pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, elf: etc.) (or perl6: for all) | irclog: irc.pugscode.org/
Set by Tene on 29 July 2008.
00:02 mmorrow joined 00:05 mmorrow left 00:11 bacek_ joined 00:13 bacek left 00:46 Exodist left 00:50 px_ left 00:55 bacek__ joined 00:57 bacek_ left 01:07 kanru joined 01:21 wknight8111 left 01:28 Limbic_Region left 01:49 samlh left 02:08 rhr joined 02:09 mj41 left 02:15 Alias_ joined
pugs_svn r22986 | lwall++ | [Cursor] factor out fate linked list generation from "ch2" code so we can 02:17
r22986 | lwall++ | refer to fates by number later when we are more dfa-ish.
02:34 |Jedai| joined 02:35 kanru left 02:51 Jedai left 02:52 kanru joined
lizsac does lwall ever come in here? 03:02
03:06 alester joined 03:16 sri_kraih joined 03:18 meppuru joined 03:25 sri_kraih_ left
TimToady beats me 03:32
alester Q: Why do you stay with your husband, Mrs. de Sade? 03:33
03:34 meppl left 03:38 johnjohn101 joined
johnjohn101 is perl6 supposed to fix the perl threading model currently in use? 03:39
TimToady in spots 03:40
johnjohn101 what about the memory usage for threads?
TimToady there's much less global state, so probably less 03:42
03:43 stephenos left
meppuru good night 03:44
03:44 meppuru is now known as meppl 03:46 meppl left 03:52 hercynium left 03:54 johnjohn101 left 04:02 elmex_ joined 04:03 Patterner1 joined 04:04 nacho_ left 04:14 Patterner left, Patterner1 is now known as Patterner 04:16 elmex left, elmex_ is now known as elmex 04:29 alc joined 04:50 z80ASM joined 05:03 z80ASM left 05:19 sri_kraih left 05:21 rhr_ joined 05:27 smg left 05:33 justatheory left 05:34 sri_kraih joined, rhr left 05:41 rhr_ is now known as rhr 05:53 jauaor left 06:14 advizor joined, advizor left 06:42 ashizawa joined 06:43 sri_kraih left 06:54 masak joined 06:57 smg joined 06:58 Bzek joined 07:01 sail0r joined, sail0r left 07:34 samlh joined 07:40 alester left
moritz_ TimToady: STD.pm dies with "Out of memory" on ../../t/spec/S06-traits/precedence.t (more than 2G available) 07:41
masak perl6: say log(0) 07:44
p6eval pugs: OUTPUT[-Inf␤]
..rakudo 32569: OUTPUT[-inf␤]
..elf 22847: OUTPUT[Undefined subroutine &GLOBAL::log called at (eval 119) line 3.␤ at ./elf_f line 3861␤]
masak S29 says calling with 0 is an error. 07:45
moritz_ is elf_f still the things to execute, or should I use elf_g or something instead?
07:53 mj41 joined 08:07 DemoFreak joined 08:16 emacsian joined, emacsian left 08:17 ludan joined 08:18 aindilis left 08:27 stephenos joined 08:37 drbean_ joined 08:42 ejs joined
masak arrgh, we need to start using Perl 6 on a massive scale before all these well-meaning self-proclaimed web journalists succeed in convincing the world that it's vapourware! 08:49
08:54 drbean left
pugs_svn r22987 | audreyt++ | * Support Cabal 1.6.0 (comes with GHC 6.10.1) in Makefile.PL. 08:58
r22988 | audreyt++ | * GHC 6.10 support, part 1 of 3: Adjust for Map.lookup's new signature. 09:00
r22989 | audreyt++ | * GHC 6.10 support, part 2 of 3: Adjust for extensible exceptions.
r22990 | audreyt++ | * GHC 6.10 support, part 3 of 3: Adjust for GADT's "rigid types" restriction. 09:01
fullermd Well, until it's released, it IS vapourware :p 09:03
moritz_ fullermd: that statement gives definition power to the release process. So if I release it today, does it stop being vapourware? 09:04
fullermd Yes.
(it may be completely broken crap non-vapourware, but it's non-vaprous at least ;)
09:04 idemal left
moritz_ well, parrot+rakudo has monthly releases 09:05
so by your definition it's not vapourware anymore
fullermd suspects you're reading way too much semantic rigor into offhand flip statements...
09:06 iblechbot joined
masak fullermd: the point is, words are powerful, and spreading lies can be harmful. 09:06
fullermd But still. Are those rakudo releases "perl 6", or "steps toward perl 6"? It's a definitional issue to be sure, but...
masak fullermd: stop discussing and start trying out Perl 6 for yourself instead. :)
fullermd: then you don't need to exactly define your terms. 09:07
fullermd Right, 'cuz perl 5 doesn't make me cranky ENOUGH with PHP :p
masak fullermd: I didn't understand that last part. what does Perl 5 have to do with PHP?
moritz_ fullermd: without semantic rigour discussions are pointless, and just contribute to FUD.
masak my sentiments exactly. 09:08
moritz_ anyway, I should focus on that freakin' tight binding hamiltonian now...
fullermd shrugs. 09:09
I don't intend to spend paragraphs on precise term delineation when I'm explicitly blurring terms to have fun with language.
masak fullermd: well, if enough people think it's "fun" to go around calling Perl 6 vapourware when it actually isn't, it won't matter much that Perl 6 exists today and can be used to write code. people simply won't try it out, because they've been misinformed. 09:10
fullermd Whistling in the dark isn't meant to be the same as 'discussion'.
09:11 tomyan joined
fullermd Don't worry. It won't happen again. 09:11
09:11 fullermd left
masak another happy acolyte! yay! 09:11
09:11 idemal joined 09:16 cosimo joined
masak is there a way to ask people not to call Perl 6 bad names without intimidating them? 09:21
literal ask nicely? :)
masak literal: I don't think I was rude in the discussion above, but I do think that fullermd got a bit scared anyway. 09:22
literal I guess he was miffed because he wasn't really being serious when he called it vapourware 09:23
masak aye.
the discussion was more serious than he was. 09:24
but then again, his claims to not being too serious were in response to moritz_ rigourously proving him wrong. I don't think he would have claimed to be unserious if we hadn't objected. 09:26
09:42 stephenos left, masak left 09:47 agentzh left 09:53 agentzh joined
szabgab where is this vaporware article? 10:00
moritz_ 21cdb.wordpress.com/2008/10/23/perl...re-really/ I think 10:02
lambdabot Title: Perl 6 isn’t vaporware? Really? « 21st Century Digital Boy, tinyurl.com/6chbgf
10:07 meppl joined
szabgab lovely 10:16
10:29 masak joined 10:30 zamolxes left 10:34 ruoso left 10:40 ruoso joined 10:55 ejs left 10:58 ejs joined 11:02 schmalbe joined
ruoso @tell pmurias I've been thinking about the AdhocSignature, and I think we can move on with it, even with the need to compare signatures. I realized that Signature is not a native-type, therefore, at most the AdhocSignature will need to be expanded to provide all the API of the standard Signature type, but it can still be completely ad-hoc... 11:11
lambdabot Consider it noted.
11:42 smg left 11:54 masak left 12:23 masak joined 12:26 iblechbot left
masak rakudo: sign(0+1i) 12:31
p6eval rakudo 32570: RESULT[1]
masak it might possibly be considered a bug that &sign accepts Complex instances 12:32
12:35 ashizawa left 13:08 Lorn joined
Lorn ppl, someone ask at perlmonks about Perl6::Form 13:09
masak Lorn: url?
this one? perlmonks.org/?node_id=723128 13:11
lambdabot Title: Perl6::Form verbatim won't do what I want
Lorn masak: www.perlmonks.org/?node_id=723128
lambdabot Title: Perl6::Form verbatim won't do what I want
Lorn masak: yes
masak :)
Lorn: Perl6::Form is written in Perl 5, though.
13:13 smg joined 13:21 iblechbot joined 13:25 alexn_org joined 13:28 cosimo left 13:29 meppl left 13:32 sri_kraih joined 13:35 ChrisDavaz joined 13:43 bacek_ joined 13:45 cognominal left 13:47 z80ASM joined 13:50 alc left, bacek__ left
pmichaud std: my $a; @$a; 14:00
p6eval std 22847: OUTPUT[cp: cannot create regular file `lex/STD/EXPR': Permission denied␤parsed␤]
masak std: my @a[10]; my @b.[10]; 14:01
p6eval std 22847: OUTPUT[parsed␤]
masak I think that's wrong.
the second form is not allowed.
moritz_ please check it locally as well, p6eval's std: target isn't 100% reliable 14:02
masak moritz_: oki, how do I check it locally?
moritz_: PS: I don't have Perl 5.10.
moritz_ masak: then you need to check somewhere that has perl 5.10 14:03
or I can do, wait a sec
14:03 z80ASM left
masak moritz_: thanks. 14:03
I'd like to have 5.10, but so far I haven't found any way to install it. Maybe other Mac OS X 10.5 users know how.
moritz_ yes, it parses 14:04
masak and do you agree that it shouldn't?
pmichaud I don't see where @$a gets parsed. 14:05
moritz_ masak: aye
masak @tell TimToady STD gives a false positive on `my @a.[10];`
lambdabot Consider it noted.
masak
.oO( false parsitive )
pmichaud it might be parsing as (my @a).[10] 14:06
masak is @%$a correct Perl 6?
pmichaud: still worthy of reporting, in that case.
pmichaud there are a _bunch_ of instances of @$foo in the synopses. 14:07
moritz_ std: @$a
p6eval std 22847: OUTPUT[parsed␤]
masak std: @%$a
pmichaud but I can't see where STD parses that.
p6eval std 22847: OUTPUT[parse failure␤]
moritz_ DumpMatch.pm needs updates wrt recent Cursor changes, I think 14:09
pmichaud perhaps someone could get the parse tree from std for me? (I don't have STD available on this system -- still stuck at perl 5.8)
14:22 ejs left 14:23 Lorn_ joined 14:28 bacek__ joined
rakudo_svn r32571 | pmichaud++ | [rakudo]: spectest-progress.csv update: 216 files, 4528 passing, 8 failing 14:30
r32571 | pmichaud++ | Failure summary:
r32571 | pmichaud++ | S02-literals/radix.rakudo aborted 8 test(s)
r32572 | pmichaud++ | [rakudo]: update test_summary.pl to report failures, output < 80 cols
14:35 Lorn left, d4l3k_ joined, dalek left, d4l3k_ is now known as dalek 14:37 pmurias joined
pmurias ruoso: why do you use $/ instead of newline 14:38
lambdabot pmurias: You have 6 new messages. '/msg lambdabot @messages' to read them.
14:39 bacek_ left, pmurias left 14:42 ejs joined, Lorn_ left, Lorn joined 14:43 rdice joined 14:53 Exodist joined
rakudo_svn r32573 | pmichaud++ | [rakudo]: Fix .list method on arrays (and arrayrefs). 15:00
15:00 rindolf joined 15:05 Lorn_ joined
ruoso @tell pmurias I use $/ just because it's easier to write than "\n" considering I'm using single-quote most of the time... 15:05
lambdabot Consider it noted.
masak :)
15:15 ejs left, ejs joined 15:18 alester joined 15:20 kanru left 15:25 kanru joined 15:26 mepp1 joined, mepp1 is now known as meppl 15:28 Lorn left
masak rakudo: say sin(0) 15:42
p6eval rakudo 32573: OUTPUT[0␤]
masak sin shouldn't be exported by default, according to S29.
15:44 cosimo joined 15:46 hercynium_ left, hercynium joined
Alias_ masak: So there should only be the original sin? 15:49
(ducks)
masak er.
Alias_: more like, there should only be the :Trig sin. which can then be exported using something like `use Num :Trig;` 15:56
15:57 schmalbe left
Alias_ I think your pun detector has malfunctioned slightly :) 15:57
masak Alias_: it's just a tendency I have to try to think up serious answers to people who are obviously joking. 15:58
Alias_: it's fun, you should try it. :)
Alias_ I do, but in my case it's my sarcasm detector 15:59
masak Alias_: (besides, it's only a pun if something has two interpretations. I could only deduce the punny one.)
rindolf Hi Alias_
Alias_ \o
rakudo_svn r32574 | pmichaud++ | [rakudo]: Make log(0) return a Failure (RT #60490, masak++) 16:00
masak pmichaud++
though I find the non-local exception throwing mechanism a bit disconcerting, to be honest. 16:02
16:02 masak left 16:07 rindolf left 16:11 stephenos joined
ruoso later & 16:20
16:20 ruoso left 16:32 masak joined
pmichaud (non-local) I think that's so that things like my $b = any(0..5); my $c = log($b); doesn't always blow up in your face. 16:35
masak oh. makes sense. 16:36
pmichaud and eventually the exception will report the location where it was created -- we just don't have that in Parrot yet.
16:39 smg left 16:40 jhorwitz joined
pmichaud (sin) since we don't have a good import mechanism yet, we should probably leave it global by default for now. That will likely change soon. 16:40
16:43 tomyan left
TimToady it remains to be seen exactly what gets imported into the prelude by default 16:44
lambdabot TimToady: You have 1 new message. '/msg lambdabot @messages' to read it.
TimToady but we're just going with a conservative approach, since it's much harder to undo later 16:45
masak: my @a.[10] is equivalent to my @a[10] which is is a shaped array declaration 16:47
masak TimToady: then I must have misunderstood S09
"Note that the square brackets are a compile-time declarator, not a run-time operator, so you can't use the "dotted" form either" 16:48
TimToady well, it could be that S09 is wrong
masak certainly.
pmichaud ...the synopses can be wrong? *shudder* :-)
16:48 alexn_org left
masak pmichaud: according to the authority hierarchy, it can. 16:49
:)
pmichaud TimToady: (quick question, I hope) - what parses the leading @ in something like @$foo
TimToady "Can Larry create a synopsis he cannot implement?"
pmichaud "No, but Damian probably can."
moritz_ it's not so hard. halt(&f) yields 1 if f() ever stops, 0 otherwise. Now implement.... 16:50
16:50 pbuetow joined
masak moritz_: sssh! don't give him ideas! :) 16:51
TimToady pmichaud: desigilname
or rather, desigilname is what parses everything after the @ 16:52
note that variable calls desigilname, which can call variable 16:53
pmichaud okay, that's what I missed. THanks.
I was thinking of "desigilname" as "name without any sigil", but it really means "name with one less sigil" I guess. 16:54
or, "take away a sigil"
16:57 cosimo left, justatheory joined
masak also, from S09: `@@shape = (4;2);` -- how does this work? what variable has been declared beforehand? @shape, right? can one write `my @@shape = (4;2);` ? 17:03
TimToady yes, @@shape always stores as @shape, but gives slice context instead of list context 17:10
and slice context pays attention to ; while list context doesn't 17:11
masak aha, ok.
TimToady at least, that's the theory--remains to be seen how it works out in practice
17:11 tewk_ joined 17:13 adc_Penner joined
pmichaud ....this is all still theoretical? ;-) 17:15
can't we find someone who can reduce it to pra.... oh. Never mind. 17:16
17:16 pmurias joined, ruoso joined
pmurias ruoso: hi 17:17
lambdabot pmurias: You have 1 new message. '/msg lambdabot @messages' to read it.
ruoso hi pmurias 17:19
pmurias ruoso: i find $/ a bit strange, as the input record separator isn't always a newline
* doesn't have to be a newline
ruoso I know I know... it's just lazyness...
:P
17:20 pbuetow left
pmurias you could make vim insert "\n" easily 17:20
ruoso but we could have a local $/ = "\n" at the beggining of mildew :P
pmurias or i could just fix $/ to "\n"
ruoso heh
no problem...
17:21 masak left
pmurias ruoso: we have memory leaks in smop again? 17:21
ruoso yes, we do... I don't know when it did start... 17:22
17:23 tewk left, ejs left
TimToady on my keyboard $/ is harder to type than \n, plus it's obfuscational 17:24
[particle] $/ is easier to type than "\n" 17:26
print $_,$/ for @a; # i prefer that over "$_\n"
TimToady I prefer .say 17:27
:)
but say $_ works in p5
.10
but in 5.8 I'd write print "$_\n" 17:28
ruoso [particle], it gets even more annoying when you're using single quotes before the newline...
'something'."\n"
ruoso .oO( note that we're talking about a long string concatenation, not a single call to print 17:29
TimToady anyway, you really want to get out of the habit of using $/ for that or you'll get confused in p6 :)
ruoso TimToady, how does 4 strokes can be easier than 2 strokes?
17:29 pmurias_ joined
TimToady I'm just saying I usually prefer interpolation anyway, so I usually already have quotes 17:30
ruoso ok... ok... I usually only use $/ when it's after a single-quote
17:30 pmurias_ left
TimToady but $/ *is* obfuscational to the casual reader 17:31
ruoso wonders if ␤ can be a core function in Perl 6 ;P
TimToady and anything in your dynamic scope can change $/, so if you call out and back in to foreign code, you could be hosed
␤ is even harder to type... 17:32
17:34 pmurias_ joined
ruoso but it's pretty ;) 17:34
[particle] not in my font
TimToady unfortunately, isn't a word character, so it'd have to be an operator
17:35 pmurias left
TimToady or a term:<␤ > 17:35
ruoso in my environment I just have to type Ctrl+Shift+U, 2424
pmurias_ 5 strokes 17:36
TimToady 7 keystorks, whatever a keystork is
ruoso yeah... but it's pretty ;)
pmurias_ it's strange
jnthn So...it's pretty strange? ;-) 17:37
TimToady so if we have that we should also have ones like ␇ and ␚ 17:39
pugs_svn r22991 | pmurias++ | [mildew] added suport for block in item context my $foo = {say 5} and for .()
r22991 | pmurias++ | also added a &capturize helper for creating captures
r22992 | pmurias++ | [mildew] replaced $/ by "\n" 17:41
17:42 pmurias_ is now known as pmurias
ruoso votes for unicode subs 17:44
:)
pmurias votes for ASCII
17:46 zamolxes joined
pmurias ruoso: if we had a :(@*_,%*_) in m0ld+C we could have realish signatures in p6 17:46
17:47 PerlJam left, PerlJam joined
pmurias later& 17:47
17:47 pmurias left
ruoso pmurias, I think you mean *@_, *%_... 17:47
18:03 pbuetow joined 18:09 cognominal joined, smg joined 18:11 rhr_ joined, rdice left 18:12 Alias_ left 18:22 pcbuetow joined, pcbuetow left 18:24 bacek_ joined, rhr left 18:26 rhr_ is now known as rhr
ruoso @tell pmurias the thing is that Signature is not a native-type, therefore anything that .^does(Signature) is a valid signature. I think it will be even better for the ACCEPTS and BIND code to be compiled directly than have to traverse the Signature object to do that... I could even argue that Signature is a Role, not a Class 18:32
lambdabot Consider it noted.
ruoso TimToady, what do you think about Signature being a role instead of a class?
18:33 bacek__ left 18:36 ludan left
ruoso and actually being only an interface, not even providing any implementation... 18:37
18:38 pbuetow left
TimToady seems like it might be sane 18:46
ruoso I'm quite convinced the compiler generating ad-hoc signature objects is pretty much a good idea not only for bootstrap, but as the actual way it works 18:52
18:55 smtms left 18:59 armagad joined 19:12 jauaor joined 19:14 Bzek left, smg left
TimToady demos welcome :) 19:17
wow, my trie implementation inspired 5.10 to say 19:28
Attempt to free unreferenced scalar: SV 0xaaaa058 at STD.pm line 9180.
moritz_ TimToady: try with the 5.10-maint branch, it fixed several such bugs 19:29
TimToady well, I think if my implementation were right it wouldn't trigger that...
it was trying to recover from calling a non-existent sub, after all 19:30
actually, I'm quite sure it was from feeding trash to YAML::XS :) 19:31
pugs_svn r22993 | ruoso++ | [mildew] doing s{\$/}{"\n"}g might not be a good idea if you have regexes using slashes as terminators and trying to get the end of the string 19:37
TimToady see, even the p5 lexer thinks $/ is obfuscated. :) 19:38
pugs_svn r22994 | ruoso++ | [mildew] adding a test for a signature with two positional arguments, this is going to be the initial test bed for ad-hoc signatures
ruoso TimToady, it would be nice if the VAST returned already placed the implicit "our"s in the declarations... 19:40
ruoso bbiab &
TimToady I'm a bit leary of compilers that lie to themselves... 19:44
I'm also leary of YAML implementations that are stupid about unicode 19:45
YAML::XS Error: Emit scalar '�', error: expected SCALAR, SEQUENCE-START, MAPPING-START, or ALIAS 19:46
and that exit without throwing a catchable Perl error 19:48
20:11 rindolf joined 20:22 aindilis joined
pugs_svn r22995 | moritz++ | [t/spec] clean up of S29-num/rounders.t 20:22
r22995 | moritz++ | * remove rakudo cludge, it now handles nested arrays sufficiently well
r22995 | moritz++ | * replace "$array[0]" interpolation by "{$array[0]}" for poor compilers
r22995 | moritz++ | that don't do one-pass-parsing yet
r22995 | moritz++ | * use numeric comparison for numeric values
diakopter TimToady: leery? 20:39
x2
20:39 ChanServ sets mode: +o diakopter, diakopter sets mode: +o TimToady, diakopter sets mode: -o diakopter 20:45 smg joined 20:50 jferrero joined
[particle] TimToady: how about coersion with *('file.foo') for strings to io objects? ;) 20:55
moritz_ wonders if [particle] had much exposure to fortran formats ;-) 21:00
21:05 elmex left
[particle] i still find myself reviewing algorithms in "numerical recipies in fortran" from time to time 21:11
21:16 masak joined
TimToady I'd much prefer io() or .io 21:20
though perhaps not *quite* as dwimmy as inginess would suggest 21:21
21:25 rindolf left 21:27 zamolxes left, sri_kraih left, iblechbot left, mj41 left, cpfr left, araujo left, jan_ left, PZt left, ilbot2 left, StephenPollei left, scrottie left, kane_ left, kolibrie left, Guest16015 left, hcchien left, sirhc left, Trey left, moritz_ left, allbery_b left, buubot left, IRSeekBot left, armagad left, PerlJam left, adc_Penner left, nothingmuch left, charsbar left, clkao left, ewilhelm left, pasteling left 21:29 zamolxes joined 21:34 aindilis` joined 21:43 StephenPollei joined, kcwu left 21:45 buubot joined, sri_kraih joined, iblechbot joined, mj41 joined, cpfr joined, araujo joined, jan_ joined, PZt joined, ilbot2 joined, kane_ joined, sirhc joined, Trey joined, Guest16015 joined, moritz_ joined, hcchien joined, kolibrie joined 21:46 smg left 21:47 kcwu joined, armagad joined, PerlJam joined, adc_Penner joined, nothingmuch joined, charsbar joined, pasteling joined, clkao joined, allbery_b joined 21:49 aindilis left 21:50 ewilhelm joined 22:00 pedrob joined 22:06 adc_Penner left 22:15 alester left
pugs_svn r22996 | masak++ | [t/spec/integration/substr-after-match-in-gather-in-for.t] 22:20
r22996 | masak++ | * added this one test from a bug discovered in November development
22:25 Lorn_ left 22:33 masak left
moritz_ pugs: say %*ENV[0] 22:36
p6eval pugs: OUTPUT[*** Can't modify constant item: VUndef␤ at /tmp/C8C8rDERe5 line 1, column 5 - line 2, column 1␤ /tmp/C8C8rDERe5 line 1, column 5 - line 2, column 1␤]
moritz_ pugs: say %*ENV<PATH>
p6eval pugs: OUTPUT[*** Can't modify constant item: VUndef␤ at /tmp/LYvDSpZgem line 1, column 5 - line 2, column 1␤ /tmp/LYvDSpZgem line 1, column 5 - line 2, column 1␤]
moritz_ maybe related to evalbot's server not having ghc 6.10.1 yet 22:37
rakudo_svn r32580 | moritz++ | [rakudo] more test files for spectest, and restored alphabetical order. 22:50
23:05 pjcj left 23:08 pjcj joined 23:17 smg joined 23:21 scrottie joined 23:22 spx2 joined 23:25 PZt left 23:30 iblechbot left, PZt joined 23:33 hercynium left 23:39 rakudo_svn left 23:42 pugs_svn left 23:43 Exodist left 23:45 aindilis` left 23:50 nacho_ joined, _Thoth_ joined 23:51 p6eval left, p6eval joined 23:58 aindilis joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!