6.2.10 released! xrl.us/hxnb | geoffb's column (/. ed): xrl.us/hxhk | pugscode.org | pugs.kwiki.org | paste: paste.lisp.org/new/perl6 or sial.org/pbot/perl6 | www.geeksunite.net
Set by integral on 25 October 2005.
rafl theorbtwo: ;-) 00:33
theorbtwo: It's kind of usual in germany though. night == Nacht; eight == acht; N8 == Nacht. 00:34
theorbtwo Oh. I suppose that makes sense... in German. 00:36
rafl Well, not even there, imho. 00:37
theorbtwo It's stupid still, but at least it makes sense, which is something.
luqui Holásteis! 02:45
scook0 err...hi 02:49
nice 'role' post, btw
theorbtwo A happy all-holows-eve to you too.
luqui thanks
luqui is going to be whatshisface from Clockwork Orange 02:50
theorbtwo has nothing planned, sadly.
02:54 justathe1ry is now known as justatheory
buu All holows, eh? 03:19
luqui wow... www.securetactics.com/SpySoftware 03:20
commercial trojans
nothingmuch Alex deLarge methinks 04:00
he's a very froody droog
autrijus *purr*
nothingmuch remembers why he doesn't play computer games 04:01
but it's too late again ;-)
autrijus why?
nothingmuch addiction problems
nothingmuch loses track of time
and gets nothing done
but my brain needs the rest
i've been on the brink of burning myself out for weeks now
autrijus do remember that this will pass, and get enough rest 04:02
nothingmuch aye
nothingmuch is constantly letting his people die in Rainbow Six 3
which is a bad thing
autrijus I am _very_ happy that you did not let me introduce you to nethack.
nothingmuch but I cant do any better
ongoing things that can be paused are usually not a big problem 04:03
i tend to play several minutes a day of those
but i'm still afraid of it ;-)
luqui nothingmuch, yes, that's the one
nothingmuch if we would have holloween i would dress up as the korova milk bar to complement 04:04
luqui If my desktop's hard drive were present I would probably be addicted to rag doll kung fu at this point
nothingmuch i'll need some mannequins, and milk, and a high end water cooling system ;-)
luqui on the other hand, the reviews weren't that great
nothingmuch (cpu cooling system)
luqui :-p
nothingmuch rag doll kung fu?
luqui an experimental fighting game that I've been looking forward to 04:05
luqui is a fan of fighting games... or rather would be if any of them were any good
nothingmuch hmm... no mac version?
luqui nein
nothingmuch watches the trailer instead 04:06
that would be very cool with a programming edge in it 04:09
luqui what would you program?
nothingmuch combos 04:10
luqui aa
nothingmuch really likes tekken (3) because of all the juggling
it's a very elegant way to kick ass
luqui always found that annoying, being at the kicked ass side 04:11
nothingmuch ah
luqui I think Soul Calibur is about as good a fighting game as they've made yet
nothingmuch i am always on the ass kicked side, except in tekken and unreal tournament
nothingmuch hasn't played that yet
luqui it's pretty good. But it's still attached to the realtime/buttonmash/energybar paradigm. 04:12
Glorified tekken, as it were. It's just very well balanced.
nothingmuch buttonmashing is not that fun
what I like in tekken is that you sometimes need to wait a second to get something in
if you are slightly less trigger happy than 100% you will win 04:13
scook0 ever tried virtua fighter?
buu You know what's better? Quake.
luqui heh
nothingmuch scook0: my parents didn't let me play violent games at that age ;-)
nothingmuch doesn't like quake
buu FREAK.
nothingmuch it's not as fun aasUT
buu meh
Same thing.
nothingmuch nuh uh
UT has more style
buu Does UT have strafe jumping and such?
luqui I remember when there was just Quake. No Quake 2, no quake 3 arena. When you said "I'm going to go online and frag somebody", you just went online and fragged somebody.
buu heh
nothingmuch you mean strafing in the air? or jumping sideways?
buu I remember playing q2 over a dialup 04:14
nothingmuch: Nope.
luqui (I actually never played Quake... any version. I'm not a big fan of FPSs)
By which I mean I despise them
buu Sucks to be you!
nothingmuch that wasn't a yes or no question, that was multiple choice ;-)
buu (0 || 0) == 0!
nothingmuch uh, your point being?
buu That 'no' means no to both. 04:15
luqui But 0! == 1
:-)
buu luqui: 0! is a syntax error!
nothingmuch ah
so wtf is strafe jumping if it's not one of those?
scook0 also !0 == 0!
buu nothingmuch: Strafe jumping is when you accelerate via jumping
luqui woah
that's scary
nothingmuch ah
buu ?eval 0!
evalbot_7763 Error: unexpected "!" expecting "x", "d", octal digit, "o", "b", digit, fraction, exponent, term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
scook0 those crazy physics exploits?
nothingmuch that's bad game design
buu nothingmuch: Bad.. game.. de.. what? What the hell? How can you even make that claim? 04:16
Obviously the rocket launchers in UT are bad game design, cause they shoot rockets.
I mean, hey, if we don't need logic..
luqui ?eval sub postfix:<!> ($x) { if $x == 0 { 1 } else { $x * ($x-1)! } } 5!
evalbot_7763 Error: unexpected "i" expecting ";", statements or "}" reserved word
luqui qua?
nothingmuch ?eval multi postfix:<!> (0) { 1 }; multi postfix:<!> ($n) { $n * ($n - 1)! }; 5!
evalbot_7763 Error: unexpected "0" expecting formal parameter, ":" or ")"
luqui didn't that used to work?
nothingmuch darn
buu: it's bad game design because it's stupid behavior 04:17
buu nothingmuch: You mean stupider than taking a rocket in the face and surviving?
nothingmuch someone didn't notice the fault in the physics engine and jump parameters
scook0 I'm not sure it was deliberate
buu nothingmuch: Dude, they put it in there deliberately.
theorbtwo nothingmuch: Because it's not like the real world?
buu Southen: It was deliberate after the first patch =]
nothingmuch no, because it's imbalanced
buu oh
nothingmuch it would be nicer if:
buu Obviously only one player can strafe jump.
nothingmuch you could jump around and be slow
or be a sitting duck and go fast
buu Oh wait, every player can strafe jump. Thus making it balanced. 04:18
luqui thinks that you only have bad game design when you make a game that isn't fun, because you've missed the whole point of making a game. For example, Deus Ex 2.
nothingmuch what does deus ex do wrong?
luqui Deus Ex was great. Deus Ex 2 was too caught up in being a movie to be a good game.
buu Deus Ex was so awesome. I've always wanted to play DX2, but everyone hated it =[ 04:19
scook0 interleaves IRC with rounds of counterstrike... 04:22
luqui counterstrike was a FPS that I actually liked pretty well 04:23
scook0 I never got into it in the HL1 days 04:25
nothingmuch oh god: www.boingboing.net/2005/10/28/no_tr..._stra.html
luqui scook0, likewise. I started at CS:S
theorbtwo Well, at least they came up with a reasonable result, even if their reasoning was completely bonkers. 04:26
luqui nothingmuch, is it just me, or did they completely miss the point?
nothingmuch who did? 04:27
luqui the article
about strawberries
the court
nothingmuch yep 04:28
theorbtwo The question was "can the scent of fresh strawberies be trademarked".
Their answer was "no".
Good answer.
nothingmuch i was actually "oh god"ing more at the court than the company
theorbtwo: the sad thing is that the answer was actually 'no, because strawberies have *five different* scents' 04:29
theorbtwo Yes.
nothingmuch on the other hand, grass has one scent, so that's OK
theorbtwo Not such a good reason.
Right.
I hope this is largely poor reporting.
nothingmuch somehow doubts the BBC will make it sound this bad 04:30
i trust the BBC more than I trust the judicinal and political systems that govern stuff that has to do with patents and trademarks and all that crap
stevan luqui: ping 05:14
luqui: I sent you a mail re: theory.pod, I am trying to understand where it fits in with the metamodel/object-space/runtime stuff I have been working on 05:15
I think it does all fit together to a degree, but the code in theory.pod is just so haskellish, so I am unclear of the details 05:16
anyway check your mail,.. I have to sleep now &
xinming seen autrijus 05:50
jabbot xinming: autrijus was seen 1 hours 47 minutes 50 seconds ago
xinming by the way, Is continuation a bit like Rule in perl 6? 06:17
luqui welll... uh 06:22
no
rule execution may be implemented using continuations, but the similarities end there
unless there is a similarity I am not seeing
xinming luqui: thanks, need more reading on haskell tutorial. 06:23
autrijus proceeds to go to the airport (to brussels) 07:09
see you in a few hours... safely in liz's place and ready to hack 07:10
<- almost finished PickAxe2
which fills my mind with weird constructs like
my $x = if 3 then { 4 } else { 5 };
(instead of)
xinming autrijus: did you meet Larry? :-P
autrijus my $x = do if 3 then { 4 } else { 5 };
xinming: sure did
worked on the tuple syntax a bit 07:11
xinming my $x = 3 ?? 4 !! 5 ;
why not use this?
autrijus sure
but still
my $x = given $y { when 3 { 4 } default { 5 } };
(in ruby all statements are expressions)
they don't have a Stmt form, it's all Exp
xinming autrijus: hmm, did Larry consider construct the object using class name in some situation? 07:12
oh,ok
autrijus xinming: I don't know... you mean like
my $dbh = DBI(dsn => $dsn, user => 'foo');
?
xinming yes. 07:13
autrijus I think it makes sense. the "new" method though is then implicit in this case
which may or may not be a good thing
<- coding monkey, not really designer
xinming hmm, sometimes, we don't need an instance to be held. 07:14
autrijus :)
autrijus goes packing some more
xinming class A { has $.a; method show { $.a.say } }; A( a = 5 ).show;
if we use A( a => 5 ).new.show, 07:15
autrijus oh, singleton classes
?
xinming the .new is a bit annonying
if there are many classes like this. :-P 07:16
autrijus I tend to agree :) 07:17
xinming hmm, most times, the small program will only hold only an instance of a class
page( document => "path/to/doc", save_to = "path/to/cache" ).process; cache( file = "path/to/cache" ).show; 07:19
that's what I think useful :-) 07:20
autrijus nodnod... otoh you can do
xinming what does otoh mean? :-S 07:21
scook0 xinming: I can see Larry's point about wanting Classname($foo) to be a cast rather than a constructor though
autrijus on the other hand
class A { sub A is export { call ./new } };
or something like that
xinming hmm, Yes, I know, we can make method to be an constructor... 07:22
autrijus: But, If there are method... This is not handful..
scook0 but if we're going to have Haskell-style data types, we'll need a shorter construction form anyway 07:23
xinming if there are lots of methods.
scook0 A:( a => 5 ).show # woo, colon!
autrijus _everybody_ wants colons! 07:24
scook0 because they're so versatile
xinming hmm, why not make `my A: $o .= new( ... )` ?
autrijus off to airport... bbl :) *wave* &
scook0 and because everything else is already taken
xinming: what are you suggesting? 07:25
xinming and use A( ... ) as a constructor?
for now, we use `my A $o .= new( ... )` to get an instance.
hmm, and I saw Larry replied about my question, 07:26
So, I just suggest using `my A: $o .= new` instead of `my A $o .= new`
scook0 xinming: oh, I didn't realise you hadn't read it
I don't think that was his objection though 07:27
xinming certainly, I won't be able to think as far as you can do, So, Just show my Humble opinion. :-( 07:28
www.nntp.perl.org/group/perl.perl6....uage/23874
?eval class A { has $.a; }; new A( a => 1 ); 07:39
evalbot_7763 Error: cannot cast from VObject (MkObject {objType = (mkType "Class"), objAttrs = <ref>, objOpaque = Nothing, objId = 51}) to Pugs.AST.Internals.VCode (VCode)
xinming the perl 5 style doesn't works. :'( 07:40
anyone here can tell me if `class.new a => 1` would work? IMHO This is a bit like a function call without parenthesis. 07:59
Detonite Morning 08:04
xinming afternoon here. :-P 08:05
Detonite Well the point comes across still I hope 08:06
rafl_ What tests do I need to run to chech if parrot embedding works? 09:03
scook0 xinming: `class.new:a(1)` should work 09:11
`class.new a=>1` won't work because method calls need either parens or adverbs 09:12
xinming scook0: thanks 09:16
scook0 I think someone (Juerd?) was trying to get rid of the method-call-parens restriction, but I don't know if that went anywhere 09:17
xinming hmm, by the way, how about `new class: a(1)` ? 09:18
indirect object syntax can be used in method calls, how about the constructor?
though, It's a special method. 09:19
scook0 I think `new MyClass: a => 1` or `new MyClass: :a(1)` would be the syntax 09:23
`new MyClass: a(1)` would call &a with arg 1 09:24
xinming thinks perl 6 might be the most complicated language in the world... 09:36
Juerd scook0: I want method calls to be possibly parenless
Just like sub calls 09:37
It just takes everything until the end of the expression,
scook0 what was the original motivation for them being different?
xinming so do I, as method is kind of sub...
Juerd That the signature is unknown at compile time, and Perl cannot know how many arguments to take
lc $foo, $bar is actually lc($foo), $bar 09:38
While split $foo, $bar is split($foo, $bar)
scook0 do one-arg subs automatically become prefix-unary ops? 09:41
Juerd It works like that in Perl 5
And isn't specified to change, so yes.
scook0 I just wonder how that interacts with multis 09:43
Juerd The same 09:50
scook0 (there are a lot of things I wonder about multis) 09:51
like being able to pass positional args using named-arg syntax 09:52
for singular subs, it's easy enough 09:53
for multis, it's *possible*, but I doubt it can be done efficiently
r0nny yo 10:09
rafl_: alive ?
rafl_ r0nny: Sure. 10:26
r0nny rafl_: any idea hol long it will take to make your sixpan tools run under gentoo ? 10:29
hol = how
rafl_ They already do, imho. 10:30
Install dpkg and dpkg-deb and go for it.
r0nny hmm 10:32
rafl_: is there a svn repo up allready ?
rafl_ r0nny: kane uses p4. I'll set up an svk mirror soon. 10:34
r0nny p4 ? 10:35
rafl_ perforce.
r0nny sad - its comnerecial
well - going to read the license 10:36
rafl_ Well, you don't need to pay for a usable version.
I also like it, of course.
r0nny atm im a monotone fan 10:37
rafl_ Well, pugs uses svn/svk. I'll check it in its repository. 10:38
r0nny i allready installed p4 10:40
repo url ? 10:41
rafl_ I don't have my laptop handy. I guess p4.elixus.org or something like that.
r0nny argh - wont work :/ 10:43
xerox Howdy. 10:44
rafl_ r0nny: ask kane or look at the wiki.
r0nny: BrainStorming. There's a p4 url somewhere.
r0nny ok
rafl_: cant do a checkout :/ 10:48
rafl_ Ask kane.
r0nny where is he ?
rafl_ [email@hidden.address] or on irc.perl.org sometimes. 10:49
drbean h /server 11:05
r0nny hmm 11:06
hes online, but not answering :/
guess hes busy 11:07
Juerd ARGH 11:41
The term "auto reference" really clouds vision and understanding
It makes things seem MUCH more magical than they are.
An array doesn't "auto reference" "when needed". 11:42
It returns a reference to itself in scalar context, that's all.
And yes, Array context is a form of scalar context
The only thing that's a bit magical is automatic *de*referencing 11:43
THAT has the "when used as" thing. 11:44
"when used as" an array, an arrayref dereferences automatically.
But it's not so that an array "when not used as" an array, automatically references.
Or, alternatively, call +@foo auto-elemsing 11:45
And ~@foo auto-stringification
Which it is NOT. It's not auto, it's explicit because of context.
xerox Do you know in which debian package are uu{encode,decode} ?
Juerd apt-cache search uuencode 11:46
Pick the one beginning with s.
rafl_ apt-file search bin/uuencode
xinming xern: try apt-cache search uuencode 11:51
Juerd xinming: Now there's a novel idea.
xinming sharutils
Juerd xinming: Glad you thought of it.
Otherwise he'd never have found it!
xinming Juerd: :-) what I use apt-cache most time, as it might find some good related softwares. :-) 11:52
Juerd xinming: I think I have to agree.
(is blindness bliss too?) 11:53
xinming apt-cache search uuencode|grep uuencode
aish - ish/base64/uuencoded_file converter
sharutils - shar, unshar, uuencode, uudecode
Juerd You're so smart. 11:54
xinming ... 11:55
Juerd I just wish I had thought of that 11:56
theorbtwo (apt-cache)++
xinming apt-cache is a better tool most time in my opinion, apt-file is better for searching using absolute path for a file. 11:57
Juerd I use both 11:58
But apt-cache much more often
rafl_ as well.
But for the given task (finding a binary) apt-file seems to be a better tool. 11:59
xinming nods.
Juerd rafl_: Well, yes. But uudecode is often used within larger tasks, for which apt-cache might reveal integrated solutions
xinming really likes perl 6. 12:02
?eval "asdfg/" ~~ rx:perl5{^(\w+)?/(\w+)?}; $1 ?? "true".say !! "false".say ; 12:04
evalbot_7763 true bool::true
xinming Is this a bug?
$1 get nothing, But it is true.
s/get/gets/
rafl_ ?eval "asdfg/" ~~ rx:perl5{^(\w+)?/(\w+)?}; $1.perl
evalbot_7763 "\\MkMatch \{matchOk = True, matchFrom = 0, matchTo = 0, matchStr = \"\", matchSubPos = [], matchSubNamed = \{}}"
xinming I know there is a property can be set using `0 but true` 12:05
But, Why here is true?
rafl_ ?eval "asdfg/" ~~ rx:perl5{^(\w+)?/(\w+)?}; "$1" 12:06
evalbot_7763 ""
rafl_ I think it's a bug.
xinming hmm, I don't know which to choose. :-)
we can say that it is a bug, because null string should be false. 12:07
rafl_ ?eval ""
evalbot_7763 ""
rafl_ ?eval ?""
evalbot_7763 bool::false
xinming But, if we need the match status of the match, this might be right behaivour
rafl_ Well, write a test in t/pugsbugs/
xinming Ok, >_< 12:08
theorbtwo Write a test when you don't know what the correct behavior is?
rafl_ Hm, good point. 12:10
xinming I think, This might be better to ask on the mailing-list. 12:12
Juerd ?eval "foobar" =~ rx:perl5/(foo)(bar)/; "--$0--$1--"
evalbot_7763 Error: Can't modify constant item: VStr "foobar"
Juerd ?eval "foobar" ~~ rx:perl5/(foo)(bar)/; "--$0--$1--" 12:13
evalbot_7763 "--foo--bar--"
Juerd Simple case works.
Re (\w+)? not matching anything: its variable should be false 12:14
Not because it matched a null string, because it cannot match a null string.
But because it didn't match, because of the + condition. 12:15
xinming so many because... >_< 12:26
the test is written,
after the bug is fixed, Where should this test be?
Ok, I think I know, rules 12:27
theorbtwo rules/p5, or something?
svnbot6 r7764 | yiyihu++ | add a bug test for numbered match variable.
xinming maybe 12:28
I wonder, what will `` be used for in perl 6?
still the same as perl 5? 12:29
scook0 xinming: I'm not sure, but I think `` will be reserved for the user 12:32
(for P5 ``, use q:x//)
xinming Larry might think he used too much... :-P leave a ascii for user to play with 12:37
theorbtwo qx isn't useful enough to take a rare ASCII character for. 12:38
scook0 bah, all Perl 6 users should just go out and buy APL keyboards :-P 12:39