6.2.11 released! | pugs.blogs.com | pugscode.org | pugs.kwiki.org | paste: sial.org/pbot/perl6
Set by wolverian on 12 February 2006.
putter has there been any discussion on p6l about general structure matching (eg, $a = [1,2,3]; $a ~~ foo/$head:$tail/) beyond just <,>. say that example was some nested hashes. 00:08
svnbot6 r9027 | Darren_Duncan++ | r2488@darren-duncans-power-mac-g4: darrenduncan | 2006-02-15 16:15:19 -0800 00:18
r9027 | Darren_Duncan++ | merged /ext/Rosetta-Engine-Native into /ext/Rosetta, renamed ::Native to ::Example
Khisanth hmm
Juerd "you need to install two or an even number of memory modules to use this feature" 01:01
rafl Juerd: What feature? 01:54
Kattana rafl: sounds like dual channel 05:00
putter stevan: this thread www.perlmonks.org/?node=227847 suggests runtime @ISA modification currently works just fine (and has for years?!?). with local() even :) So no need for C:MOP's aspersions on such eminently reasonable activities. ;) 06:33
ingy seen audreyt 07:39
GeJ ingy: ~10hrs ago 07:53
[22:27] <audreyt> g'nite *wave* 07:54
is there a RT/bugzill/whatever for p5? I bumped into something, and I'd like to check if it's been already reported and/or fixed 07:55
integral rt.perl.org, GeJ. perlbug(1) is an automated bug submission tool. 08:13
GeJ integral: thank you very much 08:16
now I have to find somebody with a 5.8.8 setup
dduncan I can't help you yet ... still 5.8.7 08:18
GeJ no problem, thanks anyway :) 08:20
azuroth spam of the week: "UPSET BCOZ OF UR SHORT DlCK? LONGER 2" WITH THIS antacid"
dduncan yeah, like those two kinds of things have any relation 08:35
svnbot6 r9028 | audreyt++ | * DrIFT.YAML: Show full Haskell type for unimplemented asYAML instances. 12:16
r9028 | audreyt++ | * VHash is now emitted correctly as a Haskell Map.
xinming audreyt++ 12:20
xinming audreyt: hmm, How much slower pugs is compared with perl 5? 12:33
still 100 times?
audreyt xinming: not sure... havn't benchmarked lately 12:34
xinming audreyt: How to benchmark it? using times?
audreyt xinming: yeah, or maybe mandel.p6, etc 12:35
ingy:
$ perl -MYAML -MYAML::Syck -e 'YAML::Load(YAML::Syck::Dump("\n\t"))'
YAML Error: Inconsistent indentation level
ingy: seems to be a YAML.pm bug
audreyt gaal_: 12:40
p <- Str.readFile "p.yml"
Right (Just yml) <- parseYamlFS p
Right str <- emitYamlFS yml
hPut stdout str
xinming real 4m30.325s
user 3m31.869s
sys 0m0.720s
for pugs. :-) 12:41
real 0m0.514s
user 0m0.044s
sys 0m0.004s
That's for perl 5.
bbl, test it on my laptop 12:42
svnbot6 r9029 | audreyt++ | * Data.Yaml.Syck: taguri roundtripping now works correctly. 13:01
svnbot6 r9030 | audreyt++ | * more YAML-emitting tweaks: 13:13
r9030 | audreyt++ | - the problemous literal-indenting style is inhibited.
r9030 | audreyt++ | - maps are now always dumped using sorted keys.
r9030 | audreyt++ | - loading of undefs is restored.
svnbot6 r9031 | gaal++ | * Data.Yaml.Syck - use unsafePackAddress to hoist string literals to top-level 13:35
r9031 | gaal++ | ("what do you mean, _Haskell has globals_!")
svnbot6 r9032 | audreyt++ | * Loading/Dumping the Prelude.pm precompilation yaml stream 13:38
r9032 | audreyt++ | is now roundtripping safely in the YamlNode level.
svnbot6 r9033 | gaal++ | Data.Yaml.Syck: NOINLINE the literals from r9031 to make them really global 13:47
lypanov win 4 13:53
oops
svnbot6 r9034 | audreyt++ | * faster emission using concat instead of append.
szbalint heh, good ol' irssi :) 13:57
Juerd rafl: dual channel memory 14:08
rafl: The weird bit was "two or an even number". Two *is* an even number.
Limbic_Region Juerd - I would argue that 2 is an odd number 14:12
it is the only even prime
that's pretty odd
Juerd Some people say it's not a prime. 14:13
Limbic_Region really? So do those same people not agree on the definition of what is prime (a number that is divisible evenly only by 1 and itself)? 14:14
audreyt gaal: 14:23
make moose.hs ; make src/Data/Yaml/Syck.hs ; ghc -O2 -optc-O3 --make -isrc/cbits -Isrc/cbits -isrc moose.hs dist/build/src/syck/*.o src/cbits/*o ; ./a.out
stevan putter: Class::MOP places no restrictions on changing ISA, only Class::C3 does that 14:26
heya audreyt :)
rafl Juerd: That doesn't make the statement wrong. 14:31
audreyt hey stevan 14:32
stevan how is .il treating you?
audreyt very happily 14:33
the food here is delicious
stevan mmmm falafels, gaal has raved about them
audreyt but most of the time are spent 1)adjusting jet lagged brain and 2)hackingwith gaal
stevan cool 14:34
audreyt: I split out the C3 algortihm BTW (search.cpan.org/~stevan/Algorithm-C3)
I got tired of copy pasting it all the time :)
audreyt ok :) 14:35
stevan audreyt: have you seen the .NET CodeDom stuff? 14:36
audreyt stevan: ASTs? yeah 15:00
havn't seen them in action
they are the counterpart of our Exp structure 15:01
stevan www.ondotnet.com/pub/a/dotnet/2003/...tml?page=1
a little tedious, but interesting none the less
audreyt they are in any case better than the Syn nodes :) 15:12
svnbot6 r9035 | gaal++ | * DrIFT.YAML: remove a spurious read of something 15:24
r9035 | gaal++ | that's already a String
r9035 | gaal++ | * Data.Yaml.Syck -
r9035 | gaal++ | - literals that are used as string parts should not have their
r9035 | gaal++ | trailing NULs counted
r9035 | gaal++ | - spotted a missing "~" -> _tildeLiteralFS. updated.
kolibrie rejoices in his non-seg-faulting pugs (thank you putter and audreyt for readline, ghc 6.4.1 tips)
audreyt gaal: www.haskell.org/ghc/docs/6.4.1/html...hesPJ.html 15:27
spinclad Limbic_Region: but how can 2 be prime when 2 itself doesn't divide it (since it's odd)? 15:46
(and doesn't that definition make 1 a prime?)
svnbot6 r9036 | fglock++ | pX/Common/iterator_engine.pl - added comments 15:54
fglock is it somehow unsafe to use 'illegal' subroutine names in perl5? like: 16:07
perl -e ' *{"<word>"} = sub { print "$_[0]\n" }; &{"<word>"}("hello"); '
stevan fglock: not at all, the only issue is that you cant use strict, and you dont get compile time typo-checking 16:09
fglock thanks 16:10
Juerd rafl: Shall I book a place to sleep in .be for the three of us? I'm assuming you arrive a day before the event, and that we spend the extra day we have in Belgium, not Germany. 16:15
Limbic_Region: Indeed 16:16
rafl Juerd: Timo can't come to FOSDEM and GPW. 16:19
Juerd: I'd like to have a place if it's not to expensive. 16:20
sili places are definitely expensive 16:21
Juerd rafl: Oh, that's too bad 16:22
rafl: What is "not to expensive"?
rafl Juerd: not *too* expensive. converging towards zero. 16:23
Juerd: What's the typical price for a night there?
LeTo rafl: I'm at GPW and could share room with you 16:24
rafl LeTo: That would be great.
LeTo It's currently as single room ordered but there are usually to beds, that would be EUR 95/2 then 16:25
rafl per night? 16:26
LeTo yep
www.wald-und-golfhotel.de/
rafl WTF.. that's much for a student.
LeTo yep
1.3 - 3.3 (2 nights) 16:27
rafl LeTo: Would you mind me making the decicion later? 16:30
svnbot6 r9037 | fglock++ | iterator_engine.pl - rules in p5 namespace: &{'rule::<word>'}(...)
LeTo rafl: np
Juerd rafl: I thought you had arranged something in bochum? 16:35
rafl Juerd: I thought that as well, but it didn't work out. 16:36
Juerd I see 16:37
rafl: re .be, found via the FOSDEM site a link to a hostel called Bruegel
Their cheapest option is 17,60 per night, in a multiperson bedroom
Including breakfast and bedsheets
Oh, it says something about membership 16:38
Let's see what that is
Juerd Oh, the 17,60 is for members 16:42
Non-members pay 20,60
In a shared 4-person room
Juerd rafl: If you're okay with this, I'll book them. 16:46
rafl: At 4 nights, it'd be 82,40 eur each.
rafl Juerd: OK 16:52
Juerd rafl: Consider it done 16:53
putter does p5 have lexical imports? 16:57
broquaint Nope. 16:59
putter broquaint: my thanks. 17:00
putter goes off to see if a module could do both its usual thing, and be a source filter...
Juerd rafl: We have to be at the hostel before 16:00, feb 24. 17:01
rafl: What will your mode of travel be?
rafl: And could you book something in bochum.de? Preferrably, <= 30 euro per night 17:03
ingy hi audreyt 17:11
putter hi ingy 17:13
you saw audreyt's note a while back re a possible YAML.pm bug?
colabti.de/irclogger/irclogger_log/...sel=36#l76 17:14
ingy putter: yeah but it doesn't barf on YAML-0.58 YAML-Syck-0.22 17:16
maybe a newer YAML::Syck messed it
putter k 17:17
ingy I'm not sure where audreyt keeps the YAML-Syck code these days
hmmm. I just synced 0.33 17:18
will try
putter where did it turn out to be? 17:19
ingy svn.kwiki.org/audreyt/YAML-Syck (if that's the latest)
putter didn't realize at one point that audreyt was the author of Syck... remembers complaining about sloppily unfixed warnings... sigh... 17:20
tnx
ingy yeah that was it 17:22
putter tnx
ingy YAML::Syck picks a really goofy serialization
but it is correct and a YAML.pm bug indeed 17:23
sili so can you use yaml to transform yaml? 17:25
putter So, perl5 *can* have lexical imports. :) Because Filter::Simple apparently gets along with Exporter, a module can just slap code with lexical at the top of the source. You might even be able to get __LINE__ correct (using #line?). 17:41
svnbot6 r9038 | fglock++ | iterator_engine.pl - better rule composition syntax, like: 17:46
r9038 | fglock++ | *{'rule::<word>'} = rule::greedy( \&{'rule::<word_char>'} );
gaal ingy: svn.kwiki.org/audreyt/YAML-Syck 17:57
ingy gaal: ^^ :) 17:58
gaal oh right
gaal has goofy vision 17:59
Limbic_Region has audreyt abandoned journaling or did it change locations again? 18:01
svnbot6 r9039 | audreyt++ | * Strictify the whole yaml dumping process. 18:07
putter LR: I think it's just been a lonnnngg pause :/
Limbic_Region ok - I hadn't been able to hang out in the channel much this past couple of weeks cause of having a newborn home and all and wondered if perhaps I just missed something 18:08
putter audreyt: feel free _not_ to point to the (ugly) p5 spike code... it's going to be moving anyway... ;)
Limbic_Region playing catch-up on all the projects you are interested in (but not actively participating in) after a 2 week hiatus is a PITA when you have to squeeze time in at $work to do it 18:09
putter LR: don't think so (re missed something). audreyt is currently hacking in .il with gaal. $work was absorbing before the travel. 18:10
indeed 18:11
pugs well on the way using yaml as the un/dumping form of compiled p6. which should fix the precompiled prelude slowness issues which have kept it small. also the yaml dump of the parse tree is coming along. 18:13
in other amusements, an experimental p5 regexp engine, doing backtracking recursive descent parsing, built from Regexp::Parser, was passing much of the p5 re_tests. and supported <subrules>. so a throw-away toy p6 parser was built on it. 18:16
what else has been going on... that's the last 1/2 week or so. but 2 weeks... I'd have to changelog for that ;) 18:17
Limbic_Region well, I will just look for the next release of Pugs change release but thanks for the quick synopsis 18:21
putter np
how goes the newborn et al
Limbic_Region oh, exciting and exhausting 18:25
Limbic_Region needs to wander off for $work
putter & 18:27
:)
rafl Juerd: Aye 18:34
Juerd: By train to bruessel
Juerd: Yes, I'll do so.
svnbot6 r9040 | fglock++ | iterator_engine.pl - uses iterators and functional backtracking; 19:53
r9040 | fglock++ | optimization (backtracking caching) was temporarily removed
putter ingy: here's an Acme module for someone to write... Acme::PolyglotInline... perl -MAcme::PerlPolyglotInline foo.c bar.py ; and perhaps perl -MAcme::PerlPolyglotInline -e 'require "foo.c";...' ;) 20:02
fglock putter: I found it is possible to use iterators and still cache things for speed. The engine is now mostly functional 20:06
dduncan following my report a day or 2 back about Pugs link errors ... 20:07
I did a make realclean and rebuilt today, and no problems occur now
but even before, I did say it probably wasn't a big deal 20:08
svnbot6 r9041 | fglock++ | iterator_engine.pl - added tests for a bug 20:32
putter fglock: so what's the next step? running re_tests? :) 20:36
kolibrie putter: thanks for the readline/ghc 6.4.1 tips yesterday - my pugs no longer seg faults 22:02

Powered by Cro and the Raku® Programming Language.

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

Thank you!