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.
00:08 pbuetow joined 00:09 mako132_ joined 00:10 rashakil joined 00:15 CardinalNumber joined
Juerd_ masak: Tiu, kiu volas featherkonton, skribu al mi retposxtmesagxon kun vera nomo, dezirata uzantonomo, kaj la kialo por voli la konton. 00:19
masak: Kaj ne iuj nomoj, sed siaj nomoj :) 00:21
luqui esperanto? 00:24
00:37 vel joined 00:38 mako132_ joined 01:03 b00t joined 01:17 weinig joined 01:21 penk joined 01:35 Debolaz joined 01:37 ProperNoun joined 01:53 CardinalNumber joined 01:57 lyokato joined 02:03 ashleyb joined 02:04 meppl joined 02:12 ProperNoun joined 02:54 mncharity joined 02:55 mncharity is now known as putter 03:12 kanru joined 03:39 luqui joined 04:07 leed joined 04:31 xinming joined 04:57 Alias__ joined 05:51 kanru joined 06:18 Alias__ joined, luqui joined 06:19 spx2 joined 06:28 iblechbot joined
masak juerd: bone. mi pensu iom pri la kialo k poste sendos al vi retposhtajhon 06:28
06:47 Yappo joined 06:49 GabrielVieira2 joined 07:13 b00t joined 07:22 prism joined 07:33 rashakil joined 07:34 rhizo joined
yves is back (gone 14:34:16) 07:41
07:42 dduncan joined 07:52 b00t joined 07:54 elmex joined 08:04 ozo_ joined, jql joined 08:11 BooK joined 08:17 sili joined 08:29 shar joined 09:21 Juerd_ is now known as Juerd
Juerd luqui: Yes 09:22
10:00 mr_ank joined 10:10 iblechbot joined 10:16 marmic joined 10:21 fglock joined 10:28 ruoso joined
jql ?eval %x = ('a' => 1); %y = ('b' => 2); %x.push(%y); %x; 10:36
evalbot6 Error: Can't modify constant item: VUndef
10:38 dduncan left
Juerd I don't think you can push onto a hash 10:40
10:42 cmarcelo joined
jql I would've been surprised if it worked 10:43
10:57 spo0nman joined 10:59 integral_ is now known as integral
Juerd I wouldn't. I have been wanting HASH.push since Perl 4 :) 11:02
Simply because writing @hash{keys %foo} = values %foo involves both a temporary hash, and ugly code :) 11:03
jql I'm not saying I would mind :)
Juerd Fortunately it's easily added to Perl 6 :)
Though perhaps I would name it .add instead of .push, as .push does sort of suggest sequential order
Maybe the ,= operator 11:04
Though killed and burried at least a dozen times :)
jql I presume there's no lazy hash? If so, .push becomes meaningful 11:05
jql has disturbing thoughts this late
integral There would only be a lazy hash if someone actually sat down a wrote it. 11:08
fglock I wrote a lazy hash for pil-run - I was told not to do that :) 11:10
jql push %lazy.specs, /^something/ => -> { generate_value($^a) };
where do I commit my example? :) 11:11
fglock jql: language-related stuff is usually discussed in the p6-language mailing list 11:13
pasteling "evalbot_r15090" at 194.145.200.126 pasted "Pugs build failure" (429 lines, 24.1K) at sial.org/pbot/22492 11:15
11:34 Lorn joined 11:37 ruoso joined, chris2 joined 12:05 xinming joined, TimToady joined 12:11 fglock joined
fglock 6 days left until vacations 12:16
christmas.whipnet.net/kids/timeleft.html
lambdabot Title: How many days left until Christmas?
12:23 buetow joined 12:24 VanilleBert joined 12:26 nferraz joined
nferraz hello, fglock! :) 12:27
12:28 lisppaste3 joined
fglock nferraz: ola 12:29
jql: I think the result of the discussions at that time was that you can have lazy hash, but you have to declare it explicitly 12:36
nferraz i was thinking on what you told me last night, about using grammars... 12:37
i thought defining something like: regex articles { (a|an|the) } 12:40
seems more elegant than the old-fashioned way... 12:41
fglock token articles { a | an | the } -- more efficient 12:44
12:46 jiing_ joined, _marco joined 13:05 ruz joined 13:07 capixaba joined, bonesss joined 13:08 elmex joined 13:25 Limbic_Region joined 13:29 diakopter joined 13:42 gnuvince joined 13:47 kanru joined 13:49 rashakil joined
fglock someone asked why wasn't 6-on-6 being written with pugs 13:56
now I'm thinking why not
it would be easier for kp6 to catch up to an existing compiler, rather to a planned one 14:01
rather then
hmm - rather than 14:03
14:03 gnuvince_ joined
fglock I wonder why nobody seems to be interested on 6-on-6 14:06
lunch &
Ziggy6 wow 14:09
you mean no 6-on-6 project exists now?
Limbic_Region fglock - orgies are so passe
14:19 falseep joined, vel joined
Steve_p I thought a selfhosting 6-on-6 is what pugs was 14:24
Ziggy6 it seems so obvious that 6-on-6 should be written in pugs 14:27
[particle] um. it'll be written in p6, won't it? ;)
Ziggy6 right, but now, pugs seems to be the closest to p6, so... 14:28
i noticed object.HOW.methods doesn't work 14:29
HOW returns something though
is it the same return value for all objects? 14:30
[particle] yes, object metainfo is largely unimplemented
14:30 marmic joined
Ziggy6 that would seem like a nice place to start hacking for me :P 14:30
[particle] look for tests, if there aren't any, write em :) 14:31
are you using svn, or svk?
14:31 penk joined
Ziggy6 svn for now 14:31
i'm still waiting for a svk2 ebuild 14:32
[particle] you may want to install App::Ack from cpan
it's a grep replacement, and works great in svn working copies
(eg. ignores .svn dirs)
Ziggy6 hehe, i ignore them visually :P 14:33
[particle] it also has colorful output... i' think you'll like it C<cpan App::Ack> 14:34
14:36 xinming joined
wolverian argh, I need a printer. 14:43
14:50 elmex_ joined 14:54 fglock joined 15:02 elmex__ joined 15:09 triplef joined
triplef Shouldn't "my $class = Class->new; my $thread = Threads->new(Class::function, $class);" send $class to Class::function? 15:10
[particle] what is Class::function here? 15:12
do you want to invoke that, and pass the result to new?
gaal if this is p5 and you want to send the coderef, better use \&Class::function
Ziggy6 ack is cute, 10x for the tip :) 15:13
gaal otherwise "Class::function" is evaluated and the result is treated as the first argument to new
[particle] anytime, Ziggy6
triplef gaal: Yes this IS perl5! I forgot to look at the channel title :-/ . But \$Class::function seems to work. Thanks alot! 15:14
gaal \&, not \$ :-)
15:14 mj41 joined
triplef gaal: Yes, that was what i was *thinking* :) \&Class::function 15:15
15:18 Lorn_ joined
nferraz what is the new recommended way to pass parameters to a sub? 15:30
I know sub foo( $param ) will work, but $param will be read-only???
fglock nferraz: either make foo a method/role and then $param.foo, or sub foo( $param is rw ) 15:33
nferraz mmm 15:37
does the $sentence ~~ s:[foo] = " "; work???
15:43 Lorn joined 16:03 buetow joined 16:15 rashakil joined, ofer1 joined 16:22 hexmode joined 16:40 stevan_ joined
yves is away: Occupōæ½xE9 16:48
nothingmuch is here, otoh 16:49
statico is here in spirit 16:55
allbery_b is sorta halfway here 17:00
[particle] "wherever you go, there you are, man."
17:02 VanilleBert joined
TimToady The Doctor is In 17:08
Psychiatric Help 5Ā¢
17:12 juerd_ joined
TimToady the main problem with doing 6-on-6 in pugs is that nobody has bothered to port a P6 regex engine to Haskell... 17:13
< nferraz> does the $sentence ~~ s:[foo] = " "; work??? 17:15
this is an ambiguous question
on the surface, it doesn't work because of that colon there
it would have to be s[foo] = " " 17:16
17:17 bernhard joined
TimToady if you are asking whether pugs supports it 17:17
I don't think pugs has the syntax yet. 17:18
if you are asking about the relative precedence of ~~ and =, the answer is that the = is not a real =, and binds more tightly to the s[foo] than the ~~
17:19 prism joined
TimToady nferraz: see above 17:19
lumi It's the Ā« s = Ā» operator?
[particle] if it's not a "real" equals, what is it? 17:20
is it more like a <== ?
TimToady a delimiter between parts of a macro
nanonyme hmm
TimToady it's real enough to have += variants though
[particle] can it be hyperized? 17:22
TimToady or maybe it's real, but s[] is a bit like a listop in having a different precedence on one side than the other
no
s[] = is really mostly a macro 17:23
s[foo] = "bar" is sugar for .subst(/foo/, { $_ = "bar" }) or some such 17:25
17:26 theorbtwo joined, ashleyb joined
[particle] does that list of variants exclude ::= 17:28
TimToady if you look at what it desugars to, $_ ::= "bar" is rather nonsensical
it would have no effect on the s[] 17:29
[particle] yeah, i thought so
just making sure
probably that excludes =>, too
(looking over the assignment ops in S03) 17:30
TimToady nferraz: I'm wrong, it is implemented in pugs 17:35
?eval my $_ = "f00"; s:g[0] += 2; .say 17:36
evalbot6 OUTPUT[f22ā¤] Bool::True
TimToady though it doesn't recognize x= there... :( 17:37
?eval my $_ = "f00"; s:g[0] x= 2; .say
evalbot6 Error: Internal error:ā¤ Unknown pseudo-assignment form:Syn "=" [Var "$_x",Val (VInt 2)]ā¤Please file a bug report.
TimToady $_x ??? 17:40
I'll bet it's desugaring to $_x=2 instead of $_ x= 2 17:41
17:45 rashakil joined
TimToady fixed 17:46
17:49 fglock joined
TimToady ?eval my $_ = "f00"; s:g[0] x= 2; .say 17:51
evalbot6 Error: Internal error:ā¤ Unknown pseudo-assignment form:Syn "=" [Var "$_x",Val (VInt 2)]ā¤Please file a bug report.
TimToady twiddles thumbs...
fglock TimToady: re P6 regex engine to Haskell - mp6 engine could be reused
Ziggy6 what's mp6? 17:52
lumi Why is evalbot nude from version?
s/from/of/
fglock Ziggy6: a minimal 6-on-6 compiler
nothingmuch it's the new EvalBot Pro0n edition
we need an op to /kick all the underage perl6 hackers
TimToady I think that's most of us 17:53
Ziggy6 mp6 = svn.pugscode.org/pugs/v6/v6-MiniPerl6 ?
nanonyme feels old
lambdabot Title: Revision 15092: /v6/v6-MiniPerl6
nothingmuch not just iblech?
fglock Ziggy6: yes
17:54 Debolaz joined
TimToady Ziggy6: btw the current plan for implementing .HOW is to stitch in the MO code that is in (I think) src/MO 17:54
fglock: yes, I think mp6 could pretty easily be translated, given someone who has learned Haskell more times than I have... 17:55
17:55 rashakil_ joined 17:56 triplef joined
TimToady fglock: did you notice that we changed the meaning of | again to collect all DFAable prefixes, not just constant prefixes? 17:57
Ziggy6 oops, i had a shock, seeing all that Haskell code :)
fglock TimToady: yes 17:58
TimToady thinking of it as deciphering a transmission from Alpha Centauri.
fglock I'm figuring out how to implement
Ziggy6 so the rule engine from mp6 is complete?
fglock Ziggy6: it's not complete, but it is written in perl 6 17:59
Ziggy6 you were thinking about porting it to Haskell?
fglock and it compiles itself
allbery_b misreads "rule enigma"
TimToady basically instead of /foo \d*/ returning just "foo", it returns "foo \d*", and all those prefix patterns get or'ed together into one DFA where the accepting states tell you which subpattern matched
Ziggy6 how did you bootstrap mp6?
fglock Ziggy6: pugs could translate mp6->haskell 18:00
18:00 xinming_ joined
fglock Ziggy6: it was bootstrapped by a compiler chain: lrep -> v6.pm -> mp6 18:00
lrep was bootstrapped by hand
TimToady: ah, so the underlying engine is still the same 18:02
it just adds a further check for longest token
TimToady yes, essentially 18:03
it would be nice not to reparse but that's an optimization
fglock problem is, if it finds a return block, for example
hmm - not a problem, really
TimToady {...} is defined to terminate "pure" region 18:04
18:04 ludan joined
fglock ah, ok 18:04
TimToady see most recent S05 changes
Juerd feather is reachable by its IP address
The DNS server is broken, and because of the storm, there is no way to reach Amsterdam right now :)
feather's IP is 194.145.200.126
[particle] /topic update? 18:05
Juerd Hm, there may be another way... Just a sec. 18:06
TimToady that's funny, I have DNS for it--just checked in a patch in fact
Juerd Probably cached by your ISP :)
Yay, the DNS machine is xen-ified
I can "xm console" it remotely
TimToady yeah, I get a non-authoritative answer 18:07
18:19 nferraz left 18:31 autark__ joined
fglock TimToady: re S05 - I'm trying to understand this part - "but not including any subrule more than once" 18:32
kolibrie lumi: evalbot didn't join the channel when feather came back up yesterday, so I started one and invited it here. 18:33
TimToady recursion will cause the token collector not to terminate 18:34
kolibrie I didn't find docs explaining all the versioning stuff, so it doesn't have that
18:34 svnbot6 joined
kolibrie anyone who knows how to fix evalbot can kill this one, start a new one, and document how to do it 18:35
TimToady X>>X>>X>>+<<X<<X<<X would be a legal token
fglock TimToady: ok
audreyt kolibrie: all I did was: "perl examples/network/evalbot/evalbot-loop"
I've summoned a new evalbot here; please drop the evalbot6 once it's here (should be in 5min or so) 18:36
TimToady thanks. how's backlogging going? :)
18:36 VanilleBert left
audreyt done with the syns 18:36
kolibrie audreyt: so you don't run the pugs step from the README, I guess
audreyt kolibrie: mm? 18:37
was confused a bit with S02 because in my mind Squid isa Fish
(in Chinese they share the word root)
TimToady heh
audreyt esp. you have a clause saying that "returns" can usefully be more specific than "of"
but then the example is showing two wholly incompatible types 18:38
kolibrie audreyt: examples/network/evalbot/README.pod says: pugs evalbot.pl evalbot6 irc.freenode.net:6667
audreyt which is mildly confusing given the Chinese foobar :)
kolibrie: README.pod predates evalbot-loop. fix it?
TimToady we just need a perfect language to categorize everything for us properly. ;)
rgs ask john wilkins ! 18:39
audreyt CosmicOS!
TimToady he's dead--I asked Umberto Eco instead...
18:39 _marco joined
audreyt Lincos, rather 18:39
kolibrie audreyt: so just run 'perl examples/network/evalbot/evalbot-loop', and then how do you invite to #perl6
18:40 TimToady joined
audreyt kolibrie: it's the default 18:40
18:40 buetow joined
audreyt see evalbot-loop and grep for #perl6 18:40
kolibrie audreyt: ok, updating README.pod
audreyt kolibrie++
18:43 evalbot_r15092 joined 18:45 TimToady joined
audreyt ?eval 1+1 18:47
evalbot_r15092 2
18:52 _marco left
svnbot6 r15093 | kolibrie++ | evalbot - added documentation for starting the autoincrementing bot 18:52
diakopter ?eval 'r15092' + 'Christmas'
evalbot_r15092 0.0
diakopter ?eval 'r15092' && 'Christmas' 18:53
evalbot_r15092 "Christmas"
kolibrie audreyt: thanks for starting up the svnbot6 too
audreyt np :) 18:55
diakopter is it possible to set up a way of inputting tests to the evalbot, who would then add failing ones to svn? 18:57
audreyt sure :) wanna write one? :) 18:59
TimToady ?eval my $_ = "f00"; s:g[0] x= 2; .say 19:00
19:00 evalbot_r15092 is now known as evalbot_r15093
evalbot_r15093 OUTPUT[f0000ā¤] Bool::True 19:00
19:07 seano joined 19:17 __Ace__ joined
__Ace__ hey folks 19:18
so.. in beta stage with Perl6 now eh?
;)
19:18 thepler joined 19:19 nipra joined 19:20 GabrielVieira joined
__Ace__ I truly hope its not another waporware... 19:20
audreyt TimToady: so the reason of accepts/rejects instead of ACCEPTS/REJECTS is ... ?
19:23 Ziggy6 left, Ziggy6 joined
__Ace__ I was also wondering... how much build in OO is there gonna be with Perl6? Kinda like the 5.next idea.. (although I wonder why the h-ll they dont have 5.prev in Ruby) 19:23
wolverian __Ace__, see spec.pugscode.org for the specs
__Ace__ ok 19:24
19:25 ludan joined 19:26 luqui joined
TimToady audreyt: because it's also expected to be called directly, not just automagically? 19:26
luqui gidday 19:28
TimToady but if anything can be a "pattern", then maybe uppercase would be a good idea for collision avoidance
luqui: howdy do
luqui am I to infer that there are audreyts around?
TimToady shh! be vewy vewy siwent 19:29
__Ace__ uh, having "loop" instead of "for".. isnt that just confusing? 19:30
TimToady it's there just to confuse you! :)
luqui it saves "for" for the useful "for" 19:31
... it has been a really, really long time since I've used C-style for in perl code
__Ace__ I bet you guys know what you are doing but sometimes things are best as they are since they are so used..
TimToady that was how Perls 1..5 were designed
Perl 6's mandate is to break everything that could use breaking
__Ace__ so, now you wanna be inventive by making some new stuff up? 19:32
TimToady Hubris it is.
__Ace__ aha
TimToady well, and Laziness, since for is shorter than loop, and used more often
__Ace__ same goes with using ~ for concatination instead of . ?
luqui TimToady, and it works on infinite lists :-)
__Ace__ ~ is 2 buttons instead of 1 with the . :) 19:33
TimToady yes, method calls are more important than concatenation
__Ace__ haha, true
luqui __Ace__, that's because we spelled the three-button -> with one button: .
TimToady especially since most concatenation is done via interpolation
__Ace__ just as I thought, you know what you are doing ;) 19:34
[particle] how dare you make such accusations!
TimToady we don't really know what we're doing, we just know what we're trying to do.
__Ace__ TimToady: And how is that going then? 19:35
TimToady things still seem to be converging on a solution
[particle] such is the path of architects. making a quick succession of suboptimal decisions.
19:36 prism joined
__Ace__ Me and my buddy did a simple javacompilator... and we thought that was hard.. Perl code cant be easy to break down.. 19:36
TimToady if you want some of the original suboptimal reasoning, you can look at the old apocalypses too.
that's why we need a powerful pattern matching engine in Perl 6 to do it for us. :) 19:37
__Ace__ hehe
and we did this in Java.. not C
19:38 fglock joined
__Ace__ so, seriously, noone knows how much is left of all this before some release? 19:39
some Ruby/Python/blahblah folks say it will never come out...
TimToady our aim is convergence, not a particular area under the curve. 19:40
as long as the area is not infinite
__Ace__ would be nice to see how you respond to such accusation
TimToady precisely how I'm responding to it now. Just keep peddling...
19:41 _marco joined
TimToady whether it comes out or not has little to do with people who aren't contributing, only with people who are. 19:41
__Ace__ so, its not about finishing it, but to work on it?
19:41 autark_ joined
TimToady it's about working on finishing it. 19:41
__Ace__ :)
TimToady convergence on a solution is important 19:42
__Ace__ I would gladly help.. but I need some realy work where I get money first.. ;)
*real
TimToady yeah, know the feeling, believe me. 19:43
most of us here do
__Ace__ bet you _do_ get money from it... some way
or no?
TimToady I get some royalties from Perl 5 book sales
__Ace__ haha, thats sad.. I mean, only that.. ?
TimToady but aside from some early grants from the Perl Foundation, I'm entirely self funded
__Ace__ no sponsor money? 19:44
TimToady and about $250k more in debt on my house than when I started
__Ace__ "Hey, this is some good stuff. Here's $1000000"
TimToady would be nice, but not counting on it. 19:45
[particle] it'd be nice to get a million pennies, even
__Ace__ isnt it funny though.. some other folks make a living working with Perl code
Ziggy6 timtoady do you have a real job? 19:46
TimToady yes. should be going there 'bout now...
I work across the street from Google
Ziggy6 wow
__Ace__ cool
with that?
Ziggy6 in seattle?
__Ace__ *what
TimToady Mountain View
company called Netlogic Microsystems 19:47
Ziggy6 did you apply there?
TimToady I was invited by an acquaintance who works there.
they are working on pattern matching in hardware, and they thought I knew something about that...
obra grins at timtoady 19:48
kolibrie Perl 6 regexes in hardware ???
TimToady but it does chew up a significant amount of my time
Ziggy6 lol
standard 8hrs/day?
luqui really??you seem to be here most of the day...
TimToady they expect me to spend part of my time working on Perl
Ziggy6 ah, you are helping them to build faster firewalls or something 19:49
TimToady yeah, things like that.
[particle] ziggy6: are you in seattle? 19:50
Ziggy6 nope
TimToady romania?
fglock btw, audreyt: Segula never paid your sponsorship :(
[particle] 'cause last time i checked, google's in california
Ziggy6 hehe tim, how did you find out?
__Ace__ TimToady: what say you about me having done a "Desktop Search" app in perl? :)
TimToady I saw the .ro first time you checked in
Ziggy6 ah, the domain name...
__Ace__ its more than that, but anyway.. 19:51
TimToady __Ace__: er, my condolences? :)
__Ace__ hehe
I actually did alot of research before I started... was like 2 years ago
Ziggy6 you're selling it? 19:52
__Ace__ the idea was to release it as open source or something
[particle] so it's vaporware, like perl6? ;)
Ziggy6 but you didn't release it yet
__Ace__ yea kinda ;)
problem is
TimToady most things fall into the category of "open source or something"... :)
__Ace__ I have ideas with it all the time I wanna implement, and even if its Perl (where things goes pretty much faster than doing the same in C), it takes time 19:53
TimToady Hopefully even faster in Perl 6, someday... 19:54
__Ace__ yea
Ziggy6 ?don't_eval "next" xx * ~ " christmas" 19:55
lambdabot Unknown command, try @list
TimToady anyway, you're welcome to help however you like.
__Ace__ when I started research I wanted among other things: 1. make it like a standalone. one run and go file (like PAR) 2. have modules that I needed.. :)
19:55 thepler joined
__Ace__ and other things I cant remember 19:55
TimToady: so, where do one start? 19:56
dl the source of perl6 and understand it?
TimToady most people start by playing with pugs, the prototype implementation
and then branch out into writing tests, or docs, or whatever
luqui __Ace__, it could be a couple years until you understand it 19:57
19:57 thepler joined
TimToady putter was working on a diagram the other day of various places that people can contribute 19:57
luqui I'm not actually sure how long, since i started a couple years ago and don't understand it
TimToady doesn't stop any of us... :)
luqui but not in a perl5-this-is-the-most-horrible-thing-i-have-ever-seen kind of way 19:58
__Ace__ I suppose the code is devided into "modules" aswell eh?
[particle] hopes putter sends a list email about that
TimToady anyway, you can get a commit to pugs really easily.
luqui but more in a perl6-this-is-the-most-abstract-thing-i-have-ever-seen kind of way
Ziggy6 hehe, i started 2 days ago, and i hope it won't take that long...
__Ace__ *divided
TimToady if you /msg me your email, I can send a pugs commit bit to you.
luqui had to learn haskell along the way
[particle] avoided the haskell part 19:59
mp6++
Ziggy6 is in for haskell too
__Ace__ crap, need to register
cant pm otherwise
TimToady or obfuscate your addr here
[particle] but you can respond to somebody else's pm
TimToady or email to me
__Ace__ yea, using md5? 20:00
lol
luqui well, it has to be something we can reverse
__Ace__ yea, give me yours then
20:00 bpphillips joined
TimToady email to larry at wall dot org 20:00
__Ace__ oki
luqui or note [particle]'s comment...
nevermind
TimToady I already get tons of spam there...
__Ace__ I send right away then
luqui are you implying that __Ace__'s message will be spam :-)
Ziggy6 ah, didn't realise you were larry wall :) 20:01
20:01 BooK joined
[particle] there's more than one way to spell it 20:02
TimToady I scarcely realize it myself. :)
__Ace__ "you got mail" 20:04
hopefully it wont see gmail as spam 20:05
TimToady commitbit sent 20:06
after svn/svk checkout it's customary to add yourself to AUTHORS as first checkin as a test.
might be a little trouble today because the svn server was having DNS issues 20:07
dunno if that got resolved
__Ace__ www.wall.org/ <- is this like a family page?
luqui updated okay...
lambdabot Title: The Wall Nuthouse
TimToady yes, though most of the kid links are pretty old
[particle] TimToady: you still have home automation running w/ perl5?
TimToady you bet
my family would now be very unhappy if I took that down 20:08
[particle] great. next time we get you up to seattle i'd like you to talk about that
TimToady they like hearing the computer say who is calling on the phone
[particle] :)
does it tell you how much you weigh? 20:09
TimToady nope
[particle] well, room for improvement :)
TimToady it only tells me how smart I am. 'Course it lies...
__Ace__ hey, anyone here wanna try my "shell" Im working on aswell? 20:10
made in perl ofcourse
TimToady the first application is actually still one of the most important; the drier in the garage plays a noise throughout the house since we can't hear it directly when the garage door is closed...
__Ace__ TimToady: btw, is there a way to change the nick to "Ace"? 20:11
luqui can only imagine how annoying that would be
__Ace__ usename I mean
TimToady probably, if not I can give you a new invite 20:12
__Ace__ only reason I use __Ace__ is since the others were taken :)
20:12 autark__ joined
TimToady yeah goto preferences 20:12
__Ace__ although it suits nice with __END__
20:13 autark__ is now known as autark
TimToady A meme that is dying with Perl 5 though... 20:13
luqui (turned into =pod ?)
TimToady among other things
depending on the nature of __MUMBLE__
__Ace__ yea, Im dying.. 20:14
TimToady file and line are just compiler "constants"
__Ace__ why are those gone btw?
allbery_b there are cleaner ways to do them
__Ace__ like? 20:15
[particle] it's now spelled $?FILE
TimToady $?FILE can be interpolated
__FILE__ can't
20:15 diotalevi joined
TimToady =pod streams can have multiple of them, unlike __DATA__ 20:16
things like that
[particle] see dev.perl.org/perl6/doc/design/syn/S02.html#Names
lambdabot Title: Synopsis 2: Bits and Pieces - perl6:
__Ace__ when you say interpolated you mean? I think I know what it is... with gfx and images..
20:16 idiotmax joined
__Ace__ I mean, combining between 2 values.. 20:17
20:17 ekelund joined
TimToady die "Error at $?FILE line $?LINE you dinglefungus!" 20:17
[particle] (descriptive errors)++
TimToady string interpolation, not point interpolation 20:18
more like "gnostic interpolation"
__Ace__ so, what do that do exactly?
TimToady same thing as Perl 5's die "Error at " . __FILE__ . " line " . __LINE__ . " you dinglefungus!" 20:19
but they parse as variables rather than symbols 20:20
so you can interpolate them like variables
diotalevi is currently wishing Perl 5 didn't die with strings but rather with objects that happened to stringify.
allbery_b has wished that a few times as well
TimToady Um, it does, if you give die an object
[particle] they're no longer preprocessor macros, they're variables now
allbery_b hm, but isn't that quasisupported in 5.6/5.8? 20:21
TimToady It's even quaisier supported in Perl 4
__Ace__ hey ekelund
cant pm.. need to reg..
diotalevi TimToady: yes, but the error $foo/0 thrown by the interpreter isn't an object.
ekelund __Ace__: hi 20:22
TimToady true nuff
diotalevi $foo/0 looks odd. I read it like "$foo, the function with zero arity."
__Ace__ TimToady: ok, think I got it now :)
luqui too
(to diotalevi )
__Ace__ so ekelund.. Perl sucks, and Ruby doesnt? 20:23
:)
allbery_b ruby looks a lot like perl5 with many of the wartier things removed.
of course, so does perl6
luqui allbery_b, plus continuations
(the only reason I have ever programmed in ruby) 20:24
(or rather, the reason I started)
allbery_b aruably that's one of the removed warts :)
20:24 b_jonas joined
__Ace__ well, ofcource ruby is "better".. its from 1995.. perl is from 1987 20:24
luqui Java is from... 1991? 20:25
ekelund __Ace__: I havnt said that :) but I like the ruby syntax. But never used ruby and perl for any big application, besides our compiler
TimToady wanders off in a daze &
allbery_b heh
luqui wonders how we started talking about ruby and other pls 20:26
20:33 _marco joined
ekelund will there be any way to talk between ruby and perl 6? any supported ways 20:45
allbery_b parrot is a sort of .net-like thing which hosts perl6 and could easily host a ruby implementation if it doesn't already 20:47
[particle] the implementation of ruby on parrot is codenamed "cardinal" 20:48
ekelund will parrot use object based assembler as in .Net?
[particle] yes, parrot has an oo-assembly language
__Ace__ parrot is supposed to support alot of languges no? 20:49
[particle] yes
__Ace__ you can do things in diff lang. and they can then interact
[particle] that's a goal, yes
20:49 davidra joined
allbery_b s/\.net/CLR/ 20:49
__Ace__ sounds like a cool thing..
[particle] i want to use perl's nifty test modules in my tcl code 20:50
20:50 davidra is now known as autark
ekelund [particle]: how big is the execution time lost when running C/C++ in parrot? 20:50
[particle] you mean, benchmarks of c vs parrot?
for the same task?
ekelund and whats the benifits to use parrot instead of say Mono 20:51
[particle]: yes
allbery_b for one, mono is officially dead now, isn't it?
__Ace__ it is? 20:52
allbery_b as of a couple weeks ago
[particle] parrot is a register-based vm, unlike java & ms stack-based vms
parrot is designed for dynamic languages
wolverian alioth's shootout has parrot entries
__Ace__ mono dead? thats interesting..
ekelund [particle]: is it because of the ms - novell thing? 20:53
[particle] here's some details on the benchmark question: shootout.alioth.debian.org/sandbox/...;lang2=gcc
lambdabot Title: Parrot PIR benchmarks | Debian : AMD&#8482; Sempron&#8482; The Sandbox, tinyurl.com/25nw99
allbery_b thought he'd herd work on it had stopped
__Ace__ and the reason is?
wolverian parrot still loses to c sometimes ;) 20:54
[particle] mono has pretty restrictive licensing
but compare parrot to perl5
ekelund hmm wasnt mono GPL?
or LGPL
wolverian mono is GPL, LGPL and MIT
"Mono's C# compiler and tools are released under the GNU General Public License (GPL), the runtime libraries under the GNU Lesser General Public License (LGPL) and the class libraries under the MIT License." 20:55
20:55 bpphillips left
wolverian that might be considered more restrictive than the artistic license, perhaps. 20:55
20:55 rashakil__ joined
ekelund wolverian: I see it as something good, then it will be free even if someone fork it 20:56
GPL will continue to be free after a fork I mean
wolverian GPL will be forked? :) 20:57
ekelund no you know what i ment 20:58
the software licensed with GPL will continue to be free
wolverian right. 20:59
ekelund wolverian: but that is when no one break the license ofcourse. 21:04
__Ace__ Checked out revision 15093. 21:13
nice :)
21:14 Aankhen`` joined 21:17 bonesss joined 21:18 larsen_ joined
ekelund __Ace__: of perl? 21:27
__Ace__ yea, perl6.. pugs
damnit, ghc 6.6 isnt in the list in ubuntu.. only 6.4 21:28
Ziggy6 congratulations, __Ace__, you are looking at an 10 hour compilation ordeal
__Ace__ heh
need to get this haskell thing to work first
ekelund __Ace__: will you wont come tomorrow right? antar lƶrdag 21:29
Ziggy6 compiling haskell takes several hours
ekelund does perl use haskell?
Ziggy6 also, get -extralibs
__Ace__ eller imon?
Ziggy6 if you compile without -extralibs, pugs won't work
ekelund __Ace__: jasa vilket foredrar du. Sorry we speak swedish :/ maybe should be privmsg 21:30
__Ace__ its probably better to meet tomorrow.. party on sat suits me better I think..
Ziggy6 didn't find this piece of info before I started compiling.. :D
__Ace__ where did you compile?
ekis: tomorrow and sunday
ekelund tomorrow then. can you work late to? because Im at work first (class with tdd then out to our customer) 21:31
__Ace__ sure 21:32
we are geeks arent we?
:)
so, we can stay up long coding
and to answer your question ekelund... there is a haskell implementation made for perl6 21:33
Ziggy6 on my 1.85Mhz barton
used only 256 Mb though (vmware) 21:34
__Ace__ guess it was easy and fast to do... compared to the one in C..
ekelund hmm to run haskell in perl or run perl in haskell
21:34 nipra joined
__Ace__ perl in haskell 21:34
ekelund because you run ghc so I guess no haskell intepret.. in perl
__Ace__ so, you can try perl6 code...
eh, ja :)
21:35 Schwern joined
ekelund __Ace__: next friday will be nice with the conferance and all. Will stay one night at the hotel to. I and my coworkers will probably party in HBG to 21:36
__Ace__ would be interesting to see the presentation ;) 21:37
gotta finish up the perl thingy.. to make it even cooler ;)
allbery_b recent ghc has a library, which would allow you to run haskell embedded in perl if you had bindings for ghc-lib :)
ekelund __Ace__: yes would be good, maybe I can get someone to record it on a tape och dvd :) 21:39
__Ace__ haha, yea, would be cool
ekelund :)
__Ace__ possibly removing all that output data flooding out 21:40
21:40 Ziggy6 left
ekelund hehe :) that to ofcourse 21:41
21:41 mj41_ joined
ekelund I am thinking of running the app and record it to flash or something. I dont think my GNU installation is supporting the external tv yet. Then I can run the flash in os x and show them 21:42
21:42 Ziggy6 joined
__Ace__ vlc2swf ? 21:42
ekelund that didnt work that well last time 21:43
:/
__Ace__ well
try xvidcap then :9 21:44
xvidcap.sourceforge.net/
lambdabot Title: Xvidcap Project Home Page
21:45 jferrero joined
__Ace__ www.google.com 21:45
lambdabot Title: Google
ekelund __Ace__: see privmsg 21:50
__Ace__ Ziggy6 make -extralibs doesnt work with ghc
hmm
ekelund: yea saw it
ekelund __Ace__: nice right :) he is cool, helt off and his solo work :0 21:51
__Ace__ he's ok :)
ekelund __Ace__: he have had an concert in NY to. My friend Rebecka was leaving NY that day she had no time to see it :( 21:52
__Ace__ NY = New York? 21:53
Ziggy6 __Ace__: you need to just unpack ghc-version-extralibs.tgz in the same directory where you unpacked ghc-version.tgz 21:54
21:55 Limbic_Region joined
__Ace__ I did 21:55
21:56 [particle] joined
__Ace__ ran ./configure 21:56
Ziggy6 a simple make install shoul do it now
__Ace__ aha
ekelund NY = New York yes
__Ace__ Ziggy6: this is gonna take a few days now eh? 21:57
damn, why is there no 6.6 ready for ubuntu.. 21:58
Ziggy6 depending on your RAM and CPU 21:59
it took me a night
for both ghc6.6 and pugs
ekelund __ace__: run dist. gcc 22:00
and add some nodes to the network :) 22:01
__Ace__ yea, well 22:02
Ziggy6 i'm kind of used to waiting for compiles, since i use gentoo... :D
__Ace__ hard to set up?
Ziggy6 i did that once and it was pretty straightforward
ekelund __Ace__: the dist. gcc? 22:03
or gentoo
:)
Ziggy6 distcc on gentoo :)
ekelund hehe :) 22:04
__Ace__ dist gcc
ekelund __Ace__: do you run the puppy linux ?
__Ace__ not atm
ekelund hmm dont think dist gcc is hard if you know what you do. A person used it to compile gentoo for my old 233 by adding his desktop and laptop to distcc :) instead of a week of compiling it took one day I guess 22:05
belive not think :)
__Ace__ nothing is hard if you know what you do 22:07
ekelund :) 22:08
or you use xcode on a mac and it will find nodes by it self over bonjour. if you config it that way. Never tried but if I remember it right xcode works like that 22:09
allbery_b ...assuming you haven't killed off the distcc stuff like I did because every so often it eats the machine 22:10
ekelund allbery_b: ;)
allbery_b (there is nobody else on my internal netwok to use it, it was just sitting there chewing CPU all by itself)
__Ace__ well, Im not compiling KDE so.. I skip that for now
ekelund allbery_b: its a lost not using the Hz :) 22:11
__Ace__: :)
allbery_b enh. at home there's no point as there's only the mini (and rarely the iBook); on campus, well, I'd rather randoms didn't borrow my iBook on me :) 22:12
ekelund :) 22:13
__Ace__: see the privmsg 22:25
__Ace__ I still cant answer in pms
ekelund bah
__Ace__: see another channel we both are in 22:26
allbery_b register with nickserv. not like they charge for it...
__Ace__ and if I forgot my pass? 22:33
ekelund hmm www.youtube.com/watch?v=1Uk9va7IzaI&NR 22:35
allbery_b I think that's where you use /stats p and talk to an ircop 22:42
__Ace__ ...still compiling ghc... 22:44
obra ghc is good that way 22:45
Limbic_Region since ghc bootstraps itself it compiles twice (like gcc). 22:46
22:47 weinig joined 22:50 autark joined 22:54 Odin-LAP joined 23:15 mncharity joined 23:16 mncharity is now known as puttter, puttter is now known as putter
putter Did the POPL keynote end up being about Perl 6? Did it leave any artifacts? (video, slides, blog, etc) 23:19
lambdabot putter: You have 2 new messages. '/msg lambdabot @messages' to read them.
putter luqui: indeed, fglock. I hope to capture the knowledge of how people can help with kp6, but that second interview hasn't happened yet. 23:24
23:26 Psyche^ joined 23:28 thepler_ joined, dduncan joined 23:38 pbuetow joined 23:42 Psyche^ is now known as Patterner
putter <log mark> 23:46
What is the state of the OO AST?
Not of migrating pugs to it, but the ast specification itself?
Largely done and expected to survive pugs implementation unchanged? A rough sketch awaiting implementation to be fleshed out? 23:47
How close is it to the current pugs ast? 23:48
Largely backwards compatible? A major rewrite?
It is no doubt at least partially my own bias, but the OO AST node keeps ending up as the leftmost node on my project dependency diagram. 23:50
Thanks.
lambdabot: ?tell gaal Any thoughts on colabti.de/irclogger/irclogger_log/...mp;sel=557 ? tnx. 23:51
lambdabot Consider it noted.
putter lambdabot: ?tell audreyt Any thoughts on colabti.de/irclogger/irclogger_log/...mp;sel=557 ? tnx.
lambdabot Consider it noted.