Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2007-08-24

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:20 jhorwitz joined #perl6
00:25 weinig_ joined #perl6
00:26 daxim_ joined #perl6
00:31 nipotaway changed the nick to nipotan
00:40 BinGOs joined #perl6
00:43 pugs_svnbot r17406 | lwall++ | [metholate] switch to using $/ for current in-flight match
00:43 pugs_svnbot diff: http://dev.pugscode.org/changeset/17406
00:43 lambdabot Title: Changeset 17406 - Pugs - Trac
01:02 araujo joined #perl6
01:10 rindolf joined #perl6
01:24 mncharity joined #perl6
01:26 ofer joined #perl6
01:32 weinig_ joined #perl6
01:39 sunnavy joined #perl6
01:45 BinGOs joined #perl6
02:01 lyokato joined #perl6
02:12 justatheory joined #perl6
02:17 weinig_ joined #perl6
02:20 REPLeffect joined #perl6
02:38 c9s joined #perl6
02:43 diakopter and perl6 loses its cents
02:47 |Jedai| joined #perl6
02:49 diakopter er, no.  TimToady: what's the distinction between dollar-cents and dollar-backslash
02:50 diakopter <sigh> dollar-slash.
02:53 pugs_svnbot r17407 | rhr++ | [STD] fill out is_type a little
02:53 pugs_svnbot diff: http://dev.pugscode.org/changeset/17407
02:53 lambdabot Title: Changeset 17407 - Pugs - Trac
03:05 elmex_ joined #perl6
03:16 hkb_zz_ joined #perl6
03:24 dduncan joined #perl6
03:33 TimToady $/ is the current match, though it still doesn't actually work right for accessing its contents, but will eventually.  $¢ is basically just $/.to
03:43 diotalevi joined #perl6
03:47 pugs_svnbot r17408 | rhr++ | [STD] all the typenames I could find in S02
03:47 pugs_svnbot diff: http://dev.pugscode.org/changeset/17408
03:47 lambdabot Title: Changeset 17408 - Pugs - Trac
03:49 TimToady mostly is_type just wants to see if there's a package of that name in ::*, I suspect.
03:50 TimToady (or in the current namespace)
03:52 rhr how does one do that?
03:57 TimToady it's just a SMOHP...
03:58 rhr HP?
03:58 TimToady Haskell Programming
03:58 TimToady I suppose you could eval"$name" and see if it complains about barewords
03:59 snearch joined #perl6
03:59 TimToady though presumably that's just depending on whatever pugs's idea of is_type is...
04:00 rhr wouldn't that find subs etc. also?
04:00 TimToady well, you'd have to look at the return type, I suppose.
04:01 TimToady but it'd certainly be better if pugs actually implemented symbol tables as user-accessible hashes...
04:01 TimToady maybe there's a way to get at the info
04:01 TimToady I just know GLOBAL::{$name} doesn't work yet
04:02 rhr ah, I thought there was supposed to be some kind of hash...
04:03 rhr should is_type handle things like Int|Str?
04:03 TimToady probably not
04:04 TimToady that's just a constraint, not a real type
04:04 TimToady is_type is really for named types
04:05 rhr what about e.g. unit23?  will all of those be pregenerated in GLOBAL::?
04:05 TimToady unit23?
04:05 TimToady oh, uint
04:06 rhr yes, sorry
04:06 TimToady um, some will likely be pregenerated, I suppose
04:06 TimToady but probably not uint23
04:06 rhr but it's supposed to take any integer there, right?
04:07 diakopter heh
04:07 TimToady we only guarantee powers of two
04:07 diakopter (sorry)  just imagining really large integers
04:07 TimToady and only insofar as "supported" by the architecture, for some def or other
04:09 rhr ok.  I'll leave the current implementation in place for now, and add a commented-out GLOBAL:: impl awaiting haskell magic
04:10 TimToady some of this may straighten itself out eventually when we get MO in pugs
04:11 TimToady we've hardwired some unicode stuff too, which is fine for now
04:11 TimToady but we do need to check for user-defined types there at some point
04:12 rhr right, but that would have to go in the "action" part of STD, no?
04:12 rhr which doesn't exist yet
04:12 TimToady *nod*
04:13 TimToady except insofar as we're currently assuming pugs is good enough, whether it is or ain't.  :)
04:14 rhr :)
04:15 TimToady next I need to write something that runs inside parens and rules that runs down the .prior links and builds a real match object, somehow
04:15 pugs_svnbot r17409 | rhr++ | [STD] is_type should be a little different in the future
04:15 pugs_svnbot diff: http://dev.pugscode.org/changeset/17409
04:15 lambdabot Title: Changeset 17409 - Pugs - Trac
04:15 TimToady but pugs is not friendly toward people who want user-defined Match or Capture objects, sigh
04:16 TimToady for now about all I can get is $<foo> and $<0> out of a Hash object
04:17 rhr did you notice that at some point parsing 42+1 began failing with 'pugs: unexpected node: <Scalar:0x2b59c57e9348>'?  Not sure what the problem is or when it stopped working.
04:18 TimToady I broke it when I switched to $/
04:18 TimToady needs fixin
04:18 rhr ok
04:18 TimToady but it seemed obvious to me that $/ needed to be the real in-flight transactional object
04:19 TimToady or we were going to be doing your copy-in-copy-out trick all over
04:19 rhr yes, that makes sense.  afk a few mins &
04:25 BinGOs_ joined #perl6
04:31 rhr re: hardwired, I should get cracking on the unicode stuff again one of these days...
04:57 diakopter joined #perl6
05:30 jisom joined #perl6
05:39 laniz_ joined #perl6
05:50 Aankhen`` joined #perl6
05:55 Aankhen`` joined #perl6
06:36 sunnavy joined #perl6
06:51 BinGOs_ changed the nick to BinGOs
07:00 f0rth joined #perl6
07:08 dduncan left #perl6
07:11 pmurias joined #perl6
07:12 jisom_ joined #perl6
07:15 pmurias why does capture.t use &foo.callwith(|$capture) instead of foo(|$capture)
07:15 pmurias ?
07:40 franck___ joined #perl6
08:04 pugs_svnbot r17410 | pmurias++ | kp6:
08:04 pugs_svnbot r17410 | pmurias++ | keys,values implemented in perl6
08:04 pugs_svnbot r17410 | pmurias++ | testing join and map
08:04 pugs_svnbot r17410 | pmurias++ | do {} around Bind
08:04 pugs_svnbot diff: http://dev.pugscode.org/changeset/17410
08:04 lambdabot Title: Changeset 17410 - Pugs - Trac
08:32 iblechbot joined #perl6
08:42 snearch joined #perl6
08:46 poletti sorry, just testing moritz_'s logger: http://www.polettix.it/ciao, link with comma
08:47 laniz_ is his code somewhere?
08:47 poletti don't know about the code, but the log is here: http://moritz.faui2k3.org/i​rclog/out.pl?channel=perl6
08:48 lambdabot Title: IRC log for #perl6, 2007-08-24
08:48 wolverian hm the vertical alignment is still off
08:48 wolverian the code is in pugs repo I think
08:48 poletti I was only ensuring that the trailing comma was kept in the auto-linkified uri
08:48 laniz_ cool
08:49 wolverian http://moritz.faui2k3.org/en/ilbot
08:49 lambdabot Title: ilbot - Irc log bot in Perl
08:49 wolverian it uses Net::IRC, though, so someone please port it to PoCoIRC :)
08:50 wolverian moritz_, by the way, http://moritz.faui2k3.org/en/ilbot claims to be xhtml valid, but is not.
08:50 lambdabot Title: ilbot - Irc log bot in Perl
08:50 laniz_ thanks wolverian!
08:51 laniz_ maybe silly to ask, but that code wont do its magic with Perl 5.8, right?
08:52 wolverian hmm? it's perl5 code afaik
08:53 laniz_ then its a perfect example to learn from :D
08:53 * laniz_ runs away with the code being all happy
08:54 wolverian except it uses Net::IRC ..
08:55 laniz_ thats why its great to learn from for me. POE is too complex for a start
08:55 wolverian well, sure, just be aware that Net::IRC is a buggy piece fo crap :)
08:55 laniz_ after all the decades of its existence??
08:56 wolverian quite.
08:56 wolverian 11:56       purl : net::irc is at http://www.funkplanet.com/irc/ or always good for a laugh or a scare, depending on if you're reading the comments or the code (= or see also  POE::Component::IRC or not supported any more and dennis tells people to use PCI. or an abomination and a crime against nature.
08:56 wolverian purl knows best :)
08:57 wolverian but, yes, POE needs some headwrapping around to make sense.
08:57 * laniz_ smiles
08:57 wolverian just remember it's a simple state machine.
09:01 franck___ joined #perl6
09:04 laniz_ seems a bit similar than DBD::mysql and DBIx::Mysql
09:04 wolverian hm?
09:04 franck___ joined #perl6
09:04 laniz_ DBIx::Mysql might be much better technically, but its too complex for a beginner
09:04 wolverian I don't think that exists
09:05 laniz_ what do you mean?
09:05 wolverian I don't see DBIx::Mysql on CPAN
09:07 laniz_ yeah, the full name is DBIx-Mysql-InformationSchema or so
09:07 wolverian oh, okay.
09:10 pmurias DBIx-Mysql-InformationSchema seems be a bit diffrent thing then DBD::mysql
09:12 laniz_ yeah, i meant more the DBIx in general
09:12 laniz_ still a beginner as you can see :)
09:13 chris2 joined #perl6
09:24 pmurias wolverian: www.funkplanet.com/irc/ is a none-working adress
09:25 pmurias s/none-working/non-working/
09:25 wolverian pmurias, quite.
09:28 wolverian pmurias, fixed, thanks.
09:28 wolverian (now it says "Net::IRC is not at http:/..." :)
09:35 pmurias ;)
09:36 cognominal_ joined #perl6
09:36 pmurias it's a bit unfortunate that if i took part in both the parrot and kp6 hackathons i would miss most of the talks :(
09:36 buetow joined #perl6
09:37 pmurias at YAPC::Europe
09:38 lumi joined #perl6
09:40 pmurias i'll probably take part in both anyway ;)
09:48 poletti changed the nick to polettix
10:00 ruoso joined #perl6
10:21 sunnavy joined #perl6
10:26 nipotan changed the nick to nipotaway
10:42 fglock joined #perl6
10:49 knewt joined #perl6
10:57 |Jedai| changed the nick to jedai
10:59 fglock pmurias: re talks, i feel like that too
10:59 pugs_svnbot r17411 | fglock++ | [kp6] touch Perl6/*.pm and recompile
10:59 pugs_svnbot diff: http://dev.pugscode.org/changeset/17411
10:59 lambdabot Title: Changeset 17411 - Pugs - Trac
11:04 fglock is Signature.arity a junction, if you have optional arguments?
11:05 rindolf joined #perl6
11:16 kanru joined #perl6
11:19 pmurias fglock: hi
11:19 fglock hi!
11:20 pugs_svnbot r17412 | fglock++ | [kp6] Signature.arity
11:20 pugs_svnbot diff: http://dev.pugscode.org/changeset/17412
11:20 lambdabot Title: Changeset 17412 - Pugs - Trac
11:23 Aankhen`` joined #perl6
11:24 pmurias is mp6 declared frozen, or is it ok to add an simplified <after> support
11:24 pmurias ?
11:25 fglock it's a junction of frozen and open
11:25 fglock it's ok :)
11:28 pmurias but not encouraged ;)
11:33 iblechbot joined #perl6
11:34 pmurias why does mp6 match literals a char at a time?
11:36 pmurias how do i recompile a file in mp6?
11:40 pmurias it emitts 'literal' ok, but splits literal into 7 seperate chars
11:48 fglock re char at a time, it didn't make much difference when benchmarking
11:50 agentzh joined #perl6
11:50 fglock for recompiling, there is util/build-perl5.sh
11:51 fglock if all goes well, you can move lib5-new to lib5
11:51 agentzh heh, there's always version-control :)
11:51 fglock then recompile again, in order to check that it bootstraps
11:52 agentzh fglock: sorry if i'm interrupting...but out of curiosity, are you going to run kp6 against the pugs test suite?
11:52 Shillo joined #perl6
11:52 Shillo Hi all
11:53 agentzh hi, Shillo
11:55 Shillo Oh. Hiya!
11:55 Shillo Glad to see -somebody- awake. :)
11:56 pmurias agentzh: there are some syntax issues blocking it
11:57 pmurias kp6 runs only t/01-sanity/01-tap.t unmodified
11:57 rindolf Hi Shillo
11:57 rindolf Shillo: it's 15:00 here.
11:58 pmurias but i for one would love to use the pugs test suit
12:01 fglock agentzh: re test suite, yes,
12:01 agentzh pmurias, fglock: thanks for the answers :)
12:01 fglock but kp6 still needs some more infrastructure before it can do that
12:02 agentzh okay
12:02 agentzh fglock: is kp6 mature enough to replace PCR?
12:02 * agentzh is looking for ways to let himself hack on kp6 and related stuff in $work...
12:03 fglock hmm - it could
12:03 fglock kp6-perl5rx is more powerful than PCR
12:03 agentzh great
12:03 fglock but it's not an easy task...
12:03 agentzh hehe, gathered so ;)
12:04 fglock i'm using v6.pm at work
12:04 * agentzh wants a lua/c++ backend for mp6/kp6.
12:05 agentzh performance is very important here...v6.pm is just not good enough to let my $boss happy ;)
12:05 fglock agentzh: nferraz wants a lua backend too - are you going to the hackathon?
12:05 Shillo rindolf: It's 2pm here, and I'm at work. :)
12:05 agentzh fglock: i'm afraid not...but i'd love to help ;)
12:06 Shillo So, few people on IRC, as Europe works while US still sleeps.
12:06 fglock if you need performance, the shortest path would be to port perl5rx back to PCR
12:06 fglock instead of waiting for kp6 to get faster
12:07 agentzh fglock: okay, will take a closer look at perl5rx :)
12:07 fglock let me find an example
12:08 agentzh k
12:08 pmurias agentzh: are you worried about regex speed or general speed
12:08 agentzh pmurias: the latter
12:10 fglock agentzh: this is an example of kp6-perl5rx - it should print a Match dump of the result
12:10 Shillo Heh, parrot backend would actually be a good start.
12:10 fglock perl kp6-perl5.pl --perl5rx <  examples/token-test-recurse.pl | perl -Ilib5regex -Ilib5
12:10 agentzh fglock: thanks :)
12:10 agentzh Shillo: parrot backend == moving target
12:10 agentzh Shillo: == pain :)
12:11 fglock agentzh: you can pipe the compiled code through perltidy, to see the generated perl5 regexes
12:11 agentzh okay
12:12 pbuetow joined #perl6
12:13 agentzh doing now
12:13 pmurias agentzh: parrot probably isn't as a moving target as it used to be
12:13 agentzh pmurias: glad to hear that :)
12:14 pmurias i may be wrong
12:14 pmurias ;)
12:15 fglock this used to work:  echo 'class Main { say "hello, World" }' | perl -Ilib5 mp6-parrot.pl | parrot -
12:16 agentzh fglock: i see an AST dump from that command :)
12:16 fglock hey, it still works!
12:16 fglock agentzh: it's a Match
12:16 agentzh cool
12:16 fglock it looks like an AST
12:16 agentzh *nod*
12:16 agentzh hehe
12:18 agentzh fglock: how are you debugging p6 regexes while hacking? is there any tracing tool?
12:18 fglock agentzh: no, it just works :P
12:18 agentzh wow...*impressed*
12:18 agentzh it always just works?
12:19 pmurias no
12:19 fglock just kidding :)
12:20 fglock i add print statements
12:20 agentzh okay
12:20 agentzh i'm going to develop a graphical regex tracer for the current PCR.
12:20 pmurias i had to look at the emited code once because mp6 emitted $$<foo> eq 'a' wrongly
12:20 agentzh the basic facilities are already there...
12:21 agentzh pmurias: right, mp6 cannot be bug free (at least for now) :)
12:21 pmurias the compile speed for me with mp6 is slow enough that it's better to think more then to add print statements
12:22 agentzh pmurias: *nod* that's a very good reason for a real regex tracer/debugger
12:22 fglock pmurias: there was a reason to have it frozen...
12:23 agentzh fglock: i know that reason :)
12:26 pmurias if kp6 obsoletes it will be very nice
12:27 pmurias obsoletes mp6
12:27 fglock yes, that's the idea
12:28 jhorwitz joined #perl6
12:30 sunnavy Hi, everybody.
12:32 sunnavy I've get some free time about 1 or 2 weeks, and I'd like to do something with P6.
12:34 fglock sunnavy: YAPC::EU would be a great place to be if you have a free week
12:36 sunnavy When I digged into the repository, wow, what a huge repository! I don't know how to start.
12:36 sunnavy fglock: I'm afraid I can't go there, I don't have enough money anyway.
12:37 agentzh sunnavy: welcome from jifty :)
12:37 sunnavy agentzh: ;-)
12:37 fglock sunnavy: t/ and ext/ contain Perl6 code (tests and modules)
12:37 agentzh sunnavy: will you write more tests for pugs?
12:37 cerridwen joined #perl6
12:38 agentzh sunnavy: or hack on p6 compilers directly?
12:38 agentzh sunnavy: or proofread the latest synopses first?
12:39 sunnavy Maybe I should proofread the latest synopses, while write more tests ;-)
12:39 agentzh sunnavy: cool
12:39 agentzh sunnavy: take a look at http://perlcabal.org/syn please
12:40 lambdabot Title: Official Perl 6 Documentation
12:40 agentzh sunnavy: it may help ;)
12:40 sunnavy thanks
12:40 agentzh np :)
12:42 rindolf Hi agentzh
12:43 agentzh rindolf: hey, rindolf
12:43 rindolf agentzh: what's up?
12:44 agentzh rindolf: looking for ways to do p6 hacking at $work
12:44 rindolf agentzh: what's stopping you?
12:47 agentzh rindolf: the current p6 compilers are not fast and mature enough.
12:47 rindolf agentzh: do you want to write Perl6 code for production?
12:47 agentzh rindolf: it makes it a bit difficult to align them up.
12:48 agentzh rindolf: aye
12:48 rindolf :S
12:48 agentzh rindolf: so i have a plan to write a lua emitter for mp6/kp6.
12:49 agentzh rindolf: so there may be more chances to get p6 aligned with our C/C++ toolchain used here in yahoo.
12:49 rindolf agentzh: are you working for Yahoo?
12:50 agentzh rindolf: yes.
12:51 rindolf agentzh: nice.
12:51 agentzh rindolf: :)
12:51 laniz_ i thought Yahoo is doing the dishes for Google now. they still have developers? :P
12:51 agentzh laniz_: ahhh...that is offensive...
12:52 laniz_ agentzh: only lame humor of a non-developer :)
12:52 pugs_svnbot r17413 | fglock++ | [kp6] fixed Sub signature (no 'self')
12:52 pugs_svnbot diff: http://dev.pugscode.org/changeset/17413
12:52 lambdabot Title: Changeset 17413 - Pugs - Trac
12:54 fglock lunch &
12:58 pmurias any idea why  $ok &&= eval { runtests("../../t/01-sanity.t") } gives
12:59 pmurias ../../t/01-sanity....../../t/01-sanity.t does not exist
13:00 pmurias stupid me
13:11 agentzh left #perl6
13:22 gilimanjaro joined #perl6
13:22 b_jonas joined #perl6
13:34 cognominal_ joined #perl6
13:57 iForget joined #perl6
14:07 stevan_ joined #perl6
14:08 amnesiac joined #perl6
14:11 snearch joined #perl6
14:27 DarkWolf84 joined #perl6
14:28 mncharity joined #perl6
14:29 turrepurre joined #perl6
14:30 buetow joined #perl6
14:40 Aankh|Clone joined #perl6
14:46 pugs_svnbot r17414 | fglock++ | [kp6] multi, arity check
14:46 pugs_svnbot diff: http://dev.pugscode.org/changeset/17414
14:49 pugs_svnbot r17415 | pmurias++ | kp6:
14:49 pugs_svnbot r17415 | pmurias++ | overelaxed the Grammar to allow ending a statement with a newline
14:49 pugs_svnbot r17415 | pmurias++ | kp6 runs TEST from the main pugs test suit listed in TESTS
14:49 pugs_svnbot diff: http://dev.pugscode.org/changeset/17415
14:51 cmarcelo joined #perl6
15:01 pugs_svnbot r17416 | fglock++ | [kp6] added Multi test; fails
15:01 pugs_svnbot diff: http://dev.pugscode.org/changeset/17416
15:04 Aankh|Clone changed the nick to Aankhen``
15:06 moritz_ re
15:09 moritz_ wolverian: re xhtml: thank you, I fixed it
15:18 justatheory joined #perl6
15:26 snearch joined #perl6
15:31 laniz_ moritz_: thanks a lot for the very well readable code of your irclog bot! very useful to learn :D
15:33 moritz_ laniz_: you're welcome ;)
15:34 moritz_ ilbot2.pl isn't up to date and may have db timeout issues on low traffic channels
15:34 laniz_ covers 3 topics at once that i want to learn :)
15:34 moritz_ which ones?
15:35 laniz_ irc connectivity, mysql access and templates
15:37 Aankh|Clone joined #perl6
15:38 laniz_ surprising to me was, that DBI tutorials have no real good example that explains how you access certain data that is stored in mysql. and i usually learn by playing with examples
15:38 laniz_ and projects like blootbot or buubot are too complex to learn from
15:38 laniz_ (if you start as beginner)
15:39 moritz_ actually the CGI code is getting a bit complex as well
15:39 moritz_ if you download earlier revisions, the CGI code is easier
15:42 zamolxes joined #perl6
15:44 laniz_ moritz_: how can i checkout earlier version? :)
15:44 moritz_ svn up -r $revision
15:44 laniz_ cool, thanks
15:46 moritz_ irc logger will be down for about an hour for server maintaince
15:46 Psyche^ joined #perl6
15:47 irclogbot_backup joined #perl6
15:47 laniz_ moritz_: hm. the syntax and variations does not work for me
15:47 moritz_ irclogbot_backup takes over, afterwards I'll  backport the changes
15:48 moritz_ laniz_: did you do a 'svn co' first?
15:48 laniz_ yeah, but maybe too long ago, hang on
15:48 moritz__ joined #perl6
15:49 moritz__ laniz_: an error message might be helpfull
15:49 moritz__ and I hope you substituted $revision with a real revision number
15:49 laniz_ [17:48:56]: laniz@castle:~/svn : svn up -r 1
15:49 laniz_ Skipped '.'
15:50 moritz__ is this the pugs root directory?
15:50 laniz_ yep
15:50 fglock pmurias: ping
15:51 fglock pmurias: Can't locate KindaPerl6/Grammar/Statements.pm in @INC
15:51 moritz__ laniz_: I'd recommend r16066 as a starting point
15:52 weinig_ joined #perl6
15:52 laniz_ oh
15:52 moritz__ r1 is _really_ old, and doesn't contain ilbot at all
15:53 laniz_ the svn syntax somehow does not work for me, but its not that important
15:54 moritz__ what does a simple 'svn up' produce?
15:54 laniz_ Skipped '.'
15:54 laniz_ but does not fetch a thing
15:54 moritz__ which means that something is wrong with your working copy
15:55 laniz_ or is "up" to upload to the repo and i need "down" or something?
15:55 moritz__ or you aren't in a svn repo at all
15:55 moritz__ no, "up" is to "update" your local copy
15:55 moritz__ upload is "commit" or "ci"
16:45 topic for #perl6 is: http://pugscode.org/ | nopaste: http://sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: http://irc.pugscode.org/
16:48 moritz_ joined #perl6
16:49 moritz_ irc logs (and the real moritz_) are back ;)
16:51 sunnavy cool :-)
16:56 Blwood joined #perl6
17:00 Blwood joined #perl6
17:06 pugs_svnbot r17421 | fglock++ | [kp6] minor fixes in Array, Multi
17:06 pugs_svnbot diff: http://dev.pugscode.org/changeset/17421
17:09 moritz_ #kp6 multi sub s($foo, $bar) { say $foo, '|', $bar }; s('a', 'b')
17:09 exp_evalbot r17421: OUTPUT[DISPATCH: calling APPLY KindaPerl6::Runtime::Perl5::Dispat​chSugar::Dispatch=HASH(0x82d7cc4) on invalid object:$VAR1 = {};␤␤ at lib5/KindaPerl6/Runtime/Perl5/MOP.pm line 29␤ main::DISPATCH('undef', 'APPLY', 'KindaPerl6::Runtime::Perl5::Dispa​tchSugar::Dispatch=HASH(0x82...') called at
17:09 exp_evalbot ..- line 3␤]
17:09 moritz_ #kp6 sub s($foo, $bar) { say $foo, '|', $bar }; s('a', 'b')
17:09 exp_evalbot r17421: OUTPUT[a|b␤]
17:11 moritz_ #kp6 sub s($foo, $bar) { say $foo, '|', $bar }; s('a')
17:11 exp_evalbot r17421: OUTPUT[a|␤]
17:15 Psyche^ joined #perl6
17:15 Psyche^ changed the nick to Patterner
17:16 fglock moritz_: i'm still implementing!
17:17 fglock :)
17:22 moritz_ #kp6 sub s(Int $foo) { say $foo; }; s(2);
17:22 exp_evalbot r17421: OUTPUT[Can't modify non-lvalue subroutine call in scalar assignment at - line 3, near ")}"␤Execution of - aborted due to compilation errors.␤]
17:23 moritz_ is it the type system that is NYI, or Int?
17:23 moritz_ it seems like kp6 parses it correctly, but emits non-functional p5
17:25 pmurias joined #perl6
17:25 pugs_svnbot r17422 | fglock++ | [kp6] Multi - more fixes, doesn't work yet
17:25 pugs_svnbot diff: http://dev.pugscode.org/changeset/17422
17:25 pmurias fglock: sorry for not svk adding Statements.pm
17:25 fglock moritz_: signatures are not finished
17:26 fglock signatures, captures, and multis are interlocked
17:26 fglock pmurias: no problem :)
17:27 weinig_ joined #perl6
17:36 TimToady hmm, "return" is the wrong verb to use for a reduce, since a rule *must* return the current parse state in addition to anything the user returns
17:36 gwall joined #perl6
17:37 TimToady spam
17:38 ruoso damn bots...
17:38 pugs_svnbot r17423 | fglock++ | [kp6] problem with t/kp6/36-multi.t - doesn't enter the for loop
17:38 pugs_svnbot diff: http://dev.pugscode.org/changeset/17423
17:39 ruoso this one is kind of smart
17:42 [particle] *shazaam*
17:43 ruoso can someone clean this shit from the irclogger?
17:43 diakopter I will
17:43 * moritz_ is doint that already
17:43 diakopter er, no I won't, since I don't have my admin credentials accessible from here
17:44 moritz_ done
17:44 moritz_ I just hope I didn't actually mark some non-spam ;)
17:44 diakopter heh
17:45 pugs_svnbot r17424 | fglock++ | [kp6] Multi matches on positional count (it's a start)
17:45 pugs_svnbot diff: http://dev.pugscode.org/changeset/17424
17:45 moritz_ just mine and ruoso++ 's notifications
17:47 * ruoso completely disturbed by that moron... and do hope that doesn't affects anyone life for real... sigh
17:47 philv joined #perl6
17:48 * moritz_ didn't even read it
17:48 * diakopter keeps /ipkb very handy
17:49 philv left #perl6
17:51 pugs_svnbot r17425 | fglock++ | [kp6] t/kp6/36-multi.t passes
17:51 pugs_svnbot diff: http://dev.pugscode.org/changeset/17425
17:52 * ruoso just hate open proxies
17:54 moritz_ $mulit.long_names might be an unfortunate name
17:54 moritz_ it contains refs to the subs, not their names
17:59 justatheory joined #perl6
18:00 fglock moritz_: i'm not sure if 'long name' in the spec is talking about strings
18:00 pugs_svnbot r17426 | fglock++ | [kp6] Multi - added TODO
18:00 pugs_svnbot diff: http://dev.pugscode.org/changeset/17426
18:01 moritz_ fglock: I didn't realize that the term "long name" was mentioned in the spec... I apologize for the fuzz ;)
18:01 fglock moritz_: i mean, it confuses me too
18:02 thoughtpolice joined #perl6
18:08 audreyt_ changed the nick to audreyt
18:12 ubajas joined #perl6
18:21 buetow joined #perl6
18:21 pugs_svnbot r17427 | fglock++ | [kp6] added a bit more calling-convention code
18:21 pugs_svnbot diff: http://dev.pugscode.org/changeset/17427
18:23 cognominal_ joined #perl6
18:30 pugs_svnbot r17428 | fglock++ | [kp6] changed $multi to &multi - I hope this is valid Perl 6
18:30 pugs_svnbot diff: http://dev.pugscode.org/changeset/17428
18:46 ubajas3 joined #perl6
18:57 pmurias joined #perl6
19:18 ubajas joined #perl6
19:37 teratoma left #perl6
20:01 ubajas3 joined #perl6
20:07 cognominal_ joined #perl6
20:13 weinig_ joined #perl6
20:17 holst joined #perl6
20:21 holst will the standard perl6 suite be able to compile the perl6 byte code into standalone executables?
20:22 moritz_ holst: probably into parrot bytecode
20:22 moritz_ holst: full perl 6 will be able to do an eval(), which requires a p6 installation
20:22 holst ok, but standalone in the sense of .pyc in perl and .class in java?
20:23 moritz_ holst: but perhaps you can avoid that by not using eval
20:23 holst hmm, thats something.
20:23 moritz_ but the term "standard perl6 suite" is rather vague
20:23 holst its just ashame that the best thing we got in the year 2007 which compiles native code is ... C
20:24 moritz_ because everything that passes the test suite is "official" and therefore "standard" ;)
20:24 moritz_ allow me to disagree.. we've got eiffel
20:24 moritz_ nobody uses that, but for OO programming that's really great
20:25 holst hmm, i've heard about it but I think the syntax was too weird for me (never used it)
20:29 moritz_ the syntax is quite oposite to perl: rather verbose, virtually no "special" chars
20:29 moritz_ and very readable
20:30 moritz_ you don't have to know the language to understand an eiffel program... kinda like pascal with "real" OO
20:30 holst I was looking at a lot of programming languages a bit back, I might have it confused with some other languge
20:31 thoughtpolice we have haskell! :)
20:31 holst maybe that was the one I was thinking off =D
20:32 * thoughtpolice is working on the new version of his irc bot
20:32 thoughtpolice the first code base was alright
20:33 thoughtpolice then I realized it was too cumbersome to really work with so I did the natural thing
20:33 holst I was looking for a languge 'which is not C++ but efficient and can talk to C'
20:33 thoughtpolice rewrite it, which is what I'm doing. :)
20:33 holst and also I wanted it to be efficient to /program in/ not only in execution
20:33 thoughtpolice and damn, I need something to blog about for DIP6. once school starts I don't know how much I'll be able to do that. :(
20:34 moritz_ holst: well, perl 5 and 6 are not C, and can/will be able to talk to C ;-)
20:34 moritz_ holst: and so can Eiffel
20:34 thoughtpolice haskell can as well. :)
20:34 moritz_ Aye
20:34 moritz_ sadly you need a major brain twist for that
20:34 thoughtpolice the 6.8 of ghc is looking nice. should be out sometime in the next month, give or take a few.
20:35 thoughtpolice moritz_: well there's the standard ffi, there's also greencard but I'm not sure how much easier it makes talking to C.
20:35 moritz_ thoughtpolice: re DIP6, are you looking for a topic?
20:35 thoughtpolice moritz_: somewhat.
20:36 thoughtpolice moritz_: if you have any ideas you can have post priviledges on it. i've been thinking about it for a while and more people posting stuff on it would make it much more useful than just I posting could.
20:36 thoughtpolice like I said, my last year of high school is starting in ~4 days or so, and when that happens, I'm not sure how much I'll be able to blog (or even code.)
20:36 thoughtpolice so having someone else in the ring would be really nice. :)
20:37 moritz_ thoughtpolice: I won't have too much time either, but I can give it a shot
20:37 holst not being able to code and being a student? thats an oxymoron
20:37 holst you will be coding like there is no tomorrow as soon as the exams are closing in =D
20:37 thoughtpolice holst: hah.
20:37 moritz_ possible ideas: rules (2), self defined operators...
20:38 holst thoughtpolice: anything to avoid studying, you know =D
20:38 moritz_ modules and classes
20:38 thoughtpolice moritz_: anything would be appreciated. :) if you want priviledges to post just gimmie an email.
20:39 moritz_ thoughtpolice: you can uses these ideas as well ;)
20:39 thoughtpolice perl 6 is a community effort, so I figure a part of those people blogging would be beneficial.
20:39 holst moritz_: will perl6 core interpreter still be written in haskell when it hits -std or will it be rewritten in C?
20:40 Tene It will be written in Perl 6.
20:40 thoughtpolice moritz_: sent. :)
20:40 holst isnt that a moment 22?
20:40 thoughtpolice moritz_: if you have any ideas of anybody else who would like to contribute, just ask and I'd be glad to give priviledges. ;>
20:41 moritz_ thoughtpolice: ok
20:43 moritz_ holst: what do you mean?
20:43 holst to implement perl6 in perl6. it would need a perl6 parser, which is itself!
20:44 ubajas joined #perl6
20:44 holst I can see how to compile a C compiler with a compiler written in C is do-able, since it's compiled and not interpreted
20:45 thoughtpolice naturally to be able to do this we already need a sufficient perl 6 implementation, which is what pugs aims to be once it's milestones are finished. :)
20:45 moritz_ holst: if perl 6 is compiled to parrot bytecode, it's the same situation as with C
20:46 moritz_ after all even perl 5 is mostly a compiler
20:46 holst maybe I was really talking about the parrot
20:46 Tene holst: perl6 is first being written in Haskell and Parrot and such with the goal of making a good enough perl 6 to write perl 6 in perl 6.
20:46 moritz_ holst: parrot is written in C
20:47 * holst withdraws from *
20:47 holst Im clearly opening my mouth about things I should not =D
20:48 holst anyways, I was just curious about the current state of perl6 and what this bytecode optimizer etc. backend meant to me as a 'native code' lover
20:49 holst if I was in my own house I would say that we have too much of this 'byte code' this day
20:49 holst but I guess there is trends in everything
20:50 thoughtpolice byte code isn't necessarily a bad thing. it helps bootstrap to other platforms. if you have some, say, ANSI C interpreter for bytecode, you've already got your implementation a lot of places.
20:50 Tene Perhaps we need a cpu that runs parrot natively. ;)
20:50 moritz_ Tene: that's a good idea...
20:50 holst like those 'java cpus'
20:50 moritz_ Tene: perhaps we can translate the parrot source code to System C to create one ;-)
20:50 holst (no they dont exist, /end of concept proof)
20:51 thoughtpolice i was thinking recently about a haskell compiler for embedded platforms; byte code would be an optimal choice in that instance.
20:51 holst but why not compile into native code with library substitutes for things the native cpu cant handle?
20:51 moritz_ what about a parrot backend for GHC?
20:51 thoughtpolice which is something I've thought of since I got my psp. :) haskell on the iphone would definately score a few points as well (although bindings for objective-c would be needed.)
20:52 moritz_ holst: I'm not a low level guru, but probably things like JIT-compiling and GC are not easily done in libraries
20:52 moritz_ holst: and you lose portability
20:52 holst the portability is in the compiler
20:52 holst like gcc
20:53 holst anyways, I dont want to drive too far off topic with that =D
20:53 moritz_ which means "recompiling for every platform"
20:53 moritz_ topics are totally overrated ;-)
20:53 holst (compiling is not a bad thing)
20:53 holst just look at gentoo =D
20:53 thoughtpolice for something like embedded platforms, you're dealing with a lot of potential architectures and binary formats, etc.
20:54 thoughtpolice if you have some haskell compiler like I was saying, it'd probably just be a lot easier to have some sort of ANSI-C runtime. if you have that, you've already got your implementation a *lot* of places, by only doing the work once.
20:54 moritz_ holst: I'd say that looking at gentoo proves that compiling _is_ a bad thing indeed ;-)
20:54 moritz_ just my 0.02€
20:56 moritz_ my first japh: http://www.perlmonks.org/?node_id=634962
20:56 holst hehe, you are preaching to the already believing
20:56 holst gentoo is insane
20:56 holst however, it would be equally insane to use a java backend for all programs in all x86 computers
20:57 moritz_ why isn't lambdabot posting the title of my URL?
20:57 moritz_ @seen lambdabot
20:57 SamB possibly more insane
20:57 SamB lambdabot is seriously ill, apparantly
20:57 SamB I saw a lambdabot2 in #haskell
20:57 holst moritz_: will that format my harddrive or something? =D
20:58 SamB ... hey ... isn't lambdabot2 in the same process as lambdabot?
20:58 moritz_ holst: I don't know what it will do on your maschine, but it left mine healthy ;-)
21:05 moritz_ time for bed
21:05 moritz_ g'night
21:05 holst nite, thanks for the heads up on perl6
21:05 holst left #perl6
21:09 obra TimToady: ping
21:09 laniz_ moritz_: sleep well :)
21:15 felipe joined #perl6
21:19 ubajas3 joined #perl6
21:21 rlb3_work joined #perl6
21:26 dmq joined #perl6
21:40 pugs_svnbot r17429 | rhr++ | [Metholated.pm] fix "called at" (maybe)
21:40 pugs_svnbot diff: http://dev.pugscode.org/changeset/17429
21:40 Limbic_Region joined #perl6
21:46 rhr is the above correct, and should other things like "Commit branch" and "didn't match at" be changed likewise?  Or is there something wrong with $¢?
22:01 TimToady $¢ should have worked there--it's set first thing in that routine.
22:01 TimToady and it should be to, not from, at least for forward ops
22:02 ubajas joined #perl6
22:03 rhr until something matches most of $/ (includint $/.to) is undef
22:03 blindfish joined #perl6
22:04 TimToady $/ should start defined; note that first call is passed MCont.new() now, not 0
22:05 TimToady however, .new could init .to to 0, I suppose
22:05 TimToady s/init/default/
22:05 rhr I guess it should just use +$¢ there instead of $¢
22:05 TimToady but in general .from is the beginning of the *previous* match
22:05 rhr oh, ok.
22:06 TimToady so would be correct only when prev match was ""
22:07 rhr so I guess I'll change all debugging use of $¢ to +$¢, does that sound right?
22:08 TimToady $¢ should always be a StrPos now, so the + should be unnecessary, assuming ~ produces the same answer
22:09 TimToady for now, if it's not really an Int, it's a bug
22:09 rhr well, it's just undef, which stringifies to ""
22:10 TimToady then it needs = 0 or // 0 somewhere
22:10 rhr how about my $¢ = +$/.to;?
22:11 TimToady again, the + should be unnecessary, and merely covers a bug elsewhere
22:13 rhr I don't see what you mean.
22:13 TimToady string positions are not integers, in the long run, they're opaque
22:15 TimToady $/.to should always return a StrPos, and a StrPos should always work anywhere a string position is necessary, without knowing whether it's measured in bytes, chars, or whatever
22:15 rhr so an undef StrPos should stringify to "0"?
22:15 TimToady It's only that pugs assumes all string positions are measured in codepoints that we use ints for now
22:16 TimToady 0 is a convenient but misleading notation for "beginning of the string"
22:16 TimToady it's the only number that doesn't care what its units are
22:17 rhr I'm confused.  The problem is that $¢ starts out undef, so you get messages like "_BRACKET called at\n".  I don't see how changing the type of $¢ would fix that...
22:17 TimToady $¢ should not start out undef, it should start out initialized to the beginning of the string
22:18 rhr OK, so the problem is that $/ starts out with ("to" => undef)
22:18 TimToady yes
22:19 rhr can that only be fixed in pugs?
22:19 rhr and does using +$¢ in the debug messages make sense in the meantime?
22:21 TimToady no, and no
22:21 pugs_svnbot r17430 | lwall++ | set proper default on .from and .to
22:21 pugs_svnbot diff: http://dev.pugscode.org/changeset/17430
22:22 rhr aha :)
22:24 pugs_svnbot r17431 | lwall++ | change .from back to .to
22:24 pugs_svnbot diff: http://dev.pugscode.org/changeset/17431
22:26 TimToady biab &
22:44 ubajas3 joined #perl6
23:14 BinGOs joined #perl6
23:18 ozo joined #perl6
23:23 flokuehn joined #perl6
23:28 ubajas joined #perl6
23:34 thorat joined #perl6
23:58 snearch joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo