pugscode.org | Beware of `make install` and Pugs, it will do bad things to your system
Set by Aankhen`` on 22 May 2007.
spinclad ?(wait with bated breath|ask the odds) 00:00
lambdabot Unknown command, try @list
spinclad ?eval ?(wait with bated breath|ask the odds)
evalbot_r16587 Error: No such subroutine: &wait
dduncan so I just made more p6l posts ... feedback welcome ... particularly there 00:10
svnbot6 r16590 | Darren_Duncan++ | ext/QDRDBMS/ : relicensed QDRDBMS under the GPL version 3 or later rather than version 2 or later ... but that takes effect after June 29th 2007; meanwhile, QDRDBMS is covered by an expiring simple permissive proprietary license 02:20
rhr ?eval my @a = [1,2],[3,4]; [~] [,] [Z] @a 03:02
evalbot_r16590 "1 32 4"
rhr should that be "1324"?
TimToady yes, but Z is still implemented the old-fashioned way 03:03
rhr ?eval my @a = [1,2],[3,4]; [~] [~]« [Z] @a
TimToady it's returning [1,3],[2,4] rather than 1,3,2,4
evalbot_r16590 Error: ␤Unexpected "\171"␤expecting dot, ":", "(", term postfix, operator or ","
rhr oh, it only does that in @@ context? 03:04
TimToady supposed to flatten in @ context, but doesn't yet
List of Capture is supposed to be sensitive to @ vs @@ 03:05
rhr what context do reduce ops provide?
TimToady and a number of routines are redefined to return List of Capture, such as map
and gather
ordinary list context, which is @ context 03:06
rhr ok
rhr does [,] flatten a multidim list? 03:07
TimToady [,] is a no-op 03:08
rhr that makes sense :) I was looking at S03 and couldn't figure out why you'd use it 03:09
TimToady but yes, it probably turns 1,2;3,4 into 1,2,3,4
I'm actually still not entirely happy with the @@ sigil... 03:10
for @@: 1..3 X 'a'..'z' is pretty darn ugly
rhr yeah, that's a little strange 03:11
TimToady for clumps 1..3 X 'a'..'z' is slightly better
for some 'clumps' or other
rhr groups? lists? 03:12
TimToady I considered "each" at one point
for each gather ...
rhr is each entirely gone now? 03:13
TimToady it doesn't quite read right though
we still have the conjectural 1 < each(@foo) < 10
as an inside-out grep
but I think a readable @@ is likely more important
it would be doubly nice if it composed into a metaoperator 03:14
so you could write for 1..3 [X] 'a'..'z'
except [] is probably not going to work there.
can't use [] as a prefix operator 03:15
since it would be mistaken for an empty array
and oddly, most of the other keys on the keyboard already mean something... 03:16
rhr :)
japhb Clearly, we need foot pedals 03:17
TimToady so for now I'm sticking with making it a funny kind of list context, and using @@: as the placeholder for that...
could make some word an alias for @@: 03:18
dduncan so I have finally put a QDRDBMS on CPAN ... but because the code isn't done yet, I made v0.0.0 a POD-only release
should be easier reading it htmlized than in pod
still, that's a step up from where I was before
does anyone know these days how often search.cpan.org is updated? 03:19
the newest thing on it now was added 2 hours ago 03:20
TimToady well, today's a Blue Moon, so I guess "once in a Blue Moon" is about how often. 03:21
japhb Cue oldies music ... 03:22
japhb is getting sufficiently annoyed at perl5 that my favorite part of 6-on-6 may be actually being able to understand the guts of the system, thus having a chance to know how to work around it when it is being stupid. 03:23
TimToady: if and when you get around to thinking about NCI ... one of those cognitive things I humbly ask you keep in mind when designing the syntax is not to make really cheap and really expensive operations look too similar. 03:26
TimToady not sure how to tell them apart, considering most of the overhead is external 03:27
Aankhen`` doesn't really understand @@ anyway.
dduncan as far as I know, @@ is just about multi-part keys
or indexes
japhb I've had quite a bit of frustration figuring out what macros/functions in the XS API just fiddle a bit here and there and which ones decide to make an entire new copy of a huge PV
dduncan same thing in a generic sense
TimToady yes, it's more about keys than values
Aankhen`` I see. 03:28
NCI == Native Call Interface?
TimToady multidimensional slices only have to go as deep as List of List, and that's what semicolon creates, and @@ unpacks
dduncan often
TimToady I keep trying to read it as Norweigan Cruise Lines... 03:29
japhb TimToady: I mean "expensive operations within the API itself". Like what I mentioned above -- it's not always clear when you are copying data and when you are slinging pointers. And MHO is that it comes from extremely huffmanized macro names that then don't give enough info about their actual operation.
dduncan that's NCL 03:30
japhb Especially stuff where deep differences in meaning are stored in capitalization ....
TimToady NCI would be glued by Perl 6 code, not by C code
japhb Aankhen``: yes, Native Call Interface
Aankhen`` Sorry for harping on this, but: for @@: 1..3 X 'a'..'z' # what does this do? 1..3 X 'a' .. 'z' creates three lists of a-z?
japhb: Thanks.
TimToady part of the reason for having an optional type system is the ability to specify the interface from within Perl 03:31
japhb TimToady: nodnod, I just meant, when designing the API, please consider making it somewhat self-documenting, not just golfer-friendly
TimToady ?eval 1..3 X 'a' .. 'c'
evalbot_r16590 ((1, "a"), (1, "b"), (1, "c"), (2, "a"), (2, "b"), (2, "c"), (3, "a"), (3, "b"), (3, "c"))
Aankhen`` Oh, hum.
Aankhen`` goes to read the Synopsis again. 03:32
Ah.
TimToady S03:1172 03:33
Aankhen`` Yeh, found it.
Hence the "Ah".
TimToady Ah
Aankhen`` -ha!
japhb Oh, and while I'm asking for things ... please can we have all NCI-related design in one S doc, as opposed to spread out over so many docs as it is now, so searching becomes a massive pain? ;-) 03:34
Aankhen`` Wait, you want it in one doc so searching becomes a massive pain? :-(
Aankhen`` ducks.
japhb s/so searching becomes/causing searching to be/
TimToady No, he wants the entire design of Perl 6 in one doc...
Aankhen`` That's easy, just use `cat`.
japhb TimToady: Thbbbpt. I know you know what I mean.
:-)
perlxs/perlembed/perlguts/... drive me insane 03:35
TimToady not really, the whole point of NCI is to do away with XS baggage
japhb steps back and thinks ....
Hmmm, perhaps the problem is a lack of imagination on my part. 03:36
Aankhen`` waits for the explosion.
japhb I literally can't imagine a situation in which a simple NCI design can possibly contain the complexity that exists in the real world. Mind you, I *can* see an 80% solution. 03:37
Or even a 99% solution.
But what I'm talking about is the other 1%. 03:38
Because of the waterbed problem
Simplifying how we treat NCI is going to be great,
Aankhen`` Hmm, waterbed problem?
japhb but essentially we're pushing down hard on a big lump in the waterbed.
And the other parts of the bed *will* go up.
Aankhen`` Ah. 03:39
TimToady only if it's full of water. XS has a lot of gas...
japhb OK, I'll buy that.
Aankhen`` Does that mean it's going to smell really bad?
japhb Except ....
Aankhen`` Or maybe cause people to pass out?
japhb When you compress a gas, you increase the pressure.
wolverian Aankhen``, yes, I think XS fits that description. 03:39
japhb So if you've got a weak point, you're going to have a blowout.
Regardless,
Aankhen`` wolverian: I figured as much. ;-) 03:40
japhb getting away from the pure metaphor:
Aankhen`` Never touched XS myself.
japhb There will be things that require deep fiddling. Like requesting zero-copy operations.
(just for instance)
japhb Hopefully these fiddly bits will have an API of their own. 03:40
Let's please not make the fiddly bits as ugly as XS. 03:41
Aankhen`` Come to the daxs side, japhb.
japhb anyway, that was really all I was trying to say. Still don't know if I was clear .... 03:42
Aankhen`` We'll find out eventually!
It'll probably be too late then, but at least we'll know.
japhb LOL
Aankhen`` Have no fear. 03:43
japhb ph34r
TimToady for the sounds of the daxs side, see www.daxo.de/
lambdabot Title: start
Aankhen`` That's alright, you can have all the ph34r you want.
Just not fear.
japhb heh
japhb goes back to trying to figure out how to keep his perl5 process from ballooning up an order of magnitude bigger than all of the readily apparent live variables add up to .... 03:46
Aankhen`` TimToady: What... is that?
daxim japhb, <search.cpan.org/dist/Devel-Leak>? 03:47
japhb "This little utility module was part of Tk until the variable renaming in perl5.005 made it clear that Tk had no business knowing this much about the perl internals." 03:51
heh
Aankhen`` japhb: LOL. 04:01
gaal spinclad: thanks, but still orz. shift-insert works as long as I don't change screen window, I forgot to mention. 05:11
meppl good morning 06:05
dduncan good morning 06:30
Tene Sure is. 06:33
thoughtpolice i suppose :p 06:42
svnbot6 r16591 | renormalist++ | - cperl-mode: token/regex (indenting+highlighting) 07:15
mj41 Hi, are there any disadvantages of PHP's hashs as arrays. I know that internals are different, but from users point of view? Any url? 07:29
wolverian mj41, PHP doesn't have arrays. 07:33
imagine perl didn't have arrays.. and fear :) 07:35
dduncan, I see you uploaded QDRDBMS, yay, looks good. 07:36
dduncan yes 07:37
it took a full year too, since the last Rosetta release, which it rewrites
wolverian btw, I think s/rationales/rationale/ is more accepted
dduncan are you talking to me? 07:38
wolverian yeah, sorry.
that's in QDRDBMS.pm 07:39
dduncan I wasn't sure if you were refering to a #perl6 thread
I'll look into what you mentioned ...
so "providing rationales" should end in rationale? 07:40
sure, I can do that
wolverian r^Hyes.
s/r//
dduncan okay, the next commit will have that fix 07:41
I also changed a similar statement in Language.pod 07:42
thank you
wolverian you're welcome
mj41 hmm, not real arrays at all? what about performance? 07:44
wolverian I don't know about the implementation, but the semantics are those of hashes 07:45
mj41 I found spyced.blogspot.com/2005/06/why-php-sucks.html ... O(1) vs. O(lg n) -> O(n) 07:50
lambdabot Title: Spyced: Why PHP sucks: Jonathan's Python Blog
wolverian mj41, right. there's really no way around that. 07:53
(except, I suppose, special casing integer-keyed hashes, but that is a heuristic, and heuristics break.)
dduncan generally speaking, I find that posting something on cpan leads me to re scanning the now htmlized documentation, and I notice small mistakes that I didn't before 07:58
what you pointed out I wouldn't have noticed, but I did find a runoff sentence further below
er, a sentence without an ending
wolverian dduncan, "define the balance" is awkward. I don't know what it means. (QDRDBMS.pm) 07:59
dduncan looking ...
found it 08:00
wolverian also, s/saavy/savvy/
a bit later: "... nor have objects in separate threads try to access..." would that be better with "trying"? I can't quote a rule why it's wrong, it just sounds wrong as it is :) 08:02
dduncan its basically saying that the database API is partly in the list of Perl routines provided by QDRDBMS.pm and AST.pm, and is partly in the data you define with the ASTs
using an analogy of another DBMS ...
say SQLite for example
part of its API is the C functions you invoke, and part of its API is expressed in the SQL strings you pass in 08:03
wolverian right. I get it now
dduncan in this case, my SQL replacement is the main api, and the perl classes provide the balance
okay
wolverian I think it is more opaque because that's the first time you speak about ::AST
dduncan looking up savvy
wolverian but I'm just nitpicking, really 08:04
dduncan savvy is now fixed 08:05
that's fine ... its good to know if something even initially confuses someone 08:06
of course, like with reporting bugs, it can also help if an alternative is suggested
for now I won't change it
pending such a suggestion
fyi, if you're looking through all the files, I suggest visiting in this order: README, QDRDBMS.pm, AST.pm, Language.pod, SeeAlso.pod, whatever 08:07
Language is the longest 08:08
wolverian s,prominant,prominent, 08:10
(same file)
s,langugaes,languages, in Language.pod 08:12
Aankhen`` s,wolverian,lolcat, 08:13
Har har.
I kill me.
wolverian I has speeling aids! 08:14
Aankhen`` ROTFL.
I CAN HAZ :STDIO PLZ
wolverian the : is brilliant 08:15
Aankhen`` Oh, wait. s/PLZ/?/
And s/ \?/?/
thoughtpolice KTHXBYE
Aankhen`` lolcode++
thoughtpolice definately 08:18
wolverian definitely 08:19
thoughtpolice a lolcode implementation would be fun to write :)
Aankhen`` Heh. 08:20
gaal there are a few already
I think there's one in python on lolcode.com
thoughtpolice yeah
there're several
.net, python, perl, there's a BNF
Aankhen`` Perl 6: Because we need a platform to build language interpreters upon. =) 08:21
dduncan you know, I suppose if I actually ran a spell-checker on my docs it would catch some things
thoughtpolice one guy is actually working on getting it to run on the JVM
which would be pretty awesome imo
Aankhen`` LMAO, that would be funny.
dduncan but I usually count on having good spelling in my brain, or if I'm not sure about something, i google it at time of writing
rgs AWSUM THX <- at least, a polite language
Aankhen`` From Java to lolcode.
thoughtpolice :)
if lolcode's turing complete: why not?!
Tene Would be nice to have a "Build your own language interpreter on Perl 6!" document. 08:22
thoughtpolice hah.
Aankhen`` Since it looks like eric256 has abandoned the BASIC implementation, I was thinking of working on it.
thoughtpolice just take the haskell route; make it a DSL with a LOLmonad like dons showed :) 08:23
gaal yeah 08:24
thoughtpolice I would probably try to write an implementation in perl 6 or haskell, but I probably would suck at writing a grammar or somesuch. :/ I could always have fun with Parsec, though. 08:27
Aankhen`` <thoughtpolice> just take the haskell route; make it a DSL with a LOLmonad like dons showed :) # wha? 08:34
I suck at writing grammars too, and I don't even know BASIC. I figure I'll learn as I go along. :-)
thoughtpolice :)
maybe I should try this as a little haskell side project. :) 08:35
gaal looks like it has a pretty straightforward ast 08:36
dduncan so I'm now running a spell-checker 09:02
gaal nothingmuch: ping 09:39
svnbot6 r16592 | Darren_Duncan++ | ext/QDRDBMS/ : various spelling and other small fixes in 7 pod files 11:08
DrSlump hiho 11:20
DrSlump is there a "push" operator? like @a=qw/a b c/; @b=qw/d e f/; @a ,= @b wich would be @a = (@a, @b) so @a = 'a', 'b', 'c', 'd', 'e', 'f'??? 11:34
or the same as push @a, @b 11:36
moritz DrSlump: I think you can use a <== notation or something
moritz DrSlump: but I'm not sure about it 11:37
DrSlump moritz: oh ok, i'm having a look about it in the archives 11:38
wasnt it <~? 11:40
moritz S03:1418
moritz "Feed operators" 11:40
<<== is the appending form unless I'm mistaken here 11:41
DrSlump well, looking at dev.perl.org/perl6/doc/design/syn/S03.html , it looks like it for me too 11:43
lambdabot Title: Synopsis 3: Perl 6 Operators - perl6:
renormalist do feed operators work in Pugs yet? Last time I checked it seemed not. 2 weeks ago 11:48
moritz ?eval my @a = <a b>; @a <<== <c d> 12:52
evalbot_r16592 Error: ␤Unexpected "<<=="␤expecting operator
moritz ?eval my @a <== <a b>
moritz pokes evalbot_r16592 12:53
svnbot6 r16593 | moritz++ | implemented problem97.t 13:17
thoughtpolice mornin' 14:26
moritz hi thoughtpolice ;)
nearly tea time ;) 14:27
thoughtpolice :)
here it's only 9:30am 14:28
and I haven't slept since like, yesterday at 5:00pm or so. so if for some reason during today I'm talking or something and I suddenly stop I probably walked off to do something and fell asleep without meaning to. 14:29
moritz ;-)
thoughtpolice it's happened before. pretty annoying to be honest. 14:32
svnbot6 r16594 | daxim++ | implemented problem46.t 15:06
TimToady ?eval ?eval ?eval "6" 15:30
moritz is there a canonical way how a true value stringifies? in ~?$expression for example 15:32
?eval ~?5
TimToady: did you read the perlmonks thread on linguistic influences on perl? 15:33
TimToady I think evalbot is still trying to figure out what my @a <== <a b>
means
moritz ;)
www.perlmonks.org/?node_id=618673 that is
TimToady I wonder if it parsed it as <= =<a b>
lambdabot Title: The Germanic language form
TimToady no, I usually bl before pm 15:34
moritz bl? 15:35
TimToady backlog 15:36
moritz ok
svnbot6 r16595 | andara++ | [runpugs] - Settings now go in $HOME/.webtermrc (using Config::General). 16:06
r16595 | andara++ | Look at webtermrc for example settings.
r16595 | andara++ | - Prompt detection now more robust.
r16595 | andara++ | - Easier back-end testing with command-line REPLs (see /bin).
svnbot6 r16596 | audreyt++ | * Pugs.Internals: Massive cleanup of the import/re-export directives 16:18
r16596 | audreyt++ | for better readability/maintainability; no semantic changes.
r16597 | audreyt++ | * Pugs.Prim: More concise, warning-free and point-free definition of perlReplicate.
r16598 | audreyt++ | * Pugs.hs: Restructure a series of boolean tests to be more lazy
japhb audreyt: Have you done the O(1) pad and MO merges yet? Or are they still pending? 16:34
TimToady O(1) pad is I think in, MO is not 16:43
if not, O(1) is close
gaal > concat $ replicate 5 " <3 audreyt! }:)" 17:45
lambdabot " <3 audreyt! }:) <3 audreyt! }:) <3 audreyt! }:) <3 audreyt! }:) <3 audreyt...
daxim cute 17:48
gaal hee, r16596 is like renaming the project to XModule 17:53
re: perlReplicate, too bad Integer hasn't got infinity 17:54
then we could easily do what Chas was looking at 17:55
w/o a cast try to VNum 17:56
ugh, isn't there monadic golf to be played for the "<- and return if ok" pattern in 16598? 17:59
(can't think of any, but the pattern looks ugly now)
lumi Wow, didn't know you could do that (as X) 18:12
svnbot6 r16599 | fglock++ | kp6-Perl6-roadmap.txt - added "development tracks" 18:14
gaal for some reason, this kills the build on my snapshot 6.7 :( (doesn't like the # in unsafeCoerce#) 18:17
trying to work around
the clever thing isn't the as X, it's the overlap in X :) 18:22
lumi That's what I meant, I didn't know you could do the overlap 18:30
lumi But I guess it's only natural, normally the overlap is on epsilon 18:31
svnbot6 r16600 | gaal++ | * Unbreak the build 18:33
moritz gaal: what a cool commit message ;-)
lumi gaal: I have a monadic golf for that pattern, but it's actually one line longer :P
Well, it's 4 lines of setup, and then a punchline 18:34
gaal CPS? :-p
lumi Where do we nopaste nowadays?
gaal moritz: it's a pretty common one if you check the logs :)
lumi nopaste?
Er
@paste
gaal perlbot nopaste
lambdabot Haskell pastebin: hpaste.org/new
perlbot Paste your code at sial.org/pbot/perl and #perl will be able to view it
lambdabot Title: sial.org Pastebot - pasteling 18:35
lumi hpaste is down
pasteling "lumi" at 192.115.131.248 pasted "Monadic pseudogolf" (7 lines, 246B) at sial.org/pbot/25306
moritz gaal: nevertheless it's cool ;)
renormalist gaal: does the commit msg mean I can do make install again? 18:36
lumi renormalist: The build, not the install...
gaal it means you can make again. :-)
lumi Err, it's actually borkened I think, but you get the gist 18:37
The 'a x' should be just 'a'
gaal lumi: needa Just
lumi Right, and tryWith should return (b, Just $ f x) 18:38
gaal uh, the /topic is FUDful. make install does need fixing, but that means someone should actually fix it.
you don't Just the result of the last application 18:39
lumi pugscode.org | Brave souls needed to fix 'make Install'! Non-brave souls should avoid it for now 18:40
lumi Better? 18:40
gaal lumi++
moritz lumi++ # indeed!
gaal what happened to everything else that used to be on the topic, anyway?
like, paste, moose, etc.
lumi gaal: Which last application? 18:41
gaal findHelper xs is not Justified 18:42
lumi gaal: That's recursion, I don't think it wants a Just? 18:43
gaal yes, what I'm saying
lumi I don't follow
moritz gaal: Juerd++ swept the topic away when he announced that feather was up again ;) 18:44
or so I think I recall
gaal lumi: ok, sorry, my mind has melted in the last heat wave and hasn't really reformed correctly
off for a film &
lumi I'm not sure if it works, would depend on the associativity of orElse 18:46
gaal (also prolly better find another name since orElse is an stm thing. re&) 18:49
renormalist does anyone know what's wrong with the make install or can describe it a bit?
Is there a chance to fix it or do we need special skills? Which ones? 18:50
rindolf Hi all.
Hi gaal
rindolf gaal: can you get a testimonial from the previous workplace of yours (the one with the net monitoring) about Perl? 18:50
Tene rindolf: I'm curious about what you're asking for and why. 18:52
rindolf Tene: I'm looking for short Perl testimonials.
Tene: preferably with longer ones.
offby1 I testify: perl is kewl
can't beat that for shortness
Tene rindolf: what for? 18:53
rindolf Tene: for perl-begin.org/ 18:54
lumi offby1: Perl rocks!
lambdabot Title: Perl Beginners' Site
rindolf Every good technology site should have testimonials. 18:55
Tene: have you read www.sparkthis.com/2006/02/slides_the_hack.html ?
lambdabot Title: Spark This: Slides: The Hacker's Guide to Marketing
Tene nope
offby1 lumi: OK, OK; you win 18:56
lumi gaal: How about `failingThat` ?
Oh, he's gone
rindolf Tene: you should - it's not very long. 18:57
Tene I will once I get internet access again. :) 18:58
moritz Tene: I want to be able to chat in IRC without internet access too ;)
Tene using my cellphone. incompetent ISP at the office. 18:59
lumi What's another way of saying "or" or "or else"?
rindolf Tene: oh.
Tene: firewall?
lumi: otherwise?
lumi: nor?
lumi Yes, but that mean "true" in Haskell
moritz aye
Tene: did't think of that, I don't have a cellphone ;) 19:00
rindolf Hi moritz
moritz: what's up?
moritz hi rindolf ;)
Tene rindolf: net access completely down. they screwed up our routing and brainwiped their equipment here.
lumi Eh, maybe I can steal otherwise for a coupla lines
moritz rindolf: hacked my first two lines of pugs haskell code the other day, and I'm still proud of it ;)
rindolf moritz: OK.
moritz: the Pugs Haskell code scares me.
I once tried to look for something there. 19:01
expert Haskell is pretty scary.
I like to use recursion and declaration a lot in my Haskell.
But people tend to use the built-ins.
moritz rindolf: I'm far form understanding that... I just fixed one builtin
rindolf moritz: Haskell built-ins. 19:02
moritz: like zip, etc.
There are tons of htem.
You can find them on zvon.org
moritz rindolf: funny enough I fixed p6's zip ;)
Tene moritz: specifically, cellphone+ssh+screen
rindolf moritz: I'm beginning to think I'm psychic. 19:03
moritz: I also called my friend just when he was thinking about me. 19:04
moritz: to help someone on #perl with an Oracle problem.
Who wrote god-damn awful Perl code.
pasteling "lumi" at 192.115.131.248 pasted "Latest attempt -- tell me if this can be improved (or if it just sucks)" (7 lines, 336B) at sial.org/pbot/25307
lumi rindolf: Wait, did your friend write awful Perl code, or did the someone, or did Oracle? 19:05
Or the problem
moritz perl code stored in an oracle db?
rindolf lumi: no it was the one with the problem. 19:06
lumi Running a wacky ecore, possibly
rindolf lumi: my friend is kind-of still experienced, but he's very intelligent and bright.
s/experienced/in\0/
svnbot6 r16601 | fglock++ | kp6-Perl6-roadmap.txt - high level steps 19:09
lumi @tell gaal Circularity problem with Meta.List :/ Meta.Class defines an instance for Boxable a => Boxable [a], and also needs it locally. 19:25
lambdabot Consider it noted.
svnbot6 r16602 | audreyt++ | * Pugs.Monads: Add MonadIO instance to the MaybeT monad. 19:27
r16602 | audreyt++ | Also simplify the `mplus` operation so the second action
r16602 | audreyt++ | is only run when the first operation returns Nothing.
r16602 | audreyt++ | (It used to run both and then needlessly discards the second.)
r16603 | audreyt++ | * Pugs.hs: Solve the monadic guard riddle with MaybeT.
lumi Aw 19:28
It's certainly a more elegant solution 19:34
svnbot6 r16604 | audreyt++ | * Pugs.hs: Further solve the monadic guard riddle by avoiding
r16604 | audreyt++ | an unneccessary boolean-test step.
r16605 | audreyt++ | * unobe++ changed %*CONFIG<perl5path> into %*CONFIG<perl5_path>, 19:41
r16605 | audreyt++ | so we need to chase that in order for pugs -CPerl5 etc to work.
rindolf audreyt: here? 19:45
TimToady: ping.
TimToady pong
or pung, if you prefer mah jong 19:47
moritz
.oO( "no pung intended" )Oo.
 19:49
fglock english question - is pugs "Perl6-in-Haskell" or "Perl6-on-Haskell"? 20:15
PerlJam fglock: it's haskell wearing a perl6 cloak 20:17
kolibrie fglock: Perl6-(written-)in-Haskell or Perl6-(running-)on-Haskell 20:30
I would go for the written/implemented in Haskell name 20:31
moritz yes, that's better because if it has a different backend, it's not "on haskell" any more - in some ways 20:32
fglock so "6-in-6" is the compiler, and "6-on-6" is the runtime?
kolibrie sounds about right 20:33
moritz 6-$foo-6 ;)
you could always use junctions 20:34
6-any<in on>-6 ;)
lumi 6-{<in on>any>-6 20:38
Erk
6-{<in on>.any}-6
moritz <in on>.pick
fglock 6..*
PerlJam pick 6? 20:39
moritz <6 6 6 6 6 6 6 6>.pick
svnbot6 r16606 | lwall++ | [STD.pm] Allow $obj.@candidates form. 21:14
moritz @karma lwall 21:15
lambdabot lwall has a karma of 82
TimToady that guy's head is getting too big for his britches...er, wait... 21:16
PerlJam What's his head doing in his britches to begin with?!?!?
diakopter reminds me of something my grandmother often said - 'my eyes were too big for my stomach.' 21:17
moritz I'd think two eyes fit easily into a stomach ;)
diakopter: we have that saying in German as well
diakopter (on her failure to eat her apportioned selections from Luby's buffet) 21:18
lumi We have a similar expression 21:20
japhb moritz: For the topic change right after moritz.faui2k3.org/irclog/out.pl?ch...01#id_l496 , the topic doesn't actually appear. And this time I checked View Source to make sure it actually was missing from the HTML. ;-) 21:36
lambdabot Title: IRC log for #perl6, 2007-06-01, tinyurl.com/yv284k
moritz japhb: you're right... investigating... 21:37
Aankhen`` for 1 .. 10 -> $x { our &::("foo" . $x) := regex { $x } }; # is this a valid way to create a bunch of regexes dynamically? Will it close over $x correctly? 21:40
TimToady well, you need ~ rather than . 21:41
Aankhen`` Er, yes, sorry.
Stupid brain is still half stuck in P5. 21:42
TimToady other than that I don't see why it shouldn't work
Aankhen`` Kewl.
moritz pugscode.org/ | Brave souls needed to fix 'make Install'! Non-brave souls should avoid it for now
Aankhen`` And since grammars can export rules, IIRC, is this also valid: for 1 .. 10 -> $x { our &::("foo" . $x) is export(:DEFAULT) := regex { $x } }; 21:42
Ugh, again with a s/ \. / ~ / 21:43
svnbot6 r16607 | moritz++ | [irclog] now topic changes should be reported correctly, japhb++ 21:50
japhb Thx, moritz++ 21:54
renormalist does the smart-match ~~ work for every type of arguments or are there argument of any kind that are forbidden to match with ~~ ? 22:11
PerlJam renormalist: nothing is forbidden in perl : ) 22:13
er, :-)
renormalist PerlJam: I know. I just have a look over an article and don't whether it's ok to say smartmatch works for every argument. 22:14
s/don't/don't know/
japhb Well, it should "work", but that doesn't mean it will actually be "smart" and DWYM. 22:15
moritz ?eval class A {} class B {} A.new ~~ B.new
PerlJam renormalist: the last section of S05 is entitled "Matching against non-strings" and says, in part, "Anything that can be tied to a string can be matched against a regex."
renormalist: and S03 has big table that tells you how smart matching works among the various standard things 22:17
(in a section entitled "Smart matching" oddly enough) 22:18
I'd be more definitive, but the information contained in the Synopses has changed since I last read them :) 22:19
renormalist I see. So I will use the term "practically everything" (not sure whether "practically" is used the same way as it is in german) 22:21
Although "practically everything" can practically mean everything. :-)
moritz renormalist: "quasi" is a nice word ;)
renormalist: in both english and german ;) 22:22
renormalist moritz: naaaaaaja :-)
renormalist TimToady: What is the official state of the STD grammar. Is "nearly complete" a correct term in an article? I want to correct the term "a large part of Perl6 is already defined...". That sounds too lame IMHO. 23:46
the "large part" term is from someone else, not me :-)
(damn, did I miss the timezone shift again?) 23:47
japhb renormalist: It is just before 5 PM in this timezone (US west coast) 23:49
[particle] renormalist: as timtoady has said before, "we're done with the first 80%, and well into the second 80%" 23:53
renormalist [particle]: then I leave it "a large part" :-) 23:54

Powered by Cro and the Raku® Programming Language.

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

Thank you!