Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2005-06-13

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 revdiablo Odin-LAP: I hate it when that channel turns into a feeding frenzy like that.
00:00 jamhed has quit IRC ("ðÏËÉÄÁÀ")
00:02 * Odin-LAP has a mental image of piranhas tearing in, and wonders why the hell he can't remember not to stick food into the Amazon.
00:03 revdiablo hmm, is there an authoritative paper on STM?
00:04 revdiablo http://research.microsoft.com/Users/simonpj/papers/stm/stm.pdf # found this, not sure if that's it
00:05 Odin-LAP http://citeseer.ist.psu.edu/978.html <-- ?
00:07 geoffb Odin-LAP, revdiablo: which channel?
00:07 revdiablo geoffb: #perl
00:07 geoffb ah
00:10 crysflame which #perl
00:10 revdiablo freenode #perl
00:23 dkf has quit IRC ()
00:31 dkf has joined #perl6
00:31 vcv- has quit IRC (Read error: 145 (Connection timed out))
00:37 aph has joined #perl6
00:39 vitaminmoo has quit IRC ("Leaving")
00:40 _meppl_ gute nacht  -  good night
00:41 vel has quit IRC (Remote closed the connection)
00:41 justatheory has joined #perl6
00:54 Khisanth has joined #perl6
00:59 _meppl_ has quit IRC (Remote closed the connection)
01:07 Qiang has joined #perl6
01:20 jhorwitz has quit IRC ("Chatzilla 0.9.67 [Firefox 1.0.4/20050511]")
01:20 svnbot6 r4575, geoffb++ | Start Pugs hacking document at lib/pugs/hack.pod (as per Autrijus)
01:20 svnbot6 r4575, geoffb++ |     * Start of source tree map: all top dirs, half of second dirs described
01:30 aph has quit IRC (Client Quit)
01:30 svnbot6 r4576, geoffb++ | More second-level dirs described
01:35 justatheory has quit IRC ()
01:46 Enveigler1 has quit IRC (Read error: 60 (Operation timed out))
01:50 kelan has joined #perl6
01:54 mugwump is there a Perl6 version of: while (<>) ?
01:55 Enveigler1 has joined #perl6
01:59 geoffb Time for dinner . . .
01:59 svnbot6 r4577, geoffb++ | Yet more second-level dirs described
02:04 sdtr443w has joined #perl6
02:04 sdtr443w Ah hah I finally found the channel
02:05 sdtr443w Is there a known problem detecting GHC in the Pug 6.2.6 Makefile.PL?  I can't get anywhere with it even with ghc sitting right in the same directory.
02:09 mugwump I would expect it looks for it in the path.  is the directory where GHC is installed in your $PATH ?
02:09 mugwump (ie, typing `ghci' at the shell should bring up interactive GHC)
02:10 vcv- has joined #perl6
02:27 SamB has quit IRC (Read error: 104 (Connection reset by peer))
02:34 goofyheadedpunk has joined #perl6
02:35 vcv- has quit IRC (Read error: 145 (Connection timed out))
02:38 mugwump anyone know where the for =$fh syntax is defined?
02:38 scw_ has joined #perl6
02:39 justatheory has joined #perl6
02:39 tewk has joined #perl6
02:42 Enveigler1 sdtr443w: Are you on windows?
02:47 sdtr443w No I'm in Linux.  I see what was happening -- I was setting it to look for ghc-6.4, not ghci
02:48 sdtr443w It looks like I have another peculiarity -- during the 'make' command, I get thrown into ghci and it's waiting for me to do something in ghci's interface.  That doesn't seem right.
02:49 xern_ has joined #perl6
02:50 xern has quit IRC (Read error: 110 (Connection timed out))
02:52 SamB has joined #perl6
02:52 jabbot has quit IRC (Read error: 110 (Connection timed out))
02:58 scw has quit IRC (Read error: 110 (Connection timed out))
03:06 goofyheadedpunk has quit IRC (Read error: 110 (Connection timed out))
03:17 Khisanth mugwump: while $foo = =<> { } ?
03:17 * geoffb bak -- mmmm, Pizza . . . .
03:43 sproingie has quit IRC (Remote closed the connection)
03:43 vcv-- has quit IRC (Read error: 131 (Connection reset by peer))
03:43 svnbot6 r4578, geoffb++ | * Remaining ext/* described, only t/* remain
03:43 svnbot6 r4578, geoffb++ | * Extra vertical space for clarity
03:47 sproingie has joined #perl6
03:48 justatheory has quit IRC ()
03:51 mugwump hmm, what would hyper mod operator assign be ?  >>%=
03:52 geoffb mugwump, sounds right
03:52 mugwump bah, not implemented it seems
03:52 geoffb what is the base class of the Code/Block/Sub/etc. heirarchy?
03:53 geoffb hmmm
03:53 geoffb ?eval my @foo = <1 2 3>; @foo >>+= 1; @foo;
03:53 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
03:53 geoffb ?eval my @foo = <1 2 3>; @foo >>+ 1;
03:53 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
03:54 geoffb ?eval my @foo = <1 2 3>; @foo >>+=<< <1 1 1>;
03:54 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
03:55 geoffb ?eval my @foo = <1 2 3>; my @bar = <1 1 1>; my @baz = @foo >>+<< @bar;
03:55 evalbot6 [2.0, 3.0, 4.0]
03:55 geoffb ?eval my @foo = <1 2 3>; my @bar = <1 1 1>; @foo >>+= @bar;
03:55 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
03:55 geoffb ?eval my @foo = <1 2 3>; my @bar = <1 1 1>; @foo >>+=<< @bar;
03:55 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
03:55 geoffb Hyper-assign not implemented at all?
04:00 dkf has quit IRC ()
04:02 mugwump ?eval my $time = eval_perl5("[localtime()]"); $time[1]
04:02 evalbot6 Error: No compatible subroutine found: "&eval_perl5"
04:02 mugwump :)
04:02 QtPlatypus has quit IRC (Read error: 113 (No route to host))
04:03 geoffb No embedding, I guess -- that would be WAY unsafe.  :-)
04:03 geoffb s,No,No Perl5,
04:06 Khisanth hrm is it >>+=<< or >>+<<= or both?
04:06 svnbot6 r4579, geoffb++ | Completed rough draft of level 1 and 2 source map
04:07 mugwump those might not always be equivalent I think
04:07 mugwump the first is in-place, the second makes a new list and then assigns the result as I interpret it
04:09 Khisanth I was thinking the first is a hyper version of += while the second is the assignment version of >>+<<, just from the looks
04:10 vcv- has joined #perl6
04:10 mugwump in other words you agree with my statement :)
04:10 Khisanth no, not really
04:11 Khisanth at least >>+<<= shouldn't make a new list
04:12 mugwump >>+<< returns a new list.  Then you assign that into the result.
04:14 Khisanth but doesn't + return a new scalar?
04:15 Khisanth which would mean += would create a new variable and assign it to the result as well?
04:15 * SamB looks at >>+<< and thinks "that looks like befunge or something"
04:15 Khisanth you still need ^ and V !
04:17 Enveigler1 has quit IRC (Read error: 60 (Operation timed out))
04:17 mr_ank brainfuck! i knew it!
04:18 hcchien $url ~~ rx:P5/http:\/\/(.+)/ doesn't work well on pugs, is anything wrong?
04:18 svnbot6 r4580, geoffb++ | Spacing, alignment, and minor wording edits of source map
04:18 Khisanth heh I guess you can't have a >v< operator
04:19 geoffb OK, at this point, Pugs source tree map in lib/pugs/hack.pod is ready for content review.  Any takers, please edit away.
04:19 geoffb Meanwhile,
04:19 * geoffb makes the sound of one man's stack popping
04:25 mugwump hcchien: s/P5/perl5/
04:25 svnbot6 r4581, geoffb++ | Oops, forgot to delete question I already answered for myself
04:25 mugwump hcchien: or is it the LTS that's breaking it?  I've just noticed that it's not interpolated...
04:26 hcchien mugwump: P5 and perl5 are the same thing, aren't they?
04:28 mugwump it was a quick guess :)
04:28 mugwump so, the escaped /'s are breaking it?
04:28 hcchien I guess so.
04:30 hcchien oh, interesting
04:30 hcchien this would be broken: say $str if ($str ~~ rx:perl5/http:\/\//);
04:31 hcchien but it's fine with: say $str if ($str ~~ rx:perl5{http:\/\/});
04:31 mugwump it looks like the parse context of the // segment is wrong; it's being passed straight to PCRE rather than being interpolated as a quoted string (I *think*)
04:32 Chewie[] !eval <f oo bar>>>.length
04:32 Chewie[] !eval <f oo bar>
04:32 Chewie[] !eval qw/f oo bar/
04:32 Chewie[] Interesting.
04:32 mugwump ?eval <f oo bar>.elems
04:32 evalbot6 3
04:33 mugwump ?eval <f oo bar>>>.chars
04:33 evalbot6 Error:  unexpected ">" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
04:33 mugwump ?eval <f oo bar>>>.<<chars
04:33 evalbot6 Error:  unexpected ">" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
04:33 SamB mr_ank: well, yes, it is valid BF, but so is everything with balanced square brackets...
04:33 mugwump ?eval <f oo bar>.chars
04:33 evalbot6 8
04:34 Chewie[] Heh.
04:34 Chewie[] ?eval qw/f oo bar/ >>.elems
04:34 evalbot6 Error:  unexpected ">" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
04:34 mugwump looks like my complaint about the interpolation is bogus.  my test script was broken :)
04:35 Chewie[] ?eval (qw/f oo bar/) >>.elems
04:35 evalbot6 Error:  unexpected ">" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
04:35 Chewie[] ?eval ("f","oo","bar")».length
04:35 evalbot6 Error: No compatible subroutine found: "&length"
04:35 Chewie[] Er.
04:35 Chewie[] ?eval ("f","oo","bar")».elems
04:35 evalbot6 3
04:36 Chewie[] ?eval (("f","oo","bar")».elems).perl
04:36 evalbot6 '3'
04:36 Chewie[] Shouldn't that return (1, 2, 3) as in S03?
04:37 mugwump congrats Chewie[]!  you've found a bug ... time to submit a failing test!
04:37 mugwump :)
04:39 Chewie[] mugwump: To be honest, I'm way too uncomfortable with the tools, lists, and trees to feel good about doing that. Plus I'm powerful tired right now.
04:42 mr_ank i'll submit it if you don't mind Chewie[]
04:42 mugwump I'm about to commit it mr_ank :)
04:43 mr_ank :) okay
04:44 Chewie[] Thanks, guys. I really need to read through the design docs and start pulling weight.
04:44 mugwump no pressure dude, enjoy learning :)
04:45 Chewie[] Heh. I love learning. Love school, love programming, love reading technical reference manuals.
04:45 Chewie[] I mean, call me odd, but damn am I odd.
04:47 Chewie[] Personally, I like the current Parrot implementation more than the current Perl 6 implementation. Not even PIR, just PASM. Reminds me of my assembly days back on DOS.
04:48 Khisanth DOS does not sound like a step forward ...
04:48 svnbot6 r4582, mugwump++ | Add test for .perl on elems reported by Chewie[]++; unresearched
04:49 mugwump http://utsl.gen.nz/perl6-intro/html/  # talk I'm giving tonight ...
04:49 Chewie[] Khisanth: x86 assembly on a 486. That was fun.
04:50 hcchien I should add some tests on rx_perl5
04:50 SamB was it more fun than on a Pentium?
04:51 Chewie[] mugwump: Where did you get that larger version of the Perl crocodile-moth-lizard-moose-bird thing?
04:51 Chewie[] SamB: It felt a lot more manual, and it was more fun to me.
04:51 Chewie[] My background (ha) is in engineering. The closer to metal, the better.
04:52 SamB hmm, you probably wouldn't like House
04:52 * Chewie[] wonders if a 15-year-old can have a 'background'.
04:52 SamB duh, yeah.
04:52 mugwump Chewie[]: google images for "perl 6" :)
04:52 crysflame yes, chewie
04:52 mugwump http://bleaklow.com/blog/archive/000018.html
04:53 Chewie[] We haven't even been around long enough to become experts in a field. I'm an expert at programming by no means.
04:53 Chewie[] mugwump: Oohhh. I found that originally on a gallery a number of months ago. I was looking for it by filename.. <giggle />
04:54 SamB even non-experts have backgrounds
04:54 crysflame mugwump: what is it?
04:55 SamB and you might know more than the average CS degree-holder
04:55 mugwump crysflame: just a rather cynical joke cover
04:56 crysflame ok
04:56 Chewie[] SamB: A week after receiving the degree, when the degree-bearer still holds the knowledge from the past year? Probably not.
04:57 mr_ank Chewie[]: Reexamining the bug/problem: im probably missing something, but why whould it return (1, 2, 3) and not 3?
04:58 mr_ank (perl representation of the number of elements in the array)
04:58 Chewie[] mr_ank: The hyperoperators operate on lists. That one takes a list and, for each element, takes the .elems. It's just like @foo = @bar >>+<< @baz;.
04:58 Chewie[] What about the native printout via .perl? That should give the array of .elems'.
05:00 Chewie[] See also http://dev.perl.org/perl6/doc/design/syn/S03.html#Hyper_operators
05:00 sproingie has quit IRC (Connection reset by peer)
05:00 SamB Chewie[]: have you any idea what an average CS student has learned by graduation? me neither. and I would like very much never to find out...
05:01 mr_ank wait
05:01 Chewie[] SamB: It's not a matter of knowledge gained and lossed by graduation. It's a matter of retention.
05:01 Chewie[] Lossed? That's the sleep deprivation ghost speaking. Lost.
05:01 mr_ank Chewie[]: should there be a ">>" char somewhere in your test eval?
05:02 Chewie[] mr_ank: I used the Unicode right French quote character.
05:03 mr_ank okay then, I guess my IRC program missed it - I see no Unicode chars
05:03 Chewie[] mr_ank: I tried it with the ASCII representation. That also produced an error. An interpretation error, too, not an inaccuracy in interpretation as is the case with the msot recent one.
05:03 Chewie[] Most, even.
05:03 mr_ank (i keep getting French and German quotes confused - will get used to them eventually)
05:03 mr_ank mugwump: in your test i see no french or german quotes
05:04 Chewie[] That's quite a presentation, mugwump. Very nice.
05:04 mugwump ok.  feel free to add them.  I'm off to give this talk ...
05:04 mugwump thanks for the feedback Chewie[] !
05:04 mugwump any other suggestions before I go?  :)
05:04 mr_ank Good luck with the talk :)
05:05 Chewie[] mugwump: Don't find any bugs live during the demonstrations. :-p
05:05 crysflame haha
05:05 Chewie[] s/p$/P/
05:05 crysflame it makes it more realistic when you do
05:05 crysflame so if it does happen, demo how easy it is to fix bugs in pugs
05:05 mugwump "never do a live demo" ... I'm not autrijus :)
05:05 Chewie[] True as that may be, it's all about the image.
05:05 Chewie[] mugwump: Giggle.
05:06 * Chewie[] will be back in about 10 minutes.
05:06 mr_ank Chewie[]: so what you were trying was (("f","oo","bar")>>.elems).perl (s/>>/$french_quote_char/) ?
05:06 Chewie[] Have fun, mugwump.
05:06 Chewie[] mr_ank: Correct.
05:06 mr_ank cool
05:07 Chewie[] It also really didn't like <f oo bar>>>.elems :-D
05:07 mr_ank yes
05:07 mr_ank and i tried it again with the quotes in place and it doesn't work either way
05:09 alinbsp has joined #perl6
05:10 mr_ank hey, do i need an svn commit account to add a few tests?
05:11 mr_ank (been working on anonymous svn)
05:11 Chewie[] You probably do. That's something I need to find out how to get in the near future.
05:12 mugwump yes you will, ask any of the admins as listed on the openfoundry project page
05:12 geoffb Ping a meta commiter, they will send you an invite email
05:12 mr_ank okay
05:15 iblechbot has joined #perl6
05:15 geoffb mugwump, perhaps add this to the docs/talks/ ?
05:17 * mugwump hmms
05:17 vcv-- has joined #perl6
05:18 geoffb mugwump, you show slides the convert a P5 while into a P6 loop, then a P5 for loop into a P6 while.  That will confuse
05:19 mugwump or amuse
05:19 geoffb heh
05:19 mr_ank okay, should I register in openfoundry and then ping an admin from there?
05:19 geoffb I'm basing that comment on a presumed beginner level viewer
05:20 mugwump mr_ank: you don't need to, the invitation will ask you to sign up
05:20 geoffb mr_ank, I believe you just ping and they invite you to OF
05:21 geoffb mugwump, is there a large version of the image on slide37 somewhere?
05:21 mr_ank ping == just email one of them? or a particular OF feature i'm not seeing?
05:24 geoffb as in, when one of them is around here, /msg them with your request and email address
05:24 mr_ank ah okay :)
05:29 geoffb OK, time to amble slowly towards bed . . .
05:29 geoffb g'night, all
05:29 mr_ank g'night :)
05:32 gaal_ morning!
05:33 Khisanth geoffb: that sounds like a good way to get people to ask "What is the point of Perl 6?" :)
05:34 sdtr443w Ah well I give up getting the new Pugs working.  I'm sticking to 6.0.13 for now.  Good night.
05:34 sdtr443w has quit IRC ("Leaving")
05:36 autrijus yo mr_ank. your email? :)
05:39 autrijus got it. welcome aboard!
05:39 mr_ank thanks!
05:40 autrijus mugwump: url to your slides?
05:40 autrijus I must've missed it
05:40 mr_ank <mugwump> http://utsl.gen.nz/perl6-intro/html/  # talk I'm giving tonight ...
05:43 autrijus k thanks
05:51 Shachaf has quit IRC ("''")
05:57 hcchien autrijus: saw my rx tests?
05:57 gaal_ mugwump, did you find the =<> you were looking for?
05:58 autrijus hcchien: no, where?
05:58 gaal_ op1 "=" and lib/pugs/run.pod <- -n and -p switches
05:58 hcchien autrijus: like these:
05:58 hcchien 12:30 < hcchien> this would be broken: say $str if ($str ~~ rx:perl5/http:\/\//);
05:58 hcchien 12:31 < hcchien> but it's fine with: say $str if ($str ~~ rx:perl5{http:\/\/});
06:00 autrijus you mean the rx interpolation is bad?
06:00 autrijus yeah, looks like so.
06:00 hcchien I guess so. what do you think?
06:00 autrijus write a t/rules/escape.t please?
06:00 autrijus and commit it
06:00 autrijus I'll take a look
06:01 hcchien or rx_perl5_match.t ?
06:02 autrijus it's not _match
06:02 autrijus so rx_perl5_escape
06:03 hcchien k
06:12 autrijus mr_ank: remember to add yourself to AUTHORS
06:16 autark-jp the qfInterpolateBackslash of rxP5Flags (and maybe rxP6Flags as well) should probably be QB_Single and not QB_No.
06:17 autrijus autark-jp: ooh. commit ahead :)
06:17 autark-jp however, that makes rx:perl5/http:\q[//]/ legal
06:17 autark-jp \me scratches head
06:25 autrijus may not be a bad thing :)
06:31 BCoffman has joined #perl6
06:33 elmex has joined #perl6
06:34 BCoffman has quit IRC ("Chatzilla 0.9.67 [Firefox 1.0/20050524]")
06:34 BCoffman has joined #perl6
06:36 alinbsp has quit IRC ()
06:36 svnbot6 r4583, autark++ | * make '\' protect delimiters in rx
06:41 autrijus hcchien: fixed :D but still test welcome
06:49 larsen has joined #perl6
06:50 elmex has quit IRC ("leaving")
06:50 alinbsp has joined #perl6
06:50 iblechbot has quit IRC (Read error: 54 (Connection reset by peer))
06:52 vitaminmoo has joined #perl6
06:56 BCoffman has quit IRC ("Chatzilla 0.9.67 [Firefox 1.0/20050524]")
06:59 hcchien autrijus: sure
07:01 gaal_ hello. i've made a list of candidates for Preludification.
07:01 gaal_ http://pugs.kwiki.org/?PreludificationCandidates
07:01 scook0 has joined #perl6
07:01 gaal_ Please hack :)
07:07 Aankhen`` has joined #perl6
07:12 Nattfodd has joined #perl6
07:26 osfa has joined #perl6
07:40 q[pdcawley] ?eval class Array { method foo { "Foo" } }; [].foo
07:40 evalbot6 Error: No compatible subroutine found: "&foo"
07:40 svnbot6 r4584, autrijus++ | * Test and document for a new bug that prevents s:P5{}{{...}} from
07:40 svnbot6 r4584, autrijus++ |   interpolating.
07:40 q[pdcawley] ?eval class Array { method foo { "Foo" } }; my @foo = (10); @foo.foo
07:41 evalbot6 'Foo'
07:41 q[pdcawley] Where do I send a test patch to?
07:42 autrijus q[pdcawley]: invitation sent.
07:42 autrijus welcome aboard!
07:42 autrijus as we're nearing preflight, please add :todo<bug> to the patch.
07:43 q[pdcawley] And does this test classes or data types or what? ie, what file should I stick it in?
07:43 gaal_ autrijus, any chance to get the intermediate accessor bug fixed before this release? :)
07:43 autrijus gaal_: what's that bug again?
07:43 gaal_ ?eval caller.subname
07:43 evalbot6 Error: No compatible subroutine found: "&subname"
07:43 gaal_ ?eval my $c = caller; $c.subname
07:43 evalbot6 Error: No compatible subroutine found: "&subname"
07:43 gaal_ oops!
07:44 autrijus ...
07:44 autrijus but anyway. to get that to work
07:44 q[pdcawley] is &?CALLER_CONTINUATION or whatever it's called a full continuation yet?
07:44 autrijus declare your sub as "returns Code".
07:44 vcv- has quit IRC (Read error: 110 (Connection timed out))
07:44 gaal_ ?eval sub x { my $c = caller; $c.subname } sub y{x} y
07:44 evalbot6 \'&main::y'
07:45 gaal_ ?eval sub x { caller.subname  } sub y{x} y
07:45 evalbot6 Error: No compatible subroutine found: "&subname"
07:45 autrijus q[pdcawley]: you can try... I think it's full, not sure
07:45 autrijus gaal_: "returns Code"
07:45 q[pdcawley] Heh. I'll get that non-deterministic Sudoku solver written and see if it breaks :)
07:45 gaal_ autrijus: on what? subname is really $.subname
07:46 autrijus gaal_: on caller()
07:46 autrijus q[pdcawley]: fwiw, pugs currently discards no scope
07:46 autrijus q[pdcawley]: so eval"" will always work
07:46 q[pdcawley] Whee!
07:46 q[pdcawley] Don't tell chip :)
07:46 autrijus otoh, pugs is also not very fast nor memory friendly
07:47 autrijus I make no promise of what will become when we compile to parrot.
07:47 q[pdcawley] Bah!
07:47 autrijus otgh, I like the eval_in_caller_scope idea.
07:47 autrijus in fact, I desperately need something like that.
07:47 autrijus (for eval_is() to work)
07:47 q[pdcawley] Exactly.
07:47 autrijus so if you can get the syntax defined for p6
07:47 autrijus on p6l
07:47 autrijus please do so
07:47 autrijus maybe in another thread, maybe not
07:47 q[pdcawley] Will try.
07:48 autrijus danke.
07:48 q[pdcawley] I don't mind that it's not built in, I do mind that I can't write it.
07:48 q[pdcawley] Which might well be my Perl 6 motto.
07:49 * autrijus mumbles something about inline PASM
07:49 gaal_ autrijus, i'm building in the background, but could you explain *why* is Code should fix the problem? caller is supposed to return an object, not code.
07:49 autrijus gaal_: oh, it's an object?
07:49 autrijus then declare it as such.
07:49 autrijus returns Control::Caller
07:49 autrijus or something like that.
07:49 gaal_ yes, it's supposed to return a Contorl::Caller
07:49 gaal_ yeah.
07:49 autrijus the reason is in chained method calls
07:49 gaal_ okay, still: why does this matter?
07:49 autrijus $a.foo.bar
07:50 autrijus currently the MMD resolver for .bar has no access to $a.foo's actual results
07:50 autrijus but only the EXp
07:50 autrijus so it's a expression-based dispatch
07:50 * gaal_ nods
07:50 autrijus that is arguably wrong.
07:50 autrijus and needs fixing
07:50 autrijus but not for this release :)
07:50 gaal_ :)
07:51 autrijus also, if you have a bit of tuits, I need help on ChangeLogging :)
07:51 * autrijus is triaging the tests
07:51 gaal_ sure - do you have the r number for the last release?
07:53 autrijus 4424
07:53 gaal_ thanks
07:53 gaal_ hah! sub foo () returns X is A; # ok
07:54 gaal_ sub foo() is A returns X; # error
07:54 autrijus that's right.
07:54 autrijus fixes... welcome :)
07:54 gaal_ "not for this release"
07:55 autrijus lol
07:55 jamhed has joined #perl6
07:56 Aankhen`` has quit IRC (Read error: 104 (Connection reset by peer))
07:56 gaal_ autrijus, your "returns" fix yields a very strange error:
07:57 gaal_ oh wait no
07:57 q[pdcawley] Mmm... tramp.el++
07:57 autrijus "the fact that it works is the strangest error"
07:57 gaal_ no, it just doesn't do anything :(
07:57 gaal_ heh hrh
07:58 gaal_ oh! look:
07:58 gaal_ ..... caller.subname # error
07:58 gaal_ caller().subname  # ok
07:59 gaal_ well, this is an improvement, so i'm checking it in
08:00 autrijus k
08:00 q[pdcawley] ?eval my @foo; @foo.foo
08:00 evalbot6 Error: No compatible subroutine found: "&foo"
08:00 autrijus meanwhile I unbroke map {3 => 4;} 1,2,3
08:00 gaal_ but i'm correct in calling this a bug, yes?
08:00 autrijus gaal_: you are.
08:00 q[pdcawley] ?eval my @foo = (1,2,3); @foo.foo
08:00 evalbot6 Error: No compatible subroutine found: "&foo"
08:00 gaal_ is now known as gaal
08:01 * autrijus wonders what is q[pdcawley] trying to do
08:01 q[pdcawley] ?eval class Array { method foo { "Foo" } }; my @foo; @foo.foo
08:01 evalbot6 'Foo'
08:01 q[pdcawley] ?eval class Array { method foo { "Foo" } };  [].foo
08:01 evalbot6 Error: No compatible subroutine found: "&foo"
08:01 q[pdcawley] Checking stuff again before writing the test.
08:01 q[pdcawley] ?eval class Array { method foo { "Foo" } };  my $foo = []; $foo.foo
08:02 evalbot6 'Foo'
08:02 q[pdcawley] Sorry, should really do this on feather shouldn't I? Wow. Expected that to break.
08:02 Aankhen`` has joined #perl6
08:02 autrijus nah, that's fine. awating your test.
08:02 q[pdcawley] Writing it as we speak. Post to p6c?
08:03 autrijus no, commit it!
08:03 autrijus you got an invitation mail after all :)
08:03 q[pdcawley] Err... I haven't a clue what my perl id or key is.
08:03 autrijus hm? it's openfoundry thing
08:03 autrijus you can commit into http://svn.openfoundry.org/pugs
08:03 autrijus as soon as you click on the url in that mail
08:03 q[pdcawley] Ah. Okay.
08:04 autrijus and agree to turn in your first unborn parent, or something
08:04 q[pdcawley] Might take a while, I'm greylisting.
08:04 autrijus 'kay. then it's easier just to hop on http://rt.openfoundry.org/
08:04 autrijus and click "register"
08:04 autrijus and tell me your userid.
08:04 q[pdcawley] Got it.
08:05 autrijus and by the time you check in your test it'd be passing.
08:05 q[pdcawley] :)
08:05 autrijus pugs is strange like that.
08:05 hcchien hmm...., make the pugs failed
08:06 autrijus hcchien: how?
08:06 dada has joined #perl6
08:06 svnbot6 r4585, autrijus++ | * fix a regression that made `map { 1 => 2; } @x` break -- Noop
08:06 svnbot6 r4585, autrijus++ |   is now restored at an empty trailing semicolon.
08:06 hcchien linking failed, a sec. let me try to paste it on nopaste
08:07 autrijus q[pdcawley]: also remember to add yourself to AUTHORS.
08:07 q[pdcawley] So, I'm in feather, I've got an svk thing checked out, how do I check it back in (sorry for being dumb).
08:08 autrijus svk ci
08:08 hcchien autrijus: http://irc.csie.org:8888/110
08:08 autrijus if you see "committing back" then it all works
08:08 q[pdcawley] And it asks for id and password?
08:08 autrijus right
08:08 q[pdcawley] So much nicer than CVS :)
08:08 autrijus that'd be your openfoundry id/passwd
08:09 autrijus hcchien: are you using gcc 3.3 on a clean tree?
08:09 autrijus hcchien: `gcc_select 3.3`
08:09 q[pdcawley] Fecking zsh, refusing to honour stty erase...
08:10 hcchien oh, hate gcc 4. :p
08:12 * q[pdcawley] mutters... still haven't got GHC building on tiger. Cleaned everything out and it bitches about out of date syntax or something during hte compile.
08:12 svnbot6 r4586, gaal++ | caller().subname works (though caller.subname still doesn't). autrijus++
08:12 svnbot6 r4587, autrijus++ | * test fixups.
08:12 svnbot6 r4588, autrijus++ | * pdcawley noted that `[].method` did not call `Array::method`;
08:12 svnbot6 r4588, autrijus++ |   neither does `{}.method` call `Hash::method`.  Fixed.
08:12 svnbot6 r4589, autrijus++ | * I fail to see where .elems would return 1-based index;
08:12 svnbot6 r4589, autrijus++ |   changed to test .keys, which works.
08:13 q[pdcawley] Ooh, forgot to add a check for hashes.... I'll do that now.
08:15 q[pdcawley] What about pairs made with the cunning pair constructor whose syntax I've forgotten?
08:16 q[pdcawley] ?eval :head<tail>
08:16 evalbot6 ('head' => 'tail')
08:17 q[pdcawley] ?eval :head<tail>.key
08:17 evalbot6 undef
08:17 q[pdcawley] ?eval :head<tail>.value
08:17 evalbot6 Error: cannot cast from VUndef to (Pugs.AST.Internals.Val,Pugs.AST.Internals.Val)
08:17 q[pdcawley] ?eval (:head<tail>).value
08:17 evalbot6 Error: cannot cast from VUndef to (Pugs.AST.Internals.Val,Pugs.AST.Internals.Val)
08:17 autrijus [=>] <head tail>
08:17 autrijus ?eval [=>] <head tail>
08:17 evalbot6 ('head' => 'tail')
08:17 autrijus ?eval [=>] <head tail foo bar>
08:17 evalbot6 ('head' => ('tail' => ('foo' => 'bar')))
08:17 q[pdcawley] ?eval :head<tail>.head
08:17 evalbot6 Error: No compatible subroutine found: "&head"
08:17 svnbot6 r4590, autrijus++ | * hardtodo'ed sprintf_and_as.t
08:17 autrijus ?eval ([=>] <head tail foo bar>)key
08:17 autrijus ?eval ([=>] <head tail foo bar>).key
08:17 evalbot6 Error:  unexpected "k" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
08:17 evalbot6 'head'
08:17 autrijus ?eval ([=>] <head tail foo bar>).value
08:18 evalbot6 ('tail' => ('foo' => 'bar'))
08:18 autrijus ?eval ([=>] <head tail foo bar>).value.value
08:18 evalbot6 ('foo' => 'bar')
08:18 autrijus ?eval ([=>] <head tail foo bar>).value.value.value
08:18 evalbot6 \'bar'
08:18 q[pdcawley] ?eval :head<tail>.value
08:18 evalbot6 Error: cannot cast from VUndef to (Pugs.AST.Internals.Val,Pugs.AST.Internals.Val)
08:19 q[pdcawley] ?eval (:head<tail>).value
08:19 evalbot6 Error: cannot cast from VUndef to (Pugs.AST.Internals.Val,Pugs.AST.Internals.Val)
08:19 q[pdcawley] ick.
08:19 autrijus ?eval (head => 'tail').value
08:19 evalbot6 Error: cannot cast from VUndef to (Pugs.AST.Internals.Val,Pugs.AST.Internals.Val)
08:19 autrijus I think I know why.
08:19 q[pdcawley] Oh good :)(
08:19 autrijus it's being parsed as
08:19 autrijus value(:head<tail>)
08:19 autrijus as in, a named argument 'head'.
08:19 autrijus so, a :todo<bug> test welcome
08:20 q[pdcawley] Add :todo<bug> to the commit message right?
08:20 autrijus no, the test
08:20 autrijus is('foo', 'bar', :todo<bug>);
08:20 q[pdcawley] Oh fuck. I've written things a test method in the hash...
08:20 q[pdcawley] Not an issue with pair thing of course.
08:23 q[pdcawley] is there a quick and dirty way of running just one test file?
08:23 q[pdcawley] ie, a perl6 version of prove.
08:24 autrijus env HARNESS_PERL=./pugs prove ...
08:24 autrijus env HARNESS_PERL=./pugs prove t/foo/bar
08:24 q[pdcawley] coo.
08:24 q[pdcawley] cool.
08:24 autrijus make subdirs; ./pugs -Iblib6/lib t/foo/bar
08:24 autrijus etc.
08:24 autrijus (Test Anything Protocol)++
08:32 q[pdcawley] cool innit?
08:33 svnbot6 r4591, autrijus++ | * test cleanup, round #2.
08:33 gaal autrijus: ChangeLog 87% done
08:35 gaal is lvalue substr a feature or a bugfix? :)
08:37 mr_ank ?eval <A B> ~~ ('A',/B/)
08:37 evalbot6 bool::false
08:37 mr_ank Shouldn't that be true?
08:38 q[pdcawley] cool innit?
08:38 q[pdcawley] Oops, wrong window.
08:38 gaal ?eval <A B>
08:38 evalbot6 ('A', 'B')
08:39 gaal (1,2) ~~ (1, 2)
08:39 gaal ?eval (1,2) ~~ (1, 2)
08:39 evalbot6 bool::true
08:39 gaal ?eval (1,2) ~~ (1, 2.0)
08:39 evalbot6 bool::true
08:39 mr_ank regexes should be allowed there afaik
08:39 gaal ?eval (1,2) ~~ (1,/2/)
08:39 mr_ank (s/regexes/rules/ heh)
08:39 evalbot6 bool::false
08:40 mr_ank look at this
08:40 gaal i think the definition of ~~ isn't... creepy
08:40 mr_ank ?eval <a> ~~(/a/)
08:40 evalbot6 pugs: src/events.c:265: init_events_first: Assertion `rc == 0' failed.
08:40 gaal ie for lists, it doesn't recursively ~~ in the list
08:40 gaal brb
08:41 mr_ank im looking for more info about that...
08:41 mr_ank i remember reading it somewhere other than the perl 6 & parrot essentials book
08:42 gaal look at the big ~~ table in A04 i think it is
08:43 mr_ank yep
08:43 mr_ank list    list      recursive disjunction*   match if any($a) =~ any($b)
08:43 mr_ank but it isn't quite that
08:44 svnbot6 r4592, pdcawley++ | Added test for calling methods on arrays and hashes, including anonymous ones.
08:44 scw_ is now known as scw
08:45 mr_ank darn, this is going to drive me crazy
08:46 gaal does `as` work now?
08:48 nothingmuch anybody got a nice project for someone just learning computers?
08:48 ingy hola
08:48 nothingmuch a 1-2 hour job
08:48 ingy want to see something cool?
08:48 nothingmuch if it's not obscene
08:49 gaal ingy, we always want to see your *modules*
08:49 ingy http://doolittle.kwiki.org/
08:49 ingy there you go gaal
08:49 gaal this is why i like #perl6
08:49 nothingmuch #touch-my-monads, #touch-my-modules
08:49 gaal instant gratification
08:50 gaal http://doolittle.kwiki.org/index.cgi?HomePage # is sucking
08:50 mr_ank yes oretty much
08:50 mr_ank *pretty
08:51 mr_ank http://doolittle.kwiki.org/index.cgi?action=search;search_term=Spoon-Command # is sucking too
08:53 mr_ank brb
08:53 gaal http://perlcabal.org/~gaal/smoke_xp.html # r4587
08:53 q[pdcawley] ?eval try { die "Foo"; CATCH { "Caught Foo" } }
08:53 evalbot6 undef
08:53 gaal ChangeLog updated.
08:53 q[pdcawley] undef? weird.
08:54 q[pdcawley] ?eval try { die "Foo"; CATCH /.?/ { "Caught Foo" } }
08:54 evalbot6 undef
08:54 svnbot6 r4593, pdcawley++ | Added a test of method calls on a bare pair
08:54 svnbot6 r4594, gaal++ | ChangeLog updates (to 4592)
08:54 svnbot6 r4595, autrijus++ | * Turn p5 regexes back to QB_No again (not QB_Single) because
08:54 svnbot6 r4595, autrijus++ |   we need to preserve the double \\ into PCRE.
08:54 svnbot6 r4595, autrijus++ |   (this needs further investigation.)
08:54 q[pdcawley] ?eval try { die "Foo"; CATCH { return "Caught Foo" } }
08:54 evalbot6 undef
08:54 gaal my favorite part of ChangeLogging is this line:
08:54 gaal * Some new test and several tests refactored, we now have 7650+ tests
08:55 q[pdcawley] How is that undef?
08:55 gaal brb
08:58 Aankhen`` ?eval die "Foo"
08:58 evalbot6 Error: Foo
09:02 Aankhen`` has quit IRC ("New DVD burner and UPS [Time wasted online: 1hr 11secs]")
09:02 svnbot6 r4596, autrijus++ | * test fixes, round #3.
09:03 q[pdcawley] Is there an array slicing syntax yet?
09:03 nothingmuch test runs are ~5 times longer with each passing month... I wonder if that is a good thing ;-)
09:05 cognominal nothingmuch, thx so much about you Forth stuff
09:06 nothingmuch wow! 4th or 5th good comment =) i'm overjoyed
09:06 nothingmuch i'll try to get chapter 8 out of the door today
09:06 cognominal it really feels a gap to bootstrap oneself to become a lambda camel
09:08 autrijus ?eval my @a = 1..10; @a[4,5,6]
09:08 evalbot6 [5, 6, 7]
09:08 nothingmuch any feedback?
09:09 autrijus q[pdcawley]: there you go
09:09 nothingmuch also, how far did you get so far?
09:09 q[pdcawley] Great. Ta.
09:09 q[pdcawley] Now, what's a quick way of writing 'all_distinct(@some_array)'?
09:10 autrijus ?eval my @a = (1,1,2,3,5,8,13,2,3,5); @a.uniq
09:10 evalbot6 (1, 2, 3, 5, 8, 13)
09:10 q[pdcawley] So @a.uniq == @a should do the trick?
09:10 nothingmuch @a.=uniq eventually
09:10 autrijus sure
09:11 autrijus ?eval my @a = (1,1,2,3,5,8,13,2,3,5); @a .= uniq;
09:11 evalbot6 [1, 2, 3, 5, 8, 13]
09:11 q[pdcawley] Cool... Makes this sudoku solver read easy to write...
09:11 nothingmuch autark-jp++
09:11 q[pdcawley] Admittedly, it won't be quick, but it'll be dead easy to write :)
09:11 * autrijus ponders renick to "autarijus"
09:11 nothingmuch autrijus++
09:11 autrijus because "ijus" sorts better than "k-jp"
09:11 nothingmuch oops ;-)
09:12 vcv- has joined #perl6
09:13 autark-jp hm, here's a better solution
09:13 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0,1][0,1];
09:13 evalbot6 [[1, 2, 3, 4], [1, 2, 3, 4]]
09:13 autark-jp is now known as jp-autark
09:13 q[pdcawley] Shame!
09:13 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0,1]<<[0,1]>>;
09:13 evalbot6 Error: Cannot cast into Hash: VRef <Array::Slice>
09:13 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0,1]>>[0,1]<<;
09:13 evalbot6 Error:  unexpected ">" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
09:13 q[pdcawley] Curse...
09:14 Nattfodd has quit IRC ("Leaving")
09:14 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0][0..3]
09:14 evalbot6 [1, 2, 3, 4]
09:14 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0][0..2]
09:14 evalbot6 [1, 2, 3]
09:14 Nattfodd has joined #perl6
09:14 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); @foo[0..4][0..2]
09:15 evalbot6 [[1, 2, 3, 4], [1, 2, 3, 4], [1, 2, 3, 4]]
09:15 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); [@foo[0..4]][0..2]
09:15 evalbot6 [[1, 2, 3, 4], [1, 2, 3, 4], [1, 2, 3, 4]]
09:15 nothingmuch jp-autark++
09:15 autrijus jp-autark: er. thanks :)
09:15 autrijus q[pdcawley]: deref isn't yet hyperised.
09:16 q[pdcawley] NP. I can write a map. Now must go help gill with a crossword.
09:18 autrijus hm. here's a docs task for people who have a bit time
09:18 autrijus refactor the pod part out of lib/pugs.pm into lib/pugs.pod
09:18 autrijus and in pugs.pod mention both the pugs command line and the pugs.pm module
09:18 autrijus as well as provide links to pugs::run and pugs::hack
09:18 autrijus in the same style as "man perl"
09:18 autrijus anyone got a few seconds to work on this?
09:20 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); [map {$^a[0]}, @foo[0..3]]
09:21 evalbot6 Error:  unexpected "[" expecting ";", statements or end of input
09:21 q[pdcawley] Hmm...
09:21 q[pdcawley] What's wrong with that?
09:21 q[pdcawley] ?eval my @foo = ([1..4], [1..4], [1..4], [1..4]); [map -> $a {$a[0]}, @foo[0..3]]
09:21 evalbot6 [1, 1, 1, 1]
09:21 osfameron_ has joined #perl6
09:21 q[pdcawley] Ah.
09:22 autrijus ?eval {$^a}
09:22 evalbot6 Error:  unexpected "\\" expecting "::"
09:22 autrijus ?eval {$^a[0]}
09:22 evalbot6 Error:  unexpected "\\" expecting "::"
09:22 autrijus it's I think a evalbot bug
09:24 ingy hi autrijus
09:25 autrijus hi ingy
09:25 ingy autrijus: did you see doolittle?
09:26 ingy gugod helped me write the plugin in a couple hours
09:26 ingy http://doolittle.kwiki.org/
09:26 autrijus no.
09:26 ingy anyway, if you want i can put all your modules under it
09:27 ingy and let the community doc them
09:27 autrijus sure
09:27 ingy the changes are saved right back into the modules
09:27 ingy so you can review with svk
09:27 ingy and keep the stuff you like
09:27 ingy it's awesome
09:27 q[pdcawley] Hmm... seems &?CALLER_CONTINUATION doesn't work yet...
09:28 q[pdcawley] Or I've not implemented choose/give_up correctly.
09:28 autrijus q[pdcawley]: t/unspecced/cont.t gives status quo
09:28 q[pdcawley] okay.
09:29 q[pdcawley] So the continuations we have are only escape continuations.
09:29 osfa has quit IRC (Read error: 145 (Connection timed out))
09:29 webmind_ is now known as webmind
09:30 autrijus there's plan to make it full one in my paper :)
09:30 q[pdcawley] How long, oh autrijus, how long?
09:30 autrijus feel free to write :todo<feature> tests
09:30 autrijus I trade feature for tests, as you know.
09:31 q[pdcawley] You asked for it :)
09:32 * q[pdcawley] mutters to himself "So, how to reduce this to the minimum 'useful' example...
09:34 nothingmuch ingy: i'd like that doolitle stuff at least for Test::TAP::...
09:34 nothingmuch you didn't mention before that it's read/write... that's a wonderful idea!
09:34 lightstep has joined #perl6
09:35 nothingmuch well now that i'm dressed, a beer, and i'm off
09:35 nothingmuch ciao guys
09:36 nothingmuch beer is like gatorade, only it tastes good
09:36 nothingmuch has quit IRC ()
09:37 Khisanth has quit IRC (Connection timed out)
09:38 Khisanth has joined #perl6
09:42 q[pdcawley] Ooh, now that's weird...
09:43 q[pdcawley] ?eval sub callcc (Code &block) {&block(&?CALLER_CONTINUATION)}; my $cnt; my $counter=0; callcc -> $c {$cnt = $c}; $counter++; $cnt(5) unless $counter == 3; $counter
09:43 evalbot6 5
09:43 svnbot6 r4597, autrijus++ | * round #4 of test fixes.
09:44 q[pdcawley] ?eval sub callcc (Code &block) {&block(&?CALLER_CONTINUATION)}; my $cnt; my $counter=0; callcc -> $c {$cnt = $c}; $counter++; $cnt(5) unless $counter == 3; $counter
09:44 evalbot6 5
09:44 gaal autrijus, i've got a smoke or roud 3 coming up (in t/pugstun/ now). i'll let it finish, ja?
09:44 gaal * of round
09:44 autrijus gaal: sure
09:45 q[pdcawley] The continuation doesn't even bounce back to the right place...
09:45 autrijus I'll review checngelog now
09:45 gaal oh, make sure to change the very first line
09:45 autrijus sure
09:46 lightstep has quit IRC (Remote closed the connection)
09:46 q[pdcawley] The continuation doesn't even bounce back to the right place...
09:47 q[pdcawley] Argh! Wrong window. Again.
09:51 q[pdcawley] Can I check in a non working example to show the sort of thing that can be done with full continuations?
09:51 svnbot6 r4598, pdcawley++ | Added a simple continuation based loop to test full continuations.
09:51 svnbot6 r4599, autrijus++ | * change hw2005.txt to hw2005.tex
09:52 autrijus q[pdcawley]: sure, check it in as examples/continuation/
09:52 autrijus and clearly mark it as broken
09:53 q[pdcawley] Wilco. I've written a non deterministic brute-force sudoku solver and, though I say so myself as shouldn't, it's gorgeous.
09:53 gaal http://perlcabal.org/~gaal/smoke_xp.html # r4596
09:54 mr_ank non deterministic? where?
09:54 q[pdcawley] About to be checked in to examples/continuation
09:54 mr_ank cool!
09:55 q[pdcawley] It doesn't work yet.
09:55 gaal i've begun a smoke of 4600 as a pre-preflight  measure.
09:57 autrijus ditto here.
09:57 svnbot6 r4600, autrijus++ | * fix test plan numbers (for the release).
09:57 svnbot6 r4601, autrijus++ | * cleanup changelog
09:57 autrijus if looks good, we enter preflight.
09:57 gaal will take 20 more minutes to complete though.
09:57 autrijus gaal: can you handle the pugs.pod thing?
09:57 * autrijus is still relenging
09:57 * gaal backloggeth
09:57 Aankhen`` has joined #perl6
09:58 gaal sure.
09:59 nnunley has joined #perl6
10:04 Khisanth has quit IRC (Connection timed out)
10:04 svnbot6 r4602, pdcawley++ | Added an example of the kind of programming that access to full
10:04 svnbot6 r4602, pdcawley++ | continuations enables. A UI will be forthcoming when we get full
10:04 svnbot6 r4602, pdcawley++ | continuations.
10:05 gaal pugs.pm in the context of pugs docs / pugs.pod can be confusing
10:06 gaal not sure how (wehther) to place it in the SYNOPSIS section of that pod at all
10:07 autrijus oh
10:07 autrijus say that you can use it at command line
10:07 autrijus _and_ as a library in p5 programs
10:07 autrijus give one example of each
10:08 autrijus that I think is okay
10:08 gaal yes, i now have =head1 SYNOPSIS
10:08 gaal B<pugs> [OPTIONS]
10:08 gaal B<perl> -e 'use pugs; # perl 6 code here...'
10:08 gaal btw, it's cute for p6 pods to have a SYNOPSIS section ;-)
10:10 Khisanth has joined #perl6
10:10 svnbot6 r4603, pdcawley++ | A few explanatory comments. Not many...
10:27 elmex has joined #perl6
10:27 vcv-- has quit IRC (Read error: 104 (Connection reset by peer))
10:27 tcliou has quit IRC (Read error: 104 (Connection reset by peer))
10:28 meppl has joined #perl6
10:28 tcliou has joined #perl6
10:28 LCamel has quit IRC (Read error: 104 (Connection reset by peer))
10:35 vcv- has quit IRC (Read error: 145 (Connection timed out))
10:35 svnbot6 r4604, autrijus++ | * round #5.
10:36 gaal ah, which reminds me, new matrix up
10:36 gaal of r4600
10:38 autrijus oh. url again?
10:39 gaal http://perlcabal.org/~gaal/smoke_xp.html
10:39 gaal pod in a minute
10:40 kungfuftr 7666 tests... spooky
10:43 autrijus gaal: hm, line 28 of caller.t failsforyou?
10:43 autrijus gaal: weird
10:43 autrijus it worksforme
10:44 gaal i may have junk in, let me check
10:44 gaal sec
10:45 gaal (running r4604 smoke now)
10:47 gaal yes, it is failing for me.
10:47 gaal btw pugs.pod in
10:47 autrijus gaal: hm. what to do what to do.
10:47 autrijus (it worksforme.)
10:48 autrijus will you change it to skip it for now?
10:48 autrijus is it possible you have unchecked in local changes?
10:48 gaal !? when i prove, it works.
10:48 gaal no, i checked.
10:48 autrijus or Prelude.pm foo?
10:48 gaal when i make test, it doesn't work.
10:48 gaal er, make smoke
10:48 autrijus wtf. :-/
10:48 svnbot6 r4605, gaal++ | pugs POD
10:50 gaal oh, crap
10:50 gaal look at the diff:
10:50 gaal  Expected: 't/builtins/caller.t'
10:50 gaal Got: 't\builtins\caller.t'
10:50 autrijus fix it please :)
10:50 mr_ank recompiling Pugs... dum deh dum...
10:51 gaal and i ran prove in a cygwin box, but make smoke in a dos box :)
10:51 gaal :)
10:52 autrijus nvm, I attempted a fix
10:52 gaal eh, should use $Config directory separator
10:52 * gaal == stoopid
10:52 autrijus I used $?FILE
10:52 gaal heh :)
10:52 autrijus even better eh :)
10:52 gaal that's better
10:53 * gaal chuckles
10:53 gaal though if you do that you'd better add a sanity check for them both not to be undefined.
10:53 gaal (i'll add it)
10:53 svnbot6 r4606, autrijus++ | * bump version. we're now in preflight!
10:53 svnbot6 r4606, autrijus++ | * remove perl5-specific paras from pugs.pod.
10:53 svnbot6 r4607, autrijus++ | * fix platform specifity in caller.t; noticed by gaal++
10:54 gaal ooh, preflight!
10:54 mr_ank q[pdcawley]: your continuations code looks very mysterious to me
10:55 autrijus nah. sanity is tested elsewhere
10:55 gaal k
10:55 gaal got a tarball?
10:56 autrijus a sec
10:56 autrijus running into MANIFEST problems
10:56 autrijus with space in filenames
10:56 gaal in time like these, /me wishes for GHC had a global ccache-like cache.
10:56 gaal s/for//
10:57 autrijus bah. manifest doesn't support spaces in filenames.
10:57 gaal oi
10:57 autrijus so much for cleverness.
10:59 autrijus uploading
10:59 svnbot6 r4608, autrijus++ | * sadly, ignore t/builtins/system/ because MANIFEST cokes
10:59 svnbot6 r4608, autrijus++ |   on filenames with spaces.
10:59 jrieks has joined #perl6
10:59 autrijus yo!
11:03 autrijus changes topic to: Preflight #1 http://perlcabal.org/~autrijus/Perl6-Pugs-6.2.7-pre1.tar.gz | http://pugscode.org/ <Overview Journal Logs> | http://pugs.kwiki.org | paste at http://sial.org/pbot/perl6
11:03 autrijus http://perlcabal.org/~autrijus/Perl6-Pugs-6.2.7-pre1.tar.gz
11:03 autrijus please test :)
11:04 gaal e742d2ce436121ee1b259b0f6aa55b662c4950bf ?
11:04 jrieks [email@hidden.address]
11:04 autrijus gaal: yes
11:04 svnbot6 r4609, gaal++ | minor fixes
11:04 autrijus 910384751a111c918799f87918b6664a too
11:04 gaal it has two sums?
11:04 autrijus gaal: note the length, silly :)
11:05 gaal junctive sums :)
11:05 jrieks 910384751a111c918799f87918b6664a  Perl6-Pugs-6.2.7-pre1.tar.gz
11:05 autrijus jrieks: invitation sent. welcome aboard (as the ship is about to sail again)!
11:05 autrijus jrieks: also add your name to AUTHORS
11:06 gaal btw: do we still need to 'nmake
11:06 gaal before 'nmake smoke'?
11:06 gaal ew, i said nmake
11:06 autrijus probably not
11:06 autrijus ew you said nmake
11:07 gaal hey, why'd it try to buid with embedded p5? :(
11:07 autrijus you have naughty env?
11:08 gaal err
11:08 gaal yes
11:08 * gaal bonks himself on the head
11:08 mr_ank is src/Pugs/Parser.hs the main grammar description?
11:08 mr_ank (or something of the sort?)
11:11 autrijus mr_ank: aye
11:11 Aankhen`` Aye aye!
11:11 * Aankhen`` syncs up and tests.
11:12 mr_ank cool
11:13 mr_ank Is there a standard/automated way of working with large parsers? or do you guys just work on the file directly?
11:13 autrijus that file is not generated
11:13 autrijus if that's what you mean
11:14 mr_ank well, i only have experience with the Dragon Book
11:14 mr_ank so I am outdated
11:14 autrijus I think Parser.hs is actually easy to understand :)
11:14 autrijus (more so than the dragon book that is)
11:15 mr_ank It looks really really cool - I am definitely putting more effort into my abandoned haskell wannabe-skills
11:18 autrijus :D
11:18 autrijus mr_ank: want a committer bit?
11:18 gaal (moving to ext/ tests)
11:19 mr_ank oh i haven't tried commiting yet - do i need it?
11:19 autrijus oh, you're already committer.
11:19 autrijus didn't see your name in AUTHORS.
11:20 autrijus try adding yourself to it?
11:20 mr_ank I haven't commited anything yet... and until then i am not adding my name to the list
11:20 autrijus alright... your choice then :)
11:20 mr_ank thanks anyway :)
11:23 gaal spooky! libwww-perl tests... contact the net! :)
11:24 autrijus it's Alive!
11:24 svnbot6 r4610, jrieks++ | added german perl 6 introduction talk, held at the austrian perl workshop 2005
11:24 gaal (moved to t/; smoke ETA 15 minutes)
11:24 autrijus sofarsogood?
11:24 gaal yup
11:25 gaal oh, eliminating the spaces tests wil no doubt make things faster :)
11:26 autrijus not that I wanted that :)
11:26 gaal say, how "erratic" is async? if it does work for simple things maybe i can hack cross-platform parallelability onto the smoke process?
11:27 autrijus on fbsd here it's pretty nonerratic
11:27 elmex has quit IRC ("leaving")
11:27 autrijus I thought smoke is already parallelisable?
11:27 gaal -j is disabled on windows because p5's fork isn't good enough
11:27 autrijus jrieks: how's your smoke running?
11:28 autrijus ah. you can try to do it with p6.
11:28 autrijus it may be better than perl5's.
11:28 gaal yes: hence, async. :)
11:28 autrijus I think it could work :)
11:29 autrijus but then you want corion for details :) (or just try it out)
11:29 gaal yes, Corion's missed! i wonder what he's up to.
11:29 gaal brb
11:34 gaal still lookin' good.
11:34 iblechbot has joined #perl6
11:42 gaal http://perlcabal.org/~gaal/smoke_xp.html # preflight 1
11:42 gaal 7341 test cases: 7341 ok, 0 failed, 986 todo, 492 skipped and 5 unexpectedly succeeded
11:42 autrijus mm 5 unTODOs!
11:42 gaal hate it when that happens :)
11:42 gaal mm 0 fails! :)
11:43 autrijus love it when that happens :)
11:44 gaal (cut'n'pasted test descriptions)--
11:45 autrijus it failsforme.
11:45 gaal oh. :( i was just about to ci
11:46 autrijus probably don't. :(
11:46 gaal what env?
11:46 autrijus fbsd
11:46 autrijus I think I'll skip for now and investigate after release.
11:46 autrijus worksforyou?
11:46 gaal huh!? it's win32 tests
11:46 autrijus no it's testing File::Spec::Win32
11:46 gaal why are you even running that? :)
11:46 gaal ahhhh
11:46 gaal ok
11:46 autrijus but I bet there's platform specificness.
11:47 gaal yes, well, i'd call it exactly a bug that hasn't been fixed yet then. :-/
11:47 svnbot6 r4611, autrijus++ | * skip errartically succeeding tests in 20_win32 until after release.
11:47 autrijus right.
11:47 autrijus otherwise all fine?
11:47 gaal yes!
11:48 autrijus good.
11:48 * autrijus waits for jrieks to finish smoking
11:49 gaal incidentally, there's something i haven't told you
11:49 autrijus oh?
11:49 * autrijus braces for impact
11:49 gaal i am not left^H^H^H^Husing ghc 6.4
11:49 gaal 6.5 snapshot
11:49 autrijus that's fine
11:49 autrijus shouldn't matter.
11:50 Aankhen`` Left...?
11:50 gaal http://www.google.com/search?hl=en&amp;lr=&amp;q=%22i+am+not+left+handed&amp;btnG=Search
11:50 Aankhen`` Ah. :-)
11:51 gaal google golf tip: you can search for "precise expression without closing the quotes
11:51 Aankhen`` :-o
11:52 gaal and precise expression*s* as well!
12:06 jrieks autrijus: finished! no failure
12:06 jrieks just this: t/builtins/io/system.t
12:06 jrieks /bin/sh: line 1: program_that_does_not_exist_ignore_errors_please.exe: command not found
12:06 gaal that's okay
12:06 jrieks yep, okay :-)
12:06 Limbic_Region has joined #perl6
12:10 meppl has quit IRC (Remote closed the connection)
12:10 vcv-- has joined #perl6
12:10 vcv- has joined #perl6
12:12 scw has quit IRC (Read error: 60 (Operation timed out))
12:14 gaal is a pf#2 cooking?
12:18 autrijus no, a release is coming :)
12:19 gaal :)
12:19 * gaal tails PAUSE
12:19 autrijus so have you tried trunk's win32 t?
12:19 autrijus just running that file alone
12:19 gaal sec
12:19 autrijus and make sure it's skipped instead of todo but passes
12:20 gaal oooh, new talk
12:21 gaal yes, it's good.
12:22 autrijus ok. fly!
12:22 gaal woohoo
12:22 spinclad has quit IRC (Read error: 110 (Connection timed out))
12:26 pelagic has joined #perl6
12:26 svnbot6 r4612, autrijus++ | * This be Pugs, version 6.2.7.
12:27 spinclad has joined #perl6
12:31 Juerd re!
12:31 gaal Juerd! long time no.
12:31 Juerd Hi
12:31 svnbot6 r4613, iblech++ | * docs/talks/perl6-apw2005, lib/pugs/hack.pod, examples/continuation, t/:
12:31 svnbot6 r4613, iblech++ |   Usual svn properties.
12:31 svnbot6 r4613, iblech++ | * docs/talks/README: Added filenames of the talks.
12:31 svnbot6 r4613, iblech++ | * lib/pugs/hack.pod: Reindented so fewer lines wrap on a 80 char display,
12:31 svnbot6 r4613, iblech++ |   filled SYNOPSIS section.
12:32 Juerd I've been away this week end
12:34 Aankhen`` All tests successful (Win32, MSVC 7).
12:34 autrijus good!
12:34 * autrijus uploadeth
12:34 * gaal cheereth
12:35 Aankhen`` Yowza.
12:35 gaal PAUSE sendeth email(s) about uriid[A/AU/AUTRIJUS/Perl6-Pugs-6.2.7.tar.gz]
12:35 Aankhen`` Sudden flurry of added files?
12:35 gaal a talk, A.
12:35 Aankhen`` That I saw. :-)
12:35 Aankhen`` Umm... I compiled r4563.  Is that very old?
12:36 Aankhen`` Er, wait.
12:36 * Aankhen`` wonders where he got that number.
12:36 gaal kinda old.
12:36 Aankhen`` Hrm, right, I got that from Pugs, nevermind. :-
12:36 Aankhen`` P
12:36 gaal we're at 4613
12:36 Aankhen`` Apparently so...
12:36 autrijus it's ancient :)
12:37 Aankhen`` Heh.
12:37 gaal '63 is this late this morning, i think.
12:37 Aankhen`` What can I say?  60 commits since I synced...
12:37 Aankhen`` Hrm.
12:37 Aankhen`` It took 1 hour 20 minutes to build and test.
12:37 Aankhen`` O_O
12:38 Aankhen`` And I ran `nmake unoptimized` this time...
12:38 gaal on your cuspy new box, A?
12:38 gaal aaaah, that's why
12:38 Aankhen`` Yep.
12:38 Aankhen`` Indeed.
12:38 gaal if you make test/smoke, optimize pays off.
12:38 gaal takes about 45 mintes here.
12:38 autrijus ditto here.
12:38 gaal minutes.
12:38 Aankhen`` I know, I dunno why I decided to do the unoptimized version this time.
12:39 gaal you were eager for tests to show up?
12:39 Aankhen`` Not really.  I just suddenly suffered from the delusion that it would speed up the whole process.
12:39 gaal "make faster"
12:39 Aankhen`` Ah well.
12:39 Aankhen`` I'll let it run again.
12:39 pelagic has left
12:40 Aankhen`` Hmm.
12:40 Aankhen`` Strange.
12:40 Aankhen`` I wonder how come I have 4563... I coulda sworn I synced up just before building.
12:41 gaal svn info?
12:41 * Aankhen`` wonders if it picked up the old pugs.exe in C:\Perl6\bin.
12:41 Aankhen`` No point now, I resynced already.
12:41 iblech has joined #perl6
12:41 * Aankhen`` scrolls up to check.
12:42 gaal i recall there's some amount of lieage in the code that embeds the r number in the executable.
12:42 Aankhen`` Ah.
12:42 Aankhen`` I was at r4609.
12:42 calvin_ has joined #perl6
12:42 Aankhen`` (that was the last revision before "* Aankhen`` syncs up and tests")
12:43 cognominal has quit IRC (Remote closed the connection)
12:44 iblech autrijus: http://linide.sf.net/pugs-svngraph-6.2.7.png and http://linide.sf.net/pugs-livecd-6.2.7.png (will be up in ten minutes or so)
12:44 Aankhen`` Doesn't look like there were any changes to Pugs itself between then and now.
12:45 gaal iblech, what's the green line stand for?
12:45 Aankhen`` I was just about to ask that. :-
12:45 Aankhen`` P
12:45 gaal i think i ask this every time and then every time forget :)
12:46 gaal would be nice to also have number of test cases on that graph, but that isn't trivial to implement...
12:49 iblech The green line is the num of developers, but scaled *50
12:50 iblech Yeah, num of test cases would be neat
12:50 autrijus iblech: thanks! feel free to post followup to p6c :)
12:50 autrijus but it's livecd.iso right
12:50 autrijus not .png :)
12:50 iblech Err, yes :)
12:51 gaal [+] grep { /plan (\d+)/; $1 }, qx<grep -r plan t/>; # modulo bugs
12:52 Juerd s/grep/map/
12:52 Juerd I think
12:53 Juerd Hm, both.
12:53 gaal neither, this shouldn't shell out to gnu grep :)
12:53 gaal = on an iterator that recurses down t from perl
12:53 cognominal has joined #perl6
12:54 Juerd [+] map { /plan (\d+)/ ?? $1 :: 0 }, =slurp glob 't/**/*' then? :)
12:54 Aankhen`` gaal >> C<^plan > might be better...
12:54 Juerd Assuming a glob that recurses for **
12:54 gaal indeed
12:54 Aankhen`` Or:
12:54 Aankhen`` C<^\s*plan >
12:54 Aankhen`` Oops, didn't Backspace enough.
12:55 Aankhen`` ^\s*plan
12:56 gaal autrijus, the title of your release message refers to 6.2.6 :)
12:56 autrijus oh wow.
12:56 autrijus ;)
12:56 larsen has quit IRC ("later")
12:57 gaal oh - also it's %ENV<>, not $ENV :-/
12:59 Juerd $ENV =
12:59 Juerd $ENV = %ENV;
12:59 Juerd Now it is :)
12:59 gaal ?
13:00 Aankhen`` Heh.
13:00 Juerd gaal: $ENV<FOO> can work too, as long as $ENV is a hashref ;)
13:00 gaal i don't get it, but that's probably just a sign i should go have lunch
13:01 autrijus gaal: Juerd assigns %ENV into $ENV.
13:01 autrijus gaal: he's basically rogering with your mind :)
13:01 gaal ah. but pugs doesn't make that assunction.
13:01 autrijus you can do it in pugs too
13:01 autrijus it's just you have to write it by hand.
13:01 autrijus like:
13:01 autrijus $*ENV = %*ENV;
13:01 Juerd Taking the world as it is doesn't fit a programmer's mind
13:02 gaal $ENV := sub {%*ENV} is rw;
13:02 Juerd We change the world until it's right :)
13:02 * gaal eats lunch &
13:02 Juerd See? You do that too
13:02 autrijus o/~ freedom defined is freedom denied o/~
13:02 Juerd Indeed
13:03 hcchien finally pushed
13:03 autrijus hcchien++
13:04 Aankhen`` Er... autrijus, small nitpick: it's "coerce:", not ":coerce".
13:04 autrijus Aankhen``: fix it :)
13:04 autrijus (I didn't add that changelog entry, fwiw)
13:04 * Aankhen`` goes back in time and warns autrijus before the mail is sent!
13:04 Aankhen`` OK.
13:04 * Aankhen`` edits.
13:05 * Aankhen`` wonders why it mentions HTTP::Message as a new bundled module twice.
13:05 svnbot6 r4614, hcchien++ |  r11843@Mahler:  hcchien | 2005-06-13 19:27:12 +0800
13:05 svnbot6 r4614, hcchien++ |  add some tests for rx_perl5_(escape|match)
13:06 scw has joined #perl6
13:08 scook0 has quit IRC ("Leaving")
13:08 Aankhen`` Was parsing speed increased by 100% or 200%?
13:09 knewt me-- # not checking autrijus' key fingerprint whilst we were both in vienna
13:15 Aankhen`` ?eval 1^..10
13:15 evalbot6 Error:  unexpected "\\" expecting digit, fraction, exponent, term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:15 svnbot6 r4615, Aankhen++ | * ChangeLog: fixed ":coerce" -> "coerce:", fixed another small typo
13:15 svnbot6 r4615, Aankhen++ | * removed "HTTP::Message" from the list of new modules in 6.2.7, since it was already listed under 6.2.6
13:15 svnbot6 r4616, scw++ | Update debian control file to v6.2.7
13:15 Aankhen`` ?eval ^1..10
13:15 evalbot6 Error:  unexpected "^" expecting term
13:15 Aankhen`` How does ^.. work?
13:16 mr_ank ?eval (1..10).perl
13:16 evalbot6 '(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)'
13:16 Aankhen`` ?eval 1..10
13:16 evalbot6 (1, 2, 3, 4, 5, 6, 7, 8, 9, 10)
13:16 Aankhen`` ?eval 1^..^10
13:16 evalbot6 Error:  unexpected "\\" expecting digit, fraction, exponent, term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:17 Aankhen`` ?eval 1 ^..^ 10
13:17 evalbot6 Error:  unexpected "\\" expecting operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:17 Aankhen`` ?eval infix:<+>(1, 3)
13:17 evalbot6 4
13:18 Aankhen`` ?eval infix:<^..^>(1, 5)
13:18 evalbot6 Error: No compatible subroutine found: "&infix:\\^..\\^"
13:19 mr_ank ?eval (1,2,3).perl
13:19 evalbot6 '(1, 2, 3)'
13:19 mr_ank <1,2,3>.perl
13:19 mr_ank ?eval <1,2,3>.perl
13:19 evalbot6 '\'1,2,3\''
13:19 Aankhen`` ?eval <1, 2, 3>.perl
13:19 mr_ank hm
13:19 iblech Aankhen``: It's the old-evalhelper-doesn't-cope-well-with-^
13:19 evalbot6 '(\'1,\', \'2,\', \'3\')'
13:19 mr_ank is that right?
13:19 Aankhen`` iblech >> Ah.
13:19 Aankhen`` mr_ank >> <1,2,3>.perl? Yes.
13:20 Aankhen`` <1,2,3> turns into a list with a single element: the string '1,2,3'.
13:20 Aankhen`` ?eval 1..10.ref
13:20 evalbot6 ()
13:20 cognominal_ has joined #perl6
13:21 mr_ank ah damn you are right
13:21 mr_ank silly me
13:21 mr_ank and:
13:21 mr_ank ?eval <1 2 3>.perl
13:21 evalbot6 '(\'1\', \'2\', \'3\')'
13:22 mr_ank hmm
13:22 jrieks_ has joined #perl6
13:23 iblech ?eval Pugs::Safe::safe_print(<1 2 3>.perl)
13:23 evalbot6 ('1', '2', '3')bool::true
13:23 jrieks_ ?eval "hello, world!".say
13:23 evalbot6 Error: No compatible subroutine found: "&say"
13:23 iblech ?eval <1 2 3>   # evalbot already .perls the result
13:23 evalbot6 ('1', '2', '3')
13:23 Aankhen`` ?eval "foo".perl.perl.perl
13:23 evalbot6 '\'\\\'\\\\\\\'foo\\\\\\\'\\\'\''
13:23 jrieks_ ?eval <1 2 3>xx99
13:23 evalbot6 Error:  unexpected "9" or "x" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:23 mr_ank heh
13:24 jrieks_ ?eval <1 2 3>x99
13:24 evalbot6 Error:  unexpected "9" or "x" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:24 jrieks_ ?eval (1,2,3)x9
13:24 jrieks_ ?eval (1,2,3)xx9
13:24 evalbot6 Error:  unexpected "9" or "x" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:24 evalbot6 Error:  unexpected "9" or "x" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:24 iblech jrieks_: Most of IO is disallowed in safemode/evalbot, only Pugs::Safe::safe_print works (which always writes to STDOUT)
13:24 Aankhen`` ?eval (1, 2, 3) xx 99
13:24 evalbot6 (1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3
13:24 jrieks_ ah okay, thx
13:25 iblech And, as you see, the output was truncated -- i.e. "x" xx 100000 will only output 500 or so chars
13:25 iblech ?eval while 1 {} # and there's the CPU time limitation
13:25 evalbot6 (no output)
13:26 iblech ?eval my $x = "x" x 100000000; bytes $x # and RAM is limited, too
13:26 evalbot6 pugs: out of memory (requested 1048576 bytes)
13:26 jrieks_ nice :-)
13:26 autrijus :)
13:27 jrieks_ what about forking?
13:27 jrieks_ ?eval while 1 { async {while 1 {}} }
13:27 evalbot6 pugs: user error (Cannot create OS thread.)  
13:27 Limbic_Region ahh - another bird brain - welcome jrieks_
13:27 jrieks_ ah okay :-)
13:28 jrieks_ bird brain? :-p
13:28 Limbic_Region Parrot
13:28 jrieks_ ah, yes of course :-)
13:29 jrieks_ ?eval for 1... {}
13:29 evalbot6 pugs: out of memory (requested 1048576 bytes)
13:30 jrieks_ thats a but, no?
13:30 mr_ank ?eval my Array of Int @ints
13:30 evalbot6 Error:  unexpected "o" expecting variable name or "("
13:31 Aankhen`` ?eval my @ints of Int;
13:31 evalbot6 Error:  unexpected "o" expecting "=", ":=", "::=", ";" or end of input
13:31 iblech You mean it should die because of CPU time limitation instead of mem? Yep, we don't have lazy lists currently
13:31 jrieks_ yes, okay
13:31 jrieks_ ?eval sub infix:?>?? ($a, $b) { $a > $b ?? $a :: $b }
13:31 evalbot6 Error:  unexpected "i" expecting subroutine parameters or block
13:32 jrieks_ ?eval sub infix:<<>?>> ($a, $b) { $a > $b ?? $a :: $b }
13:32 evalbot6 Error:  unexpected "?" expecting bare trait, subroutine parameters, trait or block
13:32 jrieks_ ?eval sub infix:<< >? >> ($a, $b) { $a > $b ?? $a :: $b }
13:32 evalbot6 Error:  unexpected "?" expecting bare trait, subroutine parameters, trait or block
13:32 jrieks_ is it possible to define an operator containing < or > without using unicode?
13:32 jrieks has quit IRC (Read error: 110 (Connection timed out))
13:33 jrieks_ is now known as jrieks
13:33 iblech It should work
13:33 iblech ?eval sub infix:«>?» ($a, $b) { $a > $b ?? $a :: $b }
13:33 evalbot6 Error:  unexpected "i" expecting subroutine parameters or block
13:33 Aankhen`` ?eval sub infix:'>?' { ($a, $b) ? }
13:33 evalbot6 Error:  unexpected "i" expecting subroutine parameters or block
13:33 Aankhen`` Oops.
13:33 Aankhen`` ?eval sub infix:'>?' ($a, $b) { $a > $b ?? $a :: $b }
13:33 evalbot6 Error:  unexpected "i" expecting subroutine parameters or block
13:34 cognominal has quit IRC (Read error: 110 (Connection timed out))
13:35 jrieks ?eval sub infix:<<<?>>($a,$b){$a > $b ?? $a :: $b}
13:35 evalbot6 undef
13:35 jrieks ?eval sub infix:<<>?>>($a,$b){$a > $b ?? $a :: $b}
13:35 evalbot6 Error:  unexpected "?" expecting bare trait, subroutine parameters, trait or block
13:38 kungfuftr someone send me a kotatsu!
13:38 jrieks ?eval sub infix:<<more?>>($a,$b){$a > $b ?? $a :: $b} sub postfix:<!> ($x) { [*] 1..$x more? 1 } sub e {my $ret = 0;for 0..1000 -> $i { my $old = $ret; $ret += 1 / $i!; return $ret if $ret-$old < 1e-50 } } "e={e()}";
13:38 evalbot6 'e=2.7182818284590452353602874713526624977572'
13:38 lumi_ kungfuftr: You're getting cold feet?
13:39 jrieks ?eval "test"
13:39 evalbot6 'test'
13:39 jrieks ?eval "/quit test :-)"
13:39 evalbot6 '/quit test :-)'
13:39 jrieks ?eval "\n/quit test :-)"
13:39 evalbot6 ' /quit test :-)'
13:40 mr_ank hmm
13:40 * jrieks thinks that the bot is safe enough ;-)
13:40 iblech Won't work, as you'd have to send a "QUIT msg" to quit the IRC connection, but evalbot always prepends "PRIVMSG #perl6 :", so the reply reaches us
13:40 mr_ank i've just tried evaluating this: (1).kv and pugs never returned
13:41 jrieks ?eval (1).kv
13:41 evalbot6 pugs: out of memory (requested 1048576 bytes)
13:41 mr_ank bug?
13:41 mr_ank ?eval (1,2).kv
13:41 evalbot6 (0, 1, 1, 2)
13:42 mr_ank ?eval ().kv
13:42 evalbot6 ()
13:42 Limbic_Region interesting
13:42 Limbic_Region ?eval (1,2,3).kv
13:42 evalbot6 (0, 1, 1, 2, 2, 3)
13:42 mr_ank i'm just bored and playing around a bit with odd constructs :)
13:42 Limbic_Region the out of memory one is likely a bug
13:42 mr_ank ?eval ('a').kv
13:42 jrieks ?eval (1..9999).kv
13:42 evalbot6 pugs: out of memory (requested 1048576 bytes)
13:42 evalbot6 (0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, 12, 12, 13, 13, 14, 14, 15, 15, 16, 16, 17, 17, 18, 18, 19, 19, 20, 20, 21, 21, 22, 22, 23, 23, 24, 24, 25, 25, 26, 26, 27, 27, 28, 28, 29, 29, 30, 30, 31, 31, 32, 32, 33, 33, 34, 34, 35, 35, 36, 36, 37, 37, 38, 38, 39, 39, 40, 40, 41, 41, 42, 42, 43, 43, 44, 44, 45, 45, 46, 46, 47, 47, 48, 48, 49, 49, 50, 50, 51, 51, 52, 52, 53, 53, 54, 54, 55, 55, 56, 56, 57, 57, 58, 58, 59, 59,
13:43 Limbic_Region ?eval (42).kv
13:43 evalbot6 pugs: out of memory (requested 1048576 bytes)
13:43 jrieks ?eval ('perl', 6).kv
13:43 evalbot6 (0, 'perl', 1, 6)
13:43 mr_ank looks like something of an off-by-one error
13:43 jrieks ?eval {'perl'=> 6}.kv
13:43 evalbot6 ('perl', 6)
13:43 jrieks ?eval (0..9).kv
13:43 evalbot6 (0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9)
13:44 mr_ank hm
13:44 mr_ank ?eval (0..1).kv
13:44 evalbot6 (0, 0, 1, 1)
13:44 mr_ank ?eval (0..2).kv
13:44 evalbot6 (0, 0, 1, 1, 2, 2)
13:44 jrieks ?eval (-1..1).kv
13:44 evalbot6 (0, -1, 1, 0, 2, 1)
13:44 mr_ank wtf
13:44 jrieks ?eval (-2..2).kv
13:44 evalbot6 (0, -2, 1, -1, 2, 0, 3, 1, 4, 2)
13:44 jrieks (index, value) pairs
13:45 Limbic_Region ?eval -2..2
13:45 evalbot6 (-2, -1, 0, 1, 2)
13:45 mr_ank yeah, but it's inconsistent
13:45 jrieks why?
13:45 jrieks ?eval 2..-2
13:45 evalbot6 ()
13:45 mr_ank look at this:
13:45 jrieks ?eval 2..-2 :by(-1)
13:45 evalbot6 Error:  unexpected ":" expecting operator, "does", "but", "is", "^..^", "cmp", "<=>", "^..", "..^", "..", postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:45 mr_ank ?eval (-1..1).kv
13:45 evalbot6 (0, -1, 1, 0, 2, 1)
13:45 svnbot6 r4617, Aankhen++ | * ChangeLog: typo fixes, reworded here and there.
13:46 mr_ank ?eval (0..2).kv
13:46 evalbot6 (0, 0, 1, 1, 2, 2)
13:46 mr_ank how come?
13:46 jrieks where is the problem?
13:46 jrieks (0, -1), (1, 0), (2, 1)
13:46 jrieks first is the key (array index), the second is the value
13:46 jrieks key/value
13:46 mr_ank ah
13:46 mr_ank AH
13:46 mr_ank i got it now
13:47 mr_ank was blinded by seeing (a a b b c c)
13:47 jrieks ?eval ("a"..'c')xx2
13:47 evalbot6 Error:  unexpected "2" or "x" expecting term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
13:47 jrieks ?eval ("a"..'c') xx 2
13:47 evalbot6 ('a', 'b', 'c', 'a', 'b', 'c')
13:48 mr_ank so (1).kv should return (0,1) - should we write test cases for these things that hang the compiler?
13:48 elmex has joined #perl6
13:48 jrieks just do it
13:48 jrieks I've not time atm
13:49 mr_ank okay
13:49 nnunley_ has joined #perl6
13:49 jrieks cya!
13:49 mr_ank byTe
13:50 gaal oh, Aankhen``, mea culpa regarding ChangeLog spellos.
13:50 jrieks has quit IRC ("Error 666: connection reset by satan")
13:51 Odin- has quit IRC ("Zoink.")
13:51 Odin- has joined #perl6
13:55 kungfuftr lumi_: yar
13:57 cognominal_ is now known as cognominal
13:57 nnunley has quit IRC (Read error: 113 (No route to host))
14:03 jhorwitz has joined #perl6
14:10 fireartist has joined #perl6
14:18 mr_ank i'll ask again just in case:
14:19 mr_ank Should I write test cases for things that hang pugs? (like (1).kv )
14:20 mr_ank (i just wrote one and put it in t/data_types - just not sure whether to commit it or not)
14:21 PerlJam mr_ank: Can you wrap that in an eval or something with a timeout so that it doesn't hang indefinitely?
14:21 vel has joined #perl6
14:21 mr_ank PerlJam: sure
14:22 mr_ank I'll give it a shot
14:22 PerlJam If so, then I think you should definitely write tests for things like that.
14:22 mr_ank is there a way to easily wrap it in a timeout?
14:23 mr_ank cause pugs just "goes on a trip" when it tries to evaluate (1).kv
14:27 mr_ank or perhaps I should just put it in t_disabled/? or is that for something else?
14:27 Aankhen`` has quit IRC ("Reboot [Time wasted online: 4hrs 30mins 4secs]")
14:28 lhooq has joined #perl6
14:29 geoffb mr_ank, t_disabled appears to be for things that aren't (yet) relevant, not merely things that are broken.
14:29 geoffb Non-standard Perl dialects, for instance
14:29 mr_ank ah
14:30 mr_ank I have no idea how to implement a pugs-level timeout
14:30 geoffb try t/pugsbugs/, I would think
14:31 geoffb Haven't looked myself, but perhaps mentioned in the networking examples?
14:31 mr_ank i was just looking at that
14:32 iblech I'd write "fail "(1).kv hangs forever"; #is ~(1).kv, "0 1", "(1).kv works";" or something like that
14:33 iblech As far as I know, there's currently no way to implememt timeouts
14:33 mr_ank right. They seem to be doing it with loops and counters
14:33 mrborisguy has joined #perl6
14:33 mr_ank if(time() - $last_traffic >= $live_timeout) { # and things like that
14:34 mr_ank iblech: okay
14:34 mr_ank in t/pugsbugs then
14:34 fireartist has quit IRC ("Leaving")
14:34 Shachaf has joined #perl6
14:35 geoffb what does 'UU' before a filename (as opposed to just 'U') mean in 'svk pull' output?
14:35 sapper has joined #perl6
14:35 iblech Yep. But this code still has a problem: The readline() of the socket to the server is blocking, that means, that Net::IRC looses control once it readline()s. (BTW, this is the reason why sometimes svnbot has a fairly long delay.)
14:35 clkao content and property changed
14:35 autrijus both content and metadata.
14:35 geoffb gotcha, thx
14:35 gaal it's short for "uberupdate".
14:36 dada has quit IRC (Read error: 60 (Operation timed out))
14:39 q[pdcawley] has quit IRC (Read error: 113 (No route to host))
14:40 * gaal reads perlintro for inspiration and surrenly realizes that the description "line noise" is getting clear to fewer and fewer newcomers.
14:41 wwalker has left
14:43 * PerlJam really likes slide #6 of apw
14:46 geoffb iblech++ # help with hack.pod
14:46 svnbot6 r4618, geoffb++ | New examples/continuation directory added to source map
14:46 svnbot6 r4619, mr_ank++ | Added failing test for (1).kv and ('a').kv (will hang Pugs, so it's commented)
14:46 geoffb There are a number of XXXX's in hack.pod, any help to excise those would be appreciated
14:47 wolverian agh, I have to cron this. I'm two hundred revisions behind again.
14:49 geoffb wolverian, make it part of your morning routine, it will instill good practices for a lifetime of work.  :-)
14:49 wolverian :)
14:50 svnbot6 r4620, iblech++ | * Removed empty docs/advocacy/, and updated pugs::hack accordingly.
14:50 svnbot6 r4620, iblech++ | * Usual svn props added to t/rules/rx_perl5_escape.t and
14:50 svnbot6 r4620, iblech++ |   t/pugsbugs/one_elem_list_kv.t.
14:52 gaal okay, i think i know how to fix the "is trait + returns" bug, but it looks hacky to me
14:54 gaal basically after the SubParameters there can be many Traits and possibly one BareTrait "returns"
14:54 gaal but they can be interleaved
14:54 gaal so that means two more lines: a typ''' and a traits'
14:55 gaal isn't there a more elegant way to express this?
14:55 dkf has joined #perl6
14:56 mr_ank iblech: properties i should be setting on new files are svn:mime-type and svn:eol-style right?
14:56 svnbot6 r4621, mr_ank++ | Fixed silly typo in test
14:56 gaal also, shouldn't multiple "returns" be an error? eg sub foo returns X (params) returns Y
14:56 gaal currently taken to mean returns Y (X silently dropped)
14:56 iblech mr_ank: Yep, see docs/other/development_tests
14:57 iblech mr_ank: But it's no problem if you forget to do so
14:57 mr_ank okay :)
14:58 geoffb Why are the docs in lib/ not in docs/ instead?
14:59 iblech I think two reason are that perldoc pugs::run works and that the docs are nicely formatted on search.cpan.org
14:59 geoffb ah!
15:00 geoffb Though that just makes me think perldoc and search.cpan.org need to have their horizons widened
15:01 fireartist has joined #perl6
15:01 iblech Yep, search.cpan.org should index blib6/ and docs/, too
15:01 gaal blib6/ ? does that exist in the source distro?
15:02 geoffb gaal, nope
15:03 fireartist anyone know what 'mode change "+e" for user fireartist by services' does / why it's been done?
15:04 fireartist (that's an irc message btw)
15:04 integral +e is the flag you get when you identify to nickserv
15:05 fireartist integral: I'd just never noticed it before, if it's normal though, I'll not worry!
15:06 Enveigler1 has joined #perl6
15:06 * geoffb is beginning to get an 'svk pull' twitch -- whenever he's spacing out, his fingers enter 'svk pull' into an xterm apparently by themselves
15:07 integral has quit IRC (brown.freenode.net irc.freenode.net)
15:09 geoffb clkao, is the plan that eventually svk will be able to interact with all open-source source control systems?
15:09 integral has joined #perl6
15:10 clkao if people want to, yes. but i don't have much incentive for that
15:11 geoffb What's working so far -- svn, cvs, and darcs?
15:11 clkao s/darcs/perforce
15:11 geoffb ah
15:12 kungfuftr clkao: nihonji?
15:13 Aankhen`` has joined #perl6
15:13 geoffb clkao, so were the current choices "commonly used systems that insist on having live access to a server"?
15:15 clkao ya. but i might have git support rsn
15:16 geoffb mmm, git
15:16 geoffb via cogito (or whatever that morphed into, haven't kept up), or directly?
15:16 clkao directly.
15:16 clkao cogito is a stinky pile of bash script
15:17 geoffb ah!
15:17 geoffb Well, I only ever read the summaries, so I missed the details
15:17 clkao i missed too, so i didn't know that until yestetrrday i actually try to use it
15:18 * geoffb wishes the kernel-traffic summaries had just a tad more focus on technical details, rather than just arguments on various topics
15:21 Boogie_ somebody knows, if it's normal, that Pugs cannot 'return($sth)', where $sth is an SQL statement handler from DBI--perl5? Or, to be more precise, it can give it back, but say $sth.rows, or $sth.finish() will give a segmentation fault, while it works fine in the subroutine
15:22 autrijus hrm, it's not normal at all
15:22 autrijus write... tests! :)
15:24 * autrijus goes back writing AST for PIR
15:24 Boogie_ yes, but it's hard to write a test for it, when there's no MySQL around (e.g. on your machine). Anyway, you're right.
15:24 autrijus try using other XS objs.
15:24 autrijus internal ones.
15:24 Boogie_ it's over my knowledge, but I'll try it.
15:25 autrijus like (I don't know) Digest::MD5 object.
15:25 Boogie_ ok.
15:25 autrijus or IO::Handle, or something. improvise :)
15:25 Boogie_ thx
15:25 geoffb Improvise?!?  We'd never do that, autrijus!  This is strictly a waterfall development model here.
15:26 geoffb I mean, REALLY.
15:26 geoffb The NERVE of such a suggestion . . . .
15:26 mr_ank so... autrijus has the biggest bucket?
15:26 geoffb Man, I am clearly PUNCHY
15:26 geoffb mr_ank, LOL
15:26 autrijus lol
15:26 autrijus yeah sure.
15:27 * autrijus DECREES that geoffb MUST IMPROVISE!
15:27 autrijus there. better?
15:27 mr_ank and that's why i'm holding a tea cup!
15:27 geoffb ROFL
15:27 geoffb MUCH
15:27 autrijus you're like a forth machine!
15:27 autrijus (only accepting command words in ALL CAPS)
15:28 * geoffb is still laughing
15:28 mr_ank really?
15:28 mr_ank you need sleep geoffb!
15:28 geoffb yes, up too late and too early.  Damn warm nights
15:29 mr_ank i haven't slept yet ... need to finish a C++ XS API *WITH THREADS*
15:29 geoffb ewww
15:29 mr_ank (the thread bit im not touching for now)
15:30 * geoffb always dreads "Oh, you know XS?  Here, I need some help with this . . . ."
15:30 geoffb good thinking, mr_ank
15:30 mr_ank it's sorta like a DBI driver
15:31 mr_ank but a lot simpler ... and i cheated -- instead of returning a handle, i just wait till the operation is done - threading is nicely encapsulated... im not going there!!!
15:31 geoffb heh, well, that's one way to do it
15:31 autrijus "our interface encapsulates 16,384 objects"
15:31 autrijus "it's easy to do that because they aren't there"
15:32 mr_ank something like that... i hate these weird C/C++ mixes SO MUCH
15:32 autrijus bbiab... need !irc to hack :)
15:32 geoffb nod
15:34 * geoffb away too . . . about time I woke up the family by changing the cat litter, bleah
15:36 ninereasons has joined #perl6
15:39 andynu has joined #perl6
15:50 fireartist has quit IRC ("Leaving")
15:50 svnbot6 r4622, andras++ | XS object will be bad after returning it - test
15:51 Boogie_ somebody can check my "t/unspecced/p5/return.t" if it's really a test (I mean, the file runs fine)?
15:54 pasteling "Aankhen``" at 203.101.4.113 pasted "t/unspecced/p5/return.t" (6 lines, 251B) at http://sial.org/pbot/11013
15:54 autrijus checking
15:54 autrijus Aankhen``: not built with embedperl5?
15:54 autrijus Boogie_: you need to check p5 embeddedness just like other p5/* tests.
15:55 autrijus see eg. basic.t for example of how to use skip_rest.
15:55 Aankhen`` autrijus >> Affirmative.
15:57 iblech Boogie_: Fixed (note that I had to put the check and plan() in a BEGIN block, so Pugs doesn't die because it couldn't load the Perl 5 module (on non Perl 5-embedded builds))
15:57 svnbot6 r4623, iblech++ | * t/subroutines/sub_ref.t -- Tests that macros are Codes, Routines, and Macros.
15:57 svnbot6 r4623, iblech++ | * Pugs.Parser -- Parse anonymous macros.
15:57 svnbot6 r4624, iblech++ | t/unspecced/p5/return.t -- Usual svn props and test that Perl 5 is embedded.
15:58 Enveigler1 Are  there any papers/discussion around showing STM being used on boxes with 1 or 2 cpus rather than massively parallel setups?
15:58 Boogie_ iblech: you was faster than me... :) I've tried to commit it and saw that somebody did it before. :)
15:58 Boogie_ autrijus: ok
15:59 Boogie_ s/was/were/
15:59 vcv- has quit IRC (Read error: 145 (Connection timed out))
15:59 autrijus Enveigler1: the SMP capability is currently in GHC trunk
15:59 autrijus Enveigler1: so I don't think benchmarks will make sense until say 6.4.1's release
16:00 autrijus Enveigler1: but in other VMs there may already be SMP support
16:01 Enveigler1 I was looking for more general discussion on STM rather GHC specific. Everything I've found so far show use with 20, 40, 60 + cpu's, which is noce in that world, but I'm wondering how it performs in a more everyday context?
16:07 autrijus nod. I havn't seen literature about that either.
16:07 autrijus maybe google scholar a bit more can dig up stuff
16:07 goofyheadedpunk has joined #perl6
16:07 ninereasons ?eval @ARGS[0] ~~ /(nothing)/; +@ARGS
16:07 evalbot6 Error: Can't modify constant item: VUndef
16:08 iblech ?eval @ARGS   # @ARGS isn't available in safemode
16:08 evalbot6 \undef
16:08 iblech ?eval @ARGS[0]   # @ARGS isn't available in safemode
16:08 evalbot6 Error: Can't modify constant item: VUndef
16:08 ninereasons thanks, iblech.
16:09 Enveigler1 Autrijus: Okay. I'll keep digging. Thanks.
16:09 ninereasons ?eval my @a; @a[0] ~~ /(nothing)/; +@a
16:10 evalbot6 pugs: src/events.c:265: init_events_first: Assertion `rc == 0' failed.
16:10 wolverian hmm, I want a Str.words iterator that gives lvalues to the original string so that I can do "foo bar".words.=map:{ .ucfirst }; without doing a .split or so that loses whitespace
16:10 larsen has joined #perl6
16:11 ninereasons ?eval  my @a=(); @a[0] ~~ /(nothing)/; +@a
16:11 evalbot6 pugs: src/events.c:265: init_events_first: Assertion `rc == 0' failed.
16:12 ninereasons hmm.
16:12 ingy hola
16:12 iblech wolverian: Yeah! That'd be nice!
16:12 Chewie[] 'foo bar'.words >>.ucfirst would work, and it's shorter.
16:12 wolverian yay, someone else likes it too :) iblech++
16:12 iblech ninereasons: It seems Parrot doesn't work well when limited to 20MiB RAM or so
16:12 Chewie[] Yeah. It's a nice idea.
16:12 wolverian Chewie[]: yes.
16:13 wolverian I'm not sure if it's been discussed on p6l or something, as I don't think I'd just come up with that by myself
16:13 Chewie[] Perhaps also .tokens to split on given rules.
16:13 ninereasons iblech, I didn't realize that evalbot was targetting parrot already.
16:13 Chewie[] (If that would even be necessary with the PGE.)
16:13 iblech ninereasons: No, but it uses Parrot for Perl 6-rules
16:13 ninereasons oh.
16:13 iblech ?eval my @a = (); @a[0] ~~ rx:Perl5/(nothing)/; +@a
16:13 evalbot6 1
16:13 ninereasons there you go.
16:14 iblech ?eval my @a = (); +@a
16:14 ninereasons that's what I wanted to show.
16:14 evalbot6 0
16:14 iblech Right. Write a test? :)
16:14 wolverian ?eval my @a = (); @a[0] ~~ rx:Perl5/(nothing)/; @a.perl
16:14 ninereasons ok :)
16:14 evalbot6 '[undef]'
16:14 wolverian ah.
16:15 mrborisguy has quit IRC ("leaving")
16:19 lhooq has quit IRC ("zu spaet um makellos zu sein")
16:21 justatheory has joined #perl6
16:22 xerox has joined #perl6
16:30 chip [email@hidden.address]
16:32 autrijus Data::Dumper::Dumper(\@a)
16:32 autrijus "canonical string representation", man :)
16:32 Limbic_Region Data::Dump::Streamer is head and shoulders better than Data::Dumper when it comes to closures and code refs
16:33 autrijus very true.
16:33 autrijus DDS is the way of the future
16:33 gaal i thought .perl was? :)
16:33 gaal or .yaml
16:33 autrijus .perl is the way of the far future
16:33 autrijus ;)
16:34 gaal .pugs is the way of the mear to medium range future.
16:34 Limbic_Region oh - it handles weak references too
16:34 autrijus .pugs is for fun :)
16:34 * autrijus launches the fun optimiser
16:35 autrijus "fun-ction-al programming"
16:35 autrijus 'fun' as in exciting
16:35 autrijus 'ction' as in ka-ching, big money
16:35 autrijus 'al' as in 'all cool peopl are doing it'
16:35 iblech FWIW, I remember a discussion on p6l about renaming .perl to .pretty(:lang<Perl>), and then extending .pretty to :lang<YAML>, :lang<Perl5>, etc. Not sure what the final ruling was, though
16:35 autrijus (I stole that from some perl monger's presentation)
16:35 autrijus iblech: yeah. I think it's sane
16:36 autrijus iblech: I further think it may be saner to have Language::* API.
16:36 autrijus that unifies .pretty, .eval, .compile etc
16:37 iblech ala Language::Perl6.new(source => "...").compile?
16:37 gaal which reminds me!
16:38 gaal we need to move eval over to the prelude
16:38 Limbic_Region autrijus - did you see the weird .kv error
16:38 Limbic_Region ?eval (42).kv
16:38 evalbot6 pugs: out of memory (requested 1048576 bytes)
16:38 autrijus iblech: SomeThingLikeThat yes. although I'm not sure about the .new.
16:38 autrijus Limbic_Region: yes, and I know why too
16:38 autrijus the Prim "kv" is infinite looping because it can't find a ref
16:38 autrijus I think "pairs" too
16:38 autrijus iblech: would you be so kind to investigate a fix? :)
16:40 Nattfodd has quit IRC ("Leaving")
16:40 iblech autrijus: I tried one earlier today, but then stopped, as I thought: (42).kv should mean 42.kv, which should give a "wasn't able to find &Int::kv method"
16:41 gaal should it not give (1, 42)?
16:41 gaal ?eval my @a=<42>; @.kv
16:41 evalbot6 Error: Undeclared variable: "$?SELF"
16:42 gaal ahem.
16:42 gaal ?eval my @a=<42>; @a.kv
16:42 evalbot6 (0, '42')
16:42 gaal yeah, i meant 0, 42. :)
16:42 iblech gaal: Yep. So we've to rename Prim.hs eval to Pugs::Internals::eval_perl6, then make Prelude::eval safe, so evalbot can eval Perl 6 code, and then fix all tests using eval_perl5 etc., right?
16:43 iblech gaal: Yep, an array should understand .kv -- i.e. [42].kv should work, as well as (my @a = (42)).kv, but $some_num.kv should not, I think
16:44 autrijus iblech: you know...
16:44 autrijus iblech: that was because in Prim it's decl'ed as "kv"
16:44 gaal iblech, re eval - yes, exactly; though there's a lot to s/// in this move.
16:44 autrijus if you change all those "kv" into Array::kv etc
16:44 gaal heh heh heh
16:44 autrijus (look at Scalar::split)
16:44 autrijus then it will Just Work (tm)
16:45 iblech autrijus: ah! /me tries
16:45 gaal but then kv won't work for lists, right?
16:45 gaal ?eval (1,2,3).kv
16:45 evalbot6 (0, 1, 1, 2, 2, 3)
16:45 gaal the above is what i expect
16:46 autrijus actually, no,
16:46 gaal iblech: however, eval lang=>something not perl6 should be unsafe
16:46 autrijus if you declare "kv" as "List::kv"
16:47 autrijus and then add one line to Pugs.Eval.Var line 315
16:47 autrijus that declares all (Syn ",") is of type "List" in invocant position
16:47 autrijus evalExpType (Syn "," _) = return $ mkType "List"
16:47 iblech gaal: Yep. Should I add a $?SAFEMODE so Prelude::eval can test if safemode is on?
16:47 autrijus then it will Still Just Work
16:47 * autrijus praises tagged unions, pattern matching, and sanity in general
16:48 gaal hmmmm, actually no i was being silly, just leave the Pugs::Internals::eval_$not_perl6 unsafe
16:48 iblech gaal: Ah, yes, of course.
16:49 gaal s/silly/not-thinking-it-through-ful/
16:50 gaal anyway: how do we apply this big s///? need also to change Test.pm and everything in ext/
16:50 ninereasons iblech, what is the formula again, for applying the "usual properties" to a new addition?
16:50 iblech ninereasons: cat docs/other/development_tips
16:50 ninereasons thanks
16:51 autrijus gaal: time to think of a refactoring browser? :)
16:51 gaal :-)
16:51 autrijus (apply parser to source, get token ranges, change those ranges)
16:51 autrijus you wouldn't need a full prettyprint even
16:51 gaal right-click Refactor > Rename...
16:51 autrijus right.
16:51 autrijus should be a simple matter of walking Exp and change the original String.
16:51 autrijus by analysing Pos chunks
16:52 autrijus and emit a set of "Edit" calls
16:53 Limbic_Region autrijus - I am not sure I saw your reply, if there was one, so I will ask again
16:54 Limbic_Region it is obvious that there are a lot of significant changes to Parrot as a result of the hackathon
16:54 Limbic_Region what is coming down the pipe for Pugs?
16:55 autrijus Limbic_Region: PugsAST is the single larges change
16:55 svnbot6 r4625, iblech++ | * Pugs.Prim, Pugs.Eval.Var -- Made (42).kv die and (42,).kv work, autrijus++.
16:55 svnbot6 r4625, iblech++ | * t/pugsbugs/one_elem_list_kv.t -- unTODO/unCOMMENT, fix, and added a new test.
16:55 svnbot6 r4626, iblech++ | Moved t/pugsbugs/one_elem_list_kv.t to t/builtins/lists/one_elem_list_kv.t, as
16:55 svnbot6 r4626, iblech++ | the bug is resolved.
16:55 autrijus Limbic_Region: instead of evaluating the parse tree directly
16:56 autrijus Limbic_Region: Eval.hs will deal with a much smaller set of nodes
16:56 gaal soooo.... i'll do the eval change now
16:56 autrijus Limbic_Region: basically all the reduceSyn things will be moved to a separate compile pass.
16:56 autrijus as well as other things like constant folding etc.
16:56 gaal unless iblech you started already? never knowing with the speed demon :)
16:56 autrijus that will result in a much faster interpreter-level performance.
16:56 iblech gaal: No, go ahead :)
16:56 autrijus and will also open the door of other language targetting that PugsAST.
16:57 autrijus (it's docced in my paper)
16:57 autrijus currently it's parse -> run
16:57 autrijus it will become parse -> compile -> (emit|run)
16:58 autrijus so we won't need to deal with all the Syn constructs in Compile.Parrot and Compile.Haskell anymore
16:58 autrijus because they'll be normalised and gone
16:59 Enveigler has joined #perl6
16:59 autrijus also if I did pull it off and GADTise PugsAST and PIR_AST, then it will also make detection of bogus programs easier and optimisation safer to write.
16:59 autrijus s/did/do/
16:59 autrijus oh, and maybe, just maybe, PugsAST can retarget C-- or LLVM or something else entirely.
17:00 autrijus so we retain Perl6's retargettability
17:00 autrijus which I always think as a good thing.
17:00 autrijus (eof)
17:01 autrijus this is in scope comparable to the GADT-ITypes grand rewrite
17:01 svnbot6 r4627, ninereasons++ | smartmatch_autovivifies.t: a pugsbug.  When a non-existent array element, $n, is tested for a match, the array is auto-dimensioned to $n+1 size. The autovivified elements are undef.
17:01 autrijus which took me two full days :)
17:01 autrijus so I need to, like, concentrate.
17:01 Forth where should I look into if I want to generate java bytecode?
17:02 Limbic_Region autrijus - sounds like good stuff
17:02 autrijus Forth: first, get a good Haskell library that can emit java bytecode from some low level AST
17:02 autrijus Forth: then hook it up with the PugsAST I'll be producing this week
17:02 autrijus Forth: you can look at Pugs.Compile.Parrot for some ideas right now, but it's very primitive at this moment
17:03 autrijus (and yes, Java and Mono retargettability are worth goal too)
17:03 autrijus what I'll do is make the PugsAST abstract and narrow enough
17:03 autrijus but the emitters for Java/Mono/LLVM/etc is probably left to other interested lambdacamels
17:04 autrijus I'll first try targetting Parrot because I can lobby the authors more efficiently :)
17:04 Enveigler1 has quit IRC (Read error: 60 (Operation timed out))
17:08 autrijus hey. if I rename the current src/IMC* to src/PIR*
17:08 autrijus will people be mad because the P<Tab> completion will break?
17:08 * autrijus is thinking of a nonconflicting name
17:09 nnunley_ has quit IRC ("This computer has gone to sleep")
17:10 * autrijus settles for Emit.PIR
17:11 G2 has joined #perl6
17:13 goofyheadedpunk has quit IRC (Read error: 110 (Connection timed out))
17:14 arcady has joined #perl6
17:19 gaal if i have sub foo(?$x = $CALLER::_), and call it foo(undef), will $x be undef or the caller's $_ ?
17:19 svnbot6 r4628, cognominal++ |  r7@cac94-1-82-67-232-89:  stef | 2005-06-13 19:18:08 +0200
17:19 svnbot6 r4628, cognominal++ |  added a nice Haskell book
17:20 iblech gaal: I think $x should be undef, as you're explicitely passing an argument to &foo
17:20 gaal i hope you're right!
17:21 autrijus I assure you're right.
17:21 gaal :)
17:21 Darren_Duncan has joined #perl6
17:22 gaal okay, eval(.... :lang<Haskell> or <haskell>? or make it incasitive?
17:23 autrijus I like that word
17:23 * chip ponders CaseInsensitiveHash
17:23 autrijus make it incasitive, I think.
17:23 gaal avva coined it
17:27 obra hello
17:31 Darren_Duncan right now I revisit Locale-KeyedText in an effort to get the whole thing executing
17:31 Darren_Duncan and so I may report a few things that Pugs needs to make them work
17:31 Darren_Duncan fyi
17:32 eric256_ has joined #perl6
17:32 svnbot6 r4629, gaal++ | * eval and related builtins are now implemented in the Prelude
17:33 autrijus Darren_Duncan: cool!
17:34 Darren_Duncan now hopefully things won't be breaking all over the place with this on-Parrot push
17:34 autrijus Darren_Duncan: we have 8k regressions :)
17:34 autrijus I won't break things all over the place unless they are somehow not testing things as they should
17:34 Darren_Duncan question ...
17:35 Darren_Duncan do you plan to keep Pugs with its standalone option for a long time, or will it be phased out in a few months when Parrot support is working well?
17:35 Darren_Duncan so that Pugs only works with Parrot and not alone
17:35 autrijus Darren_Duncan: to answer that question: no, interpreter core will Always Be There.
17:36 Darren_Duncan okay
17:36 autrijus otherwise we will be making the P6C mistake.
17:36 autrijus "With this design, each term's reduction rules are hard-coded in PIR assembly,
17:36 autrijus making it prohibitively difficult to reason about the execution.  Because P6C
17:36 autrijus cannot check the resulting assembly code for correctness, subtle bugs often
17:36 autrijus go unnoticed for a long time.
17:36 autrijus "
17:37 Darren_Duncan but when Pugs is recoded in Perl 6, that won't be the case, right?
17:37 autrijus changes topic to: 6.2.7 released! | http://pugscode.org/ <Overview Journal Logs> | http://pugs.kwiki.org | paste: http://sial.org/pbot/perl6
17:37 autrijus Darren_Duncan: yeah but you're talking about 6.283185...
17:38 autrijus and we're just at 6.2.7 :)
17:38 autrijus they're five milestones apart.
17:38 gaal the question that hasn't come up yet is whether by that time, pugs will also be able to compile haskell.
17:39 autrijus not neccessary pugs.
17:39 vcv- has joined #perl6
17:39 autrijus but yeah, whether we can as well compile pugs itself to run on parrot, or on perl6, or something else entirely, etc.
17:39 gaal yes, it'll be called Star
17:39 gaal the any-to-any compiler
17:40 autrijus heh :)
17:40 chip that's the great thing about bits
17:40 chip no matter how you recycle them, they're still just ones and zeros
17:40 autrijus but anyway. I see the current priority is getting the features in
17:40 autrijus and get the operational (and some denotational) semantics defined
17:40 autrijus with a well defined semantics we can freely retarget
17:40 autrijus with a reasonable hope of keeping portability
17:41 autrijus without a well defined semantics it's just a random bunch of OP codes that you have to special case to, and that changes from version to version, just like the perl5 mess.
17:41 autrijus i.e. you can't look at a piece of program and reason about what it will do, without knowing the exact brokenitude of the perl5 interpreter that is to run it.
17:42 Limbic_Region chip - http://perlmonks.org/index.pl?node_id=463675 http://perlmonks.org/index.pl?node_id=441152
17:42 Limbic_Region there are several other links I meant to show you, but you weren't around
17:43 Limbic_Region oh, and http://perlmonks.org/index.pl?node_id=460197 too
17:44 autrijus if Python (PyPy) can do it, there's no reason why perl6 can't :)
17:44 Darren_Duncan away from comp for a bit while Pugs recompiles with latest patches
17:44 eric256_ are the current milstones on  http://svn.perl.org/perl6/pugs/trunk/docs/01Overview.html still valid? or have they changed? if so does that mean improved OO is top priority right now?
17:44 autrijus eric256_: they're still valid.
17:45 autrijus eric256_: and yes, OO is the current priority (over rules and macros)
17:45 eric256_ ;) cool ;)
17:47 chip Limbic_Region: thanks.  I'm having some net issues here.  I can get to perl.org just fine, but comics.com seems not to like Austria.  Darn it.
17:49 Limbic_Region chip - got a feather account?
17:52 chip Limbic_Region: yup.  chip@
17:54 pelagic has joined #perl6
17:57 Limbic_Region I was just going to suggest attempting to fetch the comics from there
17:57 Limbic_Region has quit IRC ("tiuq\")
17:58 iblech gaal: Should we change evals :lang<Parrot> to :lang<PIR>?
17:59 cjeris has joined #perl6
17:59 gaal dunno? we can make the aliases too
17:59 Limbic_Region has joined #perl6
18:01 Aankhen`` "Classes and traits" -- does that include stuff like `where` and `but`? *hopeful look*
18:02 Aankhen`` "Can Pugs work with Perl 5 libraries?" seems to be a little outdated...
18:04 gaal afk
18:05 arcady has quit IRC (Connection reset by peer)
18:08 pelagic has quit IRC ("CGI:IRC (EOF)")
18:11 osfameron_ has quit IRC (Read error: 111 (Connection refused))
18:12 ninereasons what is the bug in wizard.p6 ?
18:14 eric256_ a new bug in there? it was working last time i left it
18:14 Darren_Duncan compile done
18:15 ninereasons eric256_, it isn't running for me.  perhaps a problem on my end.
18:15 eric256_ let me update and run it
18:15 eric256_ well take a while. ;)
18:16 * geoffb is really enjoying having autrijus in a US/Pacific-friendly timezone . . . the channel is actually busy during my awake hours
18:16 ninereasons eric256_, "*** No compatible subroutine found: "&new" at examples/games/wizard.p6 line 142, column 14 - line 145, column 2"
18:17 iblech gaal: Aliasing PIR to Parrot sounds good for me, committing...
18:17 eric256_ ninereasons -- just starting nmake unoptimized ow. ;)
18:19 Aankhen`` iblech++; gaal++;
18:19 iblech gaal++ # Did all the work
18:20 saorge has joined #perl6
18:24 autrijus :)
18:24 svnbot6 r4630, iblech++ | Prelude -- Make eval("...", :lang<PIR>) call eval_parrot().
18:24 svnbot6 r4631, gaal++ | You can't just make shit up and expect the computer to know what you mean, Retardo!
18:25 eric256_ hmmm weird error..
18:26 eric256_ okay i don't even know where to start with that....is it haveing trouble processing the class? or something else.
18:27 Darren_Duncan gaal, want to explain that last one?
18:27 cdpruden has joined #perl6
18:28 ninereasons Darren_Duncan, checking the svn diff explains it all :-)
18:28 Darren_Duncan yes, but asking is less work
18:28 Darren_Duncan for me
18:28 Darren_Duncan forget it
18:28 ninereasons + die "can't eval language $lang";
18:29 autrijus eric256_: shockingly
18:29 autrijus eric256_: because the OO system works
18:29 autrijus your program breaks
18:29 autrijus eric256_: try changing all your "Object" into "Subject"
18:29 autrijus and your program will run fine.
18:30 autrijus because by saying "class Object { has...}" you're saying every single object now has additional attributes
18:30 autrijus now how actually that breaks things needs to be investigated
18:30 autrijus but that's probably buggy at one level or another
18:32 clkao cool, eval('...', lang:<perl5>) !
18:33 ninereasons interesting, autrijus.  Just as you said s/Object/Subject/g fixes it.
18:33 Aankhen`` Isn't that: eval('...', :lang<perl5>)
18:34 eric256_ ahhh
18:35 eric256_ and the object part is new
18:35 cjeris has left
18:36 pelagic has joined #perl6
18:36 * eric256_ doesn't like have the Object class taken like that...shouldn't internal classes like that have some better name?
18:36 eric256_ Perl6::Object or something like that?
18:36 pelagic has quit IRC (Client Quit)
18:37 eric256_ i just don't like the idea of internal objects and user objects colliding in such a fashion, it also means if you add internal classes in the future they might clash with existing and working code, seems limiting even if you never plan on adding new classes
18:38 wixii has joined #perl6
18:38 q[merlyn] has joined #perl6
18:38 iblech 1. The idea is, I think, that the Object class isn't thaat much internal that it'd need some internal_use_only prefix. 2. If we went with Perl6::Object, we'd have to rename Array, Hash, etc. to Perl6::Array respectively Perl6::Hash, too, even though these are fairly "normal" classes
18:38 q[merlyn] is building from SVN failing on OSX expected?
18:38 pelagic has joined #perl6
18:39 iblech I.e. it'll be relatively normal to (for example) inherit from Hash if you want to have your tied hash class
18:39 dkf q[merlyn]: it is if you're using gcc 4.0
18:39 PerlJam iblech: Object might just be shorthand for Perl6::Object you know
18:39 q[merlyn] is there a pastebot here
18:39 q[merlyn] no - I rolledback to 3.3
18:39 Aankhen`` q[merlyn] >> http://sial.org/pbot/perl6
18:39 ninereasons perlbot nopaste
18:39 perlbot Paste your code here and #<channel> will be able to view it: http://sial.org/pbot/<channel>
18:39 autrijus q[merlyn]: you need to 3.3 on a clean tree
18:40 eric256_ iblech sure. but that doesn't address the concern of adding future classes to the language that could conflict with current user defined classes
18:40 q[merlyn] I did sudo gcc_select 3.3
18:40 q[merlyn] and "make clean"
18:40 chip q[merlyn]: welcome
18:40 q[merlyn] and rebuilt the Makefile
18:40 eric256_ it means you either have to tell those people too bad, or never create new internal classes....neither of those options seems like the right thing to do
18:40 pasteling "q[merlyn" at 70.212.105.232 pasted "failure on OSX 10.4.1" (10 lines, 286B) at http://sial.org/pbot/11021
18:40 q[merlyn] hey chip
18:41 svnbot6 r4632, iblech++ | sub_ref.t -- Fixed test according to p6l and added a comment.
18:41 q[merlyn] what the hell is a _fprintf$LDBLStub? :)
18:41 PerlJam eric256_: What do you mean by "internal classes"?  Why must they exist?
18:41 q[merlyn] I got 15 minutes before I have to leave CVG
18:42 iblech eric256_: Right. But I'd expect that, for example, if a new core class is introduced in Perl 6.1.0, you'd have to "use v6.1.0" to make it visible
18:42 eric256_ Object, Hash, Array, theses are all *internal* classes.  Meaning they are defined by the language
18:42 eric256_ or just use Perl6;
18:42 eric256_ iblech why paint ourselves into this obvious restriction on future growth though?
18:43 eric256_ i mean you should just say then, don't define your classes in the base level, becuase we might step on them in the future.
18:43 eric256_ but then the inconvienence is put on the user instead of on the language where it should be
18:43 q[merlyn] any ideas what to try nextg?
18:44 autrijus q[merlyn]: it's possible that "make clean'¢ì¢÷¢í¢û¢" doesn't really.
18:44 iblech I don't think having to add one line ("use v6.1.0") is a hard restriction :) The documentation will simply say "Don't forget to "use v6.1.0" if you want to use the new FooBar class or the new "/" sigil."
18:45 autrijus q[merlyn]: so check out a fresh tree, and make unoptimised with 3.3?
18:45 autrijus eric256_: once upon a time internal classes are called
18:45 autrijus OBJECT HASH CODE ARRAY REF
18:46 autrijus then they get to become regular cased and conflict with user classes. fancy that.
18:46 Limbic_Region do p6 array elements have knowledge of there own index?
18:46 Limbic_Region for @array -> $elem { say $elem.index }
18:46 pelagic has quit IRC ()
18:47 autrijus (I agree with your point entirely. it should be Perl::Object for all I'm concrebed)
18:47 q[merlyn] how do I tell svn to toss anything broken
18:47 autrijus Limbic_Region: .kv
18:47 q[merlyn] I don't have time to do a complete download
18:47 autrijus q[merlyn]: svn st then grep for ?
18:47 q[merlyn] OK
18:47 autrijus '^?'
18:47 chip zero or more beginnings of line
18:47 Limbic_Region for @array.kv -> $i, $elem { say "$i $elem" }
18:47 q[merlyn] Hmm.  only two files in that category
18:48 q[merlyn] src/Pugs/Run/Perl5_stub.{c,h}
18:48 autrijus zero or one, actually. </pedantic>
18:48 Aankhen`` What about having C<class Object {...}> define main::Object, and C<::*Object> refer to the builtin class?
18:48 autrijus q[merlyn]: hm. that doesn't sound right. are you using .dmg GHC builds and 3.3?
18:48 eric256_ should i post a request to the mailing list then or what? or is this still all in flux and don't worry about it at the moment?
18:48 Limbic_Region so if you wanted to know if you were on the last loop, you could do something like
18:48 autrijus eric256_: no, this is The Priority now
18:48 autrijus eric256_: so please post to p6l.
18:48 wixii has quit IRC ()
18:49 q[merlyn] I built GHC from source
18:49 Limbic_Region for @array.kv -> $i, $elem { say $elem; say "done" if $i == @array.end }
18:49 q[merlyn] version 6.4
18:49 autrijus Limbic_Region: you do a LAST{} block.
18:49 autrijus Limbic_Region: see S04
18:49 Limbic_Region autrijus - ok, so bad example
18:49 autrijus q[merlyn]: hm. so it's the same gcc that builds ghc and builds pugs?
18:49 q[merlyn] I got five minutes.
18:49 q[merlyn] yes
18:49 q[merlyn] yes
18:49 q[merlyn] in theory. :)
18:49 Limbic_Region autrijus - let's say you wanted to do something every 3rd element
18:49 autrijus q[merlyn]: ok, then it's not linking -lc
18:50 q[merlyn] why? :)
18:50 Limbic_Region you could use .kv and then check $i for % 3
18:50 Aankhen`` $i + 1, actually.
18:50 q[merlyn] What Do I Type To Make It Work? :)
18:50 autrijus q[merlyn]: no idea... but you can goto Makefile and look for ^pugs
18:50 Aankhen`` (assuming the index starts from 0)
18:50 Limbic_Region Aankhen``:  yeah yeah - but the point is that it will DWIM
18:50 Aankhen`` Yeh.
18:51 autrijus then change the ghc line to say
18:51 eric256_ for @array -> $elem,$b,$c { say $elem; say "done" if $i == @array.end }
18:51 autrijus -lc
18:51 autrijus before -threaded
18:51 PerlJam Limbic_Region: I wonder if there would be any use in an EVERY block.  :-)
18:51 Limbic_Region ?eval my @array = 1..5; @array.kv
18:51 evalbot6 (0, 1, 1, 2, 2, 3, 3, 4, 4, 5)
18:51 G2 has quit IRC (No route to host)
18:51 Aankhen`` ?eval <foo bar baz>.kv
18:51 autrijus q[merlyn]: I also hear that the upcoming GHC 6.4.1 will be more powerpc friendly
18:51 evalbot6 (0, 'foo', 1, 'bar', 2, 'baz')
18:52 autrijus q[merlyn]: I further hear that the .dmg GHC works for other people; I'll try to convince those people to build binaries for you
18:52 Aankhen`` ?eval for <foo bar baz>.kv -> $i, $elem { say $elem; say "done" if $i == 2 }
18:52 evalbot6 Error: No compatible subroutine found: "&say"
18:52 Aankhen`` Oops.
18:52 Limbic_Region ok - thanks - will keep this in my back pocket for the next SoPW at the Monastery for something like this
18:52 Aankhen`` ?eval for <foo bar baz>.kv -> $i, $elem { $elem }
18:52 autrijus q[merlyn]: but in the meantime (that is 3min) I probably can't help a lot. :-/ maybe play on feather at the moment
18:52 evalbot6 undef
18:52 dkf autrijus: I'm running into the same problem using the .dmg GHC
18:53 autrijus dkf: even gcc_select 3.3?
18:53 autrijus but in all fairness, Tiger came out after GHC 6.4
18:53 Aankhen`` ?eval my (@foo, @bar); @foo = <foo bar baz>; for @foo.kv -> $i, $elem { push @bar, @elem; push @bar, "done" if $i == (@foo - 1); } @bar;
18:53 evalbot6 Error: Undeclared variable: "@elem"
18:53 autrijus so the fix for it will appear in 6.4.1 which is already branched, I think
18:53 Aankhen`` ?eval my (@foo, @bar); @foo = <foo bar baz>; for @foo.kv -> $i, $elem { push @bar, $elem; push @bar, "done" if $i == (@foo - 1); } @bar;
18:53 evalbot6 ['foo', 'bar', 'baz', 'done']
18:53 dkf autrijus: Yep. But I just checked out a clean tree and I'm building it as we speak...
18:54 G2 has joined #perl6
18:54 Limbic_Region Aankhen`` why @foo - 1 and not @foo.end ?
18:54 Aankhen`` ?eval my (@foo, @bar); @foo = <foo bar baz>; for @foo.kv -> $i, $elem { push @bar, $elem; push @bar, "done" if $i == @foo.end; } @bar;
18:54 evalbot6 ['foo', 'bar', 'baz', 'done']
18:54 pmichaud has joined #perl6
18:55 Limbic_Region salutations pmichaud
18:55 q[merlyn] OK, so I'll just be pugs-disabled for a bit
18:55 pmichaud greetings, lr
18:55 q[merlyn] weirdly, the CPAN release worked
18:56 dkf q[merlyn]: A fresh tree does build with the .dmg GHC.
18:56 svnbot6 r4633, iblech++ | t/ -- First part of test fixes (is eval '...' -> is eval('...')).
18:56 Limbic_Region pmichaud - have you had a chance to consider my last response WRT patches to the Synopses?
18:56 pmichaud l_r:  I'm just getting to it now, actually :-)
18:57 Limbic_Region okie dokie
18:57 * Limbic_Region will go back to not feeling well
18:57 pmichaud would've done it over the weekend, but this past weekend was my last chance to see the kids for a month so I wanted to take care of that first :-)
18:58 Limbic_Region pmichaud++  # family first
18:58 q[merlyn] I hate dmg because they tend to install in mixable locations
18:58 q[merlyn] I use ghc from darwinports
18:59 q[merlyn] because I can remove it
19:01 Limbic_Region PerlJam - there is an EVERY thingy - just not called EVERY
19:01 autrijus hey pmichaud!
19:02 pelagic has joined #perl6
19:02 pmichaud hiya, autrijus
19:02 autrijus pmichaud: got time to review my paper?
19:02 Limbic_Region several of them in fact
19:02 pmichaud sorry I didn't get a review of your document back to you before your deadline -- I was out of town (in the middle of nowhere in New Mexico) and didn't get it until late on the 9th
19:02 pmichaud I'll be glad to review anything you'd like me to review now, though :)
19:02 pmichaud (very glad, even)
19:03 autrijus please do; in 2 weeks I'll know if it's accepted
19:03 autrijus (which I think it will)
19:03 autrijus then I need to prepare a final version
19:03 pmichaud shall I just review the draft you sent or is there an updated version?
19:03 autrijus so your review is still essential
19:03 autrijus pmichaud: it's in the pugs tree as docs/talks/hw2005.tex
19:03 autrijus or you can review the pdf at http://perlcabal.org/~autrijus/hw2005.pdf
19:03 pmichaud okay, I'll review that one.  Should do it a bit later this afternoon
19:03 autrijus thanks a lot!
19:04 * autrijus is doing this unenviable work of writing ASTs for both PugsAST and PIRAST
19:04 pmichaud unenviable indeed
19:04 pmichaud what do you have so far on the PIRAST?
19:05 autrijus pmichaud: I'll check it in later in pugs as Emit.PIR
19:05 autrijus src/Emit/PIR.hs
19:05 autrijus pmichaud: but part of it is in the paper
19:05 pmichaud autrijus: that would be excellent
19:05 pmichaud I look forward to seeing it
19:05 autrijus it's basically a typechecked AST for the imcc parse tree
19:06 autrijus not terribly advanced
19:06 autrijus and only for emitting purpose
19:06 autrijus so I get to gloss over a lot of parsing
19:06 pmichaud advanced systems often arise from simple starts -- that's the way to go :)
19:06 Enveigler has quit IRC (Read error: 60 (Operation timed out))
19:06 autrijus nod. the plan is to have a mid-level PugsAST
19:06 autrijus similar to scope of the intended PAST but much more constrained
19:06 autrijus almost ANF-level constained
19:07 autrijus and do PugsParseTree -> PugsAST -> PIRAST -> PIR -> PASM -> PBC
19:07 svnbot6 r4634, iblech++ | Pugs.Prim -- Repair Pair::kv.
19:07 goofyheadedpunk has joined #perl6
19:07 svnbot6 r4635, iblech++ | t/ -- Remaining test fixes WRT eval.
19:07 svnbot6 r4636, iblech++ | hw2005.tex -- Very minor grammar fix.
19:08 autrijus pmichaud: so, unidiff against hw2005.tex is appreciated -- or you can just commit back; I can hand you a commit bit for that :)
19:09 pmichaud I'll do a unidiff -- I'm a bit more comfortable with that
19:09 autrijus 'k
19:09 autrijus thanks in advance :)
19:09 pmichaud it'll also make it slightly easier for you to see what I propose/change :)
19:09 autrijus nod
19:12 ninereasons has quit IRC (""back soon"")
19:13 Darren_Duncan one missing feature I think, 'is constant' with variables
19:13 Darren_Duncan eg, my Str $xy is constant = 'AE';
19:13 Darren_Duncan not a show stopper yet though
19:13 ninereasons has joined #perl6
19:14 Darren_Duncan this does not compile
19:14 autrijus danot really missing
19:14 autrijus my $x := 888;
19:15 Darren_Duncan I started out using that syntax, but changed to 'is constant' later on recommendation that it is more readable
19:15 Darren_Duncan or self-documenting
19:16 Darren_Duncan they may not have exactly the same meaning either
19:16 Limbic_Region doesn't the latter allow rebinding while the former wouldn't?
19:16 Darren_Duncan or rather, I started out with '::=' but changed to 'is constant'
19:16 Darren_Duncan the '::=' is compile-time binding
19:16 Aankhen`` ?eval my $foo is rw;
19:16 evalbot6 Error:  unexpected "i" expecting "=", ":=", "::=", ";" or end of input
19:16 Aankhen`` ?eval my $foo is Str;
19:16 evalbot6 Error:  unexpected "i" expecting "=", ":=", "::=", ";" or end of input
19:17 Aankhen`` Doesn't look C<is> parses.
19:17 Aankhen`` Er, look like.
19:17 * Aankhen`` wanders off to play some more Battlefield 2.
19:17 Darren_Duncan ?eval my $foo is constant = 3;
19:17 evalbot6 Error:  unexpected "i" expecting "=", ":=", "::=", ";" or end of input
19:17 Darren_Duncan yes
19:18 Darren_Duncan that has been the situation for a few weeks now
19:18 autrijus right, because vars don't take traits yet
19:18 autrijus did it work before?
19:19 justatheory has quit IRC ()
19:19 Darren_Duncan also, 'trusts Foo' inside a class declaration doesn't work yet
19:20 Darren_Duncan ?eval class Foo trusts Bar;
19:20 evalbot6 Error: No compatible subroutine found: "&trusts"
19:20 Darren_Duncan ?eval class Foo { trusts Bar; }
19:20 evalbot6 Error: No compatible subroutine found: "&trusts"
19:20 autrijus hrm. I got this mail saying my 45min I18N talk got accepted for Euroscon.
19:21 autrijus curiously, none of the other 3 perl6/pugs related talks did.
19:21 Darren_Duncan now, assuming that private attributes are being enforced as such, this is a show stopper for my module, since Translator needs to access Message vars
19:21 iblech has quit IRC ("school tomorrow, off &")
19:21 autrijus Darren_Duncan: not enforced.
19:21 Darren_Duncan though I suppose I could work around it for now using the Message accessors, but the final version shouldn't
19:21 autrijus Darren_Duncan: you can comment the "trusts" out.
19:22 Darren_Duncan it is commented out in the checked in version
19:22 Darren_Duncan will add comment about not enforced
19:22 autrijus ok
19:22 autrijus oh wait you mean $:foo
19:22 autrijus sorry. I read incorrectly
19:22 autrijus no accessors are generated for them, yeah.
19:23 Darren_Duncan yes I mean $:foo
19:23 autrijus you need to use $.foo for now. is there a "trusts" test?
19:23 autrijus I can maybe do it later this week
19:23 Darren_Duncan don't know
19:23 autrijus not seeing one
19:23 autrijus ah. it just takes time for it to flow thru... I got another accepted talk
19:23 * autrijus wonders what the other two will fare
19:23 Darren_Duncan I don't know if this goes against your principles, but can you make it so 'trusts' compiles, even if it currently doesn't do anything?
19:24 autrijus I have no principles... :)
19:24 autrijus I'm just this implementation monkey
19:24 pmichaud heh, autrijus is like me
19:24 pmichaud I don't have any principles, I just want to get things to work :)
19:24 autrijus right.
19:24 * autrijus praises monkeyhood
19:24 pmichaud if that means implementing broken things for a while, so be it :)
19:25 Darren_Duncan when I say "principles", I mean that "if it compiles, it behaves correctly"
19:25 autrijus oh. no.
19:25 autrijus pugs compiles.
19:25 autrijus but there's still loads of TODO tests
19:25 autrijus 1187 TODOs, actually
19:25 Darren_Duncan on the other hand, if things can compile and not run, then I would like for somethings to at least compile asap
19:25 autrijus out of some 7600
19:26 Darren_Duncan namely the 'trusts Foo;' and 'is constant'
19:26 autrijus Darren_Duncan: yes, and I implemented the parsing.
19:26 Limbic_Region pmichaud - ok, I think the biggest problem with people providing patches to the Synopses currently is that only about 1/3 of them are written - so in many cases - there is nothing to patch
19:26 Limbic_Region I will see what I can do about using the Perl6 Timeline by Apocolypse to frame all of them
19:26 autrijus Darren_Duncan: committing in. I would love a "trusts" test in t/oo/trusts.t
19:26 pmichaud Limbic_Region: well, I'm just about to add S17.pod into the repo
19:26 Limbic_Region will send them directly to you to confer with the cabal about uploading them
19:26 autrijus mmm Threads!
19:26 Limbic_Region and then people can fill in the blanks as they come up
19:27 autrijus @Larry got threads done?
19:27 pmichaud no, it's a placeholder
19:27 Limbic_Region and those more ambitious can google for old answers
19:27 autrijus oh. bah. :-/
19:27 Darren_Duncan can you do a 'my $foo is constant = 3;' parsing also?
19:27 pmichaud sorry :-)
19:27 Limbic_Region pmichaud - sound good?
19:27 pmichaud but I'm sure if someone wanted to draft up a document on threads it would get comments much more quickly
19:27 pmichaud actually, now that I think about it, it may be that @Larry (or $Larry) already had some pretty strong ideas about threads
19:27 autrijus right, except we don't know :)
19:28 pmichaud right
19:28 autrijus I know he doesn't want a 5005 share-all-global-by-default model.
19:28 autrijus but that's the extent.
19:28 Limbic_Region well - kinda hard to know what is left to talk about without knowing what hasn't been filled in yet
19:28 pmichaud so, perhaps what we need is to go ahead and put up frameworks, and for those sections where @Larry or $Larry have claimed they want a hand in drafting it we can put that as a placeholder
19:28 pmichaud so at least people will know if @Larry is working on it or if it's just being waited upon
19:28 autrijus sure.
19:28 autrijus worksforme.
19:28 autrijus yes.
19:28 autrijus semaphores.
19:29 pmichaud lemme propose this to @Larry and make it so
19:29 autrijus Darren_Duncan: enjoy.
19:29 Darren_Duncan autrijus, I just added an empty file t/oo/trusts.t to my working dir as a reminder to make the file a bit later
19:29 Limbic_Region pmichaud - $larry didn't yell at me (since I already pretty much said it was a good idea)
19:29 autrijus r4637
19:30 Limbic_Region pmichaud - see http://perlmonks.org/index.pl?node_id=465629
19:30 Darren_Duncan updating ...
19:30 Limbic_Region could be that he was waiting to talk to you - but he could have just as easily said "we are not looking for framework until we are ready to talk about all the details"
19:30 Limbic_Region btw - for the frames, I will be following http://perlmonks.org/index.pl?node_id=332117
19:31 pmichaud Limbic_Region: did Larry respond at all to your post?  I have the impression he's been a bit busy lately, because he hasn't responded to a few of my messages to p6l
19:31 larsen has quit IRC ("br")
19:31 larsen has joined #perl6
19:31 Darren_Duncan back shortly
19:32 larsen has quit IRC (Client Quit)
19:32 Limbic_Region pmichaud - TimToady = $larry
19:32 larsen has joined #perl6
19:32 pmichaud oh, thanks -- I wasn't aware of that connection
19:32 Limbic_Region he didn't reply directly to me, but to a reply in that thread
19:32 pmichaud got it
19:32 Limbic_Region S09 looks really interesting too
19:32 pmichaud my reading of his response is that frameworks for the documents is a good idea
19:33 Limbic_Region yeah - me too
19:33 pmichaud okay, so here's what I'm going to do
19:33 svnbot6 r4637, autrijus++ | * parse for "trusts" and "my $foo is randomtrait", as suggested
19:33 svnbot6 r4637, autrijus++ |   by Darren_Duncan++
19:33 Limbic_Region pmichaud - out of curiosity - there are several headings in S17 I didn't put in as they looked to be covered by other RFCs
19:33 Limbic_Region can you give the list a sanity check
19:33 Limbic_Region using http://perlmonks.org/index.pl?node_id=332117 as reference material
19:33 pmichaud l_r:  sure, I'll do as much of a sanity check as I can
19:34 pmichaud but I'll go ahead and check in framework documents, and I'll post an invitation to @Larry (and others) that if there are any specific parts of the synopses they wish to "claim" then let me know and I'll put a marker in to that effect
19:34 Limbic_Region ok - so tomorrow providing tuits, I will frame out all the remaining synopses
19:34 Limbic_Region and we will mark reserved headings as such
19:34 pmichaud exactly
19:35 Limbic_Region the next task would be to go over the existing synopses and find missing headings
19:35 Limbic_Region since the S now means both synopsis and spec - there is stuff left out
19:36 Limbic_Region or rather, some stuff is left out which made sense when the S only stood for synopsis but now it doesn't
19:36 pmichaud yup
19:36 Limbic_Region ok - nice doing business with you ;-)
19:37 pmichaud hey, anything that advances the state of Perl 6, I'm for :-)
19:37 Limbic_Region me too, but I see myself having a much less significant role
19:38 pmichaud well, when I see what autrijus is accomplishing, I feel pretty insignificant myself at times :-)
19:39 * Limbic_Region wonders how many days remaining until Xmas anyway
19:39 * autrijus casually notes that autrijus is only responsible for 1806 out of 4637 pugs commits.
19:39 autrijus so that's <40% :)
19:40 autrijus (and parrot is almost twice the number of commits.)
19:40 autrijus so significance is very much overrated :)
19:40 pmichaud it's not just the number of commits, it's the speed :-)
19:41 autrijus well, yes, amphetamines has its uses...
19:41 autrijus ...but I think diet coke is good enough for most of my needs
19:41 Darren_Duncan okay, 'trusts' now compiles
19:41 Limbic_Region right - keeps people calm
19:42 Limbic_Region s/people/ADD/
19:42 Limbic_Region err - well, you get the point
19:42 autrijus so some of it is the speed, but not a significant part :)
19:42 Limbic_Region I am not sure it has anything to do with the amount of code or the speed - even though they are both impressive
19:42 Limbic_Region I think it has more to do with - I am not going to complain or suggest things - I am going to do something
19:43 pmichaud there is that :-)
19:43 Limbic_Region and oh by the way - I am not going to work in a vaccuum, everyone is invited to come along and play
19:43 autrijus like, optimising for fun :)
19:44 autrijus gawd, this PIR AST is 40 node types and counting
19:44 Limbic_Region it creates a synergetic energy that makes people want to help
19:44 pmichaud not much point in doing this if it's not fun.  :-)
19:44 autrijus whomever that want to target PIR directly from parse tree is, I repeat, mad
19:44 autrijus (as pugs currently does, that is.)
19:45 Darren_Duncan and 'is constant' now also compiles
19:46 G2 has quit IRC ("oops")
19:50 Aankhen`` What about making C<new Proxy: STORE => {...}, FETCH => {...};> compile? :-D
19:50 * Limbic_Region wonders if there is a breakdown of tests for passing/todo/bug/failing ?
19:50 autrijus Aankhen``: er, it's just adding the Proxy class by default.
19:50 autrijus Aankhen``: you can stuff it into src/Pugs/Context.hs yourself.
19:50 Limbic_Region specifically how many of the tests are todo features
19:50 Aankhen`` :-o
19:51 * Aankhen`` wanders into Haskell-world.
19:51 autrijus Limbic_Region: I don't think "make test" tallies them
19:51 autrijus Limbic_Region: but you can grep anyway
19:51 autrijus Aankhen``: with luck you can even learn to write Proxy::new into Prim.hs and hook it up with ScalarProxy!
19:51 autrijus (then you get a full Proxy implementation)
19:52 Aankhen`` O_o_O
19:52 Limbic_Region autrijus - I am just trying to balance out how much of known design is unimplimented against unknown (with 2/3 of the design docs not published)
19:53 Limbic_Region IOW - of the 1/3 we know about, how far along are we
19:53 Aankhen`` So this would involve putting it in the big list in initTree?
19:53 autrijus Limbic_Region: we're 7000ish pass with 1000ish todo
19:53 autrijus Aankhen``: yes
19:53 Aankhen`` OK.
19:54 autrijus so in the known space, I'd say we're 70% there
19:55 pmichaud and I'm not entirely sure that the remaining 2/3rds is "unknown space".  Lots of it will be similar to perl 5
19:55 autrijus problem is that there's 2/3 not published, so we're more like 25%
19:55 autrijus pmichaud: aye but similar ne defined
19:55 autrijus and semantics undefined is verifications denied, or something :)
19:56 pmichaud well, even in the sections that are "defined" the definitions sometimes change from time to time :)
19:56 Limbic_Region so assuming all things are equal we can expect around 30_000 tests
19:56 autrijus Limbic_Region: that sounds like correct.
19:56 pmichaud out of curiosity, how many tests does p5 have?
19:57 Darren_Duncan committed v4638 which uses 'trusts' and 'is constant'
19:57 Aankhen`` W00t!
19:57 Aankhen`` I added it. :-D
19:58 Shachaf !seen ingy
19:58 Darren_Duncan seen ingy
19:59 Khisanth the bot is not here
19:59 pdcawley has joined #perl6
19:59 Darren_Duncan hi
19:59 autrijus pmichaud:
19:59 autrijus 70_000.
19:59 pmichaud seems like p6 ought to have a similar number of tests :-)
19:59 Darren_Duncan indeed
20:00 pmichaud when it's done
20:00 autrijus but Perl 5.6.x has < 12_000
20:00 autrijus so... :)
20:00 pmichaud ahhh
20:00 pmichaud seems like p6 ought to have around 30_000 tests... when it's done.  :-)
20:00 svnbot6 r4638, Darren_Duncan++ | Locale-KeyedText update to 6v0.1.2 : uncommented 'trusts ...' in KeyedText.pm, uncommented all 'is constant' in all t_LKT_*.pm and LKT_30_TrnMsg.t ; Pugs can now parse those 2 features
20:00 svnbot6 r4639, Aankhen++ | * added Proxy to the builtin classes.
20:00 Darren_Duncan but one of the big advantages of 5.8 is the much larger test suite
20:00 Limbic_Region and how many of those tests are generated versus written by hand
20:00 Darren_Duncan over 5.6
20:00 autrijus right.
20:01 pmichaud either way, 30_000 seems like it might be a lower bound
20:01 Limbic_Region meaning pugs/p6 is likely to have far more tests if we go the generated route
20:01 Aankhen`` Hmm.
20:01 Skilled_sniper has joined #perl6
20:01 Aankhen`` C<new Proxy:> still doesn't parse.
20:01 pmichaud well, we're also going to have the parrot tests as well
20:01 Skilled_sniper quit
20:01 Skilled_sniper has quit IRC (Client Quit)
20:02 ingy hola
20:02 geoffb hola, ingy
20:02 Darren_Duncan I also hola
20:02 geoffb friendliness++
20:04 Aankhen`` autrijus >> Actually implementing a Proxy class seems to be a little beyond me. :-)
20:04 autrijus Aankhen``: that's fine :) at least write tests.
20:05 Aankhen`` A'right.
20:05 elmex has quit IRC ("leaving")
20:07 pelagic has quit IRC (Remote closed the connection)
20:08 Aankhen`` There are already tests in t/subroutines/lvalue_subroutines.t
20:08 autrijus ok. good
20:08 Aankhen`` I'll add one or two.
20:12 goofyheadedpunk has quit IRC (Read error: 110 (Connection timed out))
20:12 svnbot6 r4640, autrijus++ | * first cut at Emit.PIR -- almost enough to express mandel.p6
20:12 svnbot6 r4640, autrijus++ |   -- but not quite.  need to push the PugsAST side now --
20:12 svnbot6 r4640, autrijus++ |   it's just like playing Sokoban...
20:16 * Limbic_Region googles for Sokoban
20:20 geoffb Limbic_Region, it's a game of pushing squares around a puzzle maze to reach an exit point
20:20 svnbot6 r4641, Aankhen++ | * added more tests for lvalue subs.
20:20 geoffb It's actually damned addictive . . . .
20:20 geoffb It's always the simple game play that keeps 'em coming back
20:21 pdcawley Turing complete as well.
20:21 * geoffb mumbles "Man I wish I had a Robotron arcade machine . . . ."
20:21 geoffb pdcawley, really?  Cool!
20:22 autrijus I think I need to sleep a bit so I don't fall into underclocking mode :)
20:22 autrijus bbl... *wave*
20:22 Limbic_Region sleep well
20:22 pdcawley Yeah. There's a proof somewhere.
20:22 geoffb Have a good rest!
20:23 pmichaud later, autrijus
20:23 autrijus :)
20:23 Aankhen`` See ya autrijus.
20:24 meppl has joined #perl6
20:29 Southen[a] has joined #perl6
20:29 * Limbic_Region heads homewards
20:29 Limbic_Region has quit IRC ("tiuq\")
20:39 sapper has quit IRC ("bye")
20:39 svnbot6 r4642, Aankhen++ | * cleaned up t/subroutines/lvalue_subroutines.t somewhat.
20:46 Southen has quit IRC (No route to host)
20:53 cdpruden has quit IRC ("Leaving")
21:11 jhorwitz has quit IRC ("Chatzilla 0.9.67 [Firefox 1.0.4/20050511]")
21:11 vcv-- has quit IRC (Read error: 104 (Connection reset by peer))
21:15 dkf has quit IRC ()
21:22 pdcawley_ has joined #perl6
21:22 Aankhen`` G'night.
21:22 Aankhen`` has quit IRC ("Error: Mouse not found. Do you have cat? (Y/N) [Time wasted online: 6hrs 9mins 59secs]")
21:25 eric256_ do i have to be subscribed to the mailing lists to send a message?
21:25 wolverian no.
21:26 eric256_ hmm.. well i'm sure i  sent it in before tese other messages. odd
21:27 pmichaud eric256_: which mailing list did you send to?
21:27 wolverian eric256_: the first mail takes longer to arrive with the perl6-* lists, anyway.
21:27 knewt has quit IRC (Read error: 131 (Connection reset by peer))
21:27 svnbot6 r4643, ninereasons++ | perl6.vim: add a couple Types keywords
21:27 svnbot6 r4643, ninereasons++ | progressive_powerset.p6: s:g/return/last/ fixes a test
21:27 pmichaud I've noticed that the perl6-compiler list often "holds" messages for a few hours before delivering them
21:27 pmichaud I don't know why
21:28 wolverian the lists (-language, anyway) are moderated and you're whitelisted after the first mail.
21:28 wolverian or so I seem to remember someone saying.
21:28 eric256_ perl6-language
21:28 pmichaud wolverian: yes, that's true also.  However, I know that I'm whitelisted and my posts still sometimes take a while to come back
21:28 pdcawley_ I think they all are, but I'm not sure.
21:28 pmichaud eric256_: if you haven't posted to p6l in a while, I suspect it's waiting for moderator approval
21:28 pmichaud after that you'll be whitelisted again (I think the whitelisting is good for 30 days or something like that)
21:29 vcv- has quit IRC (Read error: 110 (Connection timed out))
21:29 wolverian pmichaud: I had a problem when my mail envelope had the wrong FROM field. that's against the spec and awfully wrong (tm) and it made the whitelisting process fail.
21:29 eric256_ ahh okay
21:29 wolverian things like that can happen when you're, well, me. :)
21:30 pmichaud wolverian: well, that might explain some of my problems too.  But I've even noticed that with the weekly summaries the p6c copy of the summary arrives significantly after the p6i and p6l ones
21:31 iblechbot has quit IRC (Read error: 131 (Connection reset by peer))
21:31 wolverian pmichaud: oh, okay. I haven't participated in p6c, can't compare myself.
21:32 justatheory has joined #perl6
21:33 knewt has joined #perl6
21:38 pdcawley_ has quit IRC ()
21:39 vel has quit IRC (Remote closed the connection)
21:39 G2 has joined #perl6
21:39 pdcawley_ has joined #perl6
21:46 pdcawley_ has quit IRC ()
21:52 pmichaud has quit IRC ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
21:52 svnbot6 r4644, Darren_Duncan++ | Locale-KeyedText updates : uncommented all tests in MsgProps.t and TrnProps.t, removed 2 tests from MsgProps.t and 1 from TrnProps.t (the same tests were removed in the Perl 5 version), reorganized t_LKT_Util.pm, split up a test in Compile.t to match the Perl 5 version, cleaned up several KeyedText methods that fetch properties ; there are now no runtime deaths, but many wrong answers are still returned
21:56 Darren_Duncan I'm stopping for today
21:58 Darren_Duncan my main problems right now seem to concern Pugs transforming one container or variable or value type into another type
22:05 justatheory has quit IRC (Read error: 145 (Connection timed out))
22:06 Darren_Duncan question: if you say 'return;' inside a constructor method new(), is the calling code supposed to receive an object or undef?
22:07 Darren_Duncan and if that's true, then how do you indicate that a constructor has failed due to bad input, without throwing an exception?
22:08 Darren_Duncan in Perl 5, the undef is returned by 'return;' and a caller will know if the constructor failed by seeing that; seeing an object meant success
22:08 Darren_Duncan Pugs worked this way a few weeks back but now it seems to always return an object
22:08 Darren_Duncan testing more ...
22:15 Darren_Duncan okay, that may not be the problem, ignore what I said
22:15 G2 has quit IRC ("oops")
22:16 alinbsp has quit IRC ()
22:16 Enveigler has joined #perl6
22:16 Darren_Duncan has left
22:21 Limbic_Region_ has joined #perl6
22:22 Limbic_Region_ is now known as Limbic_Region
22:27 * geoffb hopes Limbic_Region ate or something in between there, and doesn't face a nearly 2 hour commute . . . .
22:38 jamhed has quit IRC ("ðÏËÉÄÁÀ")
22:46 justatheory has joined #perl6
22:52 vcv-- has joined #perl6
22:53 vcv- has joined #perl6
22:57 Enveigler has quit IRC (Connection timed out)
22:59 wolverian hmm, got my friend to learn haskell
22:59 geoffb wolverian++
23:00 wolverian :)
23:00 wolverian is there syntax for constructing a lazy list in perl6 specced?
23:00 wolverian er, rephrase that as needed. I suck at 2am
23:01 geoffb I recall seeing something, but I also recall seeing something saying that it changed.  Useless answer, I know, but what I mean is "I don't know for sure, but I wouldn't necessarily trust AES on this one."
23:02 wolverian right. I recall a thread on p6l on this subject without a final decision from anyone.
23:02 * wolverian digs it up
23:02 geoffb nodnod
23:02 * geoffb is reading Perl5 test library to see if there are techniques we can steal^Wborrow^Wclone^Wuse
23:03 geoffb Judging by the amount of crap special-casing VMS testing, this comment is appropriate
23:03 geoffb # VMS will avenge.
23:03 geoffb Yep.
23:03 wolverian wow, did google change back to the before-beta version of the google groups interface? yay!
23:03 eric256_ yea my message posted to p6l ;-)
23:06 SamB wolverian: ooh, cool!
23:06 SamB why did they even change it?
23:06 * SamB wonders why it went "beta" all of a sudden in the first place
23:07 wolverian SamB: I don't know. the beta version SUCKED.
23:07 SamB yeah
23:07 geoffb Someone thought they were being clever.
23:07 revdiablo the beta seemed to be a knock-off of the gmail interface
23:07 geoffb Google Maps -- now that's clever.
23:07 geoffb Maybe done by the same people, before they started Gmail?
23:08 revdiablo maybe at the same time, part of a Grand Unified Scheme
23:08 wolverian my @fibs = 1, 1, [+] zip @fibs, @fibs.tail; # I want this to work correctly (lazily) :)
23:08 geoffb BTW, from reading these Perl5 tests, I think I can definitively say "If Win32 is the least sane platform you've ever worked on, count yourself LUCKY."
23:09 geoffb wolverian, ooh, cute.
23:09 wolverian geoffb: yes. :) this is why I want to know how laziness works on the syntax level. to compare with haskell!
23:09 wolverian (the above was given to me in haskell as an example of its elegance.)
23:10 wolverian (well, as an example of _something_. I don't know what the purpose was, really.)
23:10 geoffb *chuckle*
23:13 cognominal in http://svn.openfoundry.org/pugs/src/Emit/PIR.hs, what the following means?
23:13 cognominal {-
23:13 cognominal ...
23:13 cognominal -}
23:14 kelan cognominal: multi-line comments in haskell
23:15 kelan actually inline comments because you can put them inside expressions
23:16 cognominal thx
23:17 SamB {-# FOO #-}, on the other hand, would be a pragma...
23:18 mugwump whereas {-< >-} is a Tie Fighter
23:18 SamB no, the TIE cockpits are spherical
23:19 SamB whereas that has a diamond shaped cockpit
23:19 Boogie_ has quit IRC (Read error: 110 (Connection timed out))
23:19 mugwump {-( )-}   {- happy? -}
23:20 geoffb Well,  |-o-| might be better, since Tie Fighters have flat sides, only Tie Interceptors are curved on the sides
23:20 geoffb I can't believe I even have that in my brain
23:21 eric256_ lol. ;)
23:22 eric256_ hmm... circumfix operators could be used to create a multiline comment ;)
23:22 wolverian hmm. shouldn't ?SUB work as well, besides &?SUB?
23:23 * mugwump notes you can get all these craft and more in Celestia from the www.celestiamotherlode.net
23:30 * eric256_ searches for the spec on circumfix operators
23:31 mugwump eric256_: it's only in the relevant apocalypse iirc
23:32 mugwump A06
23:33 mugwump also A12
23:34 larsen has quit IRC ("later")
23:35 eric256_ thanks
23:35 eric256_ i would never have checked A06
23:39 _meppl_ has joined #perl6
23:40 Enveigler has joined #perl6
23:43 meppl has quit IRC (Read error: 145 (Connection timed out))
23:48 _meppl_ is now known as meppl
23:51 Limbic_Region seen chip
23:51 Limbic_Region he's probably sleeping
23:55 dkf has joined #perl6
23:58 Limbic_Region has quit IRC ("tiuq\")

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

Perl 6 | Reference Documentation | Rakudo