pugscode.org/ | nopaste: sial.org/pbot/perl6 | ?eval [~] <m oo se> | We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by diakopter on 11 July 2007.
00:05 polettix left 00:06 theorb joined 00:07 devogon left 00:19 theorbtwo left 00:23 weinig left 00:28 Schwern joined 00:31 rfordinal left 00:35 REPLeffect left 00:36 davidfetter joined
davidfetter hello 00:36
i'm packaging parrot RPMs and i'm trying to figure out how to get it to create a perl5 Parrot:: package. anybody have experience with this kind of thing? 00:37
Tene davidfetter: you might have a bit more luck in #parrot on irc.perl.org, if nobody responds here. 00:38
Oh, you're already there. Nevermind.
davidfetter heh 00:39
great CD, that
Tene CD? 00:40
davidfetter Nervermind. it's by nirvana :)
Nevermind*
00:47 BinGOs left, BinGOs joined
Juerd Big network maintenance tomorrow 0:00 - 4:00 CET 00:56
feather will be offline a few times. 00:57
(online intermittently...)
00:58 baest left 01:01 nipotaway is now known as nipotan 01:03 lyokato joined 01:26 felipe left 01:44 BinGOs left 01:48 BinGOs joined, weinig joined 02:06 vel joined 02:20 felipe joined 02:24 felipe left, DarkWolf84 left, hkb_zz left, thepler left, Gothmog_ left, rhizo left, cerridwen left, rafl left, szbalint left, takanori_ left, weinig left, BinGOs left, cognominal__ left, ting left, lisppaste3 left, fglock_ left, knewt left, exp_evalbot left, rgs left, Maddingue left, cj left 02:25 felipe joined, weinig joined, BinGOs joined, DarkWolf84 joined, cognominal__ joined, hkb_zz joined, ting joined, lisppaste3 joined, thepler joined, Gothmog_ joined, fglock_ joined, knewt joined, exp_evalbot joined, rhizo joined, rgs joined, Maddingue joined, cerridwen joined, rafl joined, szbalint joined, cj joined, takanori_ joined 02:33 ajelenak joined 02:34 ajelenak left 02:35 weinig left 02:40 vel left 02:41 Schwern left 02:48 araujo joined 02:55 Lunchy left 03:02 jonadab joined 03:04 Schwern joined
Schwern How does one rollback a change? svk merge -c1234 //local/project says it did something, but appears to have had no effect. 03:04
Hey, this isn't #svk!
cads!
03:06 jonadab left 03:12 perlmonkey2 joined, perlmonkey2 left 03:13 vsoni joined 03:26 bsb_ joined 03:27 bsb left 03:32 weinig joined, weinig left, weinig joined 03:37 vsoni left 03:44 Schwern left 03:47 Lunchy joined 03:53 penk left 04:02 penk joined 04:15 amnesiac joined 04:17 jhorwitz left 04:20 hkb_zz is now known as hkb_lab 04:50 penk left 05:03 penk joined 05:30 Ashizawa joined 05:42 Ashizawa_ joined, Ashizawa left 05:52 rfordinal joined 06:04 devogon joined 06:09 liffey joined 06:11 DarkWolf84 left 06:52 ft left 07:03 fglock_ left 07:04 amnesiac left, iblechbot joined 07:23 riffraff joined 07:27 elmex joined 07:33 Aankhen`` joined 07:36 jisom joined 07:41 franck___ joined 07:43 ting left, ting joined 08:03 drrho joined 08:40 rgs left 08:41 rgs joined 08:46 dduncan left 08:52 riffraff left 09:12 lyokato left 09:33 riffraff joined, riffraff left 09:36 fglock joined 09:37 bsb_ left 09:39 lyokato joined
pugs_svnbot r17966 | Aankhen++ | * added *.fasl to svn:ignore. 09:42
r17967 | Aankhen++ | * kp6 Lisp: added package.lisp
diff: dev.pugscode.org/changeset/17967
lambdabot Title: Changeset 17967 - Pugs - Trac
09:44 c9s_ left 09:46 rhizo left
pugs_svnbot r17968 | Aankhen++ | kp6-cl: 09:48
r17968 | Aankhen++ | * normalized EOL at EOF.
r17968 | Aankhen++ | * placed all code in KP6-CL package.
diff: dev.pugscode.org/changeset/17968
lambdabot Title: Changeset 17968 - Pugs - Trac
pugs_svnbot r17969 | Aankhen++ | * kp6-cl: eliminate redundancy in Runtime.lisp 09:54
diff: dev.pugscode.org/changeset/17969
lambdabot Title: Changeset 17969 - Pugs - Trac
Aankhen`` Strangeā€¦ my changes to EmitLisp.pm don't seem to be taking effect. 09:55
avar did you do `make' ? 09:58
Aankhen`` Er, no. 09:59
One sec. :-P
Boy, it sure takes a while. 10:10
avar: I ran make, now what? 10:16
10:16 jisom left
Aankhen`` The changes still don't seem to be having an effect. :-| 10:20
AVAR WAKE UP 10:27
10:28 xinming left
avar uh, I don't know 10:28
what did you change 10:29
10:30 xinming joined
Aankhen`` Added a line to CompUnit::emit_lisp's output. 10:30
avar don't see it after my svk pull 10:31
Aankhen`` I haven't checked it in, wanted to test it.
One sec.
Okay, pull now. 10:34
pugs_svnbot r17970 | Aankhen++ | kp6-cl:
r17970 | Aankhen++ | * import CL package.
r17970 | Aankhen++ | * use #: notation in DEFPACKAGE.
diff: dev.pugscode.org/changeset/17970
lambdabot Title: Changeset 17970 - Pugs - Trac
avar why use #: ? 10:37
Aankhen`` To avoid interning a bunch of symbols in the KEYWORD package. 10:38
It's not required, just good form.
avar I don't see anything new in emit_lisp
ah, defpackage 10:39
Aankhen`` You ought to see ~ ' (:use #:cl))' ~ Main::newline() at the very least.
avar $ echo '[ 1, 2 ]' | perl kp6-mp6-perl5.pl --lisp
(defpackage #:Main
(:use #:cl))
(in-package #:Main)
works here
Aankhen`` Then why not for me? :-S
avar ;; Compiling file /var/src/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Runtime/Lisp/Array.lisp ... 10:40
WARNING: Cannot redefine #<BUILT-IN-CLASS ARRAY> with a different metaclass #<STANDARD-CLASS STANDARD-CLASS>
*** - (SETF FIND-CLASS): cannot redefine built-in class #<BUILT-IN-CLASS ARRAY>
Aankhen`` I'm trying the exact same code as you, but it gives the same output as before.
Yup.
The code defines some classes with the same names as builtin classes. 10:41
avar yes you're not regenerating for some reason
Aankhen`` Yeah.
avar what system is this?
what version of make?
Aankhen`` 3.81
pugs_svnbot r17971 | avar++ | r63583@t: avar | 2007-09-20 10:41:03 +0000 10:42
r17971 | avar++ | generated files with EmitLisp.pm changes
diff: dev.pugscode.org/changeset/17971
lambdabot Title: Changeset 17971 - Pugs - Trac
Aankhen`` I'm going to try running `make` again.
After that I'll start mangling the class names.
Actually, I can do that while it's building, since it takes so long.
Oh, now it works. 10:44
And it only compiled EmitLisp.pm this time.
pugs_svnbot r17972 | Aankhen++ | * kp6-cl: name mangling to prevent conflicts with builtin classes. 10:48
diff: dev.pugscode.org/changeset/17972
lambdabot Title: Changeset 17972 - Pugs - Trac
avar you also have to change the emitter 10:49
Aankhen`` Indeed. :-) 10:50
pugs_svnbot r17973 | Aankhen++ | * kp6-cl: name mangling in EmitLisp.pm too.
diff: dev.pugscode.org/changeset/17973
lambdabot Title: Changeset 17973 - Pugs - Trac
avar forgot 'make' :) 10:51
pugs_svnbot r17974 | Aankhen++ | * kp6-cl: missed a couple names to mangle. 10:53
diff: dev.pugscode.org/changeset/17974
lambdabot Title: Changeset 17974 - Pugs - Trac
Aankhen`` Bleh, missed another one too.
pugs_svnbot r17975 | Aankhen++ | * kp6-cl: regenerated files with EmitLisp.pm changes.
diff: dev.pugscode.org/changeset/17975
lambdabot Title: Changeset 17975 - Pugs - Trac
Aankhen`` Back in a few, lunch. :-) 10:57
11:07 lyokato left
avar echo 'say { foo => "bar" }' | perl kp6-mp6-perl5.pl --lisp | perl -pe 's/apply \$global::code_/apply kp6-cl::/i'|cli\ 11:14
sp
pugs_svnbot r17976 | fglock++ | [kp6-kp6] added some debugging code 11:17
diff: dev.pugscode.org/changeset/17976
lambdabot Title: Changeset 17976 - Pugs - Trac
Aankhen`` avar: ? 11:21
11:29 rindolf joined
avar Aankhen``: try that 11:29
*** - FIND-CLASS: KP6-STR does not name a class
Aankhen`` Yeah, I've added the classes to the export list locally.
avar :)
Aankhen`` I'm more concerned with the invocant bit though.
How does the $GLOBAL::code_say get in there?
Or rather, isn't there any way to just get the name of the routine being called? 11:30
I'm just doing an s/// for now.
Or perhaps not. Bleh. 11:31
Trying to use s/// makes it blow up.
fglock Aankhen``: the routine is a Code object; GLOBAL::Code_say is the name of the variable that contains the Code object
lambdabot fglock: You have 1 new message. '/msg lambdabot @messages' to read it.
Aankhen`` fglock: Ah. 11:32
How would one mangle that into kp6-cl::say? 11:33
fglock can you use namespaces? GLOBAL is required, if possible 11:34
Aankhen`` kp6-cl is the namespace in kp6-cl::say, natch. I could change it to GLOBAL::say if you'd prefer, though. 11:35
fglock how about kp6::GLOBAL::say 11:36
or kp6-GLOBAL::say
Aankhen`` The latter seems good to me.
avar CL has no concept of sub-namespaces right? 11:37
Aankhen`` avar: Right.
You could have a module named kp6::GLOBAL.
But that would be the entire name, not a hierarchy.
pugs_svnbot r17977 | Aankhen++ | kp6-cl:
r17977 | Aankhen++ | * export types.
r17977 | Aankhen++ | * don't use for-side-effects loop in SAY.
r17977 | Aankhen++ | * load runtime before other code.
r17977 | Aankhen++ | * regenerated EmitLisp.pm files.
diff: dev.pugscode.org/changeset/17977
lambdabot Title: Changeset 17977 - Pugs - Trac
fglock Aankhen``: i'll change it 11:38
Aankhen`` Kewl.
11:38 buetow joined 11:39 drrho left
fglock how about kp6-GLOBAL::kp6-say, such that both the namespace and the variable name are mangled? 11:41
Aankhen`` Sure. 11:42
The AST needs the variable name to be mangled?
11:46 drrho joined
fglock no, just at the emitter (almost done) 11:46
Aankhen`` Ah. 11:47
pugs_svnbot r17978 | fglock++ | [kp6-lisp] &GLOBAL::say mangled to kp6-GLOBAL::kp6-Code_say 11:49
diff: dev.pugscode.org/changeset/17978
lambdabot Title: Changeset 17978 - Pugs - Trac 11:50
11:54 Ashizawa_ left
avar ;; Compiling file /var/src/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Runtime/Lisp/GLOBAL.lisp ... 11:57
*** - SYSTEM::%FIND-PACKAGE: There is no package with name "KP6-GLOBAL"
hrm, wtf
fglock GLOBAL.lisp declares itself as #:kp6-cl 11:58
Aankhen`` I'm working on it. :-) 12:00
Should a KP6-UNDEF class go in Undef.lisp?
Or is there a better place for it?
12:01 Schwern joined
pugs_svnbot r17979 | fglock++ | [kp6-lisp] declare package name in GLOBAL 12:02
diff: dev.pugscode.org/changeset/17979
lambdabot Title: Changeset 17979 - Pugs - Trac
fglock Aankhen``: i think so
12:03 vel joined
pugs_svnbot r17980 | Aankhen++ | * kp6-cl: added Undef.lisp 12:03
diff: dev.pugscode.org/changeset/17980
lambdabot Title: Changeset 17980 - Pugs - Trac
avar ;; Loading file /var/src/pugs/v6/v6-KindaPerl6/lib/KindaPerl6/Runtime/Lisp/GLOBAL.fas ... 12:05
*** - FIND-CLASS: KP6-CODE does not name a class
is that an issue with it being in #:kp6-cl ?
avar can't use sbcl ;/
12:06 vel left 12:11 qewer4 joined
qewer4 hi, i use Wordpress 2.2.2 Classic Theme 1.5. is it possible to show the treatment-field seperated on a site? 12:12
avar you're in the wrong channel 12:13
qewer4 ive also tried it in #wordpress but its deserted 12:14
avar You should probably try their mailing list instead of joining random unrelated channels asking about wordpress 12:15
Aankhen`` What makes you think anyone here will be able to help you? :-S
qewer4 so can you help me? 12:18
avar: are you avatar? 12:20
avar qewer4: no 12:21
12:22 qewer4 left, cmarcelo joined
moritz_ re 12:24
12:26 rindolf left 12:27 ruoso joined 12:31 nipotan is now known as nipotaway 12:50 hoelzro joined
hoelzro I have a quick question about the ..., ???, and !!! operators in Pugs. I don't mean to sound ungrateful or anything, but I was just wondering why these haven't been implemented yet. It seems like they'd be pretty easy to implement. Are they just not a high priority? 12:52
moritz_ hoelzro: there could be many reasons, one of them priority 12:53
hoelzro: other possible reasons are "no test case", or "too new in the specs"
hoelzro I wasn't aware the feature was that new, but then again, I've only been reading about Perl 6 for a few weeks now =P 12:54
moritz_ hoelzro: I don't know if "too new" applies here, it's just a general guess 12:55
hoelzro oh ok
moritz_ hoelzro: btw do you have a commit bit? 12:56
hoelzro no, but I'd like one; I'd like to try my hand at implementing those operators and the test cases for them 12:57
I figure those three operators would be a good starting point
12:57 xinming left
moritz_ hoelzro: very good idea 12:57
sadly I don't have my password around, so we need to wait for somebody else ;) 12:58
fglock, diakopter, Juerd: are you around? hoelzro wants a commit bit ;)
ruoso hoelzro, are you meaning "??" and "!!" 12:59
the ternary if?
hoelzro no, the yada-yada operator and friends
the ones that call fail, die, and warn
ruoso ah... ok
moritz_ ruoso: do you take the honor of inviting hoelzro? I don't have my password at hand ;) 13:00
ruoso moritz_, and I don't remember the url ;)
moritz_ ruoso: commitbit.pugscode.org/ # I hope ;) 13:01
lambdabot Title: Welcome to CommitBit for PugsCode
ruoso hoelzro, what's your email addr?
hoelzro [email@hidden.address]
ruoso moritz_, hoelzro, oops... I can't invite...
ahmmm 13:02
I think I can
I was just lost in commitbit interface
moritz_ ok ;)
ruoso done 13:03
moritz_ ruoso++
hoelzro thanks much
moritz_ hoelzro: did you read Pugs::Doc::Hack in docs/? 13:05
hoelzro not yet
avar hoelzro: ??? !!! aren't implemented because pugs (the hs implementation) is pretty much abandonware 13:06
moritz_ hoelzro: don't listen to avar ;) 13:07
avar the "na na na na na na" method?:) 13:08
moritz_ yes ;)
avar: or what do you mean by "abandonware"? that it will finally be replaced by a 6-on-6 implementation and doesn't need much sugar?
avar Development of pugs pretty much dropped to a standstill when audrey wasn't active anymore 13:09
at least I haven't seen any substancial .hs commits since early this year 13:10
moritz_ avar: that's why we need to encourage people to work on it ;)
avar Yes by all means work on it, I'm just offering an IMO more realistic explanation for why that and a bunch of other stuff isn't implemented 13:11
moritz_ avar: ok ;)
I think it still makes sense to work on everything but the object model 13:12
hoelzro so there hasn't been much work on the Haskell portion of pugs?
moritz_ hoelzro: right 13:13
hoelzro is that because not enough devs know Haskell?
moritz_ hoelzro: yes, mostly 13:14
hoelzro more work for me then =P
moritz_ hoelzro: audreyt has been seriously sick, and didn't commit very much since then
avar sick + doing other stuff
syck?:)
moritz_ yes
ruoso fglock, we need to hunt doewn the bind/assign problem 13:15
I just got another accidental circular reference
which makes a deep recursion that eats the memory in 10 seconds
moritz_ hoelzro: btw it is a cutom to add your name to AUTHORS in the first commit, to test the commit bit 13:17
ruoso hoelzro, and you can always check your karma ;)
(is karma alive?)
moritz_ @karma ruoso 13:18
lambdabot ruoso has a karma of 96
moritz_ @karma moritz
lambdabot moritz has a karma of 106
hoelzro karma? 13:19
moritz_ it's really just a programmer's jok 13:20
s/jok/joke/
avar @karma avar
lambdabot You have a karma of 46
moritz_ when somebody writes hoelzro++, your karma is increased by one
avar I GROW STRONGER
hoelzro oh
I assume it won't work if I do it? =D 13:21
moritz_ avar: but compared to ruoso and me you still suck ;) SCNR
hoelzro: right
avar moritz_: I challange you to unicycle jousting
ruoso hoelzro, it'll probably decrease
hoelzro haha
pugs_svnbot r17981 | ruoso++ | [kp6] preventing circular references in Visitor/Token.pm
diff: dev.pugscode.org/changeset/17981
lambdabot Title: Changeset 17981 - Pugs - Trac
moritz_ hoelzro: and when you commit something to the svn repo, pugs_svnbot will automatically increase your karma ;)
avar the longer the log message the more KARMA 13:22
ruoso avar, hehe... I've never realised that
ruoso will start to write longer messages
haeheh
avar start using svk too, implicit karma
ruoso one char by lien
s/lien/line/
hoelzro so should I make a commit with just my change to AUTHORS then? or should I do actual work, then commit? 13:23
avar hoelzro: We have a very strict commit rule: there are none:) 13:24
but commit frequently after each logical change:)
hoelzro I'm a fan of committing often =D
avar *and after
moritz_ hoelzro: that's good
hoelzro which is something my co-workers don't really practice...
but they're using VSS, so can you blame them? 13:25
13:25 jasonmay joined
moritz_ wtf is VSS? 13:26
hoelzro visual sourcesafe?
be happy you never heard of it
moritz_ ok, I am ;)
hoelzro it's Microsoft's RCS
and it's TERRIBLE
avar it's more crappy than that, amazingly
moritz_ is it meant for Word files? ;-)
hoelzro no, plain text 13:27
but that doesn't stop us from using it for .docs... 13:31
13:32 TJCRI joined
allbery_b ah, visual sourcesafe, the source motel 13:32
hoelzro what a great analogy!
13:40 Blwood joined
ruoso fglock, found the problem in the !before token 13:41
it was returning $MATCH, instead of $MATCH.bool
hmm
Match.true should return Match.bool.true
hmm
interesting
it wasn't ;) 13:42
the bool object always evaluate as true...
hmm
there are some things weird in this object system
moritz_ kp6: say (3!=3) 13:43
exp_evalbot r17981: OUTPUT[ā¤]
moritz_ kp6: say (3==3)
exp_evalbot r17981: OUTPUT[1ā¤]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } 13:44
exp_evalbot r17981: OUTPUT[syntax error at position 0, line 1 column 0:ā¤( ( do { class MyClass { has $.bool; method true { return $.bool } }; my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } } ).perl ).saā¤^ HEREā¤]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" };
exp_evalbot r17981: OUTPUT[syntax error at position 0, line 1 column 0:ā¤( ( do { class MyClass { has $.bool; method true { return $.bool } }; my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" }; } ).perl ).saā¤^ HEREā¤]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } };
exp_evalbot r17981: OUTPUT[syntax error at position 0, line 1 column 0:ā¤( ( do { class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a = MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } }; } ).perl ).saā¤^ HEREā¤]
ruoso hrm
kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } }; 13:45
exp_evalbot r17981: OUTPUT[syntax error at position 0, line 1 column 0:ā¤( ( do { class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = 1; if ($a) { "ok "} else { "not ok" } }; } ).perl ).saā¤^ HEREā¤]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = 0; if ($a) { say "ok "} else { say "not ok" } };
exp_evalbot r17981: OUTPUT[not okā¤]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = 1; if ($a) { say "ok "} else { say "not ok" } }; 13:46
exp_evalbot r17981: OUTPUT[ok ā¤]
ruoso hmmm
fglock ruoso: i think .bool should just be renamed to .true 13:51
ruoso fglock, the problem is that the context is not propagated 13:54
kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); $a.bool.bool = 0; if ($a) { say "ok "} else { say "not ok" } }; 13:55
exp_evalbot r17981: OUTPUT[syntax error at position 73, line 1 column 73:ā¤class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); $a.bool.bool = 0; if ($a) { say "ok "} else { say "not ok" } }ā¤
.. ^ HEREā¤DISPATCH: calling APPLY on invali...]
ruoso kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); ($a.bool).bool = 0; if ($a) { say "ok "} else { say "not ok" } };
exp_evalbot r17981: OUTPUT[no method 'p5landish' in Class 'MyClass'ā¤]
pugs_svnbot r17982 | ruoso++ | [kp6] problem with !before solved. It was returning a MATCH object to the bool value... 14:00
diff: dev.pugscode.org/changeset/17982
lambdabot Title: Changeset 17982 - Pugs - Trac
14:06 xinming joined 14:09 hkb_lab is now known as hkb_aw
pugs_svnbot r17983 | fglock++ | [kp6-lisp] GLOBAL loads ok 14:09
diff: dev.pugscode.org/changeset/17983
lambdabot Title: Changeset 17983 - Pugs - Trac
pugs_svnbot r17984 | fglock++ | [kp6-lisp] export method names ??? 14:18
diff: dev.pugscode.org/changeset/17984
lambdabot Title: Changeset 17984 - Pugs - Trac
14:18 f0rth_ left 14:19 cmarcelo left 14:21 f0rth joined
fglock hmm - Lisp (or me) is confusing the namespaces: "The class #<STANDARD-CLASS KP6-STR> has no slot named KP6-GLOBAL::VALUE" 14:22
the accessor is getting a namespace... 14:23
14:24 amnesiac joined 14:25 ak__ left
avar rest assured that we're screwing things up:) 14:26
packages are apperently a reader-type thing not an eval thingy, see the pdf I gave you
Aankhen`` fglock: Where is the namespace in that? 14:27
fglock: KP6-GLOBAL::VALUE is this symbol read in the KP6-GLOBAL package: 'value 14:28
allbery_b @remember avar rest assured that we're screwing things up:)
lambdabot Done.
ruoso wonders if it wouldn't be a better idea to implement the namespace management above lisp, instead of using it...
avar true
well, maybe not. I don't know how all this stuff works. Maybe it's needed for classes.. 14:29
14:29 cmarcelo joined
fglock no, classes are anonymous things 14:29
pugs_svnbot r17985 | fglock++ | [kp6-lisp] a "hello" program for debugging
diff: dev.pugscode.org/changeset/17985
lambdabot Title: Changeset 17985 - Pugs - Trac
Aankhen`` You can implement a lot of Perl 6 by simply using Lisp features, but I do think it might be a better idea to implement namespaces separately. 14:30
The Lisp package model is quite different from the Perl 6 package model.
ruoso Aankhen``, and that's a path already made in mp6
moritz_ are lisp lists lazy?
ruoso which proved non-viable for a more complete implementation
fglock Aankhen``: using a hash?
avar moritz_: no
But yes, lisp packages are purely a reader-level thing for your convenience. You don't really need that when doing emitting 14:31
We probably want to use it in the compiler runtime itself though..
ruoso avar, the compiler runtime must be the normal runtime 14:32
remember about "eval"
14:33 kyrbe joined
pugs_svnbot r17986 | Aankhen++ | * moved (DEFPACKAGE :KP6-GLOBAL ...) into package.lisp. 14:33
diff: dev.pugscode.org/changeset/17986
lambdabot Title: Changeset 17986 - Pugs - Trac
fglock i committed svn.pugscode.org/pugs/v6/v6-KindaPe...hello.lisp for debugging the runtime; the compiler can be fixed when it works
14:34 kyrbe left
Aankhen`` fglock: Yeah, using hashes. 14:34
fglock: Might I prevail upon you to use lowercase symbols? :-) 14:35
avar and a sane indenting style?:) 14:36
Aankhen`` Heh.
All I want is lowercase symbols. 14:37
Proper indentation only comes through a decent editor.
moritz_
.oO(lisptidy?)
Aankhen`` LOL.
14:38 pmurias joined
avar lisptidy is just (read) and (print) in your implementation 14:38
pmurias ?eval '...'
fglock Aankhen``: sure
pmurias pugs: ... 14:39
lambdabot pmurias: You have 1 new message. '/msg lambdabot @messages' to read it.
exp_evalbot OUTPUT[*** Cannot use this control structure outside a 'routine' structureā¤ at /tmp/ZiKWD7CxGj line 1, column 1 - line 2, column 1ā¤]
Aankhen`` fglock: One question, probably due to my lack of understanding of kp6. Since you're mangling the names for Lisp, is it necessary to use a name like kp6-Code_say instead of kp6-code-say?
avar Aankhen``: no, either way works
does _ screw things up?
the names aren't being mangled .. yet
Aankhen`` avar: No, it's just aesthetic. :-)
Eh, anything that goes from foo to bar-foo_baz is mangling as far as I'm concerned. 14:40
I guess my terminology is incorrect.
avar they'll be mangled more:)
Aankhen`` Like how?
fglock not fully mangling yet, it doesn't do lowercase/uppercase yet
Aankhen`` I see. 14:41
avar sh-3.1$ perl -wle 'print "(defun kp6-", join q[-], map { ord } split //, shift' 'infix:<~>'
(defun kp6-105-110-102-105-120-58-60-126-62
something like that
Aankhen`` Oh.
ruoso fglock, wouldn't it be a good idea to implement namespaces and pads in upper level, since there's anoter backend being developed? 14:42
moritz_ I hope nobody tries to read that code ;)
fglock ruoso: makes sense
avar moritz_: there'll be a comment giving the real name:)
Aankhen`` moritz_: LOL. 14:43
avar but yeah, don't expect emitted code to be readable
ruoso: definetely
Aankhen`` Is this correct? `grammar` ISA `class` ISA `module` ISA `package`
ruoso Aankhen``, I don't think so 14:44
classes can be outside packages
like anon-classes
Aankhen`` But aren't they anonymous packages then?
ruoso no... the would be inside someone's else package
avar packages are a different thing from classes alltogether
Aankhen`` Ah.
fglock ruoso: this will be a big change, expect some problems...
ruoso :) 14:45
fglock ruoso: maybe i should work on it in a branch
ruoso nah...
better break everythih
so more people try to fix it
;)
avar ruoso++ 14:46
pmurias avar: there is no point in mangling \w 14:50
ruoso I think there's a exponential graphic of performance from faster to slower: mp6 -> kp6-mp6 -> kp6-kp6... where kp6-kp6 is in the very low end of the curve 14:52
it takes years to parse say "Hello!"
user 0m5.588s 14:53
time echo -n 'say "Hello!"' | perl -Ilib-kp6-kp6-p5 kp6-kp6.pl
gah
moritz_ so kp6-kp6 _can_ parse 'say "Hello"'? 14:54
ruoso moritz_, almost
it's failing with no method 'p5landish' in Class 'Match'
but already at the end of the match
pmurias that's not good
ruoso which means some token is setting the bool of its parent match with the match itself
instead of a literal value 14:55
(the ones which implement p5landish)
that's because kp6 doesn't propagate the context
kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); ($a.bool).bool = 0; if ($a) { say "ok "} else { say "not ok" } };
exp_evalbot r17986: OUTPUT[no method 'p5landish' in Class 'MyClass'ā¤]
ruoso pmurias, this is the code to reproduce it
pmurias kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); ($a.bool).bool = 0; if ($a) { say "ok "} else { say "not ok" } };kp6: class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); ($a.bool).bool = 0; if ($a) { say "ok "} else { say "not ok" } }; 14:56
exp_evalbot r17986: OUTPUT[syntax error at position 192, line 1 column 192:ā¤class MyClass { has $.bool; method true { return $.bool } }; module Main { my $a := MyClass.new(); $a.bool = MyClass.new(); ($a.bool).bool = 0; if ($a) { say "ok "} else { say "not ok" } };kp6: class MyClass { has $.bool; method true {
..return $.bool } }; module Main { my $a := MyClass.new(); $a.bo...]
pmurias argh
ruoso: i see 14:57
pugs_svnbot r17987 | Aankhen++ | * added KP6-VALUE accessor to all classes that had a VALUE slot.
diff: dev.pugscode.org/changeset/17987
lambdabot Title: Changeset 17987 - Pugs - Trac
Aankhen`` Should all those classes be inheriting from some sort of base type? 14:58
ruoso they do
but p5landish only makes sense to classes that *can* be mapped to p5 values
like, int, bool, str
like: int, bool, str
Aankhen`` Er, I'm talking about the CL backend. 14:59
ruoso ah
ok
Aankhen`` D'oh, bad commit message.
Is "backend" even the right word?
pmurias fglock used [kp6-lisp] in his commit messages
backend is the right word if you mean the lisp backend 15:00
Aankhen`` Yeh.
I've been using kp6-cl, guess I'll switch to kp6-lisp.
fglock Aankhen``: re base class, i think CLOS takes care of that 15:01
Aankhen`` fglock: Whatcha mean?
ruoso are you planning to use CLOS as-is and not implement a MOP on top of it?
fglock the base class is a lisp class, right?
ruoso: yes, at least for the prototype impl 15:02
Aankhen`` fglock: Not following you. Which base class?
We have a bunch of classes, kp6-int, kp6-str, etc.
They don't share any superclass.
fglock Aankhen``: kp6-object
ruoso every kp6-lisp object would be a lisp object
Aankhen`` Okay, so that ought to be added to the list of ancestors.
15:02 ixolit joined
pugs_svnbot r17988 | ruoso++ | [kp6] 'no method' confess instead of dies... \n[kp6] kp6-token-debug shows a letter hint for the level, and detects closure exits... 15:02
diff: dev.pugscode.org/changeset/17988
lambdabot Title: Changeset 17988 - Pugs - Trac
Aankhen`` Which is what I was asking about. :-)
ruoso but I still think that's not a nice path to follow... 15:03
15:03 ixolit left
ruoso you'll probably not be able to implement all p6 semantics directly on CLOS 15:03
Aankhen`` ruoso: Such as?
fglock ruoso: it's still an experiment
Aankhen`` ruoso: Hmm, actually, you're right. 15:04
ruoso Aankhen``, I don't know CLOS enough... but it's just a feeling..
pmurias ruoso: are you still looking for the token which sets .bool to Match?
ruoso yep... and I don't have any clue
pmurias i'll try 15:05
ruoso I changed the die to confess when the method is not found
Aankhen`` fglock: There's a comment in Str.lisp saying "TODO: subclass Value". Is Value == kp6-object?
ruoso pmurias, and there's a call to "join" in the stack trace
avar u.nix.is/cl-class.png 15:06
avar experiments
fglock Aankhen``: it's Object -> Value -> Str/Int/Num/Code, and Object -> Container -> Scalar/Hash/Array
Aankhen`` fglock: Ah. 15:07
pmurias ruoso: adding debuging statements in the code generated in Visitor/Token.pm might be a good idea
ruoso there are some already
try running with the KP6_TOKEN_DEBUGGER env var set to 1
pmurias seen it 15:08
buubot pmurias: Sorry, I haven't seen it
pmurias ruoso: i mean adding check if bool was just set to Match
and printing the rule ast/code if it was
ruoso pmurias, following the stack trace 15:09
it's not in the bool
it's in the join impl
avar Aankhen``: Check out my leet subclassing
ruoso Array.pm 104
avar fglock: What's the best reference for the methods those should contain? I'm planning on writing a bunch of lisp objects that do what they do
Which'll be fast and will make the emitter a lot easier later on when FETCH and such can just be method calls 15:10
Aankhen`` avar: Wait, you just added subclasses?
I did too. :-P 15:11
fglock avar: not fully specified, but the Perl 5 version has what kp6 needs (Runtime/Perl5/*)
avar I'm working on different code, it'll be pure-cl for a while
I'm mainly teaching myself lisp by implementing cl-perl6 :)
Aankhen`` percl6!
fglock the methods in Runtime/Perl6/* should be reusable, so you shouldn't need lisp versions 15:12
Aankhen`` avar: Did you fix up the class hierarchy? If so, you should commit it so I don't conflictā€¦ :-)
avar I'm not working on /Lisp/, feel free to commit
Aankhen`` Oh.
Sorry, I misunderstood.
avar fglock: It'll be in /Lisp/ eventually:)
pugs_svnbot r17989 | Aankhen++ | [kp6-lisp] 15:14
r17989 | Aankhen++ | * added kp6-object, kp6-value, kp6-num and kp6-container classes.
r17989 | Aankhen++ | * set up inheritance hierarchy between classes.
diff: dev.pugscode.org/changeset/17989
lambdabot Title: Changeset 17989 - Pugs - Trac 15:15
15:15 penk left
avar Everything.isa(Object) in p6 right? 15:16
ruoso pmurias, found
moritz_ avar: I don't think that Junction.isa(Object) 15:17
but I could be wrong
ruoso moritz_, everything.isa(Object)
pugs_svnbot r17990 | ruoso++ | [kp6] return match.result.str in str, instead of just match.result 15:18
r17991 | fglock++ | [kp6] KindaPerl6::Visitor::Namespace stub
diff: dev.pugscode.org/changeset/17991
lambdabot Title: Changeset 17991 - Pugs - Trac
fglock Junction is a bit special, but not much
just like containers are a bit different from normal objects 15:19
and Code, and lazy lists, and... 15:20
TimToady ?eval Junction ~~ Object 15:21
pugs_svnbot r17992 | ruoso++ | [kp6] do not call pad declaration in the compiler yet
diff: dev.pugscode.org/changeset/17992
lambdabot Title: Changeset 17992 - Pugs - Trac
ruoso WOooooooooooHoooooooooo: Now it's official... ;) kp6-kp6 can parse: say "Hello!" 15:22
pmurias wow ;)
fglock ruoso++
pmurias ruoso++
ruoso the return is still not the expected
but it finishes
TimToady say ruoso++ for 1..*
moritz_ ruoso += * 15:23
ruoso echo -n 'say "Hello!"' | ./kp6-token-debug
buubot ruoso: -n 'say "Hello!"'
TimToady pugs: Junction ~~ Object
exp_evalbot OUTPUT[Bool::Trueā¤]
TimToady pugs: Junction ~~ Any
exp_evalbot OUTPUT[Bool::Falseā¤]
moritz_ ok
TimToady and function parameters default to Any, not Object 15:24
15:24 Shillo joined
Shillo Hullo, all 15:25
pmurias Shillo: hi
moritz_ hi Shillo ;)
avar So everything is anything in perl6?:) 15:28
TimToady junctions aren't anything; they're just objects
everything that has been observed is anything, I suppose :) 15:29
at least in the Copenhagen interpretation 15:30
moritz_ is there an up to date inheritance graph somewhere?
15:31 hkb_aw is now known as hakobe
TimToady hmm, we drew a lot of them several years ago, but I'm not sure any of them is specced officially 15:34
if pugs had MO we could make it print out its version
15:34 weinig left
moritz_ ok, then the laziest path is to wait for audreyt's MO-branch ;) 15:35
pugs_svnbot r17993 | fglock++ | [kp6] some progress in namespace-in-perl6 15:36
diff: dev.pugscode.org/changeset/17993
lambdabot Title: Changeset 17993 - Pugs - Trac
15:42 mj41 joined
fglock ruoso: echo 'say "hello"; $x::y::xxx' | perl kp6-mp6-perl5.pl --do Global,Namespace --ast | perltidy 15:44
pugs_svnbot r17994 | ruoso++ | [kp6] kp6-kp6 already in the visitors...
r17995 | fglock++ | [kp6] namespace in Perl 6; missing some initialization code
diff: dev.pugscode.org/changeset/17995
lambdabot Title: Changeset 17995 - Pugs - Trac
TimToady for 'a'..* X 'a'..'zzzzzzzzzzzzzzzzzzz' -> $sub, $super { say "$sub isa $super" if ucfirst $sub ~~ ucfirst $super } 15:48
that would probably get most of them eventually
assuming builtin class are not too long and don't do camel case 15:49
[particle] or underscore
TimToady well, so far we've managed to avoid both 15:50
[particle] and somehow you've managed to avoid unicode, too :P 15:51
pugs_svnbot r17996 | fglock++ | [kp6-lisp] namespace implemented in Perl 6 land
diff: dev.pugscode.org/changeset/17996
lambdabot Title: Changeset 17996 - Pugs - Trac
fglock done :) 15:52
moritz_ fglock: wow, that was fast ;)
TimToady basically, we reserve all the ugly class names for users :P 15:54
15:58 DarkWolf84 joined
pugs_svnbot r17997 | fglock++ | [kp6-lisp] lispish .INDEX, .LOOKUP 15:58
diff: dev.pugscode.org/changeset/17997
lambdabot Title: Changeset 17997 - Pugs - Trac
15:58 Shillo left, barney joined 16:01 justatheory joined 16:03 Blwood left
pugs_svnbot r17998 | fglock++ | [kp6-lisp] more lispy calls 16:03
diff: dev.pugscode.org/changeset/17998
lambdabot Title: Changeset 17998 - Pugs - Trac
pugs_svnbot r17999 | fglock++ | [kp6-lisp] removed some semicolons 16:06
diff: dev.pugscode.org/changeset/17999
lambdabot Title: Changeset 17999 - Pugs - Trac
pugs_svnbot r18000 | ruoso++ | [kp6] new test... if without else needs to generate a p6 value, instead of leaking the implicit p5 return 16:12
diff: dev.pugscode.org/changeset/18000
lambdabot Title: Changeset 18000 - Pugs - Trac
hoelzro does $SIG{__WARN__} still work in perl 6? 16:14
I guess it'd be %SIG{__WARN__} 16:15
moritz_ or %*SIG{__WARN__}
just grep the synopsis for SIG 16:16
hoelzro I don't understand the whole deal with @* vs @, %* vs %, etc
where can I read up on twigils??
moritz_ S02 I think
hoelzro k, thanks
moritz_ $* are just global vars
and $? are vars that are known at compile time 16:17
hoelzro I see
what's taking the place of @_, again?
I thought it was @*..., but it can't be global
moritz_ and $^ are things that are auto declared (like sort { $^a <=> $^b } )
no, it's still @_ iirc 16:18
hoelzro really?
moritz_ but you can declare sub signatures
hoelzro right
moritz_ so you rarely ever need it
hoelzro sub foo ($a, $b, @*args) { ... }
moritz_ pugs: sub fac(Int $a) { [*] 1...$a }; say fac(5) 16:19
exp_evalbot OUTPUT[*** ā¤ Unexpected "$a"ā¤ expecting operator, ":", ",", term postfix or "}"ā¤ at /tmp/EI4lKjaZo1 line 1, column 27ā¤]
moritz_ pugs: sub fac(Int $a) { [*] 1..$a }; say fac(5)
exp_evalbot OUTPUT[120ā¤]
16:19 weinig joined 16:28 rindolf joined
moritz_ I'm off for vacation (and buying a new laptop :/), see you all in about two weeks 16:34
hoelzro take it easy
16:34 IllvilJa left
moritz_ if someting goes wrong with exp_evalbot, contact bloonix++ 16:34
16:39 franck___ left, TJCRI left
TimToady hoelzro> sub foo ($a, $b, @*args) { ... } 16:47
you want *@args there, which is different from @*args 16:48
hoelzro yeah, I took a look at that
TimToady the former is a slurpy argument, while the latter is global
hoelzro good thing I'm not writing Perl 6 for Pugs =P
TimToady anyway, welcome! 16:49
hoelzro thanks!
I'm trying to get my first commit done before I go to class
16:50 drrho left, REPLeffect joined 16:51 ferreira joined
ferreira The reactions to micro-articles have begun. See the comments at www.oreillynet.com/onlamp/blog/2007...ching.html 16:51
lambdabot Title: YAP6 Operator: Stitching - O&#039;Reilly ONLamp Blog, tinyurl.com/2dcsta
16:57 f0rth left, f0rth joined
avar my $x = ~foo() is string context on foo() right? 17:02
pugs: sub foo { 5 } ~foo()
exp_evalbot OUTPUT["5"ā¤] 17:03
hoelzro does anyone online have experience with hacking the parser? 17:04
ferreira pugs: class Foo { sub prefix:<~> { 'a Foo' } }; ~(Foo.new) 17:05
exp_evalbot OUTPUT["<obj:Foo>"ā¤]
ferreira What is wrong? The output should not be 'a Foo' ?
avar pugs: class Foo { method prefix:<~> { 'a Foo' } }; ~(Foo.new) 17:06
exp_evalbot OUTPUT["<obj:Foo>"ā¤]
avar maybe it's not supported
17:07 Loro_ joined
TimToady hoelzro: I presume you mean the pugs parser. I've tweaked it upon occasion, but whenever I try to do anything substantial in Haskell I get into an argument with the compiler and lose. 17:08
hoelzro ok
I was wondering if anyone knew how to do optional arguments 17:09
TimToady if you see gaal online, he's a much better Haskeller
hoelzro for the yada* operators
17:09 polettix joined
hoelzro I know Haskell, but my experience with Parsec is...well, I started this morning 17:09
TimToady it would be a list operator
so it'd work like print or warn
hoelzro yes 17:10
actually, EXACTLY like warn
as in, all I'm doing is calling warn =P
TimToady so I'd look at how warn is parsed
hoelzro I tried, but I couldn't find it anywhere 17:11
TimToady hmm
avar CL-USER> (elems *my-hash*) 17:12
0
;
CL-USER> (STORE *my-hash* "key" "value")
"value"
CL-USER> (elems *my-hash*)
1
C
whee:)
hoelzro just curious, what is Lisp used for in Pugs?
TimToady Parser/Literal.hs: doYada "???" = "&warn"
hoelzro TimToady: I'm looking at that right now
but that's the yada operator part 17:13
avar hoelzro: We're working on another compiler that uses it as an emitter backend
hoelzro sweet
avar It's currently not as complete as the .hs stuff but easier to get into, and IMO has more of a future:) 17:14
TimToady then there's the parrot perl6 compiler, which targets parrot, And my STD parser, which targets nothing (yet)
well, it's really targetting pugs at the moment
avar kp6 differs from the others in that it compiles itself which makes it a "real" compiler IMNSHO :) 17:15
TimToady anyway, it's a healthy competition, I think 17:17
hoelzro competition can be good for projects
TimToady as long as we keep talking to each other
hoelzro yeah
17:17 rhizo joined
avar totally:) 17:18
TimToady hoelzro: btw, I don't think Perl 6 will necessarily have a %SIG, or if it does, it likely won't contain __WARN__ 17:20
if warn needs a global hook, it'd be more like $*WARN_HOOK or some such 17:21
hoelzro oh that, yeah, I figured out a way around it
avar amazingly remembers what the diference between (let and (let*
..is
17:21 DarkWolf84 left
TimToady avar: just finish all your sentences with ] 17:22
[particle] TimToady: mostly it's coopetition between projects, except when we all battle with the language designer :) 17:24
TimToady ah, well, he could use a good whack upside the head occasionally. 17:27
fglock moritz_: ping 17:32
davidfetter oi, fglock :) 17:35
fglock davidfetter: hey!
davidfetter what's new & good? :)
TimToady about an hour ago moritz said he's off for vacation for two weeks
fglock davidfetter: all is new and good :)
davidfetter my spies tell me you're in .pt 17:36
17:36 REPLeffect left
fglock looks around 17:36
17:37 REPLeffect joined
fglock davidfetter: it's a nice weather here - they had a cool winter back in .br 17:38
davidfetter so i heard 17:39
what brought you to .pt ?
[particle] $$
davidfetter hā‚¬h
17:40 ruoso left, ruoso joined
fglock a bunch of brazilian *-pm group leaders were hired by an open source company in .pt - it's a nice environment 17:41
davidfetter neat. can you talk about what you're doing there? 17:42
fglock and ā‚¬ > R$ too
davidfetter heh
fglock we are working on customized versions of several open source products, and contributing back to the projects 17:43
Perl 6 is one of them :)
davidfetter w00t!
fglock i've used v6.pm on a migration project
davidfetter sweet 17:44
avar sial.org/pbot/27597 # luev slime:)
pugs_svnbot r18001 | avar++ | r63624@t: avar | 2007-09-20 17:41:59 +0000
r18001 | avar++ | add a value member to the container for storage
r18002 | avar++ | r63625@t: avar | 2007-09-20 17:42:55 +0000
r18002 | avar++ | Implemented new/store/lookup/pairs. May still need to return kp6 objects and could use some cleanup of the code and API
diff: dev.pugscode.org/changeset/18002
lambdabot Title: Paste #27597 from Someone at 208.78.101.240
Title: Changeset 18002 - Pugs - Trac
avar evaluation aborded there is when I switched to the other buffer to write the function, before I loaded it into the lisp image:) 17:45
fglock davidfetter: i'm working with several people i've only met in the net
davidfetter that's common :)
avar fglock: I'm going to be making more libary runite hacks, but first $exercise :) 17:46
fglock i mean, working in the same place
17:46 IllvilJa joined
fglock avar++ 17:46
avar is LOOKUP fglockish for FETCH ? 17:47
:)
fglock lookup is the read-key-from-hash operation
ferreira I am looking for instructions on keeping in sync with Pugs and compiling it, working from feather. Anyone knows where I can find instructions for that? 17:48
fglock fetch is read-from-container
a fetch on a hash returns the whole hash
ferreira: i think it's this: feather.perl6.nl/svk.html 17:49
lambdabot Title: SVK on Feather
ferreira Fglock: thanks 17:50
fglock then follow the instructions in pugs
np
17:51 kane_ left
fglock avar: does the namespace elimination transform makes sense? 17:51
ruoso changing EmitPerl5 to return Bit(false) when an if doesn't have an else 17:52
/return/return in the else/
fglock ferreira: how about adding links to previous/next article 17:55
17:56 Patterner left
ferreira at ONLamp? I am having trouble with MovableType that seems to randomly change URLs in some updates. I am thinking of alternatives. 17:57
17:57 daxim joined 18:01 pl4fks joined
pugs_svnbot r18003 | ruoso++ | [kp6] if without else emits an else { Bit(False) } for implicit returns 18:01
diff: dev.pugscode.org/changeset/18003
lambdabot Title: Changeset 18003 - Pugs - Trac
ferreira I don't understand these instructions on SVK. I started a svk repository for saving the micro-article series. So I cannot just copy ~audreyt/config to ~/.svk. I would need to understand what mirrors this is setting.
18:01 Psyche^ joined, Psyche^ is now known as Patterner
fglock ferreira: you can add the articles under docs/articles/ 18:02
then svk add / svk ci 18:03
ferreira ok
18:05 REPLeffect left
pugs_svnbot r18004 | ruoso++ | [kp6] prefix:<%>. Still does nothing... kp6 does not implement contexts 18:05
diff: dev.pugscode.org/changeset/18004
lambdabot Title: Changeset 18004 - Pugs - Trac
18:05 pl4fks left
fglock avar: are we going to use kp6-* for method names? 18:05
18:07 REPLeffect joined
fglock are there some guidelines on lisp indentation best practices? 18:07
18:08 qmole left
pugs_svnbot r18005 | fglock++ | [kp6-lisp] Hash.pairs, Hash.elems return Perl 6 objects 18:11
diff: dev.pugscode.org/changeset/18005
lambdabot Title: Changeset 18005 - Pugs - Trac
18:11 SubStack left, SubStack joined
fglock ruoso: prefix:<%> just calls .hash 18:11
in kp6
ruoso fglock, I've just changed it 18:12
18:15 weinig left
pugs_svnbot r18006 | fglock++ | [kp6-lisp] added Undef, Value, Container to Runtime.lisp 18:20
diff: dev.pugscode.org/changeset/18006
lambdabot Title: Changeset 18006 - Pugs - Trac
18:20 buetow left
ferreira pugs build fails with error in src/Pugs/Internals/Monads.hs What am I doing wrong? pugs build is broken? 18:23
18:24 qmole joined 18:25 Blwood joined
fglock ferreira: sorry, i don't know 18:29
pugs_svnbot r18007 | fglock++ | [kp6-lisp] &GLOBAL::say is stored in the right place
diff: dev.pugscode.org/changeset/18007
lambdabot Title: Changeset 18007 - Pugs - Trac
fglock home & 18:30
18:34 TJCRI joined, barney left 18:40 calin joined 18:41 gnuvince joined 18:42 calin_ joined
gnuvince Is audreyt still active on the Pugs project? I notice that she hasn't committed anything to pugs since July. Some of her CPAN modules are still being maintained however. 18:43
ferreira I think she's busy at work
gnuvince Where does she work? 18:46
18:51 Schwern left
allbery_b she was out for a long time with hepatitis, then there were complications to her recovery 18:52
gnuvince I didn't know about that. I knew she got hep A, but not that her recovery didn't go well.
pugs_svnbot r18008 | ruoso++ | [kp6] some value checking... kp6-kp6 does emit some code... but not the right code
diff: dev.pugscode.org/changeset/18008
lambdabot Title: Changeset 18008 - Pugs - Trac
ruoso pmurias, run: echo 'say "Hello!"' | perl -Ilib-kp6-kp6-p5 kp6-kp6.pl 18:53
kp6-kp6.pl already pass through all the visitors
:)
it emits the p5 skelethon code
the ast is not as it should yet
we need a p6 dumper 18:54
does anyone wants to implement KP6Data::Dumper?
;)
TimToady KP6Data::YAML maybe...
ruoso I was thinking in the quasiquotes too 18:56
maybe KP6AST::Dumper
18:56 fglock left, gnuvince left
ruoso kp6-kp6 almost works ;) 18:59
pugs_svnbot r18009 | ruoso++ | [kp6] dumps the ast as a warning in kp6-kp6 before emitting... The ast now is highly incomplete
diff: dev.pugscode.org/changeset/18009
lambdabot Title: Changeset 18009 - Pugs - Trac 19:00
19:00 calin left
ruoso A long way to go now... Must understand why the AST is almost empty... 19:05
but...
home &
19:05 ahmadz joined
ahmadz hi all 19:06
pmurias ahmadz: hi al
l
ahmadz: hi ;)
ahmadz pmurias: ;-)
pmurias ruoso: i works here :) but takes 0m31s 19:07
ruoso heh
pmurias, I'm going out now... 19:08
do you want to search why is the ast getting out empty?
now really home & 19:09
ferreira Why I got "A checksum mismatch occurred: Base checksum mismatch on '/mirror/pugs/src/perl6/Perl-6.0.0-STD.pm" when doing "svk pull". I am at feather. 19:12
TimToady that file isn't there anymore. it was renamed to STD.pm 19:15
seems like kind of a lousy error message... 19:16
clkao: see above 19:17
19:17 jisom joined
clkao hello 19:18
aaah what version of svk is on feather
TimToady ENOKLOO 19:19
ferreira This is svk, version v2.0.2 (using Subversion bindings 1.4.4)
clkao hmmm
this is during pull... /me ponders
19:21 weinig joined 19:28 ludan joined, TJCRI left 19:45 dduncan joined
avar @tell fglock re namespace elimination transform, haven't looked at it yet, will peek mayybeees 19:51
lambdabot Consider it noted.
avar @tell fglock I'm still n00bing around with CLOS so I don't know the best names for methods, but I suppose a prefix makes sense, maybe 19:52
lambdabot Consider it noted.
avar @tell fglock You should indent lisp like emacs does it:)
lambdabot Consider it noted.
hoelzro who was the resident Haskell expert again?
19:52 amnesiac left
hoelzro began with a 'g', if I remember right 19:53
ferreira gaal ?
hoelzro that looks right
alright, yada-yada-yadas are working! 19:57
TimToady hoelzro++ 19:59
but weren't you supposed to go to class?
hoelzro I just got back
I had inspiration while listening to my professor 20:00
TimToady sorry, kneejerk reaction from spawning too many teens...
hoelzro that's fine =P
are the yadas supposed to accept lists, like this: ... ("list", "of", "things")? 20:01
TimToady yes, and the args are implicitly joined with '' 20:02
hoelzro really? I didn't know that 20:03
I supposed I'll actually have to LEARN Perl 6 after I get more coding done
TimToady that part's actually the same as Perl 5... :)
hoelzro really?! 20:04
TimToady try "warn 'foo','bar'"
hoelzro I clearly haven't been writing enough Perl...
20:05 weinig left
TimToady well, it's not all that useful of a feature, but warn and die were modeled on print 20:05
hoelzro by the way, what's the purpose of fail if we have die? 20:06
TimToady die always throws an exception
fail only does that if the caller uses "use fatal" 20:07
hoelzro ah
TimToady otherwise it returns an "unthrown exception"
pmurias The argument
is optional,
TimToady which eventutally throws an exception if you try to use it
pmurias TimToady: it might suggest that !!! takes one argument
TimToady also known as "interesting values of undef"
hoelzro I think you can do die w/o an arg
20:08 TJCRI joined
pmurias yes 20:08
TimToady but instead of "Something's wrong" or "Died" the default message for the yadas can be more specific
20:08 pmurias left
TimToady "Attempt to execute stub" or some such 20:09
hoelzro oh? I just have it calling die/fail/warn
would you like me to change it?
TimToady sure
hoelzro would you like me to change it to anything in particular, or should I be creative? =D 20:11
TimToady maybe "Unimplemented stub called"
20:11 rindolf left
hoelzro ok 20:12
what version of pugs does the evaluator in this channel use? a nightly build? 20:13
TimToady probably more frequent than that
[particle] it rebuilds when it senses a disturbance in the repo
hoelzro dang 20:14
impressive
Jedi build system?
[particle] use The::Repo;
TimToady pugs: say $*PUGS_VERSION
exp_evalbot OUTPUT[Perl6 User's Golfing System, version 6.2.13, October 17, 2006 (r17871)ā¤]
avar Updated annually?:) 20:15
wolverian heh
[particle] hasn't been released in a while... sohuld probably have a -devel on the version or something
TimToady -extended_version or _directors_cut maybe 20:16
[particle] and the last couple hundred revs have been for kp6, not pugs
deluxe_widescreen_edition
ferreira the revision number is from svn.pugscode.org? That (the revision number) differs if you use other svk mirrors, right? 20:17
20:35 ferreira left
diakopter more like -unrated_edition 20:43
[particle] diakopter: svnbot throttling doesn't seem to be working. have a look at some code i wrote up: nopaste.snit.ch:8001/11202 20:46
...it may need some fixing up, i have no good way to test 20:47
diakopter [particle]: so it always sleeps 0.51 ? 20:49
hoelzro is &die implemented as a multi sub?
I'm getting complaints from !!! about "no multi variant found for &die" 20:50
[particle] diakopter: yep 20:51
or 1s if the platform doesn't support sub-second resolution
davidfetter wonders which platforms don't offer sub-second resolution 20:52
diakopter [particle]: someone on #parrot demanded a maximum of 0.2msgs/s
[particle] it was 2msg/s 20:53
diakopter someone said 5 seconds between messages.
[particle] that's wrong.
5 minutes between runs
2 messages/second
hoelzro what are the various forms of die? 20:55
diakopter [particle]: oh yeah. I see where you said diakopter: kismet says 2 messages/second would be a good value
[particle] yep, that's it 20:56
hoelzro: see perlcabal.org/syn/S03.html#List_pre...precedence Exception generators 20:57
lambdabot Title: S03
hoelzro thank you 20:58
I'm seldom sure which synopses to look at...
[particle] that's why you should have a local repo to search :) 20:59
svn.perl.org/perl6/doc/trunk
lambdabot Title: Revision 14463: /doc/trunk
20:59 snearch joined
diakopter [particle]: there. let's see if that works. 21:01
pugs_svnbot r18010 | hoelzro++ | Added myself to the AUTHORS list, fixed a discrepancy in docs/SEEALSO, and 21:03
r18010 | hoelzro++ | implemented the yada operators (..., !!!, ???). Tests are soon to follow...
diff: dev.pugscode.org/changeset/18010
lambdabot Title: Changeset 18010 - Pugs - Trac
[particle] diakopter++ # i'll test after lunch & errands 21:05
avar is a true/false value of the Bit class or of Bool directly? 21:06
avar wondering what the hell fglock has been doing:)
21:06 IllvilJa left
Juerd avar: Should be Bool, because bits are 0/1, and bools are false/true. 21:10
That's the only distinction AFAIK. Of course, the obvious 1-on-1 mapping applies. 21:11
Juerd does not see any use for having different classes
avar different stringify and such? 21:12
21:18 xinming left 21:19 TJCRI left 21:22 kjwcode left 21:23 snearch left 21:24 ludan left
pugs_svnbot r18011 | avar++ | r63636@t: avar | 2007-09-20 21:14:25 +0000 21:24
r18011 | avar++ | XXX note
r18012 | avar++ | r63637@t: avar | 2007-09-20 21:15:55 +0000
r18012 | avar++ | Add the kp6-value introspective accessor to all containers, this gets the CL hash table, string, array etc
diff: dev.pugscode.org/changeset/18012
lambdabot Title: Changeset 18012 - Pugs - Trac
hoelzro is there POD for Test.pm? 21:27
pugs_svnbot r18013 | avar++ | r63638@t: avar | 2007-09-20 21:19:25 +0000
r18013 | avar++ | Undef isa Value and is nil by default
r18014 | avar++ | r63639@t: avar | 2007-09-20 21:20:08 +0000
r18014 | avar++ | eek, forgot C-x C-s
r18015 | avar++ | r63640@t: avar | 2007-09-20 21:20:33 +0000
r18015 | avar++ | make the (str) method use the (kp6-value) introspector
r18016 | avar++ | r63641@t: avar | 2007-09-20 21:21:19 +0000
r18016 | avar++ | camel case class names, not that it matters to lisp
r18017 | avar++ | r63642@t: avar | 2007-09-20 21:21:36 +0000
r18017 | avar++ | delicious indenting
r18018 | avar++ | test dir
r18019 | avar++ | test dir nuked
r18020 | avar++ | r63643@t: avar | 2007-09-20 21:22:21 +0000
r18020 | avar++ | camel case class names, not that it matters to lisp
diff: dev.pugscode.org/changeset/18020
lambdabot Title: Changeset 18020 - Pugs - Trac
avar urgh! 21:29
pugs_svnbot r18021 | avar++ | r63644@t: avar | 2007-09-20 21:22:32 +0000
r18021 | avar++ | camel case class names, not that it matters to lisp
r18022 | avar++ | r63645@t: avar | 2007-09-20 21:22:52 +0000
r18022 | avar++ | M-x indent-region
r18023 | avar++ | r63646@t: avar | 2007-09-20 21:24:21 +0000
r18023 | avar++ | Drop kp6-Hash-new, just use (make-instance) which now initializies the hash thanks to :initarg, Implemented DELETE and CLEAR and make pairs return key-value pairs, not value-key pairs (oops), more cleanup
r18024 | avar++ | test dir un-nuked
r18025 | avar++ | test dir nuked
diff: dev.pugscode.org/changeset/18025
lambdabot Title: Changeset 18025 - Pugs - Trac
hoelzro nevermind...I found it =P
avar how do I make svn/svk remember my bloddy password so it doesn't ask for it every commit?
svk doesn't have --password for push like svn to make it less annoying 21:30
clkao apparently there's something silly in the codepath. try to do something with svn that requires password once and have it remembered
avar yeah, I've been commiting test dirs with svn, keeps asking
iirc it cached it when i tried it some months ago 21:31
I did svn co svn.pugscode.org/pugs; cd pugs/misc/pX/avar; svn mkdir test; svn ci test
lambdabot Title: Revision 18025: /
21:33 iblechbot left
avar pugs.switch_to(Git) 21:33
version control is always greener on the other side
hoelzro does old-school eval { ... } if($@) { ... } still work? I'm trying to write tests for .../!!!/???
21:35 xinming joined
hoelzro nevermind, I think I found it 21:37
are there hooks for die, warn, and fail? 21:42
avar currently you're the dude that understands this haskell nonesense best here:) 21:43
hoelzro yay!
I was talking about Perl 6 though; I'm writing a test for the yadas
avar GLOBAL::die() and so on I think 21:44
..maybe...
21:56 Schwern joined 21:58 orevdiabl joined, revdiablo left
hoelzro pugs: say $*PUGS_VERSION 22:07
exp_evalbot OUTPUT[Perl6 User's Golfing System, version 6.2.13, October 17, 2006 (r17871)ā¤]
avar You've implemented this stuff? 22:09
pugs_svnbot r18026 | avar++ | r63661@t: avar | 2007-09-20 21:43:51 +0000
r18026 | avar++ | Be more quiet when compiling
r18027 | avar++ | r63662@t: avar | 2007-09-20 22:08:33 +0000
r18027 | avar++ | GLOBAL needs to be in the kp6-cl package too
r18028 | avar++ | r63663@t: avar | 2007-09-20 22:09:18 +0000
r18028 | avar++ | Use the kp6-Hash interface instead of manually (make-hash-table); syntax fixes
diff: dev.pugscode.org/changeset/18028
lambdabot Title: Changeset 18028 - Pugs - Trac
22:09 REPLeffect_ joined 22:11 vel joined
hoelzro which Perl 6 implementation is furthest along? 22:11
avar pugs.hs 22:14
hoelzro I figured 22:15
it doesn't seem like a lot of people are working on it
avar no, like I said earlier 22:17
:)
kp6 is the most active atm 22:18
hoelzro audrey's off doing other things?
avar yup
hoelzro =(
if I had a better idea of what needed to be done, I'd keep going
and my Lisp isn't nearly as good as my Haskell 22:19
avar maybe you'd like to hack kp6?:)
hoelzro I could learn
avar Oh the lisp backend is just a current project, the most complete one and the one that's actually working is the perl5 emitter
22:20 vel left
avar You could also write a haskell emitter, a lot of the current stuff from pugs.hs could be reused 22:20
or do @anything :)
hoelzro I think for now, I'm going to read over the specs better
pugs_svnbot r18029 | avar++ | r63669@t: avar | 2007-09-20 22:17:18 +0000 22:21
r18029 | avar++ | kp6-CLEAR does not need a key argument
r18030 | avar++ | r63670@t: avar | 2007-09-20 22:18:35 +0000
r18030 | avar++ | Load Int too
r18031 | avar++ | r63671@t: avar | 2007-09-20 22:20:52 +0000
r18031 | avar++ | Export the Hash.lisp methods, this is stupid
diff: dev.pugscode.org/changeset/18031
lambdabot Title: Changeset 18031 - Pugs - Trac
22:23 Aankhen`` left
pugs_svnbot r18032 | avar++ | r63675@t: avar | 2007-09-20 22:27:14 +0000 22:27
22:27 REPLeffect left
pugs_svnbot r18032 | avar++ | (foo) not ( foo ); (kp6-APPLY not (APPLY 22:27
diff: dev.pugscode.org/changeset/18032
lambdabot Title: Changeset 18032 - Pugs - Trac
22:30 REPLeffect_ is now known as REPLeffect
pugs_svnbot r18033 | avar++ | r63676@t: avar | 2007-09-20 22:27:48 +0000 22:30
r18033 | avar++ | emitter changes from the last few commits
diff: dev.pugscode.org/changeset/18033
lambdabot Title: Changeset 18033 - Pugs - Trac
22:33 hoelzro left 22:37 Blwood left 22:38 BinGOs left 22:54 penk joined, kaether joined 22:59 Limbic_Region joined
dduncan whats up with pugscode.org? I can't connect to it now 23:14
one moment ... 23:15
oh, maybe that thing on log.perl.org explains it
or indirectly
since that's 2 days old, and pugscode was working an hour ago 23:16
so my question stands
23:16 ft joined
TimToady I didn't have any trouble connecting to www.pugscode.org 23:17
not from work anyway, but it doesn't ping from my home machine 23:18
23:21 PerlJam left, wolverian left, Juerd left 23:23 dalek left, pugs_svnbot left
TimToady it is not a good sign that all the folks from feather dropped off at the same time... 23:23
Tene mysterious questions about problems with a server, followed by everyone on that server falling offline... sounds like the start of a horror film. 23:25
TimToady but Juerd said earlier today that: Big network maintenance tomorrow 0:00 - 4:00 CET
feather will be offline a few times.
(online intermittently...) 23:26
and that seems like about now
23:30 REPLeffect left 23:41 luqui joined 23:45 dalek joined 23:47 cmarcelo left, Loro_ left

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!