The Return of the Journal : pugs.blogs.com/ | pugscode.org | pugs.kwiki.org | paste: paste.lisp.org/new/perl6 or sial.org/pbot/perl6
Set by GammaRay on 31 December 2005.
00:28 aufrank joined 00:44 lampus joined 01:48 Amnesiac joined 01:49 shreddr joined
shreddr must go TODAY. MESSAGE ME ONLY ON MSN AT [email@hidden.address] AIM AT OGD443 or YAHOO at MCSLTD2 IF INTERESTED! 1 alienware desktop computer price $550, one alienware area51-m 5700 notebook price $550. prices include sameday shipping, case, wireless router. 01:53
01:58 binary42 is now known as binary42____ 01:59 binary42____ is now known as binary42 02:31 stevan_ joined
\xe6var ugh 03:05
I thought
if $look_mommy >= $no_parentheses { ... 03:06
was possible because you could no longer do
if %hash {whitespace_between_the_key_and_they_hash_name} { ...
so stuff like
if %hash {"am I "} {}.. 03:07
\xe6var pokes
I seem to be right;) 03:08
$ pugs -e 'my %h = (1 => 0); if %h {1} { say "foo" }'
foo
%h<> I mean 03:11
beppu ? 03:24
\xe6var indeed 03:26
04:28 stevan_ joined, stevan_ is now known as stevan 04:32 lypanov joined
meppl guten morgen 04:42
rep moin
meppl good morning rep 04:43
05:23 sapper joined 06:22 scw joined 06:24 lypanov joined 06:38 sapper left 06:52 elmex joined, ManChicken joined 06:53 Debolaz joined
ManChicken Anybody got a good reference for this stuff? 06:54
I know it's not production material yet, but I'm curious as curious can be.
Been playing with it for a few hours, and things are a bit different.
07:07 cognominal joined 07:10 Debolaz joined 07:36 xinming joined 07:37 perlbot joined 07:40 ManChicken left 07:45 azuroth joined 07:49 iblechbot joined 07:58 azuroth joined 08:39 marmic joined 08:40 nnunley joined 09:07 elmex joined
ingy seen audreyt 09:43
jabbot ingy: audreyt was seen 14 hours 41 minutes 9 seconds ago
09:46 G2 joined 09:50 bsb joined 09:53 scook0 joined 11:08 iblechbot joined 11:15 r0nny joined, chris2 joined
svnbot6 r8765 | putter++ | Parse/Rule/Core.pm - completed r8456 (the renaming of $.match_num/name to $.capture_num/name). 11:33
11:40 arcady joined 11:45 gugod_ joined 11:48 kane_ joined
svnbot6 r8766 | gaal++ | my.t - marking tests :todo<feature> 11:55
11:55 gugod joined 11:56 gugod joined 12:02 \xe6var joined
audreyt hm, I hope I don't get used to those 16-hours sleeps 12:53
because my entire body -- especially stomach -- feels terrible after waking up 12:54
I'll bbiab...
13:01 putter joined
Alias_ heh 13:02
know what you mean
putter notes more 7-hour sleeps => fewer 16-hour sleeps. ;)
audreyt how very true ;) 13:04
putter *chortles*: Oooh, audreyt is well rested!!! We're going to romp! 6.2.11, here we come! :) (took a moment to sink in) 13:05
;)
audreyt I wouldn't say "well-rested" -- "somewhat less zombie-like" is more like it 13:06
but I need to fix svk first as the svn bindings broke with perl 5.8.8 13:07
mmm yaks
s/5.8.8/5.8.8rc1/
putter lol. putter pictures "night of the living dead" zombie hackathon...
audreyt hackoween 13:08
so, I think I owe myself a really good 9pm breakfast
I'll see ya :) & 13:09
putter :) &
oh drats. putter kills a smoke. forgot to set PATH to include parrot. external!&#@!@#$ parrot. ;) anyone sure if the pugs build does/doesnt happen normally in the absence of parrot? putter contemplates rebuilding. 13:10
pasteling "bsb" at 84.177.43.70 pasted "Number.hs refactor" (45 lines, 2.2K) at sial.org/pbot/15537 13:11
bsb Hi all
putter hi bsb
putter looks at paste
bsb I need someone with more haskell clues verify my patch 13:12
I'm mainly unsure about what should be Int and Integer
I haven't tested it, compiling 13:13
putter "I haven't tested it, compiling" ;) 13:23
bsb It takes so looong 13:24
Done now. No damage to t/syntax/parsing/numbers.t at least 13:25
putter you likely know by now, but the functions are variously misnamed, eg b36DigitToInteger and b36DigitToInt
re looong, make unoptimized is your friend :) 13:26
at least until the file compiles. then its a friend which has overstayed its welcome. ;) 13:27
note that multiple '_'s are speced as allowed. but maybe that's a todo for after the refactor. 13:29
bsb Thanks. 13:31
putter err, no. never mind. my misnaming comment was bogus.
bsb There's a test in numbers.t: isnt eval('1__0'), 10, "Multiple embedded underscores fail"; 13:32
putter is puzzled. thought an excess of '_' was speced as ok. exploring...
Juerd I haven't seen any spec of that 13:33
That usually indicates Perl 5 is leading, which would mean 1__0 == 10
But there may be a spec somewhere.
bsb S02: A single underscore is allowed only between any two digits in a literal number, ... 13:34
Juerd Does it say why?
bsb No
Juerd I'll p6l it 13:35
putter S02 is clear enough... 13:37
Juerd It's clear, but I want to know WHY the change is made. 13:38
putter my perl5 is warning if there are multiple or trailing _ 13:39
so it looks like it's already in the process of p5 depreciation
bsb A02: Underscores will be allowed between any two digits within a number.
Ambiguous plural there
putter bsb: right. I was mistaken. No ambiguity. Only single '_'.
As to "WHY", it looks like that's more of a p5 question than a p6 one ;) 13:40
Juerd putter: Ah, the warning is enough for me.
bsb: No, not ambiguous, that plural, as there may be multiple underscoreS, just not without digits in between. 13:41
putter re ambiguous plural, the S02 at dev.perl.org says A single underscore is allowed only between any two digits in a literal number 13:46
('twas what I meant by unambiguous;) 13:47
Juerd: actually, the A02 phrasing _is_ ambiguous. it may well have been saying _'s are not allowed trailing or next to the decimal or "e". with no commitment on the number of _'s which can occur properly placed with digits on both ends. 13:51
Juerd We were talking about different things, but it does not matter.
putter yeah. breakfast sounds like a plan. (9am) (pugs-time == GMT-5?:) 13:56
Juerd pugstime == GMT -12 + int rand 24 14:00
14:20 Khisanth joined, dolmans joined 14:22 sapper joined 14:28 dolmans left
svnbot6 r8767 | tewk++ | t/builtins/undef.t - TODO crashing test due to missing functionality 14:47
14:54 rantanplan_ joined 15:26 Limbic_Region joined 15:37 feng joined 15:41 penk joined 15:43 penk joined 15:46 Debolaz joined 15:47 penk joined 15:54 ManChicken joined
ManChicken Is there a comprehensive collection of built-in perl6 functions? (e.g. perlfunc) 16:02
I'm having a doosey of a time finding one.
Alias_ I'm not sure they actually are built-ins at the moment
ManChicken I know this isn't production-ready or done yet, but I'd really like to play with things.
Alias_ But rather default macros that map to methods 16:03
or something like that...
ManChicken It seems like much of perl5 built-ins are still available. Is this the case?
Alias_ The style should feel somewhat similar
ManChicken It looks like Perl, just with everything I've ever had to hack out already thought of. I love it. 16:04
All it needs now is cpp macros ;)
Alias_ You know it does have macros right?
ManChicken Oh? 16:05
Alias_ Basically, a lexically sane equivalent of the source filter hacks of Perl 5
ManChicken I heard macros, but I though of the VB definition of a macro...
Alias_ I'm afraid I don't know the difference
ManChicken VBA macro is just a function that's immediately visible in a MS Office application 16:06
Alias_ oh..
ManChicken IIRC
Never played with it much.
Alias_ A Perl 6 macro is a lex-tree transformation
(although my terminology is always suspect)
16:07 Amnesiac joined
ManChicken Pardon my ignorance, but what does that mean exactly? I'm but a humble hacker. 16:07
Alias_ I believe it's vaguely equivalent to C-ish macros
But you'd want confirmation from someone that actually can keep up
:)
ManChicken Nice.
Alias_ I just hang out here because if I don't I'll never be able to keep up at all... I work higher level that the language usually 16:08
ManChicken heh 16:09
Alias_ s/that/than/
audreyt got some nice japanese food... and then am promptly dozing off again
ManChicken Ooh.
Debolaz Eww, asian food.
audreyt but I want to hack *pout*
ManChicken I'm about to go get greasy McDonalds... because I like cholesterol...
Alias_ audreyt: BTW, mst is reporting more problems with M:I
ManChicken Okay, gonna run, be back shortly to hack.... 16:10
Alias_ audreyt: I think basically if you're CPAN is sub-whatever-version and M:I wants to do the dep installs itself, and CPANPLUS isn't available, it freaks
your
audreyt Alias_: okay, bounce it to my mailbox? I _think_ my body is telling me it wants more sleep before I make it hack
Alias_ he's still analysing
audreyt Alias_: okay, then maybe we inject a minimum version check -- or bundle a copy of CPAN.pm (unthinkable?)
Alias_ yes, unthinkable
audreyt or fix it Correctly using he analysis
I'd love to hear about it tomorrow
Alias_ But I think we're heading towards forcing a CPAN.pm upgrade :/ 16:11
Dunno if we can avoid it yet
audreyt we'll see.
Alias_ indeed
Any reason M:I can't use CPAN.pm for the upgrades instead of CPANPLUS?
Or did you just go for the nicer API :)
audreyt the latter 16:12
also because I use cpanplus muself
also because it's the new core module etc
Alias_ nods
Not yet it isn't
The fight continues
audreyt :D
I'll be back later after a variable-length nap.
Alias_ Looking at 5.8.8 it would appear M:B isn't even in yet either
audreyt waves &
Alias_ later 16:13
putter & 16:17
16:21 planrantan joined
Limbic_Region the repository still down? 16:22
nope
;-)
16:30 bsb joined
gaal say, why does util/gen_prelude.pl:_BypassPreludePC use an IORef Bool? isn't a Bool enough? 17:19
17:35 DaGo joined
svnbot6 r8768 | bsb++ | r202@rex: bsb | 2006-01-21 18:30:45 +0100 17:35
r8768 | bsb++ | Merging number parsing changes back to main local
bsb Obviously I don't understand svk 17:40
ingy studies haskell 18:09
Alias_ hands ingy the paracetemol
ingy Alias_: what's that? 18:11
Alias_ um... pain tablets?
Asparin?
(sp)
18:11 elmex joined
ManChicken Asprin* 18:13
^_^
This perl6 thing is so nifty. 18:14
Makes me want to try ADA....
But I don't know why.
lypanov ada is teh funk 18:15
18:33 lypanov__ joined 18:48 lypanov_ joined
ManChicken Does "use" work the same way in Perl6? 18:52
18:53 iblechbot joined
ingy can someone help me write a little haskell function? 19:08
lypanov_ i suck at haskell, but i wouldn't mind learning? ;)
ingy ok 19:09
I have this function called fibs
Prelude> let fibs = 0 : 1 : [ a + b | (a, b) <- zip fibs (tail fibs)]
Juerd ManChicken: Nothing works exactly the same way in Perl 6 19:10
ManChicken: It's the philosophy of Perl that makes Perl 6 be still Perl, not the actual language. 19:11
ingy I want to write a function that returns a list of all the values between fibs 2000 and fibs 2100 that don't have the digits '20' in them
Juerd ingy: Really?
I can think of more fun things to do ;)
ingy well really I want the values that don't match /20\d\d/ 19:12
Juerd: yeah, but can you do that?
Juerd No, I don't speak Haskell
Years ago, I said that a good programmer can easily pick up any programming language. 19:13
But either I was wrong, or I'm not a good programmer myself.
ingy bites her tongue :P 19:14
Juerd Because I find it very hard to write useful stuff in non-imperative languages.
Don't bite your tongue :)
Don't bite anything but food.
ManChicken Juerd: Well, that makes it no fun to play with. 19:15
Juerd: Because there are no references yet :( 19:16
Juerd ManChicken: Huh? Perl 6 has references. Don't listen to dragonchild :)
ManChicken No no no no, reference materials
heh
as in documentation
gaal ingy: you can use drop and take
Juerd ManChicken: There certainly is some. 19:17
ManChicken I LONG to play in perl6, but I can't find any refs.
Juerd ManChicken: Have a look in pugs' docs tree, and at the synopses.
ingy gaal: how do I pattern patch that though?
ManChicken Juerd: I've looked at the synopses, they're a little on the vague side. I need something a bit more explicit (I'm on the dense side).
I'll check out the pugs doc tree... 19:18
Juerd ManChicken: Can you give an example of what you found vague?
ManChicken Didn't think to look there yet.
Well, primarily the code samples embedded in the documentation.
gaal ingy: oh i didn't see the end of your question
ingy: so lets start with fibs 2000 to 2100
ManChicken They don't really go to the same level of detail that perl docs have in the past.
ingy follows along at home... 19:19
Juerd ManChicken: That's because knowledge of Perl 5 is assumed, and most things map 1:1 to Perl 5.
ManChicken Oh, really?
Well, hell, I can work with that.
Juerd ManChicken: Note that these documents are meant for Perl 6 developers, not Perl 5 users :)
ManChicken Yeah...
gaal let intetresting = take 101 drop 1999 fibs
ManChicken But I'm impatient. I want to play with the new toy now.
^_^
gaal modulo off-by-ones :-)
Juerd ManChicken: It takes some getting used to, but a Perl 6 user can certainly benefit from these documents.
ManChicken: Have you installed Pugs yet? It comes with a great bunch of example code. 19:20
ManChicken Yeah.
I'm using the win32 port.
gaal then you want to treat each value as a string, so
Juerd Yuch.
ManChicken Don't have a unix box available for my personal use yet, and I doubt my office would appreciate me installing it on there.
gaal map show interesting
Juerd You shouldn't use Win32.
bsb filter 19:21
Juerd ManChicken: You can get free access to feather.perl6.nl, a Linux box that runs pugs.
ManChicken I also may be switching jobs soon....
bsb instead of map, yeah?
ManChicken Juerd: What now?
gaal then you want to filter on those that match
ManChicken Juerd: Why would anybody give out free shell access?
Juerd ManChicken: Because it speeds up development.
ManChicken Oh.
Juerd ManChicken: If you're able to try out Perl 6, you can find bugs and report them.
ingy gaal: you mean: take 100 (drop 1999 fibs) 19:22
gaal ingy: do you have parsec where you're at?
ingy gaal: no
gaal ingy: yes, or even: take 100 $ drop 1999 fibs
ingy gaal: right, thanks
Juerd ManChicken: Given the dynamics of Perl, it's impossible to design a 100% perfect language, and corner cases have to be found using actual implementations.
gaal is it /20\d\d/, or /^20\d\d/?
ingy gaal: the former
gaal I think this might work (we can assume every char is a digit): 19:23
wanted '2':'0':x:y:_ = True
wanted _ = False
errr
wanted _:'2':'0':x:y:_ = True 19:24
the other version is for /^20\d\d/
ManChicken Juerd: How does one go about gaining access to said shell?
bsb I thought !~ /[20]/
ManChicken The win32 port is a nightly build.... 19:25
Juerd ManChicken: You send an email requesting access to [email@hidden.address] and include your realname and preferred username.
ManChicken is it not?
ingy bsb: correct
gaal oh, I missed the "don't". Sorry :)
ingy = False
gaal reverse True and False then.
ingy gaal: can I somehow use that as a guard on a list comprehension?
Juerd ManChicken: Don't forget to ask audreyt for a committer bit, if you want to report bugs (reporting bugs is done by adding unit tests) 19:26
gaal ingy: probably :)
ingy does it need to be a lambda thingy?
gaal should get more comfy with those
ManChicken Juerd: Is it really a bad idea to use Jon Worthington's automated win32 build?
gaal ingy: lambda is just anonysub
ingy right
and \ is the syntax? 19:27
ManChicken The current version is from yesterday...
gaal \args ->
ManChicken I've heard that parrot really isn't ready for much playing yet.
gaal like p6 $args -> { }
ingy cool. that's all the stuffs I need to play around
contrary to Juerd, I think this is a nifty little problem :) 19:28
Juerd ManChicken: It's a bad idea to use win32, in my opinion.
ManChicken Is activestate putting any hand into the perl6 pot? 19:29
Juerd ManChicken: It's probably okay to use the win32 port if you know how to hack it.
ManChicken: Not yet, as far as I know of.
ManChicken Well, if I notice anything odd, I'd try it out on a BSD build first.
Juerd ManChicken: Whatever makes you feel comfortable.
ManChicken Or I could try using cygwin. 19:30
ingy ManChicken: as a former employee of ActiveState I can pretty much say that it will be a long time until they get involved
ManChicken I just don't feel like playing with remote development.
ingy: Why? They've got a lot they could gain from perl6.
ingy ManChicken: A) there's no $$$ in it. B) they are not really doing anything new 19:31
ManChicken weird...
C:\Documents and Settings\Michael\My Documents\Devel\perl6>pugs coin.pl
*** No such method in class Num: "&srand"
at coin.pl line 38, column 1-14
heh
That comes from: srand(time());
ingy ActiveState is a (still breathing and cashflow positive) remnant of the dotcom era 19:32
gaal haskell is really good at "connect the pipes" kinds of tasks
ManChicken What a shame.
I use their stuff all the time.
My company even bought a few licenses to their perl IDE.
Nobody uses them, but they're paid for...
Everybody still prefers VIM
heh
And those IDEs aren't cheap 19:33
ingy ManChicken: they are still maintaining their current products because they still bring some revenue
ManChicken Well, I'm glad to see the community is picking up where they left off with perl6
I use perl on win32 daily.
I'd like to use perl6 on win32 daily.
ingy I think Perl 6/ pugs doesn't need any help from AS 19:34
ManChicken My boss says I'm too much of a coder to use windows, yet he still prefers that I do.
heh
ingy: Good. Although ActiveState's perldocs were very nice. I'll miss that.
gaal ManChicken: pugs works OK on win32. do you want to make it better? I have a few ideas but no tuits for that.
ManChicken Especially their win32 API docs.
ingy ManChicken: nod 19:35
ManChicken tuits?
heh
gaal ManChicken: start by building an MSYS perl 5.8
ManChicken As I've said before, I'm but a humble perl hacker.
Hmm...
Never ported anything before....
gaal then you can compile a pugs with embedded parrot and perl5! 19:36
ManChicken Hell, I don't even know how to get MSYS running right. Every time I try, it screws up on me.
gaal so far no haskell :)
ingy Juerd, gaal: I can't msg on freenode :( 19:37
<-- to lazy to register
anyway, I am sometimes female on irc 19:38
ingy strokes her beard...
ManChicken If someone wants to walk me through getting MSYS installed, I'm more than happy to play with getting a build there. 19:39
I can't promise results, but I'm willing to dive in for a day and see if it's worth the time.
heh
ingy takes ManChicken for a long walk on $short_pier
ManChicken heh
gaal ManChicken: starting's not too bad, you install MSYS and the "developer toolkit" 19:40
I ran out of tuits trying to build a modern perl
(perl5 that is)
I think I should have tried a different make program.
Alias_ ManChicken: You could always have a shot at making PXPerl work? 19:42
gaal Alias_: doesn't it work already?
Alias_ PXPerl is SO close to being done
Only for pure Perl
gaal oh :(
Alias_ You have to hook up a compiler yourself, which isn't trivial
I've been waiting for someone to do it and send me a screencast of it :) 19:43
Because compilers are a bit beyond me
gaal "screencast"?
Alias_ buzzword of the year :)
theorbtwo gaal: Use vnc2swf to record what you're doing on screen. Then send it to somebody else. 19:44
Alias_ Go look it up
gaal ah
Alias_ Except with things like Wink you can do back and edit it
gaal "script" for a gui
Alias_ And put in callout boxes and explainations
Juerd ingy: I see. What do you prefer to be viewed as? 19:45
ingy Juerd: you mean Man or Chicken?
<-- Human Paraquat 19:46
19:46 feng123 joined
ingy (The Big Lebowski)++ 19:46
gaal: how do I run haskell commands like a Perl script 19:47
I want to 'run' a .hs file
with ghc
integral runghc 19:48
ManChicken Okay, installing mingw runtime and all that.
I'll give it a shot.
Alias_ If you get it working, please screencast it and post it to the PXPerl forums 19:49
I will be eternally greatful
gaal ingy: #!/bin/cpr errr, oh wait no
ingy :p
integral: thanky
gaal ingy: alternatively have two shells, one with an editor and one with ghci. :r reloads the file you're editing 19:51
dinner &
integral gaal's suggestion is usually much faster too
ManChicken Ooh, this is nice. 19:53
I get a real shell.
bsb ok ingy 19:54
filter (\x -> null(show x `intersect` "20") ) $ take 100 [1..]
I forgot "show" is toString
ManChicken Gotta grab gcc and all that. Can someone show me a place to grab instructions on how to grab the source? 19:58
Alias_ The webside has some of the docs 19:59
But half the problem is knowing what exactly to download
ManChicken Wow, these binaries are nice. 20:03
gaal ManChicken: there should be a "DTK" there I think 20:04
ManChicken yeah 20:05
libtool & such
ingy gaal: gracias
ManChicken they have an exe...
wonder if it's an installer
gaal ingy: de nada 20:06
ManChicken It's installing perl already 20:09
I thought that's what I was going to try to build...
Or am I building PXPerl?
gaal ManChicken: perl -v will prolly tell you 5.6.1 20:13
ManChicken yeah
So I'm building 587?
gaal yup
the reason this is desirable is that GHC on win32 was built with msys 20:14
ManChicken This is the first time I've gotten msys to work. 20:17
It's nice.
gaal i hate its default shell :)
ManChicken I prefer zsh myself...
I'll build from source shortly.
gaal feels like being on hpux ten years ago :)
ManChicken Feels like redhat
Alias_ Again, if you get a full Perl install working, please document it. k thnks bye 20:18
ingy Alias_: it's spelled 'thx' fyi
Alias_ omg wtf
ManChicken heh 20:19
Alias_ purl, omg poem
ENOPURL!
ingy chuckles
ManChicken I'm not sure if I'll be able to get one working, but I can sure give it the old college try... even though I never finished college.
Alias_ ManChicken: That's fine, others have also failed
But ActivePerl is really asking to be replaced at this point
So someone's gunna do it soon 20:20
ingy watches ManChicken amusedly
gaal Alias_: if all you want is to replace activeperl... build your own perl with ms vc toolkit 2003
I did that successfully
Alias_ gaal: Yes, but ActivePerl has the key advantage there of being an .msi you can download off the intarweb 20:21
ingy something like, mortals rush in where Mongers fear to tread...
Alias_ I want an .msi that lets you installed compiling modules direct from CPAN
install
gaal Alias_: building perl w/msys won't get you that
Alias_ Basically, what we have on linux or everywhere else
ingy ManChicken++
Alias_ gaal: Why not?
gaal won't hurt either :)
Alias_ You'd have a compiler available, and set up so perl can see it 20:22
gaal that, yes.
Alias_ And then it's up to someone to work out how to package the setup
ManChicken Alias_: I wouldn't mind helping you in that goal.
Alias_ ManChicken: PXPerl is the main effort towards that end
But the author went AWOL due to some student election
ManChicken Well, couldn't you staticly link it and distribute it that way? 20:23
Alias_ shrugs
I dunno about these compilery things
ManChicken heh
Alias_ I just know I want something like ActivePerl that installs straight from CPAN
And can do XS modules
That's all I ask
ManChicken Yeah.
ActiveState can do that.
Alias_ No it can't
ManChicken perl -MCPAN -e shell works though.
Alias_ nope
ManChicken Well, they've got the huge lib of binaries you can install. 20:24
gaal yep it does
ManChicken But in msys you can perl -MCPAN -e shell
Alias_ ManChicken: A huge library of 40% broken, mostly out of date, binaries
ManChicken heh
That I won't argue against.
Alias_ That we rely on their good graces to maintain
ManChicken Although they do have a 587 release
Alias_ Anyways, with a standard base platform that COULD compile from CPAN, at least we'd have something to test on 20:25
And then maybe actually make more modules Win32 compatible
Unlike now, where the test cycle involves waiting for PPM to catch up 20:26
ingy ActiveState has a very solid methodology for building ActivePerl
building the modules, much less so
Alias_ right
The dual modules problem is a killer too
Anything that uses Scalar::Util is dead atm
recursively
ingy why S::U? 20:27
Alias_ I talked to the new PPM maintainer, apparently their build farm has trouble recognising dual modules
So while the core one is ok, the non-core version dies at compile time
And the dependencies link back to the non-core one, so the build of THOSE shortcuts
(to fail)
And thus recursively outwards 20:28
So you have the current situation where PPI is at 0.990 in the PPM lib, and dead
ppm.activestate.com/BuildStatus/5.8-P.html 20:30
Look how much red is on that page
Also annoyingly, the only platform not quite so broken is Mac, the one their own developers use 20:31
ManChicken grrr 20:34
screwed up my msys build already
Alias_ :)
ManChicken The separation between mingw and msys is too freakin' complicated.
I install gdb and lo and behold, there is no gdb... 20:35
I'll try building from source 20:36
Juerd ingy: I'll just use "it", okay? :) 20:41
ManChicken I wonder if I can build from distribution source.
WOW! msys GCC is 3.4.2 20:42
Latest is 4.0.2 20:43
ingy Juerd: that or "thoust" 20:44
thine, etc
ManChicken Man, GNU's mirrors are slow... 20:46
Been a while since I played on one.
I'm gonna have to build the entire environment before I can even look at perl 587 20:49
What a pain.
Oh well... it's not like I have anything better to do today. heh
Alias_ heh 20:50
The tricky bit will just be packaging the .msi :)
ManChicken Oh, you want me to package it?
Wow, I don't even know where to begin on that.
heh
First things first, getting it to bloody compile.
Alias_ Well, that would be the "Payble in a vertical metre of beer" award one 20:51
ManChicken But before that even, GCC needs an updating.
Alias_: I'll consider that a promise ;)
Alias_ done
ManChicken As long as it's not budweiser.
:(
Alias_ Vertical in the sense of stacked crates/cartons/etc
ManChicken Newcastle might be nice though.
Alias_ And of your choice, as long as it's not insanely over the top premium
ManChicken heh 20:52
Wow, GCC takes a while to decompress...
heh
How dare a 30MB archive take a while to decompress!
heh
It'll be a long build.
Man... the GCC dist comes with EVERYTHING. g77, ada, java, objc, etc. 20:59
heh
msys distributes it all separately.
ingy gaal: ping 21:12
is _ greedy in haskell pattern matching?
gaal ingy: pong 21:17
I... don't know
ManChicken So... how long does it take to fully compile GCC on msys win32? heh 21:18
gaal actually it could be that _:'x' implies _ is a char
ingy It seems maybe _:'x' might never match 'x'
ManChicken Answer: For bloody ever.
gaal hmm right, it always matches *something*
ManChicken I'm not even going to get to the 587 code if this keeps up.
ingy I mean it might not match "foooooox" 21:19
gaal now I just don't know if it matches a string in your case or a char
because x:[a] means x has type a
so presumably _:'2':'0':_... no, maybe _ is special-cased here. 21:20
try it or ask people who know, ie #haskell :)
bsb I think _ only matches one thing 21:25
ingy bsb: how do I match [Char] for /20\d\d/ ? 21:26
ie /^.*20\d\d.*$/ 21:27
bsb Not sure. There is a Regex module
ingy I can keep poking at this too.
mmmm Regexp! :p 21:28
bsb Text.Regex
gaal that's cheating :) if you allow that, might as well do it with parsec :)
ingy heh
yeah, I think I should be able to do this sans Regexp 21:29
maybe I should read the Prelude for some tasty bits 21:30
ManChicken Ooh, in perl6, will it be easier to address a string as an array of chars?
bsb matchRegex (mkRegex "^(20[[:digit:]]{2})") "20123"
You probably can do it without a regex, but I'm not Haskell native so wouldn't know 21:31
I keep writing perl in Haskell
ingy :)
ManChicken I still don't quite understand the relationship between haskell and perl. Anybody care to explain? I've got plenty of time to learn while GCC compiles. 21:32
Alias_ Perl 6? 21:33
ok... Haskell is really good at writing parsers and compilers
And some of the features for Perl 6 aligned quite nicely with the Haskell feature set
So after reading a compilers book, audreyt decided to try writing a basic Perl 6 parser in Haskell 21:34
As an experiment
And got a basic expression parser for Perl 6 running in 6 days
And so the pugs stuff developed as a way of taking Perl 6 code and trying it out 21:35
By converting it to Haskell
(details more complicated)
Now, Haskell is a pretty mind bending language, while Perl isn't (in comparison)
ManChicken Fascinating. 21:36
Alias_ So all of a sudden, there's this "standard" type language that compiles to, and can load, haskell modules
And so the Haskell people now have a sort of "scripting language" on top of Haskell
I'm vastly overgeneralising, but that's the basics
ManChicken I doubt they're too upset about that.
I don't doubt that.
But I appreciate you being gentle. ;) 21:37
Alias_ It's been a huge win for both sides
integral well, it can access haskell stuff, but pugs doesn't actually compile to haskell (yet)
s/yet/currently/
ManChicken I've never really been playing around durring something like this.
Alias_ integral: Like I said, vastly overgeneralising :)
So Haskell gets this normal top level language on it, and Perl gets the benefits of the most kick-ass technical base we could ever want 21:38
Because Haskell at a technical level does kick insane amounts of ass
integral oh, well actually it appears the GHC template-haskell backend is still present! But it appears to just be the old one compiling straight off the AST, not via the PILs
ingy I think in the pattern (_:x:y:_) the first '_' matches one char and the 2nd '_' matches all the remaining chars 21:45
like in _:xs and x:_ 21:46
so the constuct for /.*?/ still evades me
integral yes. : is right-associative: ( _: (x: (y: _)) )
ingy but i get to pair with audreyt tomorrow :) 21:47
integral ingy: I'm not sure what you mean, patterns for deconstructing aren't regular, they're only linear
ManChicken This MSYS compile system is poorly packaged.
ingy integral: that is probably why I can't make this work :) 21:48
ManChicken Like, all of the install packages are out of date, or inappropriately packaged.
ingy <- zzzzzzzzzzz &
ManChicken Now I'm having to build from source half of the stuff.
Alias_ yep 21:49
I only had to look at the webpage to see that 21:50
bsb ingy: "Pattern matching" is deceptive isn't it? 21:52
Maybe you could match (a:b:c:d:rest) and then check a == '2', b=='0' etc 21:53
Then recurse with b:c:d:rest
Or ('2':'0':a:b) as the guard 21:55
actually ('2':'0':a:b:rest)
I think .*? is done by matching fn (x:xs) = ... fn xs 21:57
ManChicken Okay, I'm gonna let msys win for now. 22:01
So damned confusing. Can't get even the runtime to build from source. 22:02
Alias_ heh
ManChicken gcc won't build without the runtime.
Alias_ OK then, it may be time to post the competition officially
ManChicken MSYS needs some serious work. 22:03
More than I have the time or patience to do.
22:03 ManChicken left 22:09 chris2 joined, SamB joined 22:12 Odin- joined
wolverian hm, this java project of mine is 10% of the java size in perl6. 22:34
with no loss of typing or features 22:35
integral how does performance compare at the moment? 22:36
wolverian I don't know, I don't expect pugs to actually run it
it's heavily type annotated and uses roles
s,run,compile, 22:37
integral ooh, that sounds good. It's nice to hear that the type stuff is grokkable and usable 22:38
wolverian I'm not sure about the syntax, but it feels right :) 22:40
Foo of Bar is the same as Foo[Bar], right? 22:45
22:48 vel joined 22:57 vel left 23:40 Shachaf joined 23:48 bsb 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!