PZt left, eternaleye joined 01:46 kst joined
dalek ecza: 82e9ec8 | sorear++ | lib/ (2 files):
Add coercions from new numbers to double and bool
 01:46
01:50 bacek left 01:58 bacek joined, PZt joined 02:02 Moukeddar joined 02:16 dayangkun joined 02:17 Moukeddar left 02:48 dayangkun left 02:49 silug joined
sorear rakudo: say (6 / 3).WHAT 02:56
p6eval rakudo 4398a2: OUTPUT«Rat()␤»
sorear Is this correct?
02:56 aks joined
sorear welcome, aks. New here? 02:57
Garda hugme: hug sorear
no hugme bot?
03:01 lue joined, aks left 03:04 dalek left 03:05 p6eval left 03:06 silug left 03:08 sorear left, diakopter left, TimToady left 03:14 sorear joined 03:15 diakopter joined, TimToady joined, alester joined 03:16 daxim_ left 03:19 p6eval joined, ChanServ sets mode: +v p6eval 03:22 gbacon_ left 03:23 Su-Shee_ joined 03:26 am0c left, Su-Shee left
sorear my biggest complaint with C# is a failure to borrow C++'s functional cast syntax 03:31
03:33 tyatpi left 03:38 am0c joined 03:41 thou left 03:45 zorgnax joined
sorear rakudo: say 5i % 2i 03:55
p6eval rakudo 4398a2: OUTPUT«modulus (Complex % Complex) not defined␤»
sorear rakudo: say 5i % 2
p6eval rakudo 4398a2: OUTPUT«modulus (Complex % Int) not defined␤»
04:04 kst` joined 04:05 kst left 04:07 kst`` joined 04:08 kst` left 04:24 tyatpi joined 04:29 tyatpi left
moritz \o 04:32
04:33 hugme joined, ChanServ sets mode: +v hugme
moritz hugme: hug me 04:33
hugme hugs moritz
sorear odd 04:38
dalek quit for no reason
04:38 dalek joined, ChanServ sets mode: +v dalek
dalek ecza: 9788237 | sorear++ | lib/ (3 files):
Secondary fixes revealed by (regression testing)++
 04:54
sorear pugs: FatRat 05:05
p6eval pugs: OUTPUT«*** No such subroutine: "&FatRat"␤ at /tmp/pEkApilTGL line 1, column 1 - line 2, column 1␤»
sorear pugs: Rat
p6eval pugs: ( no output )
sorear no /FatRat/ in my mu checkout either 05:07
moritz FatRat is much newer than pugs' last patches
sorear does any implementation except niecza have it yet?
moritz no 05:08
sorear \o/ 05:09
05:09 satyavvd joined
sorear I guess? 05:09
sorear still needs to make niecza's FatRat and Complex types usable 05:10
05:23 mtk left 05:25 alester left
dalek ecza: b48708f | sorear++ | lib/Builtins.cs:
Make -, *, /, % type-aware
 05:29
sorear niecza: say 39549590295082340582305802338 + 29349934
p6eval niecza v5-37-g9788237: OUTPUT«39549590295082340582335152272␤»
05:33 envi joined, mtk joined
dalek ecza: 0346395 | sorear++ | lib/Builtins.cs:
Type-aware comparison operators
 05:46
ecza: 4c03af8 | sorear++ | lib/ (3 files):
Remove many superfluous uses of Num from the C# code
 05:59
06:04 silug joined 06:05 jaldhar left, f00li5h left, jaldhar joined 06:08 nymacro joined 06:20 jaldhar left 06:23 jaldhar joined
tadzik phenny: tell jnthn t/p6regex/01-regex.t fail: wklej.org/id/533239/ 06:32
phenny tadzik: I'll pass that on when jnthn is around.
06:42 birdwindupbird joined 06:47 birdwindupbird left 06:51 _jaldhar joined 06:52 kuku_ joined, jaldhar left, kuku_ is now known as birdwindupbird 06:55 silug left
sorear niecza: say ((1 / 0x1_0000_0000) * (1 / 0x1_0000_0000)) 06:57
p6eval niecza v5-40-g4c03af8: OUTPUT«5.42101086242752E-20␤»
sorear niecza: given ((1 / 0x1_0000_0000) * (1 / 0x1_0000_0000)) { .say; .WHAT.say } 06:58
p6eval niecza v5-40-g4c03af8: OUTPUT«5.42101086242752E-20␤Num()␤»
sorear niecza: given ((FatRat.succ / 0x1_0000_0000) * (1 / 0x1_0000_0000)) { .say; .WHAT.say }
p6eval niecza v5-40-g4c03af8: OUTPUT«1/18446744073709551616␤FatRat()␤»
sorear \o/
07:00 _jaldhar left, _jaldhar joined
sorear moritz, diakopter: Would you mind if I ran a continuous integration script on host04? (it would basically be like what p6eval is already doing, except also running spectests and the bootstrappability test and reporting the results) 07:05
07:08 Trashlord left 07:12 molaf_ left 07:14 Trashlord joined 07:17 Garda left 07:21 Yappoko__ left, Yappoko__ joined 07:22 satyavvd left, satyavvd joined 07:36 Trashlord left 07:37 Trashlord joined 07:48 satyavvd left, _buno_ joined 07:49 _buno_ left 07:51 nymacro left, nymacro joined, Trashlord left 07:54 Su-Shee_ is now known as Su-Shee, Trashlord joined 07:59 Garda joined 08:12 wamba joined 08:21 satyavvd joined 08:24 satyavvd_ joined, satyavvd left
moritz sorear: just take care of load/memory usage 08:37
dalek ecs: e1801df | moritz++ | S05-regex.pod:
[S05] fix a fossile (though the whole section looks pretty fossile-ish)
 08:38
ecs: b52825b | moritz++ | S32-setting-library/Exception.pod:
[Exceptions] stub out Failure and Backtrace
08:42 satyavvd_ left 08:43 kjeldahl joined 09:01 Garda left 09:09 envi left 09:14 skangas left 09:37 skangas joined 09:38 silug joined 09:42 skangas_ joined, skangas_ left 09:45 silug left 09:46 silug joined 09:50 silug left, silug joined 09:52 Mowah joined 10:07 dudulz joined
tadzik hrm. Actions.pm is Nqp, not Perl6, right? 10:12
sorear right
tadzik fun. So writing action methods for Rakudo will be a bit different than for a Perl 6 module 10:13
10:14 silug left
sorear btw 10:14
10:14 silug joined
jnthn tadzik: Yeah, but it's a subset. 10:14
phenny jnthn: 06:32Z <tadzik> tell jnthn t/p6regex/01-regex.t fail: wklej.org/id/533239/
sorear niecza: my Int $x = 495394393904582058028505238; $x++; say $x
p6eval niecza v5-40-g4c03af8: OUTPUT«495394393904582058028505239␤»
sorear perl6: my Int $x = 495394393904582058028505238; $x++; say $x 10:15
tadzik jnthn: yeah, I know that. But in Actions.pm I'm no longer doing Pod6::Block.new, but some Past::Op instead, creating the Pod6::Block, si?
p6eval pugs, niecza v5-40-g4c03af8: OUTPUT«495394393904582058028505239␤»
..rakudo 4398a2: OUTPUT«Type check failed for assignment␤ Container type: Int␤ Got: Num␤ in '&infix:<=>' at line 1␤ in main program body at line 7654:CORE.setting␤»
tadzik sorear: nice!
jnthn tadzik: I can't answer that without knowing more about what you want to do. 10:16
tadzik jnthn: I need to move this github.com/tadzik/gsocmess/blob/ma...ons.pm#L40 to Rakudo Actions.pm, and likely push those Blocks to some $POD variable 10:17
sorear tadzik: it has Rat and FatRat and Complex too, though the latter two are hard to use. 10:18
jnthn tadzik: What's the intention? That these things are available at runtime?
tadzik jnthn: yes, I need to test them 10:19
test that the parsing did a good job
jnthn: I have the tests for the outside code, but I'd like to know if that still works when I moved it inside Rakudo
jnthn tadzik: Is the POD normally made available at runtime? 10:20
tadzik jnthn: yes, through a $=POD variable
jnthn tadzik: Where can I read more about it? 10:21
tadzik jnthn: S26
jnthn tadzik: OK. So $=POD is essentially per compilation unit, it seems. 10:25
tadzik I think so 10:26
jnthn std: say $=POD;
p6eval std 4b1b100: OUTPUT«ok 00:01 113m␤»
jnthn std: say $=WTF;
p6eval std 4b1b100: OUTPUT«ok 00:01 113m␤»
jnthn hm :)
sorear $=WTF is totally legit
jnthn sorear: What does it mean?
sorear it refers to a =begin WTF block in the file
or something like that
S29 was rather handwavey on the subject iirc 10:27
10:27 nymacro left
jnthn sorear: And...if there's more than one of those? 10:27
jnthn doesn't see any $=... variables other than $=POD in S26.
sorear ISTR they go into an array
tadzik S26 mentions $= 8 times, and never explains fully what it is
jnthn ah 10:28
Magical file-scoped values live in variables with a C<=> secondary
sigil.
(says S02)
10:31 silug left 10:32 silug joined 10:33 wamba left
tadzik yay, I didn't break the spectest 10:39
sorear \o/
tadzik I'm not sure '=begin item ACCEPTS()' is a valid Pod 10:46
either '=item ACCEPTS() – bla bla' or '=begin item\nACCEPTS()\nbla bla' 10:47
dalek ecza: b2c61f5 | sorear++ | / (6 files):
Implement .Rat
 10:48
tadzik "Typenames that are entirely lowercase or entirely uppercase are reserved" -- English-o? 10:49
sorear no 10:51
tadzik oh, I fail at reading 10:52
missed the 'that'
sorear perl6: my $case = 0.130442081080024.Rat; say "$case.perl() {0e0+$case}" 10:54
p6eval pugs: OUTPUT«*** No such method in class Rat: "&Rat"␤ at /tmp/MX29LwsYOq line 1, column 7-33␤»
..niecza v5-41-gb2c61f5: OUTPUT«686/5259 0.130443050009508␤»
..rakudo 4398a2: OUTPUT«776/5949 0.130442091107749␤»
sorear found a superior algorithm on wikipedia.
tomorrow: the full set of coercions .FatRat .Num .Complex; Str.Numeric supporting all numeric types; transcedental* functions 10:55
*sqrt isn't transcedental but I don't have a good inclusive term 10:56
sorear out
10:57 silug left
dalek kudo: 30ebf78 | tadzik++ | src/core/ (2 files):
Remove invalid Pod

S26 states: After the typename, the rest of the C<=begin> marker line is treated as configuration information for the block. This information is used in different ways by different types of blocks, but is always specified using Perl6-ish option pairs.
We don't parse the block configuration information yet, plus they're not needed anyway there.		 11:01
tadzik s/anyway there/there anyway/, but eh 11:02
jnthn tadzik: Hmm, will have to try and keep your changes regularly merged into nom :) 11:05
11:06 Patterner left
tadzik jnthn: the changes will be in a separate branch 11:06
11:07 Psyche^ joined, Psyche^ is now known as Patterner
tadzik I think modifying src/core/Something.pm does not cause core.pm to regenerate 11:10
11:10 REPLeffect left 11:13 REPLeffect joined 11:17 mj41 joined 11:29 donri joined 11:46 Garda joined
Garda hey 11:47
how stable is perl6
I mean, is the language likely to change alot by the time it's released properly?
donri a lot, not sure, change, surely
sbp Garda: welcome! I don't think there will be any "released properly" 11:48
tadzik :)
sbp but as time goes on, less and less gets changed, and more and more gets coded
tadzik Garda: you can look at how the specification changes and see for yourself how much does it change
sbp and I think we're very far along that curve. most of the bits you would want to use are "set"
11:48 kcwu left
tadzik Garda: see github.com/perl6/specs/commits/master 11:49
sbp only occasionally have I seen things I'd want to use that are in flux. recent discussions about numerical base conversion for example
Su-Shee how is a monthly release not "released properly"?!
sbp Su-Shee: are there Windows, OS X, and Linux binaries yet? :-)
tadzik Su-Shee: I'm sure Garda means "a complete PErl6 implementation"
Garda yea, I meant sorta like when someone blesses it and says, "it's done" 11:50
I don't know
Su-Shee sdp: strange definition of "released properly"
Garda: it's blessed plenty. download it, use it, write code. see for yourself if it's useful to you or not. 11:51
tadzik almost everyone is confused :)
Su-Shee I'm not. ;)
tadzik yeah, what Su-Shee says. Try for yourself and see if it fits, don't wait for the blessing
sbp *sbp. and we all have our little mental milestones
Su-Shee sometimes I think people _want_ design by commitee and a stamp of approval ;) (which only some languages have anyways.. :) 11:53
tadzik yeah
sbp some are more mental than others 11:54
tadzik I think masak should be our comitee, he's experienced in declaring things Official and Complete, and no one ever opposes
sbp seconded
tadzik we should ask him to declare Perl 6 Properly Released
Su-Shee well let me put it that way: if browsers and web developers would actually _wait_ until the w3c approves anything - I think we could start using html4 then this year or last year. ;) 11:55
tadzik yay, Rakudo compiles itself!
sbp you think HTML 4.01 implementation is complete? how quaint
tadzik I mean, the kind of Rakudo I mangled
sbp —
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN">
<title/GS/<body onLoad="document.f.q.focus()"<form
action="google.com/search" name=f/<p/<input
name=q< type=submit>//
—
see how many browsers properly render that. valid HTML 4.01 11:56
Su-Shee sbp: that was my way of saying that I give a shit if or if it is not. the real world is somewhere else anyways.
sbp using various SGML shortcuts that they left in for no obvious reason
I'm arguing the same point of view more vehemently
:-) 11:57
Su-Shee yes that was my point. who gives a shit if someone "officially" released perl 6 at some random level everyone defines differently for himself instead of just trying and using it?
sbp it's a tax that perl6, as the first truly new popular 21st century language, has to inevitably pay... 11:58
tadzik people who still look at Perl6 like how dog looks at a hedgehog. I think there're plenty of them actually, I can understand that
Su-Shee from a ceremonial point of view it's released very properly and very officially with a new baptism every month :)
right now, it's elitist and very much not popular. ;) 11:59
tadzik Same as people won't stop laughing at GNU/Hurd until the Hurd people say "yay, we're done!"
sbp . o O ( another mental milestone: a better package manager than gem and easy_install... )
tadzik hey, what's wrong with panda? :) 12:00
Su-Shee tadzik: I agree with most of those people, btw. If it wasn't for my overall never dying love for perl, I wouldn't have looked at perl 6 again at least 4 years ago, basically.
but I'm just a user, I look differently at languages. 12:01
sbp tadzik: good question, actually 12:02
I will attempt to go through the "doing stuff with perl6" process from scratch
tadzik sbp: I'd *very* appreciate a GH ticket for every single thing you miss 12:03
sbp let's see what bits can be improved and whatnot
sbp tries to make sure all old perl6 bits are properly removed first
12:03 skangas left
sbp . o O ( hmm, why do I have two versions of ffmpeg... ) 12:04
okay, going to start with the wiki page I made that everybody forgot about:
github.com/perl6/misc/wiki/Get-Into-Perl6
notice that February and March tarballs are missing from the Rakudo Star downloads list 12:05
did pmichaud go on holiday?
tadzik no, Star stopped being monthly released
sbp becomes the 433rd downloader of the April R* release
oh? why? and it's going again now? 12:06
"256MB is known to be too small for building Rakudo"
lucky I'm not doing this on my server...
tadzik there is an announcement somewhere. Basically, Star is released when something significant happens 12:07
Su-Shee I thought it's monthly rakudo and quarterwise rakudo*?
tadzik dalek, wakey wakey
sbp ah, okay
I'm going to do this with my normal procedure, installing into ~/usr/opt
just to be tricky
README doesn't note whether ICU 4.6 or 4.8 is to be preferred, or which versions at all are compatible and which aren't. I'll assume that 4.6.1 will work 12:08
a direct link to site.icu-project.org/download/46#ICU4C-Download in the README would be nice 12:09
(binaries for everything except OS X, heh)
sbp gets download.icu-project.org/files/icu4..._1-src.tgz
icu4c-4_6_1-src.tgz rudely untars to icu/ rather than icu4c-4_6_1-src/ 12:11
readme.html? wonderful...
"How To Download the Source Code", heh 12:12
wtf is this rubbish:
—
Depending on the platform and the type of installation, we recommend a small number of modifications and build options.
Namespace: By default, unicode/uversion.h has "using namespace icu;" which defeats much of the purpose of the namespace. (This is for historical reasons: Originally, ICU4C did not use namespaces, and some compilers did not support them. The default "using" statement preserves source code compatibility.)
We recommend you turn this off via -DU_USING_ICU_NAMESPACE=0 or by modifying unicode/uversion.h:
—
it wants me to patch the default charset on OS X, so I suppose I'll do that 12:13
patch says: 12:14
#ifndef U_CHARSET_IS_UTF8
-# define U_CHARSET_IS_UTF8 0
+# define U_CHARSET_IS_UTF8 1
#endif
but my source says this:
#ifndef U_CHARSET_FAMILY
# define U_CHARSET_FAMILY 0
#endif
12:14 MayDaniel joined
sbp oh, the patch offset is wrong 12:14
line 185 in my version. way to update the readme, 12:15
but done, anyway
"It may make sense to build the ICU code into static libraries (.a) rather than shared libraries (.so/.dll)." but it doesn't say where or when this might make sense
no instructions for install on OS X, going to use the "UNIX" instructions 12:16
phew. icu/readme.html#HowToConfigureICU says to try --help on this pre-configure script 12:17
idea seems to be ./runConfigureICU MacOSX for me
I note there's no --prefix option
wait, now it says to run make 12:18
but I didn't pass it a --prefix!
perhaps [ configurearg ... ] includes flags?
yep, this worked: 12:19
$ ./runConfigureICU MacOSX --prefix=$HOME/usr/opt/rakudo
export CPP= CC= CXX= CPPFLAGS= CFLAGS= -O2 CXXFLAGS= -O2 LDFLAGS= MAKE=
Running ./configure --prefix=/Users/sbp/usr/opt/rakudo for MacOS X (Darwin) using the GNU C++ compiler
so when they say "arg", they really mean "flag or arg"
(making rakudo use icu will be fun) 12:20
12:20 mj41 left 12:22 snearch joined
sbp skips make check 12:25
12:25 Moukeddar joined, Moukeddar left, wamba joined
sbp okay, now it says: 12:25
$ perl Configure.pl --gen-parrot
$ make
$ make install
which again has no prefix, so I'll try passing --help to Configure.pl
$ perl --version 12:26
This is perl 5, version 12, subversion 2 (v5.12.2) built for darwin-2level
no --prefix option
there's also no way to point to where ICU is, despite me installing it in a non-standard location
sbp runs what it says to run anyway 12:27
seems like it's only going to build locally, but it would be nice if it made that clear before the install lines 12:29
not clear what make install will do, whether it will use /usr/local etc. 12:30
oh, silly me. a little bit further down:
The "--prefix=" option can be provided to Configure.pl
to change the location of the install directory.
despite the fact that Configure.pl doesn't list --prefix!
so, $ perl Configure.pl --prefix=$HOME/usr/opt/rakudo --gen-parrot 12:32
Garda how can I check if a hash exists 12:33
12:33 dudulz left
Garda if exists %people{"bob"} 12:34
Su-Shee the hash or if the key exists or if the key exists and has a value? :) 12:37
sbp Garda: github.com/perl6/roast/blob/master...h/exists.t 12:39
exists %h1: 'a'
%h1.exists('a')
there's also defined(...) 12:40
but I assume you want .exists
12:41 [Coke] left
jnthn rakudo: my %h = a => 1; say %h.exists('a'); say %h.exists{'b') 12:41
p6eval rakudo 30ebf7: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<{ }>, couldn't find final '}' at line 22␤»
12:41 MayDaniel left
jnthn rakudo: my %h = a => 1; say %h.exists('a'); say %h.exists('b') 12:42
p6eval rakudo 30ebf7: OUTPUT«Bool::True␤Bool::False␤»
Garda sbp, thanks, I'm looking at it now
12:43 [Coke] joined
sbp heh: src/core/YOU_ARE_HERE.pm 12:44
Garda Su-Shee, wanted to know if a key exists in a hash I know I have created
sbp, thanks, just what I wanted
sbp :-) 12:45
perhaps it would be a good idea to get a quantitative estimate of how much of the synopses are implemented in rakudo, and then set an arbitrary large percentage figure before having TT make some kind of super announcement that perl6 is GO, GO, GO 12:48
hits 98% coverage, rakudo gets slashdotted, September of No End starts for perl6
make and make install done
lots of binaries 12:49
$ ls
./parrot-prove*pbc_merge*
../parrot_config*pbc_to_exe*
ops2c*parrot_debugger*perl6*
panda*parrot_nci_thunk_gen*ufo*
parrot*pbc_disassemble*ufobuilder*
parrot-nqp*pbc_dump*
notice that panda is in there already
sbp wonders which ones to link in ~/usr/bin, always likes to play the minimalist game
sbp plumps for perl6 and panda 12:50
$ perl6 --version
This is Rakudo Perl 6, version 2011.04 built on parrot 3.3.0
Copyright 2008-2011, The Perl Foundation
now going to try installing JSON::Tiny
sbp tries panda --help 12:51
funny, using --help made it download into ~/.panda/projects.json
phenny sbp: 12:51Z <Morbus> tell sbp news.nationalgeographic.com/news/20...-12-plane/
tadzik oh, fun
sbp $ time panda help 12:52
Usage: panda [command [args]]
...
panda help 1.98s user 0.13s system 71% cpu 2.930 total
why does it take 1.98 seconds to print the help message?
I guess it's looking online to try to update projects.json each time I run it?
tadzik how about "Rakudo is slow"? :)
sbp even if I'm just running panda help?
tadzik it shouldn't, I'll take a peek 12:53
yeah, the panda object is initialized before MAIN
sbp right. even though rakudo is slow, I'm not sure printing a few lines to stdout should take 1.98 seconds... :-)
sbp now tries $ panda list | grep -i JSON 12:54
tadzik the bin/panda script is a bit stupid. I was trying to reuse as much code as possible for the REPL and for the command-line args, and it ended up being crap :)
I'll fix it today after dinner
sbp ah, I see. thanks!
that's good service :-)
tadzik btw, I see that dalek does not report my new branch but rakudo/gsoc-podparser is ready for testing :) 12:55
sbp it took 12.91 seconds to search for JSON in the list, might be some room for improvement there too
$ time panda list | grep -i JSON
JSON::Tiny
JSON::RPC
panda list 12.91s user 0.45s system 64% cpu 20.639 total
grep -i JSON 0.00s user 0.00s system 0% cpu 20.634 total
sbp thwaps dalek
is there no way to see what packages are currently installed? 12:56
tadzik it is
list would print [installed] or [installed as a dependency] or something like this
sbp because the fact that it downloaded a projects.json file makes me wonder whether there isn't some sort of JSON package installed already
tadzik but only if it's installed by panda
JSON is a panda dependency, yes
sbp ah, so grepping for "install" should work?
but JSON won't show up because panda didn't install it?
it's just included in rakudo? 12:57
tadzik if you installed it by panda. Is that Star, or did you bootstrap panda yourself?
sbp Star
tadzik in Rakudo no, in star yes
you can run 'panda install Pies' and it will reinstall all the deps as well
now when I think about it, 'panda' is less confusing and more recognized
sbp what would happen now if I used panda to install JSON::Tiny?
tadzik it will install JSON::Tiny to your ~/.perl6/lib 12:58
sbp but perl6 won't use it?
tadzik it will, why not? 12:59
it depends what's higher in PERL6LIB
kernel 2.6.39 is acting silly
sbp how do I print the source that a module was loaded from? 13:00
can't see anything in the S11 test suite
tadzik I don't think you can
or I don't understand 13:01
sbp well, I can get the INC list:
$ perl6 -e 'say join(":", @*INC)'
/Users/sbp/.perl6/lib:/Users/sbp/usr/opt/rakudo/lib/parrot/3.3.0/languages/perl6/lib:. 13:02
but say I didn't know the INC list, or wanted to test
I do 'use JSON::Tiny'
then I want to say some attribute of JSON::Tiny to see where it's from
or something deep in the module import bowels
tadzik ah 13:04
13:05 perlmonkey joined
sbp . o O ( plan 3? ) 13:07
tadzik still far to plan 9 (ha ha!)
sbp hehe
no idea on printing a module source then?
been through all *.t files in S11-modules now
13:07 wamba left
tadzik nope 13:07
but, it's achieveable 13:08
1) File::Find the module in @*INC, 2) slurp it :)
sbp hmm. well isn't the provenance metadata saved internally? 13:09
perlmonkey hi
sbp hi perlmonkey
anyway, I'll try installing a module
perlmonkey if I'm iterating over a hash of hashes, is this correct way: for my $day (sort (keys(%{$D1}))) {
sbp github.com/colomon/Math-Prime/ seems a good one
jnthn star: use JSON::Tiny; say %*INC.perl
p6eval star 2011.04: OUTPUT«undef␤»
jnthn aw.
wait, undef?!
star: use JSON::Tiny; say %*INC<JSON::Tiny> 13:10
p6eval star 2011.04: OUTPUT«Any()␤»
perlmonkey paste.debian.net/117617/ <-- here is a more complete example
tadzik perlmonkey: I think you're writing Perl 5 code, this is a Perl 6 channel
perlmonkey :( 13:11
tadzik you may want to consider asking on #perl-help on irc.perl.org
perlmonkey ok sorry
sbp —
$ panda install Math::Prime
==> Fetching Math::Prime
==> Building Math::Prime
==> Testing Math::Prime
Unknown option: e
Tests failed for Math::Prime
—
tadzik eww, please paste somewhere else
sbp no, heh
13:11 perlmonkey left
tadzik anyway, you're running Mac OSX, aren't you? 13:11
sbp never understood that "etiquette", never going to follow it without a decent argument for it :-)
yep
13:11 satyavvd joined
tadzik I think it's because the bigger the paste, the harder it gets to follow a discussion 13:12
13:12 mtk left
sbp yep, that's certainly true when lots of people are talking or when the pastes are large 13:12
tadzik sbp: your prove, or TAP::Harnes, I never know, is probably quite old
sbp but I think it's okay to use discretion in all cases, and I don't think a paste of the size above is at all damaging to the integrity of the communication here :-) 13:13
hmm, okay. what version is likely to be required, and how can I find out what I have vs. what I need?
Su-Shee it makes people having to read and parse utterly useless stuff in the backlog. the developers read each other due to timezones differences. 13:14
sbp yep, but when a paste is clearly delimited it's not hard to skip it
and anyway, we've expended more lines now talking *about* the paste than the paste itself spanned
whereas the paste itself was on topic, and the conversation about it is not
Su-Shee and who are you to decide what's hard to skip and what not for _all_ others? 13:15
sbp as I say, it's down to discretion. and my belief based on ten years of hardcore irc usage is that a paste of that size poses no problem. if somebody has a very compelling argument to the contrary then of course I'll update my usage patterns based on that
Su-Shee well you asked for an explanation, you got several, you're refusing because you don't like them. hence the conversation. ;) 13:16
sbp the explanations conflict with my experience
they seem de jure, whereas my experience is de facto
Su-Shee well I have 15 years of hardcore irc usage if it comes down to that and _please_ just accept the culture of a channel. which your decade of irc should tell you anyways.
sbp I think it's worth going against the grain to dispell dogmata of these kinds 13:17
Su-Shee then enjoy.
13:17 Su-Shee left
sbp peculiar 13:18
(sorry about that, didn't predict that effect)
tadzik: "Unable to find module 'TAP::Harness' in the @*INC directories." (same with Harnes) 13:19
I don't know what prove is, and it's hard to search for 13:20
13:20 dual left
sbp unless you mean standard /usr/bin/prove, which actually seems likely 13:20
arnsholt which prove in a terminal prints the full path to the command
(prove is the program used to run and collate results from tests)
sbp thanks. its --version is reporting: prove v2.64, using Test::Harness v2.64 and Perl v5.10.0 13:21
arnsholt And the version of prove installed with system OS X is from roughly the lower neolithic IIRC
sbp hehe. right, my local perl installation is 5.12
so I'll try updating that, thanks
sbp looks into it
arnsholt Is it from MacPorts, or is it system Perl? 13:22
sbp the 5.10 is system perl. the 5.12 is manually built and installed from source
13:22 mtk joined
arnsholt In any case, the port you're looking for is p5-test-harness 13:22
sbp ah, I have it in ~/usr/opt/perl5.12/bin but it's not linked 13:23
tadzik sbp: yeah, I mean the standard prove
sbp one of the drawbacks of the minimalism
tadzik: when panda failed, would it have left stuff lying around that needs to be cleaned?
or when I do install again, will it be okay with the previous failed installation attempt? 13:24
arnsholt Yeah, put that in your $PATH (or replace the system one with a symlink to that, or whatever) and it'll work
sbp (reports "TAP::Harness v3.17 and Perl v5.12.2" now)
tadzik sbp: I think so
sbp: well, it can reuse this stuff anyway
arnsholt Yeah, 3.17 is less ancient
sbp okay, giving "$ time panda install Math::Prime" a go
hmm, less ancient... :-) 13:25
it's testing now, so apparently less ancient enough!
arnsholt Yeah, IIRC that option was renamed in v.3
(I had trouble of a similar nature in a personal project last year)
sbp does, though, seem to be stuck on "t/01-basic.t .. 1/?" 13:26
perhaps the tests are just intensive though. checking source...
github.com/colomon/Math-Prime/blob...01-basic.t 13:27
doesn't look like it should be taking this long, guess it's stalled
oh, wrong
Succesfully installed Math::Prime. great!
doing sanity testing now
works fine. I also note that the README does say "It's still disturbingly slow.", which explains the apparent test stall; I should have looked at the README first 13:28
tadzik: are there packages that can install scripts into bin/ too? 13:29
13:32 orafu left, kcwu joined 13:35 skangas joined
tadzik sbp: sure 13:37
like panda
13:37 huf left
tadzik I think LWP::Simple does it too 13:37
sbp got a good example of one that I could try out? modules.perl6.org doesn't seem to say which on... thanks
tadzik also, Mandelbrot, ufo, Yapsi 13:38
sbp got an error. pastebin.ubuntu.com/611410/ 13:39
Garda can confirm: perl is awesome! :) 13:42
sbp including perl6? :-)
Garda so I spent the last 30mins or something writing an elo calculator
sbp, in perl6 :)
sbp elo? this? en.wikipedia.org/wiki/Elo_rating_system 13:43
tadzik Garda: nice :)
sbp: funny 13:44
reproducable?
Garda sbp, yea, it takes a results.txt file
sbp tadzik: running again now 13:45
Garda with names listed as Foo:Bar:0.5
sbp Garda: source online anywhere?
Garda and calculates everyone elo score
sbp, give me a few mins to finish it
i'll pastebin it somewhere
tadzik Garda: you may want to make a module for this
sbp thanks
tadzik: new error this time
pastebin.ubuntu.com/611413/
tadzik something like search.cpan.org/~tbone/Chess-Elo-1....ess/Elo.pm 13:46
tadzik grumbles
sbp: have you installed anything with dependencies before?
did the tests pass for panda?
sbp I didn't run make check before installing panda 13:47
13:47 woosley joined, am0c left
sbp note that panda was just part of rakudo star, I didn't even know it was included 13:47
lemme know if there's any diagnostic I can run now 13:48
13:49 gbacon left
tadzik you can clone the panda repo and run 'ufo && make test', assuming you have ufo installed 13:49
13:49 pernatiy_ joined
tadzik Pies is now renamed to Panda 13:51
sbp yep, will do
13:51 pernatiy left
sbp failed, error message coming up 13:52
tadzik the same one?
sbp pastebin.ubuntu.com/611415/
nope. is there some specific directory I have to run that in?
it can't find t/ basically
tadzik que?
in the directory you cloned the repo to 13:53
dalek osystem: f128f99 | tadzik++ | META.list:
Pies in now panda
 13:54
tadzik bleh, typo
sbp in rakudo/, or in modules/Pies/ or what?
13:54 huf joined
tadzik git clone git://github.com/tadzik/panda; cd panda; ufo; make test 13:54
sbp doing that now. note that my panda is from rakudo star of course though, so probably older 13:55
pastebin.ubuntu.com/611416/
another fail. gotta run for a bit now, back in 30-60 mins
tadzik aaarrr 13:57
dalek ar: 4268181 | tadzik++ | / (2 files):
Track the Pies -> panda rename
 13:58
href="https://modules.perl6.org:">modules.perl6.org: e3ddd2f | tadzik++ | web/index.tmpl:
Track the Pies -> panda rename
14:01 am0c joined 14:02 silug joined
dalek href="https://perl6.org:">perl6.org: 290a262 | tadzik++ | source/ (2 files):
Pies is now panda
 14:04
14:07 silug left
Garda sbp, it's really simple: 14:07
pastebin.ca/2066882
pastebin.ca/2066883
the second file is an input file
it wants to be called results.txt 14:08
dalek ar: ff57d8f | tadzik++ | skel/build/Makefile.in:
Remove the Pies leftovers
tadzik Garda: what is the purpose of line #30 in the first one?
Garda none: sorry, leftover from something else 14:09
tadzik also, you can now @results.push(). Looks nice :)
Garda how would you have appended to an array before? 14:10
14:11 woosley left
Garda oh, wait, I did it the before way? 14:11
tadzik push(@results, $foo)... or push $foo, @results? I never know, that's why I like the method call :)
but whatever works for you, TIMTOWTDI
Garda you're right, much better 14:12
14:12 xabbu42 joined
Garda also, if there are any glaring mistakes, or ugly things like that, please point them out 14:12
14:12 xabbu42 left
Garda it's pretty much the first thing I've written in perl 14:12
tadzik there aren't any, as far as I can see 14:13
never written Perl 5 before?
Garda no
tadzik oh, you don't need a ; after a sub declaration
whoa. Where are you learning from?
14:13 Trashlord left
tadzik you can also do %player.keys, but keys %player is good too 14:14
Garda here: secure.wikimedia.org/wikibooks/en/...rogramming
also the pdf that comes with rakudo
tadzik ah, The Book
Garda tadzik, you mean in the last for loop? 14:15
tadzik Garda: aye
Garda yep, works
14:17 s1n left, s1n joined
tadzik I think using @a.sort instead of sort(@a) is something we could put in some perl6style guide 14:18
colomon +1 14:23
14:28 huf left, huf joined
sbp tadzik: back, sorry for the rush before 14:30
wasn't even looking at the error messages, knew I had to go out soon
Garda: great! interesting to see what you learned from too
thanks for posting those
14:31 flatwhatson left
sbp adds the wikibook to Get-Into-Perl6 14:31
tadzik colomon: maybe you could take a peek at the issue sbp has? I'm not able to reproduce this on Linux
sbp (error messages are 611410, 611413, and 611416 on pastebin.ubuntu.com) 14:32
aha, I have panda/ext/Shell/Command.pm 14:34
but my @*INC only contains:
/Users/sbp/stuff/tmp/panda/blib/lib
/Users/sbp/stuff/tmp/panda/lib
so it's looking in lib/, but not ext/
it's setting PERL6LIB itself of course, so it seems like ext/ needs to be added to that? 14:35
I'm not sure if that's part of "ufo" or part of "make test", checking now
it's part of make test 14:36
tadzik hold on
sbp: what does PWD=`pwd`; echo $PWD give you?
sbp in ~/stuff/tmp/panda it gives: /Users/sbp/stuff/tmp/panda 14:37
tadzik hmm
PERL6LIB=ext:lib perl6 bin/panda install . 14:38
sbp: could you try the above?
14:38 Moukeddar joined, Moukeddar left 14:39 wamba joined
sbp it's bootstrapping the modules that it needs 14:40
tadzik oh, funny
sbp for the logs, for my benefit: panda depends on File::Tools, JSON::Tiny, Test::Mock
tadzik I guess the bootstrap.sh is overcomplicated and broken on Mac
sbp all the deps are installed, it got panda, the panda tests are running now 14:42
==> Succesfully installed panda
there were some test failures. well, at least one. gonna paste
pastebin.ubuntu.com/611441/
only one failure 14:43
panda version and panda --version don't work, by the way
tadzik that's not a test failure, it's a trap 14:44
sbp which I think is required for POSIX compliance. not sure though
tadzik it's a git error message, and it's expected by the tests
sbp ah, okay! gotcha
tadzik oh, is it? I may think about it then
sbp: github.com/tadzik/panda/issues/4
sbp yeah, I wanted to run it just to check to see whether this new installation was really installed :-)
tadzik turned out to be not quite as LHF as I thought
sbp hehe 14:45
okay, trying to install LWP::Simple again
by the way, every time I invoke panda now, I get this error message twice:
File::Mkdir is now deprecated, please use Shell::Command mkpath() instead 14:46
as you'll see for yourself, because LWP::Simple install failed again
pastebin.ubuntu.com/611447/
tadzik are you sure your panda is fresh?
sbp same error as 611413, it seems, with the addition of the Mkdir one
nope, definitely not sure of that
tadzik you can update it with 'panda install panda'
14:47 Trashlord joined
sbp thanks 14:47
tadzik sbp: could you try removing ~/.panda and retry?
sbp yep 14:48
panda install panda gave "Project panda not found in the ecosystem" by the way
tadzik hmm, gimme a sec
yeah, the API page seems to not have updated
sbp well, I ran rm -rf ~/.panda; panda install panda 14:49
and that's going okay now 14:50
now I'm getting a single "File::Mkdir is now deprecated, please use Shell::Command mkpath() instead" when I invoke panda... :-) 14:51
tadzik great
update your panda ;)
the hackathoners in Russia have fixed this
sbp that was after I updated it. the update was successful
tadzik oh?
hrm
sbp trying LWP::Simple again, which is also working
tadzik maybe you're still running the Star Panda? 14:52
what does `which panda` give you?
sbp pastebin.ubuntu.com/611449/
don't think so, because the LWP::Simple installation is working now
that's the top of it, you can see the error message there
Succesfully installed LWP::Simple
my which panda is: /Users/sbp/usr/bin/panda 14:53
tadzik hrm
ah, gotcha!
try perl6 ~/.perl6/bin/panda install LWP::Simple
sbp that worked, with no File::Mkdir error message 14:54
tadzik sbp: also, I refactored bin/panda a bit, but there's no speedup, still 2.somethins seconds on help
sbp :-)
tadzik OTOH, update is almost as fast :)
0.07 seconds longer 14:55
14:55 ymasory left
sbp so that's basically all startup time? scary 14:55
now, I have lwp-download.pl and lwp-get.pl in ~/.perl6/bin
but not in ~/usr/opt/rakudo/bin, which is interesting
tadzik is it? 14:56
sbp shouldn't the successful rakduo star panda installation of LWP::Simple put them there?
or does it *always* put them in ~/.perl6/bin?
tadzik yes
or to DESTDIR
I think :)
sbp yes to ~/.perl6/bin?
tadzik yes, panda installs to ~/.perl6/bin by default 14:57
sbp hmm. what about the modules? always in ~/.panda?
tadzik no, everything goes to ~/.perl6/somewhere 14:58
.panda is a place for configs and downloads
sbp oh right, sorry, found it at ~/.perl6/lib/LWP/Simple.pm now
though I also have ~/usr/opt/rakudo/lib/parrot/3.3.0/languages/perl6/lib/LWP/Simple.pm - put there as part of rakudo star I assume? 14:59
tadzik I think so, yes
moritz yes 15:00
tadzik oh hi moritz
sbp there are no lwp-download.pl and lwp-get.pl in ~/usr/opt/rakudo/bin though, I note
hi there moritz
15:06 ymasory joined
moritz \o 15:06
jnthn o/ moritz 15:15
dalek kudo/nom: 4700c15 | jnthn++ | src/binder/s (2 files):
Bring in a couple of 6model header files.
 15:16
kudo/nom: 90aa201 | jnthn++ | src/Perl6/Metamodel/BOOTSTRAP.pm:
Signature also needs a slot for caching the named to positional mapping.
kudo/nom: 3368f83 | jnthn++ | / (2 files):
Remove P6lowlevelsig PMC, since we're just going to use one representation of signatures from now on. Also stop building perl6multisub.pmc - the logic in it will mostly move to ops.
kudo/nom: 4521157 | jnthn++ | src/ops/perl6.ops:
Toss some dynops that we'll no longer need with 6model including the 'cannot be unseen!!1!!' rebless_subcalss.
15:16 f00li5h joined
moritz bind.c:11:38: error: ../pmc/pmc_p6lowlevelsig.h: No such file or directory 15:17
bind.c: In function ‘Rakudo_binding_bind_llsig’:
15:18 Garda left
jnthn moritz: Same here. 15:18
moritz: Fixing. :)
sbp does panda come bundled with niecza? (guessing not) 15:19
moritz does apt come bundled with windows XP? :-) 15:22
jnthn uh-oh...thunder storm.
moritz jnthn: we had that yesterday 15:23
sbp hehe
moritz sbp: the comparison wasn't fair, but you get the gist :-)
sbp I was just about to say, odd to compare niecza to WinXP... :-)
moritz hopes that sorear doesn't kill him :-) 15:26
sbp or Mr. Gates. trouble from both sides
moritz I kinda doubt that Mr. Gates reads #perl6 logs :-) 15:29
moritz needs to fetch his luggage before the luggage storage closes for the night 15:30
see you later, if the hotel has free wifi :-)
ttfn
sbp enjoy
15:33 kaare_ joined
dalek kudo/nom: e53ff5c | jnthn++ | src/ops/perl6.ops:
Toss ops related to building llsig.
 15:37
kudo/nom: 4d064d0 | jnthn++ | build/Makefile.in:
Build binder into ops library rather than PMCs one - we're not going to need it from the PMCs one any more.
kudo/nom: 367e9ff | jnthn++ | src/ (3 files):
Start refactoring the binder to use Signature and Parameter classes. Untested beyond the fact that it compiles.
tadzik moritz: oh, you had storm too? 15:41
sbp: there's still lots of blockers for Panda on Niecza
15:42 Pathin_ left
tadzik jnthn: first, I thought that rebless_subcalss is a typo-joke. Then instead of class-calss I thought of calls-calss. Then, I don't know what's to be unseen :) 15:43
jnthn tadzik: There was an op called rebless_subclass. It did...great evil. :) 15:44
15:44 Pathin joined
tadzik oh, so it's the inside joke (ha ha!) 15:45
jnthn It was actually just a typo :P
One Parrot developer once looked at the op and was like "cannot be unseen" :)
It *was* evil. ;) 15:46
tadzik :)
jnthn is enjoying collapsing three different types of signature representation into just one.
tadzik em, is one supposed to define quote_EXPR? 15:47
like, at all?
jnthn You inherit it from HLL::Grammar. 15:48
tadzik that's what I thought. So there's absolutely no need to define it yourself?
jnthn No 15:49
IIRC you can cutomize it by adding extra quote_mod though. 15:50
tadzik cardinal has it written in PIR, and there becomes some fun when we migrate that to nqprx
it also has quote_term, quote_concat and other fun things 15:51
jnthn oh
I'm guessing that its PIR one won't migrate over well at all
Since it probably works with PGE's world view or something. 15:52
tadzik yeah
jnthn But does the quote_EXPR in nqprx not do the job? 15:53
tadzik no one knows. No one has ever succeeded to move a compiler from PGE to Nqprx, as far as I know 15:54
so I made nqp-rx compile the grammar and actions, but there are things broken beyond my comprehension
jnthn: nom build fails for me 15:57
wklej.org/id/533545/txt/
15:58 Pathin left 15:59 orafu joined
jnthn tadzik: That somehow looks like an incomplete nqp installation. 16:01
tadzik hmm
jnthn nqp_group is something that nqp's make install should put in place.
tadzik maybe I broke it with yesteday's mangling
yeah, I had a not-installed nqp 16:05
16:06 cdarroch joined, cdarroch left, cdarroch joined
jnthn phew :) 16:07
dalek kudo/nom: 5892b51 | jnthn++ | src/ops/perl6.ops:
Oops, was a little too eager with the delete key; we still need deref_unless_object, albeit updated for 6model. Few other cleanups of unused things in ops too.
 16:08
16:12 satyavvd_ joined 16:14 satyavvd left 16:15 icwiener joined
colomon tadzik: what's the issue? no time to backlog, parents and in-laws visiting and I'm grilling. ;) 16:16
16:18 satyavvd_ left
tadzik colomon: oh, nevermind, got solved in mid-time :) Have fun! 16:18
sbp colomon: if you mean the panda issue, it's fixed now. stuff was out of date, that's all
yeah, enjoy grilling. don't grill your in-laws!
I'll probably try to do the same installation procedure from scratch when the next Rakudo Star comes out 16:19
hope there will be one this month? would be nice to incorporate the panda updates
sorear good * #perl6 16:20
sbp hey sorear 16:21
tadzik I don't think panda updates are a motor to update Star
I think pmichaud wanted to release Star this month for the performance bumps though
sbp ah, good. maybe they should be though, always handy to have a working module manager!
just as important to have those top level bits as the bottom level bits, when you're doing stuff
I'm sure you agree. maybe pmichaud needs to hear a few more +1s in that direction? :-) 16:22
if not, perhaps a few more links to panda, and information about updating panda, could be added to rakudo/README? 16:24
sorear sbp: Once upon a time, Parrot bundled a properly configured and integrated version of icu. It was removed for political reasons; talk to jnthn. 16:25
jnthn has probably moaned quite enough about that issue already...
I just don't build with ICU anymore.
Can't be arsed to wrok out how to get it to work on win32 *again*. 16:26
sorear sbp: 'panda' doesn't just print lines to stdout; it also compiles panda, and loads perl6.pbc, both of which would easily explain 2s.
sbp aha, thanks. tadzik did look into optmising it a bit as you'll see probably further down the logs before getting back here, but it didn't speed it up much. no big deal at the moment, it just surprised me 16:27
(of course I was measuring panda performance relative to other languages' package managers)
on ICU, yeah. well I see unicode support and think "mandatory!"
but what a silly installation procedure that was. and then, when it was done, no clear process at all to include the installed ICU libraries because I was using a non-standard path 16:28
sorear sbp: some irc clients don't have the ability to split the window and keep one half scrolled up while the other tracks current discussion, so pastebins help
sbp again not really a big deal, but if somebody slavishly follows instructions like me, they'll be trying to apply broken patches to ICU before they even touch rakudo
you'll notice I used a pastebin after the little incident there... 16:29
16:29 colomon left
sbp maybe it'll take me another five years to learn the value of emoparting... :-) 16:29
I would like to know what funtionality ICU brings to parrot/rakudo 16:30
sbp looks at the docs again for that
sorear the main thing ICU brings is a complete set of character property tables 16:31
jnthn std: sub foo($y = $x, $x) { } 16:32
p6eval std 4b1b100: OUTPUT«===SORRY!===␤Variable $x is not predeclared at /tmp/ykjSbVs57o line 1:␤------> sub foo($y = ⏏$x, $x) { }␤Cannot put required parameter after optional parameters at /tmp/ykjSbVs57o line 1:␤------> sub foo($y = $x, $x⏏) {
..}␤Pote…
sorear for instance "\x3000" ~~ /\s/ is true with ICU available
(that's IDEOGRAPHIC SPACE)
jnthn std: sub foo($y = $x) { my $x }
p6eval std 4b1b100: OUTPUT«===SORRY!===␤Variable $x is not predeclared at /tmp/okachTT5sG line 1:␤------> sub foo($y = ⏏$x) { my $x }␤Potential difficulties:␤ $x is declared but not used at /tmp/okachTT5sG line 1:␤------> sub foo($y = $x) { my $x⏏
..}␤ …
jnthn Good.
sorear I think without ICU parrot treats all characters >255 as being \w
hi jnthn. 16:33
sbp hmm. so couldn't those tables be included quite simply by other means?
this sounds like a parrot issue rather than a perl6 issue though?
I mean, python for example just compiles UnicodeData.txt down into some data structure that it can use throughout the standard library. it's not having to include a whole different library for that 16:34
sorear yeah well... this is Parrot we're talking about.
16:34 lumi_ left
TimToady I've been harping on that for years, actually... 16:34
sorear (I don't have a high opinion of #parrot. Can you tell?) 16:35
sbp not possible to work around it at the rakudo level, perhaps? (or not worth it? or not politically friendly?)
16:35 lumi_ joined
jnthn Possible. 16:36
But too much to do.
Could happen at some point.
sbp if not, I might file a rakudo bug (or am at least suggesting on logs here if I don't get around to it) that the ICU instructions in README could be moved about, made more clear that it's a very optional thing, what kind of effects there are, etc. 16:37
TimToady could base it on P5's tables too, which I think are pretty up-to-date
sbp (even if it's marked pri=low, worth having on file)
TimToady but that's even less palatable in the world of talking birds
sbp . o O ( verbirdgenous ) 16:38
how does panda compare to masak's proto? 16:41
not immediately obvious from his README that there's any difference at all
panda seems more up to date, at least 16:43
jnthn sbp: panda is actively developed. proto is not, faik. 16:44
sbp thanks. masak's META.info was last updated 28 Jan 2011 16:45
but the rest seems to be from about June 2010 16:46
16:48 uniejo left 16:55 colomon joined, silug joined 17:00 baest_ is now known as baest 17:06 silug left, silug joined
dalek p: f6bda2d | jonathan++ | src/ops/nqp.ops:
Fix copy-pasto.
 17:09
tadzik sbp: re panda performance, like sorear says, it's compiling bin/panda and loading perl6.pbc, but also loading all the Panda modules, which are recompiled, but are still a big piece of code 17:10
sorear if you tried to use niecza as a drop-in replacement for Rakudo in panda, (lets pretend it magically gets all the features you need), you'd probably find it was slower 17:12
niecza's compiler, due mostly to .net weaknesses, loads significantly slower than Rakudo's 17:13
like 5-10 seconds for hello world
sbp ah, note that I was only asking about panda in niecza so that I could document it in Get-Into-Perl6 not for performance reasons :-)
PerlJam jnthn: I saw this quote on FB and thought of you: "I am a firm believer in the people. If given the truth, they can be depended upon to meet any national crisis. The great point is to bring them the real facts, and beer." --Abraham Lincoln
jnthn beer :D 17:15
A man who really understood the needs of the people. :)
17:19 silug left
dalek ecza: de0797d | sorear++ | / (4 files):
Implement .Int and .Num coercions
 17:19
17:29 wamba left
dalek kudo/nom: 710de96 | jnthn++ | src/ (2 files):
Re-work call into the signature binder. Should be a good bit cheaper than how we locate signature and capture today.
 17:36
kudo/nom: 98c43fd | jnthn++ | src/Perl6/Actions.pm:
Create lexical slots for parameters.
kudo/nom: c50c120 | jnthn++ | src/binder/bind. (2 files):
A few little fixes to the binder updates.
kudo/nom: 56a776f | jnthn++ | src/Perl6/ (4 files):
Get signature building a little further along, plus assorted fixes to trait application time.
sorear hmm, there are quite a few numeric functions I'm misisng! 17:54
17:54 jferrero joined 18:15 s1n left 18:18 Trashlord left 18:19 birdwindupbird left
dalek ast: efd5f21 | tadzik++ | S0 (4 files):
Remove pieces of invalid Pod

See github.com/rakudo/rakudo/commit/30ebf78		 18:22
18:29 s1n joined
dalek ecza: 03012ee | sorear++ | lib/CORE.setting:
Fix a few parameter names
 18:29
ast: 8657a3a | sorear++ | S (17 files):
Unfudge a few niecza tests
 18:30
18:40 y3llow_ joined 18:42 y3llow left, y3llow_ is now known as y3llow
moritz back 18:44
sorear hello moritz 18:45
tadzik yay, my first backtracking problem encountered and solvedt 18:46
sbp what happened to all the old rakudo.org blog links?
e.g. rakudo.org/2008/09/final-report-for...found.html
but anything with a date in it seems to have disappeared
sorear rakudo.org stopped working and alester tried valiantly but failed to fix it 18:47
sbp ah, shame
moritz sbp: I thikn the blog wasn't ported, most of that was just release announcements
sorear the content was migrated to a new server with a different CMS
the blog is probably still there under a different URL scheme
sbp hmm, if so rakudo.org/?s=Final+Report+for+Mozilla doesn't find it 18:48
sorear 3.97-4.06 sec "Hello, world" before my next attempt
sbp “> 1- ready binaries for different systems at the same date of releasing source, especially for windows 18:53
This will be rather easy if somebody steps up and builds the binaries. I fear that most Rakudo developers lack the expertise to do themselves.”
moritz in www.perlmonks.org/index.pl?node_id=845586
perhaps I should throw up my installation as an OS X binary
but I'm not sure I'd maintain it with much regularity 18:54
flussence (in an ideal world, the OS-dependent part of rakudo would be tiny...) 18:56
sbp incidentally, moritz seems to be in pretty much all the perl6 results I've just been randomly Googling through. see stackoverflow.com/questions/3383339 for example — ysth's comment made me laugh 18:57
18:58 kaare_ left 19:02 dukeleto left 19:03 dukeleto joined, s1n left
tadzik yay, I can parse paragraph blocks! 19:05
Well, Rakudo can
arnsholt tadzik: That's even better! ;) 19:06
arnsholt is trying to wrap a slightly weird C library with SWIG 19:07
An interesting experience
tadzik I think I should complete week #1 first, but this is just too exciting 19:08
Besides, Week #1 haven't started yet anyway :)
is <?before> there? 19:11
19:12 ymasory left
tadzik or, nevermind 19:13
19:18 lumi_ left
moritz it is, though <?after> is NYI 19:18
19:18 s1n joined 19:19 lumi_ joined
tadzik yay, I managed to parse paragraph blocks with no backtracking 19:23
19:28 ymasory joined 19:36 silug joined 19:46 wamba joined 19:47 ymasory left 19:48 jferrero left
tadzik I can write <!foobar> and token foobar { [ 'foo' | 'bar' ] }; Can I write something like ![ 'foo' | 'bar' ]? 19:49
oh, just <![ 'foo' | 'bar']>, right?
right. Or at least "the tests pass" :) 19:50
sorear no
19:50 jferrero joined
sorear <![ 'foo' | 'bar' ]> is a negative lookahead character class 19:50
you want <!beforefoo || bar> 19:51
tadzik it gets autoquoted this way? 19:52
sorear parse erro 19:53
semantic dispatch failed, rather
tadzik hmm 19:54
this way, I don't need to put 'foo' or 'bar' in quotes?
sorear correct 19:55
tadzik it looks a bit funny 19:56
sorear any \w in a regex matches itself even if unquoted
19:56 jferrero left, mj41 joined
tadzik Method 'beforebegin' not found for invocant of class 'Pod6;Grammar' 19:56
sorear you need a space after before; I typoed
tadzik nice. Thanks sorear 19:57
19:58 snearch left 20:11 meteorjay left 20:12 IllvilJa joined 20:16 Mowah left 20:18 meteorjay joined 20:24 dukeleto left, wamba left, dukeleto joined
tadzik . o O ( will the feed operator at the beginning of a line get parsed as Pod? ) 20:32
dalek p: 3339a14 | jonathan++ | src/NQP/ (2 files):
Add a way to specify extra vm libs (e.g. pmc or op libs for Parrot) to load and use.
 20:33
20:41 ymasory joined 20:48 dual joined
Tene tadzik: I've moved a few other compilers over from PGE to NQP without much trouble; iirc the PIR stuff in the compiler was just copied literally from rakudo at the time, fwiw. 20:54
20:55 kjeldahl left
dalek ecza: f1966c1 | sorear++ | / (3 files):
Reorganize nam to load much less (~75%) data; small startup time improvement.
 20:56
tadzik Tene: oh, I see
Tene tadzik: so, you should be able to use the NQP-provided quote stuff without any trouble, I expect 20:57
tadzik Tene: have you looked at the latest Cardinal Pull Request?
Tene tadzik: I haven't looked at cardinal whatsoever in quite a while.
tadzik: do you have permissions to manage the cardinal repo yourself? If you think it's good, just update it. You've done far more work than I have on it lately; feel free to take as much ownership or maintainership as you like. 20:58
sorear grumbles at the CLR not letting me form the type object* 20:59
21:01 mtk left
tadzik Tene: the thing is I'm not sure whether it's really good, I feel like a blind man in the fog when it comes to pge-nqprx transition 21:05
Tene ah, yeah, understandable.
tadzik loliblogged! ttjjss.wordpress.com/2011/05/22/the...ng-period/ 21:08
21:09 mtk joined
sorear has had a vision of a possible future of niecza 21:14
21:15 wamba joined
sorear actually I think I want the simple past "had" here 21:15
21:16 sftp left
jnthn tadzik: Nice! :) Though shouldn't the tl;dr go at the start? ;-) 21:17
sorear it involved discarding the "Perl 6 globals are CLR globals" optimization, consolidating all of the variables in each unit into a single RuntimeUnit object, and a fair amount of bounded serializaiton stuff
21:17 Trashlord joined 21:18 sftp joined 21:19 mj41 left
dalek kudo/nom: e1e7c83 | jnthn++ | src/binder/bind.c:
Fix a silly.
 21:25
kudo/nom: b72506d | jnthn++ | src/ (2 files):
First cut at ClassHOW.type_check.
kudo/nom: a05d65d | jnthn++ | / (5 files):
Top type accepts everything. Also prep for putting junction auto-threading back in later.
kudo/nom: 8b71496 | jnthn++ | / (5 files):
.WHAT, .HOW. .WHO.
kudo/nom: 97577c6 | jnthn++ | NOMMAP.markdown:
Update NOMMAP.
jnthn remembers that he's teaching tomorrow, and reluctantly puts down Rakudo nom so he can sleep 21:26
well, maybe just one more patch... 21:27
sorear sleep well... *wonders when to start studying rakudo nom*
jnthn sorear: It's quite a building site at the moment. 21:28
tadzik jnthn: I didn't want to ruin the suprise :)
benabik Fun to watch though.
jnthn sorear: Mostly I'm ripping a load of primitives up and re-doing them. :)
And playing with circularity saws.
sorear there will be some nice catching up to do after the merge. :) 21:29
in more than one sense of the word
21:30 pamera1 left 21:34 ymasory left
dalek kudo/nom: dd55769 | jnthn++ | / (2 files):
Native types should be able to inherit (of course, 'instances' cannot, but the type object can, so int ~~ Int, etc.)
 21:35
sorear wait, rakudo nom supports 'int'? 21:36
jnthn sorear: Initially just for natively typed attributes, but it's the foundation for the rest of it. 21:37
sorear: There's a bunch of other stuff needed before Rakudo can interestigly support 'int'. 21:38
But, gotta start towards it somewhere... :)
OK, really sleep & 21:40
21:50 ymasory joined 21:51 cdarroch left 21:53 kst`` left 22:06 ymasory left
dalek ecza: b8d8df9 | sorear++ | lib/CLRBackend.cs:
Localized cleanups of CLRBackend, use varargs
 22:31
22:31 daniel-s joined 22:45 rgrau joined 22:48 flatwhatson joined 22:50 icwiener left 23:08 hercynium joined 23:26 daxim_ joined, noganex_ left 23:28 noganex joined 23:35 flatwhatson left 23:39 flatwhatson joined 23:44 rgrau left 23:45 lumi_ left 23:47 lumi_ joined 23:53 flatwhatson left 23:56 daniel-s left 23:59 flatwhatson joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!