svn switch --relocate svn.openfoundry.org/pugs svn.pugscode.org/pugs/ | run.pugscode.org | spec.pugscode.org | paste: sial.org/pbot/perl6 | pugs.blogs.com
Set by avar on 16 November 2006.
00:04 Psyche^ joined, Psyche^ is now known as Patterner
fridim What can I run like perl6 program with the pugs etch debian package ? 00:05
use v6-pugs v6-alpha seem to not work with this version of pugs
00:10 lisppaste3 joined 00:56 chessguy joined, lyokato joined 01:06 BooK_ joined 01:11 Psyche^ joined, Psyche^ is now known as Patterner 01:14 elmex joined 01:42 chessguy joined 01:51 chessguy joined 02:04 [particle] joined 02:44 neonse left 02:53 Supaplex joined 02:57 weinig is now known as weinig|zZz 03:00 Supaplex left 03:34 justatheory joined 04:02 coumbes joined 05:32 Psyche^ joined 05:46 Psyche^ is now known as Patterner 05:49 kanru joined 06:00 justatheory joined 06:14 BooK joined 06:22 froggero1 joined
froggero1 anyone know anything about hmac md5? 06:23
gaal @tell fridim try the latest pugs from the source at svn.pugscode.org/pugs/ 06:25
lambdabot Consider it noted.
froggero1 bascially, I took the javascript hmac md5 script from pajhome.org.uk/crypt/md5 (md5.js) and the values that it's returning compared to backpan.cpan.org/authors/Clinton_Wong/ 's hmac::md5 module don't match up
can't figure out why... I have a random key that gets generated, and used in the hex, and that's getting transferred the same... putting it in the same order, supposed to be using the same protocol... maybe one of them is out or something? 06:26
gaal froggero1: I don't have time to look at this myself right now, but one lead is to make sure there isn't some bogus unicode conversion going on 06:27
froggero1 good idea... I know the java is in ascii, will check the mod
not much in the readme... backpan.cpan.org/authors/Clinton_Wo....02.readme 06:28
love it
in the perdoc: 06:29
SYNOPSIS
use HMAC::MD5;
hex($string, $key) returns a 16 byte digest in binary form.
hexhash($string, $key) returns a 16 byte digest as printable hex.
gaal Try out Digest-MD5
froggero1 the javascript uses hmac though 06:30
the front end for this is a website... which is in perl, but i'm using javascript to turn the form input values into a md5 hash based on the password and a random key.. unless there's a better way to do this (i don't know enough about javascripting at the moment) 06:31
i tried digest-md5 but the values were out as well.. that's what led me to hmac in the first place 06:32
afk smoke
06:37 RHainsworth joined
froggero1 aight, I gotta go... but if anyone thinks of another way to do this... send me an email -> adam [at] talsma [dot] ca 06:40
cheers guys, cya :)
06:48 takesako__ joined 07:00 Lee_ joined 07:03 takesako__ joined 07:50 marmic joined 08:06 iblechbot joined 08:07 drrho joined 08:24 mjk joined, |Jack| joined 08:45 Eidolos joined 09:16 devogon joined 09:18 ozo joined 09:23 elmex joined 09:52 ayrnieu joined 09:55 andara joined, Lorn joined 10:14 RHainsworth joined 10:20 orafu joined 10:32 fglock joined 10:47 buetow joined 11:06 spinclad joined 11:12 elmex joined 11:13 spinclad_ joined, araujo joined 12:14 chris2 joined
rgs chemical programming !! www.inria.fr/rrrt/rr-5743.html 12:28
lambdabot Title: RR-5743 : Generalized Multisets for Chemical Programming
rgs not enough paradigms yet
12:29 avaron joined
rgs some slides: upp.lami.univ-evry.fr/Documents/Rad...UPP'04.pdf 12:35
12:47 weinig joined 12:49 xinming_ joined 13:14 Limbic_Region joined 13:23 ofer1 joined 13:28 iblechbot joined 13:36 xinming_ joined 13:39 kane-xs_ joined
integral cool. 13:48
13:50 gnuvince joined 13:54 weinig is now known as weinig_, weinig_ is now known as weinig
Limbic_Region smokin 13:54
13:55 BooK_ joined, _meppl joined 13:57 _meppl is now known as meppl 14:16 RHainsworth joined 14:20 RHainsworth joined, iblechbot joined 14:21 chessguy joined 14:52 buetow joined 15:20 fglock joined, DHGE joined
svnbot6 r14891 | andara++ | [runpugs] 15:28
r14891 | andara++ | First sketch for preloading Pugs sessions.
15:41 xinming__ joined 15:43 xinming_1 joined 15:46 xinming_2 joined 15:50 devogon joined
fglock v6-alpha-0.016 uploaded to CPAN 16:03
svnbot6 r14892 | fglock++ | v6-alpha-0.016 in CPAN 16:05
16:07 spinclad joined 16:13 hexmode joined
kolibrie fglock: any changes in v6-alpha, or were you just bringing CPAN up-to-date? 16:24
16:26 wamiks joined
fglock kolibrie: this is just an upload from the repo; the CPAN version was incompatible with one of the prerequisites and no longer worked 16:30
16:30 ruz joined
fglock the errors you pointed out are known errors :( 16:30
kolibrie fglock: that's what I thought - but just wanted to make sure I wasn't missing anything
16:30 shobadobs joined
kolibrie fglock: it still seems that I can run my grammars with PCR, even though I can't run extract_pghpw.t with prove6 16:31
so I don't know if I did something weird in extract_pghpw.t, or what 16:32
fglock: MiniPerl::Grammar is being very helpful as a grammar template 16:33
and so far I haven't found any speed issues 16:34
16:36 weinig is now known as weinig|food
fglock kolibrie: MP6 is much more stable than v6, because it depends on no external modules. It's also fully tested, as it depends on itself for compiling 16:42
kolibrie fglock: it just doesn't have inheritance 16:43
nwc10 oh, yes, for 6-on-5, do we need a way to inject new 'core' subs into every namespace? 16:47
16:49 ludan joined, elmex joined
gaal too bad AUTOLOAD won't work for that 16:50
but wouldn't UNIVERSAL?
nwc10 need to con the tokeniser 16:51
to do it properly
so that they're not barewords
but my question was, do we need it?
fglock nwc10: the current implementation wraps the core subs, so that it can control the parameters/dispatch 16:52
gaal maybe a switch to mean "there are no barewords here, everything's a function 16:53
"?
nwc10 I don't know. Asking questions of me isn't necessarily a help here.
I was more asking questions of Perl 6
gaal gotta go for a bit
16:59 hexmode joined 17:05 marmic joined 17:07 chessguy joined
fridim Hi there. 17:18
lambdabot fridim: You have 1 new message. '/msg lambdabot @messages' to read it.
fridim nice bot
TimToady nwc10: if I was going to add something like that, I'd generalize it slightly to a lexically scoped macro mechanism on about the level of C macros. That would give you much of the power of source filtering without most of the problems. 17:28
nwc10 TimToady: ah OK. I was thinking more in terms of "what can't 6-on-5 do well" 17:29
TimToady A: mostly, things that require source filtering... :) 17:30
17:32 kanru joined
Limbic_Region TimToady - perl 5 question. Do you know why goto $code_ref; behaves as if you said goto &$code_ref; but performs way better? 17:34
17:36 buetow joined
TimToady no, I don't. I didn't add that functionality. 17:37
I wonder if it bypasses the code that strips the old @_ off and shoves the new one onto the stack...
Limbic_Region TimToady - that's the part I was wondering about too 17:38
TimToady might be someone was trying to make tail recursion faster.
just guessing... 17:39
Limbic_Region In case you are at all interested - perlmonks.org/?node_id=589762
lambdabot Title: named sub, subref and magic goto speed
TimToady yes, I was already following that. 17:41
17:41 justatheory joined 17:42 justatheory_ joined
TimToady afk & 17:43
nwc10 Limbic_Region: does the optree differ for the two goto examples? 17:44
Limbic_Region nwc10 - I don't have perl available ATM to find out 17:45
nwc10 - No matter what it is doing, it seems to me that it is a doc bug 17:47
17:47 andara left
Limbic_Region AFK & 17:48
17:57 weinig|food is now known as weinig 18:09 brent[] joined 18:11 fglock joined 18:18 scrottie joined 18:27 fglock joined 18:34 AzureBOT joined 18:35 jdv79 joined
jdv79 about the $^ sigiled vars - where does the rest of the var name come from? 18:36
TimToady Limbic_Region: actually, nwc10 is on the right track, I think. the &$foo has to throw in an extra opcode to possibly translate a symref to a CV. bare $foo bypasses that. 18:37
jdv79: "come from"?
18:38 rindolf joined
TimToady in my experience, the rest of the var name comes from typing more characters after the ^. :P 18:38
rindolf Hi all. 18:39
TimToady hi
Limbic_Region TimToady - I think it is underdocumented regardless. 18:40
but then again, I think the behavior of glob iterator is underdocumented too 18:41
jdv79 i'm ust reading j worthington's p6 oo presentation and he shows "$^n" but I don't get where the n part came from.
TimToady I still don't know what you mean by "came from"
wolverian I think he means it's not declared.
or rather, that is confusing him.
TimToady $^n is self-declaring, jdv79 18:42
many twigiled variables are self-declaring.
$*n is a self-declared global, for instance.
but the twigil tells it the scope. 18:43
jdv79 oh ok. is there a central doc on all the twigil variations?
TimToady so with ^ it happens to be a parameter to the current block.
sure
Limbic_Region jdv79 - they are the new $a and $b
hrm - didn't mean for that to happen 18:44
jdv79 cool i guess;)
Limbic_Region evalbot: sub hello { say $^greeting; } hello('world'); 18:45
TimToady sorry, my firefox died...
S02 talks about twigils
Limbic_Region ?eval: sub hello { say $^greeting; } hello('world'); 18:46
TimToady goto dev.perl.org/perl6 and it'll have a link to the various docs.
lambdabot Title: Perl 6 - perl6
evalbot_r14892 Error: ␤Unexpected ": sub"␤expecting "::"
TimToady ?eval sub hello { say $^greeting; } hello('world'); 18:47
evalbot_r14892 OUTPUT[␤] Bool::True
TimToady can't use it in a normal sub, only in a bare block.
Limbic_Region said that but unfortunately his client ate it
TimToady ?eval my $hello = { say $^greeting; }; $hello('world'); 18:48
evalbot_r14892 OUTPUT[world␤] Bool::True
TimToady jdv79: the exact spot for twigils is dev.perl.org/perl6/doc/design/syn/S..._Variables 18:50
lambdabot Title: Synopsis 2: Bits and Pieces - perl6:, tinyurl.com/y648lg
TimToady jdv79: in addition to being the new $a and $b, the ^ notation also lets you document the meaning of the single argument to a single-argument block where in Perl 5 you'd have to use $_ 18:55
e.g. @foo = grep { transmogrify $^element }, @bar
18:56 neonse joined, Psyche^ joined
Limbic_Region TimToady - the optrees are different. Looking at the differences now 18:58
TimToady probably an extra RV2CV op
looks like the alpha sorting semantics of $^a <=> $^b is sub-documented in the synopses...at least, I'm not grepping anything... 18:59
Limbic_Region TimToady - www.perlmonks.org/?node_id=383504
lambdabot Title: ikegami's scratchpad
Limbic_Region looks like more than just 1 extra op
TimToady wow, looks like refgen thought it needed to boil down a list or some such... 19:01
like it was trying to handle \($a, $b, $c) or some such. 19:02
ashelyb_ PCR question: how do I write a rule that will match anything up until another string in PCR subset of p6 rules?
TimToady what's the matter with /.*? string/ ? 19:04
or are you using :ratchet? 19:05
ashelyb_ cause the .* is <rule1> and 'string' is in <rule2> and I cannot get it to work that way
suppose I have "xyxyxyxyxyz" and the match is /<anything><matches_xyz> 19:06
TimToady if you're using rules then :ratchet is on by default, and .*? will probably end up too short
rules and tokens don't backtrack by default 19:07
ashelyb_ p::c::regex because calling them through p::c::regex seem work so well and all the tests are also writen with p::c::regex.
so is there some kind of negative look ahead 19:08
TimToady is <anything> written as a rule or token or regex
?
kolibrie I'm not sure that .*? is even implemented in PCR :ratchet - it is still listed in the TODO
ashelyb_ I;m currently only using Pugs::Compiler::Regex->install() 19:09
TimToady you can probably use <!before string> if you need a negative lookahead.
19:09 Psyche^ is now known as Patterner 19:10 dakkar joined
TimToady I'm not talking about the routine names, but whether the regexes themselves were declared as rules or tokens or regexes. 19:10
I really have no clue about the PCR->install interface
ashelyb_ there seems to be a P::C::regex and a P::C::rule ...
both with install methods 19:11
fglock quantifiers don't backtrack in :ratchet mode in PCR (which is wrong)
kolibrie uses the v6-alpha interface
fglock the install methods are sugar for *rule = ...
19:12 avaron joined
TimToady er, quantifiers shouldn't backtrack in :ratchet mode... 19:13
fglock oops, I thought so
19:13 wilx left
fglock so it is implemented right, after all 19:13
TimToady however, you should be able to override and force a quantifier to backtrack under :ratchet with :+ or whatever it is today... 19:14
so .*? and .*+ both force backtracking, either eager or greedy. 19:16
:ratchet just turns off the backtracking on .* .+ etc.
jdv79 thanks TimToady! 19:17
TimToady actually, I think we changed the extra + to ! or some such.
ashelyb_ my use case is that I have rule1 = /<-[ ]+>/; and rule2 matches an ip address, but sometimes there is no space between the token for rule1 and the rule2, but I don't want rule1 to gobble the ip address that rule2 is supposed to match.
TimToady becaause jfriedl claimed ++ for possessive matches. 19:18
ashelyb_ so how would I write rule1 to not match any spaces or the ipaddress?
19:19 shobadobs left
fglock you mean /<-[ ]>+/ 19:19
ashelyb_ yes I meant that.
:)
19:19 shobadobs joined 19:20 larsen_ joined, weinig_ joined
fglock how about \S+ 19:20
or \D+ 19:21
ashelyb_ hmm so rule1 = /\S+?/; ?
fglock \D means 'not a number'
ashelyb_ yeah but rule1 can match on any character but a space.
even a number
TimToady yes, we did change it to .*? and .*! to override :ratchet 19:22
ashelyb_ my problem is that the space between the rule1 and the ipaddress is optional.
fglock I've got to go - it's my birthday party :) 19:23
ashelyb_ ! happy birthday!
TimToady ditto!
kolibrie fglock: happy birthday!
Limbic_Region feliz cumpleanos
PerlJam fglock: how old are you? (Happy birthday!) 19:24
fglock thanks! have a virtual piece of cake
kolibrie mmmmm 19:25
fglock PerlJam: 43, I think
tchau & 19:26
19:26 fglock left
kolibrie ashelyb_: if rule1 is fixed width, you can say /\S**{8}/ or /\S**{8..10}/ or whatever 19:30
19:31 stevan__ joined
kolibrie ashelyb_: you can define a regex by what is matches, what it does not match, or how long it is 19:31
TimToady jdv79: dev.perl.org/perl6/doc/design/syn/S..._variables also discusses the $^ notation. 19:34
lambdabot Title: Synopsis 6: Subroutines - perl6:, tinyurl.com/pfpph 19:35
19:35 spoop joined
ashelyb_ how would I say don't match xyz? as a string not as a word class like /<-[a-z]>+/ 19:35
<-x><-y><-z> hmmm 19:36
or no that would probably be <-[x]><-[y]>... 19:37
kolibrie maybe <!'xyz'>
ashelyb_ trying.
Limbic_Region TimToady - out of curiosity, is there anything neat in the works for pack/unpack since they aren't spec'd yet? 19:39
TimToady haven't really thought about it much. mostly I think that sort of thing should be handled by serialization roles. 19:40
ashelyb_ unknown node: $VAR1 = undef;
... it doesn't like the <!'stuff'> syntax...
Limbic_Region often laments for an unpack that remembers where it left off such that you wouldn't need to keep track of x yourself
TimToady try <!before stuff>
where stuff is a re
ashelyb_ where stuff is xyz
ah
that's what you meant, trying...
Undefined subroutine &Pugs::Emitter::Rule::Perl5::not_before called at /usr/local/share/perl/5.8.4/Pugs/Emitter/Rule/Perl5.pm line 101. 19:41
TimToady then I'd say negative lookahead is not yet implemented... 19:42
Limbic_Region TimToady - so would strings come with some built in serialization roles or is everything user defined?
TimToady dunno 19:43
Limbic_Region or rather, instead of trying to force a solution on a problem let me present the problem and ask how you would solve it
in general, if I pull a record defined by some unpack format out of a string - I want unpack to remember where I was so next time I get a record I don't have to specify my skip value 19:45
TimToady I think there are several components to the solution. first, treat the normal pack/unpack templates as a language that compiles down to some internal form. 19:48
19:48 weinig_ is now known as weinig|away
TimToady second, have a way for a class serializer to return one of those compiled thingies without having to re-engineer the pack template syntax. 19:48
third, have a way for pack/unpack to refer to the current string position like regexen can. 19:49
possibly there are more unifications with re syntax down the road too 19:50
basically, think of pack templates as compiling down to a sub that could be called as a rule. 19:51
Limbic_Region ok 19:52
TimToady which basically means unpack could be returning a Capture.
and pack would just turn a Capture back into a serialized string.
I'm sure there are plenty of details around the devil. 19:53
Limbic_Region is beginning to like the sounds of this
ingy hola 19:54
pasteling "ashleyb_" at 66.201.51.66 pasted "got it -- matching before IP address, thanks TimToday and kolibrie." (43 lines, 908B) at sial.org/pbot/21784 19:55
19:58 ashleyb joined
TimToady ingy: hey! 19:59
ingy hi TimToady! 20:01
how's the season at Chez Wall? 20:02
TimToady Geneva's home from college, which is nice. 20:03
I have the crud, which is not so nice.
ingy I had the crud for the entire month of November 20:04
TimToady comes of living somewhere soggy. :)
ingy I ran the Dublin marathon on Oct 30, and got the nastiest thing on the flight home
comes from flying overseas in the recycled breath of sick people 20:05
TimToady there are an amazing number of bugs that kick you harder when you're already down for some reason.
ingy yeah, but I never had them set up camp like this before
anyway, feeling great i a very soggy Seattle at the moment 20:06
n
ooo. lucky spacebar!
TimToady need a half-space bar 20:07
20:07 Aankhen`` joined
Juerd hands TimToady a hobby knife 20:07
ingy Capitol Hill - A River Runs Through It
TimToady on some old typewriters you could fake a half-space because backspace would go back space-and-a-half while the key was down, then revert a half when the key came up. 20:08
ingy laughs
add it to the spec!
TimToady actually, I preferred the ones where the ordinary space bar went half space when you pushed it down and half space when you let it up. 20:09
ingy nice
TimToady made it really easy to insert a word that was one character longer.
ingy oh right, yeah 20:10
that took me a sec
take a half from each side
TimToady right
Juerd Makes me wonder if this was an implementation side effect, or an intended feature :) 20:11
ingy YAML forbids your typewriter
god i'm glad the halfspace never made it into ascii 20:12
TimToady why? YAML is defined only by what it ain't.
unless you're claiming that a typewriter is Markup Language
ingy just has an overwhelming sense of fud around TimToady 20:14
:)
TimToady didn't make it into Unicode either. though we do have thin space, hair space, six-per-em space, punctuation space, and zero-width space...
20:14 stevan__ is now known as stevan
ingy mostly f :D 20:15
TimToady -Ofud
ingy YAML has quite a few line endings that it allows
iirc 20:16
like NEL
yaml.org/spec/cvs/current.html#id871136 20:17
lambdabot Title: YAML Ain&#8217;t Markup Language (YAML&#8482;) Version 1.1
ingy goofy
stevan ingy!!!!!
ingy hi stevan! 20:18
it's been a bit
how you been?
stevan hums the theme to chariots of fire and pictures ingy crossing the finish line in Dublin
ingy: $work is *finally* starting to let up (holidays and all) 20:19
ingy My whole body seized up with like 300 meters to go
I almost just fell over
cause they had this fake finish line that i picked up the pace for
stevan: yay 20:20
stevan bastards!
ingy yeah, it was actually the start line with a big arch of balloons
but the finish was 1/4 mile or so more 20:21
stevan ingy: we are flying nothingmuch out here in Feb. for $work, we should try and organize an East Coast Hackathon or something :)
ingy stevan: I'm in
stevan ingy: my house is not big enough to host, but I am sure we can work something out
ingy I've started doing a serious Kwiki reorg... Kwiki 2.0
ok 20:22
stevan cool
ingy yeah I need to start writing the plugins in Perl6 though
hmm, that's a really good idea
stevan I have been hacking Ocaml lately,.. it has a really nice bridge to Perl (perl4caml) and its like Haskell only not as psychotically pure :) 20:23
ingy suprised I just thought of that
oh yeah?
what kind of bridge?
stevan ingy: its pretty low level (dealing with sv,av,hv, etc) 20:25
merjis.com/developers/perl4caml
lambdabot Title: Developers/Perl4Caml
stevan but its very easy to write a nice wrapper around modules
TimToady so is there an Ocaml to Haskell bridge? if so, we're home free. :)
20:25 penk joined
stevan TimToady: I would think there might be :) 20:25
i mean,.. it sounds enough like a PhD thesis, I am sure someone has tried it :) 20:26
TimToady maybe we should go for a record. we could send out Perl 6.0.0 and have it require five or six different runtimes all in the same process. 20:27
stevan TimToady: why stop at 5 or 6 ? 20:28
20:28 justatheory joined
TimToady I think that would be a record already... 20:28
ingy vote for 6 runtimes personally, for obvious reasons 20:29
TimToady doing more than 5 or 6 might slow adoption a tad... :)
ingy stevan: can you write Perl modules in Caml easily using that bridge? 20:30
Inline::Caml
stevan ingy: I have been thinking about that actually :)
20:30 dvorak_ joined
stevan ingy: the trick is that ocaml is very very static 20:30
ingy so is C, no? 20:31
stevan ingy: no, Ocaml has a very aggressive type inferencer which does not lend itself to much flexibility 20:32
but I am exploring some of the deeper internal modules to see what is possible
my first goal is to make it possible to subclass a perl class with an ocaml class
ingy who is coming out with nothingmuch btw?
nnunley stevan: OCaml provides some nice bits for foreign bindings, as I recall.
stevan ingy: just him, its actually for $work stuff,.. but I figure we can take a few days off for $play 20:33
ingy wow. hi nnunley!
where are you these days?
nnunley Hey, ingy.
London, for the time being.
stevan nnunley: I havent explored those yet
ingy stevan: maybe I'll come visit you early, and overlap with nothingmuch
nnunley: still with fotango? 20:34
nnunley stevan: I think all they require is an opaque pointer with function bindings for calls through to C, as well as a few function bindings for garbage collection.
stevan why is that I am always attracted to modules which have this type of caveats "Run-time support for objects and classes. All functions in this module are for system use only, not for the casual user."
nnunley ingy: For the time being.
ingy nod
20:35 penk joined
gaal mooses 20:35
20:37 polettix joined
ingy gooses 20:42
mmm gøøse
I wrote an awesome module called Script::Hater 20:43
for generating Configure type scripts from data
programming in YAML
I should call it Code::Hater 20:44
ingy isa Code::Hater
ingy isa Data::Lover 20:45
gaal call it descript
ingy how about Scriptnoxious
gaal NOXcript 20:46
ingy I hope Perl6 is the ultimate mini-language meta-language
gaal UMLML? eep 20:47
ingy Many programming tasks are well suited to having their own special programming language 20:48
So hopefully changing the P6 grammar is AllThat(tm)
20:50 stevan_ joined
nnunley ingy: How's you? 20:58
21:03 shobadobs left 21:04 shobadobs joined 21:09 dvorak joined 21:16 jan___ joined 21:21 ron_ joined 21:22 devogon joined 21:30 chessguy joined
GeJ good m􏿽xF8􏿽xF8se all 21:47
22:07 Limbic_Region joined 22:24 araujo joined 22:25 araujo_ joined 22:26 araujo__ joined 22:27 araujo joined 22:28 jan__ joined 22:33 gnuvince joined, ron_ joined 22:34 araujo joined 22:35 araujo joined 22:46 marmic joined 22:55 araujo joined 23:04 ron_ joined 23:06 nperez joined 23:18 sonorous joined 23:55 neonse left

Powered by Cro and the Raku® Programming Language.

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

Thank you!