pugscode.org/ | nopaste: sial.org/pbot/perl6 | pugs: [~] <m oo se> (or rakudo:, kp6:, smop: etc.) || We do Haskell, too | > reverse . show $ foldl1 (*) [1..4] | irclog: irc.pugscode.org/
Set by TimToady on 25 January 2008.
00:04 gbacon left, gbacon joined
meppl good night 00:08
00:10 jeffreykegler left, Schwern joined 00:11 meppl left 00:30 BinGOs_ left, BinGOs joined 00:51 BinGOs left 00:57 Schwern left 01:03 rindolf joined 01:04 cmarcelo left 01:09 devogon left 01:17 tobeya left 01:27 pasteling joined, cnhackTNT joined 01:31 dduncan joined, alester joined 01:47 japhb left 01:49 japhb joined 01:55 rindolf left 01:57 alester left 01:58 alester joined 02:05 namelessuser left 02:30 cmarcelo joined 02:37 justatheory joined 03:19 peeps_ joined 03:27 Auzon left 03:35 justatheory left 03:36 chacha_chaudhry joined 03:45 charsbar joined 03:49 rindolf joined 03:58 alester left, jjore joined 04:04 peeps_ is now known as peepsalot 04:08 cnhackTNT left 04:09 dduncan left, njbartlett left, [particle] left, zev left, smtms left, charsbar left, gbacon left, silug left, lisppaste3 left, allbery_b left, kane_ left, Khisanth left, dmq left, dmq joined, gbacon joined 04:10 charsbar joined 04:11 GeJ joined, djm joined, bphillips joined 04:12 broquaint joined, bloonix joined 04:16 chacha_chaudhry left, kane_ joined 04:17 qmole joined 04:18 mugwump joined 04:19 nnunley joined, dduncan joined, njbartlett joined, [particle] joined, zev joined, smtms joined, Helios- joined, jrockway joined, mofino joined, shachaf joined, yahooooo joined, tcliou joined 04:20 thestarslookdown joined 04:24 Khisanth joined 04:26 alester joined, allbery_b joined 04:27 OuLouFu joined 04:38 rindolf left 04:43 pbuetow left, pbuetow joined 04:44 orafu left 05:15 manfred joined 05:21 alester left 05:26 japhb left 05:48 kanru joined 05:49 cmarcelo left 05:50 lisppaste3 joined 05:54 japhb joined 06:02 macroron joined, broquain1 joined 06:03 broquaint left 06:06 alester joined 06:07 thestarslookdown left 06:17 silug joined 06:26 cnhackTNT joined 07:01 macroron left 07:16 alester left 07:22 alester joined 07:27 Auzon joined 07:35 alester left 07:54 Aankhen`` joined 08:00 IRSeekBot left 08:02 IRSeekBot joined 08:22 devogon joined 08:50 iblechbot joined 08:57 BinGOs joined 09:18 rindolf joined 09:27 cosimo_ joined 09:41 cosimo_ left 09:44 riffraff joined 10:11 cnhackTNT left 10:47 Aankh|Clone joined, blindfish joined 10:50 meppl joined 10:54 ruoso joined 10:55 cognominal__ joined, cognominal_ left 11:05 Aankhen`` left 11:08 Aankh|Clone left 11:14 chris2 joined 11:15 Aankhen`` joined 11:19 monomorph joined 11:22 barney joined 11:31 dduncan left 11:33 Jedai joined 11:34 Aankhen`` left 11:35 luqui joined 11:46 chris2 left 11:55 Belaf joined
pugs_svn r19792 | ruoso++ | [smop] Now only one subtest is failing. I still have memory leaks, but I think they come from the sm0p nodes not being destroyed. 11:57
r19793 | ruoso++ | [smop] adding two more trace messages... 12:16
ruoso now it shows every alloc and free... sums the leaks in the end
rakudo_svn r25427 | bernhard++ | [Rakudo] 12:22
r25427 | bernhard++ | Clean up src/utils/Makefile
12:22 jiing left 12:26 jiing joined 13:24 luqui left 13:30 cognominal_ joined 13:31 cognominal__ left
ruoso found a way to break one of the bootstrap infinite loops. CPS destruction won't happen for the SLIME and NATIVE objects... 13:40
13:42 Alias_ joined 13:44 pbuetow left 13:45 pbuetow joined 13:50 zamolxes joined
pugs_svn r19794 | cosimo++ | [t/spec/S16] Fudged chmod :todo tests 14:02
r19795 | cosimo++ | [t/spec/S16] More fudging of :todo tests 14:03
14:28 pbuetow left 14:29 barney left 14:37 pbuetow joined 14:49 pcbuetow joined 15:14 pbuetow left 15:31 pcbuetow is now known as pbuetow 15:57 alester joined
rakudo_svn r25432 | jonathan++ | [rakudo] Very preliminary work on inheritance. Doesn't work unless you declare classes in reverse order, due to an ordering change in the PAST -> POST translation phase. 15:58
16:07 Aankhen`` joined 16:15 Psyche^ joined, meppl left 16:17 meppl joined 16:23 xinming_ joined 16:25 pbuetow left 16:26 alester left 16:27 Patterner left, Psyche^ is now known as Patterner 16:28 Aankhen`` left 16:29 Aankhen`` joined 16:37 rindolf left
TimToady crumb, I just figgered out that binding should be <- and list assignment should be := 16:41
16:42 Aankh|Clone joined
peepsalot i'm reading the perl6 faq, and I don't understand the method that perl6 would use to interpret this example: if 1 <= $roll1 === $roll2 <= 6 { say "You rolled doubles!"; } 16:48
16:48 kanru left
TimToady that's an extension of the normal mathematical notation 1 <= $roll <= 6 16:49
with the additional constraint that both rolls have to be the same
you can string together as many comparisons as you like in p6 that way 16:50
peepsalot yeah, i think i got it right after I wrote it out
TimToady otherwise you'd have to write
okay, nevermind
I don't want to write it out... :) 16:51
although it shouldn't really be using === there when == would do
monomorph there: 1 <= $roll1 and $roll1 == $roll2 and $roll2 <= 6
peepsalot i will... so it's equivalent to: (1 <= $roll1) && ($roll1 === $roll2) && ($roll2 <= 6)
monomorph except that $roll1 is only evaluated once, i think 16:52
peepsalot ok, thanks 16:55
monomorph what was that remark about <- and := ? a joke i didn't get? 16:58
17:00 Aankhen`` left 17:01 pbuetow joined 17:02 cognominal_ left 17:03 pbuetow left
TimToady no, it's a serious proposal to do away with the arcane disambiguation rules in S03's description of assignment 17:05
monomorph this is about precedence in my @foo = 1, 2, 3 vs. my $foo = 1, $bar = 2, $baz = 3 right? 17:09
TimToady yes 17:12
the proposal is that = always be tighter than , and := always be looser 17:13
(assuming the theft of := from binding)
17:13 justatheory joined 17:14 riffraff left
rhr_ what would ::= and =:= become? 17:15
17:15 justatheory left
TimToady ::= could be <<= but I dunno about =:= 17:15
monomorph =><=
TimToady <->
oh wait... :)
actually, <-> would work, I think 17:16
it's <=> that wouldn't
on the other hand, could just be VAR($x) === VAR($y) with no loss of clarity 17:17
so we could probably just do away with =:=, cute though it is
I can be a ruthless sob at times... 17:18
rhr_ that's true, I haven't seen much use for it
TimToady biggest Q is how to get there from here, I think...
first need to invent <- 17:19
then switch all := to <-
and when all platforms are clear of :=, switch that to list assignment
monomorph <<= looks like left-shift assignment, even though it isn't in Perl6
TimToady sure, shift all these values left and bind them :)
17:19 devogon left
rhr_ <<= or <<- ? 17:20
17:20 pbuetow joined
TimToady <<- 17:20
<<= is undefined currently, though <<== and <== are
alas, we can't use <= for binding
peepsalot why do zip and each functions separate parameters with semicolons instead of commas? 17:21
TimToady so you can say zip(1,2,3;4,5,6)
pugs: 1,2,3 Z 4,5,6
exp_evalbot RESULT[((1, 4), (2, 5), (3, 6))] 17:22
peepsalot oh
pugs: zip(1,2,3;4,5,6) 17:23
exp_evalbot RESULT[((1, 2, 3, 4, 5, 6),)]
peepsalot uh
TimToady pugs is not up to date there
it had an old zip from before we invented slice contexts (S09) 17:24
peepsalot pugs: 1,2,3 Z 4,5,6 Z 7,8,9 17:25
exp_evalbot RESULT[((1, 4, 7), (2, 5, 8), (3, 6, 9))]
TimToady pugs: 1,2,3 X 4,5,6 X 7,8,9
exp_evalbot RESULT[((1, 4, 7), (1, 4, 8), (1, 4, 9), (1, 5, 7), (1, 5, 8), (1, 5, 9), (1, 6, 7), (1, 6, 8), (1, 6, 9), (2, 4, 7), (2, 4, 8), (2, 4, 9), (2, 5, 7), (2, 5, 8), (2, 5, 9), (2, 6, 7), (2, 6, 8), (2, 6, 9), (3, 4, 7), (3, 4, 8), (3, 4, 9), (3, 5, 7), (3, 5, 8), (3, 5, 9), (3, 6, 7), (3, 6, 8), (3,
..6, 9))]
monomorph stop, you're gonna break it! 17:26
peepsalot heh
TimToady pugs: die
exp_evalbot OUTPUT[*** Diedā¤ at /tmp/kj1oCGfM83 line 1, column 1 - line 2, column 1ā¤]
peepsalot brutal
TimToady pugs: die "again"
exp_evalbot OUTPUT[*** againā¤ at /tmp/iTmynkEZGZ line 1, column 1 - line 2, column 1ā¤]
monomorph poor pugs 17:27
17:29 cognominal_ joined
monomorph so what about $a < -$b, space required before minus? 17:29
TimToady yes
but it's already required before
or you have a postfix .<>
I think people will already be consistently putting whitespace around their comparisons 17:30
(and hopefully most other infix operators)
monomorph *i* do that in any language
crazy idea: if := is list assignment, make = list binding. it's available then, has the right precedence, and is consistent with how array assignment behaves in every language in widespread use except perl. 17:32
TimToady er, binding already is list binding
oh, you mean thrown out assignment :) 17:33
*throw
ain't gonna happen, I don't think 17:34
assignment is important for distinguishing value semantics from reference semantics
monomorph i mean that @a = @b could do what @a := @b does know, because @a := @b already does what @a = @b does now. not a serious proposal, just a quick idea
TimToady @a = @b should *copy* the value of @b, not make @a and @b identical 17:35
lambdabot Maybe you meant: activity activity-full admin all-dicts arr ask . ? @ v
TimToady the fact that other languages confuse them results in FAQs for those languages
monomorph that's true 17:36
i like how @foo *is* an array, as opposed to *points to* an array 17:39
17:39 zamolxes left
TimToady well, it does both, depending on the operator, just as $x can be either numeric or string depending on the operator 17:40
monomorph :)
peepsalot pugs: my @a = 1,2,3,4; say @a; 17:41
exp_evalbot OUTPUT[1234ā¤]
TimToady but that's why we can get away with saying that p6 doesn't have references; it does, but it's all hidden away most of the time
peepsalot pugs: my @a = (1,2,3,4); say @a; 17:42
exp_evalbot OUTPUT[1234ā¤]
TimToady you'd either have to use parens or := under the new regime
17:44 [particle] left
peepsalot the download link on pugscode.org is busted, what gives? 17:45
TimToady well, actually, you could also say my @a = list 1,2,3,4
dunno, something's busted on rakudo maybe 17:48
but since rakudo is now the official name of parrot's perl6 implementation, we should probably redirect pugs download elsewhere 17:49
just to avoid confusion
afk & # errands 17:52
18:03 [particle] joined 18:05 drrho joined 18:08 mncharity joined
peepsalot what's the difference between "make" and "make soon" for pugs? 18:13
18:13 sri_work joined 18:24 icwiener joined, jferrero joined 18:33 rindolf joined 18:42 Limbic_Region joined 18:43 sri_work left
spinclad so to C< for @rhs -> lhs {...} > we now give a converse, C< lhs <- rhs; ... >. this suggests C< foronce rhs -> lhs {...} >, or C< let rhs -> lhs {...} >, or would just C< rhs -> lhs {...} > be possible? 18:52
peepsalot: 'make soon' builds pugs unoptimized and unthreaded. 18:58
peepsalot ok thx
spinclad (re just C< rhs -> lhs {...} >): of course not, that would be two terms in a row. 19:01
(iiuc) 19:02
19:12 literal joined 19:19 DarkWolf84 joined, sri_work joined
cognominal_ suddemly realize that the synopsis is not a prescriptive but descriptive. It attempts to describe how the people will use the language insteads 19:28
nothingmuch spinclad: -> $foo { } is a sub declaration IIRC
cognominal_ ... of prescribing how they should use it.
s/synopsis/synopses/
TimToady++ 19:29
I suppose what he does is a capital crime in many circles. 19:30
linguist should be barred from computers... :) 19:31
19:39 pmurias joined
pmurias ruoso: hi 19:39
mncharity ok, I'm puzzled. 19:43
Juerd nothingmuch: It's a sub, not a 'declaration' iiuc 19:44
mncharity am looking over misc/pX/Common/yet_another_regex/engine and it's README.
Juerd cognominal_: My perception is that it is a bit of both
mncharity it was written just under a year ago now. and has largely been untouched since. 19:46
nothingmuch Juerd: it doesnt' declare a named sub, you mean?
spinclad nothingmuch: sure, a pointy block. appears in C<< for <expr> <pblock> >>, and C<< <expr> <pblock> >> would be two terms side by side.
mncharity it was passing perl5's re_tests, various similar p6 tests, and could be run as pugs's regex engine against the full pugs t/regex tests.
spinclad (pointy block / pointy sub)
mncharity fully backtracking, had commits, etc.
nothingmuch spinclad: that is my understanding anyway 19:47
19:47 cognominal__ joined
mncharity though still missing lots of p6 rx support. 19:47
spinclad nothingmuch: mine too
mncharity code could use refactoring. but code forking for refactoring was set up to be trivial, and was documented. 19:48
Juerd nothingmuch: It "declares" as much as [] does for an array (still iiuc)
mncharity it would seem to have been an obvious path to make STD work.
19:48 cognominal_ left
mncharity so... why hasn't it been touched since last year? 19:48
nothingmuch Juerd: erm, i don't follow 19:49
Juerd nothingmuch: IOW: yes :)
mncharity (not understanding that, I doubt my ability to predict what will actually help us make progress towards xmas)
so, yes. /me puzzled. 19:50
spinclad mncharity: it's possible once audrey got swallowed by $life, and TimToady took up the use of pugs, that noone was around and active and aware of yet_another_regex/engine. 19:51
*yet_another_regex_engine 19:53
mncharity svn.perl.org/perl6/pugs/trunk/misc/...ex_engine/ 19:54
lambdabot Title: Revision 14497: /pugs/trunk/misc/pX/Common/yet_another_regex_engine, tinyurl.com/yrdj69
19:56 edenc left
mncharity the README does suggest audrey seem to be coming back, which was part of why it got backburnered in favor of hs. but TT was working on STD at the time. so if lack of awareness is part of it, then the mechanism seems more likely to have been forgetting, rather than not originally knowing of it, no? 19:57
err, that's odd. is svn.perl.org/ currently active? it seems to have a significantly older version of README than I'm getting via svn up. 20:01
lambdabot Title: Subversion Server - perl.org
wolverian it doesn't seem to be updated. 20:02
spinclad sadly, i wasn't aware of it (mostly offline around then). can't speak for TT; ISTM he's been the only relevant party besides you (and audrey) in this.
pmurias ruoso: is sm0p grammar intended to be diffrent from kp6's 20:04
mncharity re svn.perl.org/perl6/ , yeah, 14497 vs 19795. Does anyone know how to poke or kill off svn.perl.org/perl6? 20:05
lambdabot Title: Revision 14497: /
pmurias * different 20:06
mncharity svn.pugscode.org/pugs/misc/pX/Commo...ex_engine/ <- "current" version 20:07
lambdabot Title: Revision 19795: /misc/pX/Common/yet_another_regex_engine, tinyurl.com/2d96yh
pmurias mncharity: it might have been the issue of Pugs::Compile::Rule taking the attention
s/taking/getting/ 20:12
spinclad mncharity: [while i'm looking at its log: r15355 has an interesting log entry: 'Fixed one _very_ long standing "thinko" bug with repetition. It dated back to the original regex engine spike a year ago. So there are many variants of it scattered around Common. :(' Do you recall if you or others propagated this to relevant other engines?] 20:20
mncharity re PCR, ok, that looks plausible. PCR continues to be active (cpan shows updates through Nov'07). and my fuzzy recollection is pugs uses PCR by default? so pugs provides the test suite. 20:21
20:22 macroron joined
mncharity re thinko, my very fuzzy recollection is not completely. but also the belief that it didn't matter, because none of them were being used anyway. I believe all of fglock's work was unaffected, because he never wanted to extensively use anon subs, and so ended up with an independent code base. 20:26
rakudo_svn r25446 | jonathan++ | [rakudo] Initial implementation of roles and compile time role composition. Has an ordering issue, as with the inheritance patch earlier, which may be a PAST -> POST translation bug. Simply compiles to use Parrot's role implementation directly. 20:28
20:32 lichtkind joined
mncharity re PCR, so... why hasn't STD ended up running on PCR? 20:32
ie, why isn't PCR+STD (with p6 code transliterated to p5) == full p6 parser running on p5? 20:36
lichtkind @seen fglock 20:37
lambdabot I haven't seen fglock.
20:40 lichtkind left
spinclad mncharity: good questions. answers may lie in lack of project 'management' on this side of the pugs/parrot divide. TT has written STD and shepherded it through pugs on his own. (more) 20:43
counter-question: how can we get people in place to be aware of resources (code and people) and make these connections? 20:47
mncharity tries to picture web-based project management software to support anarchistic development... 20:48
lol. re counter-question, ^^
more seriously, hmm... pondering... 20:49
spinclad parrot has #parrotsketch, the Perl 6 project as a whole has a weekly phone call... who here could take part in that? 20:51
20:51 LionWRK joined 20:52 rindolf left
wolverian LionWRK here is interested in making a perl6 compiler, though I don't know the specifics. please welcome him :) 20:52
LionWRK heh hi 20:53
Auzon Hi.
spinclad welcome LionWRK. where are you coming from, and where do you want to go?
LionWRK nowhere / somewhere
Auzon mncharity: it sounds like a wiki updated by the project leaders or maybe automatically (something like an SVN bot?)
mncharity: there's also Hiveminder, which is awesome, but we never really used it 20:54
I use Hiveminder for one of my own projects (plus my classwork) and it's great. I dunno how it'd scale with the amount of tasks that Perl 6 has, but it might be worth a look since we have a group there already. 20:55
spinclad LionWRK: are you aware of the various compilers already in what we like to think of as progress? 20:57
pugs, kp6, rakudo
20:58 araujo joined
spinclad built on haskell, mp6/perl, and nqp/parrot respectively 20:58
LionWRK spinclad, yes
what interested me was the talk about perl6 on the CLR
it's mostly out of personal interest in writing compilers, which I've never done 20:59
pmurias LionWRK: do you want do something from scratch or a backend? 21:01
21:04 devogon joined
spinclad or frontend work on adapting the P6 STD grammar to the various compilers? 21:05
21:05 macroron left, macroron joined, macroron left
spinclad kp6 has perhaps the most relevant plugin architecture currently for CLR backend work 21:06
also needs middle-end work, as it were, to make it more useable 21:08
kp6 and rakudo are largely written in P6 dialects 21:09
21:10 acmoore joined
mncharity re project management pondering, one observation is we seem to currently have a much easier task at present than originally. more a minor matter of implementation, than the original 'design a language and implement it'. even if the p6 language spec continues to change, 21:11
ruoso pmurias, hi...
sm0p grammar is just a macro expander
and it's already working :)
spinclad mncharity: agree 21:12
21:12 drrho left
pmurias ruoso: that's great 21:13
mncharity as it may well need to once there starts to be real usage experience (a 5.999+i imaginary series?), there seems enough spec around to make it plausible one could now implement a p6-like language implementation, with a reasonable expectation it could easily track subsequent evolution.
ruoso pmurias, almost all smop tests are passing now 21:14
I'm just finishing the review on the refcount usage
and I think I'll have the smop runtime bootstrapped
spinclad and we have some tools built: nqp seems to be much nicer to work with than bare parrot, and mp6/kp6 in a similar position on the perl5/lisp/[other backend] side 21:15
21:15 Aankh|Clone left
LionWRK I haven't given that level of thought to it really 21:15
spinclad that is, we have a couple such implementations going already 21:16
LionWRK I'd start from scratch most likely though
ruoso pmurias, after that, we'll be able to implement the p6opaque SMOP__ResponderInterface
which is the default object system for Perl 6
pmurias, take a look and run make -f Makefile.cvs build-world in smop 21:18
21:19 explorer__ joined, explorer__ left
spinclad LionWRK: from scratch a plan might be: look at STD and P6 regex language, build a regex engine, steal parts of existing compilers to build one you like 21:19
21:20 jferrero left
spinclad this might not be well distinguished from plugging in to existing compilers, but could have a different scope and focus 21:21
21:21 jferrero joined
wolverian do we have definitions of the various acronyms and such somewhere? 21:21
e.g. STD
obra I believe pmichaud put together just such a thing 21:22
wolverian looks on teh wiki
s/teh/the/
ruoso wolverian, obra, I think the wiki has it
wolverian www.perlfoundation.org/perl6/index....and_jargon is one, but doesn't have STD 21:23
lambdabot Title: Glossary of Perl 6 Terms and Jargon / Perl 6, tinyurl.com/ytvdje
pmurias ruoso: 1 failure here 21:24
wolverian hum, I can't edit the wiki 21:25
it stays at "Loading editor..."
I do have js enabled 21:26
ruoso pmurias, yep.. just one :)... and that's a good thing
it means that smop is almost bootstrapping its runtime 21:27
spinclad STD appears to stand for 'proposed STanDard Perl 6 grammar in standard Perl 6' (or is that STandarD ?)
ruoso and I can truly say that *everything* in smop is an object... including the interpreter... 21:28
Auzon How far along does that put smop? 21:30
wolverian could someone else add it to the wiki, since I can't? it's probably because I use ubuntu hardy, which has the development version of my browser, which isn't quite stable yet..
ruoso Auzon, far enough to start implementing a backend for it in kp6 in a very short time
Auzon so kp6 -> smop -> code runs? 21:31
ruoso Auzon, yep
spinclad smop++ 21:32
( wolverian: don't know if i have access, will try ) 21:33
wolverian spinclad, thanks!
spinclad++
it's very important to have the terminology accessible to newcomers
spinclad agreed 21:34
mncharity STD: svn.pugscode.org/pugs/src/perl6/STD.pm 21:44
wolverian fudge should be added to the wiki page too 21:47
spinclad (cheat and metholate as well, perhaps, under STD: they let STD run on pugs) 21:50
mncharity Auzon: havn't forgotten your observation... still pondering...
*good observation 21:51
spinclad (by rewriting it, i believe)
mncharity svn.pugscode.org/pugs/src/perl6/cheat svn.pugscode.org/pugs/src/perl6/metholate 21:52
link for fudge?
spinclad ack-grep -f | grep fudge 21:54
t/spec/fudgeall
t/spec/fudge
so svn.pugscode.org/pugs/t/spec/{fudge,fudgeall} 21:55
mncharity svn.pugscode.org/pugs/t/spec/fudge svn.pugscode.org/pugs/t/spec/fudgeall
spinclad yes, better: those are clickable
mncharity ok, the question was "how can we get people in place to be aware of resources (code and people) [...] ?" 22:04
a possible response is "create the svn.pugscode file code_and_people.txt". 22:05
which seems a bit unsatisfying.
in wiki variants as well.
why?
one possibility is, who's the target audience? 22:06
"people working on the perl6 project"?
Auzon I think that audience makes sense 22:07
maybe also include people wanting to help.
mncharity which raises the question, "what perl6 project?". There's the p6-language list. There's parrot. There is/was pugs. pugs created/spun off/gave life to something like a p6 project. A test suite, infrastructure (feather). There's kp6. There was redsix. 22:09
wolverian they're included in that definition, surely :)
22:10 jonathan4 joined
mncharity If STD were what a "perl6 project" consists of now, than the target audience and answer would obviously be "email TimToady". 22:10
22:10 jonathan4 is now known as jnthn, ofer left
mncharity which again, doesn't seem a very satisfying answer. 22:10
If someone wanted to do a perl6 on perl5 project, I know what I would tell them. 22:11
Here are the available people and resources. Plan on working closely with the p5 core development team if you plan a long term project. Plan on forking the core and putting significant effort into it if you have a shorter term one. 22:12
If someone wanted to do a perl6 on ruby project, again, I'd know the people and resources to suggest. And again, cautions like watch for 1.9.0 being potentially buggy, perhaps unlikely to be as challenging as the p5 case. 22:14
monomorph i guess part of the problem is that there are conflicting visions 22:15
mncharity If someone wanted to work on a perl6 project, well, there are some obvious suggestions.
monomorph some people feel that parrot is not going anywhere. others think that an opensource project in haskell is doomed from the beginning. etc.
jnthn I've kinda ended up just picking one and hacking on it. :-) 22:16
mncharity There are a couple of parsers now which do or can output ast representations, all variously difficulty to deal with. Put in some grunge time, create postprocessors for each, and emit simple, portable, common ast representation, to decouple front-end from compiler-and-runtime advancement.
spinclad well, sure, some people don't know thing one about where parrot is going
mncharity That's simple, not difficult, but not particularly interesting and fun. Great for a "perl6 project" though. 22:17
pmurias mncharity: you would have to design the common ast 22:18
mncharity But non-exciting and fun tasks only get done if their context is exciting. So the grunge bits are exciting by contagion.
The "perl6 project" isn't doesn't seem to be exciting people just now.
Though there seems an untapped reserve of interest. Potentially a large one. 22:19
So maybe it's just a lack of... project cheerleading-ness.
mncharity pauses to backlog comments...
22:21 mj41 left
mncharity re 'design the common ast', the ast's are not complex. the entire original p5 backend took something like 2 person/calendar days. one can do a p6 ast in a 100 lines of ruby code, perhaps 3x that in p5. 22:23
22:23 manfred left
monomorph mncharity: so you're suggesting more hype? parrot had quite a bit of hype back in 2003, i'm not sure if it helped... 22:23
pmurias mncharity: "orginal p5 backend" = ? 22:25
wolverian parrot didn't have the tools necessary for easy bootstrap back then
mncharity re visions, I would have said "agreed" an hour ago. now I'm getting a bit of traction by thinking of them as distinct projects. but yes - Perl 6 as a cluster of somewhat overlapping projects with only somewhat overlapping visions of how to get to xmas, and little coordination between them.
22:26 iblechbot left
mncharity re "orginal p5 backend", pugs got a p5 and js backend within a few months of its creation, near its period of peak activity. Spring of... whatever year it was. 22:27
They both took only a day or two to write. Taking pugs ast dump, reading it, emitting from the ast, and running it. 22:28
spinclad spring of ... 2005? 22:29
mncharity Not passing all that much of the test suite to start with. But design or complexity of the ast was not a big issue.
$ svn log -r 1 -> 2005-02-06 yes 22:30
spinclad (birthday coming up!) 22:31
mncharity lol :)
what's 3 in people years?
Auzon 100? :P 22:32
monomorph isn't there a perl6 operator for that?
wolverian Unknown unit 'people'
spinclad 3, maybe: it's finally on its feet, a bit unsteadily
starting to talk up a storm 22:33
mncharity :)
early rapid infant development, followed by a lot of eating and sleeping, now... 22:35
Auzon pugs: sleep
exp_evalbot OUTPUT[*** ā¤ Unexpected end of inputā¤ at /tmp/ZwcaaKn4x4 line 2, column 1ā¤]
Auzon pugs: sleep 5; 22:36
exp_evalbot OUTPUT[*** Unsafe function 'sleep' called under safe modeā¤ at /tmp/7jlYg1NY3g line 1, column 1-8ā¤]
Auzon heh.
spinclad pugs: while 1 { sleep; }
exp_evalbot OUTPUT[*** ā¤ Unexpected ";"ā¤ at /tmp/hruPccxtf0 line 1, column 16ā¤]
spinclad pugs: while 1 { sleep }
exp_evalbot OUTPUT[*** ā¤ Unexpected "}"ā¤ at /tmp/xUIMlS3I7n line 1, column 17ā¤]
spinclad durn 22:37
re glossary wiki page: my browser is like cold molasses, probably won't get to it before bed. anyone else wants to, feel free. 22:39
mncharity so I was going to say something like "so there are 142 people on the channel, and little two day projects which would obviously get us closer to xmas are left undone for years". apropos massaging pugs ast dumps into something people can less painfully play with. 22:43
but then there was pmurias question re the feasibility of doing that. and I remember all the questioning of ruby as a nice backend. 22:45
so perhaps there's an issue of "the big picture, or maybe the characteristics of some of its components, is not clearly visible". 22:47
pmurias: here's svn.pugscode.org/pugs/perl5/PIL2JS/ and its ast svn.pugscode.org/pugs/perl5/PIL2JS/lib/PIL/ 22:49
lambdabot Title: Revision 19795: /perl5/PIL2JS
wolverian maybe pugs kind of scared people off, too. 22:53
mncharity the p5 backend was pil-run. ast handling: svn.pugscode.org/pugs/misc/old_pugs...n/EvalX.pm
lambdabot tinyurl.com/3acchl
wolverian and it's only now picking up, when people are realising you don't _need_ to write in haskell to write a perl6 implementation. :)
mncharity :) hmm... 22:54
Auzon well yeah, now we can write it in Perl 6 :) 22:55
mncharity there does potentially seem an issue of frozen-ness. perhaps of negativity? in order to explain the "there lots of people interested, but there aren't lots of people puttering about making small improvements". 22:56
22:57 jjore left
pmurias sleep& 22:58
22:58 pmurias left
mncharity eg, we've so lowered the barrier to entry (well, a (better) wiki might dramatically lower it further for some types of work), that coming up with explanations for why there aren't more people past the barrier and bouncing around, needs a bit more creativity. 22:59
g'night pmurias
started to say "I don't think I really understand it", but that should be a question instead. "Why don't *you* spend more time on helping with p6?". hmm, that might make an interesting survey. 23:00
it's about that time of year where kids taking courses for which they have to conduct surveys are starting to think of topics. perhaps we could get someone to survey current perl developer community attitudes towards p6, its development effort, and contributing to same. 23:02
s/to/towards/ 23:03
no, to. sigh. 23:04
23:07 justatheory joined
peepsalot i'm guessing all this [ol][li] crap is not supposed to be part of the perl examples? www.programmersheaven.com/2/Perl6-F...pothetical 23:12
Juerd It's not 23:13
peepsalot k, guess i will try to email the webmaster or somethign
23:13 rwt joined
rwt How do I get a current subroutine name and can I call subroutine by name? 23:15
23:19 alester joined
Juerd rwt: For what version of Perl? 23:20
rwt 5.8.8
Juerd We're building a future version of Perl here :) 23:21
rwt Sorry.
Juerd It's still somewhat hypothetical :)
Try #perl
rwt :)
jnthn peepsalot: Erm, I work at that site. 23:22
I'm *sure* I checked a fix for that bug in a while back...
As in, quite a while back. 23:23
peepsalot jnthn, well, its baaaack ;-)
jnthn peepsalot: Or never got deployed... 23:24
peepsalot set the building on fire!
that'll teach em to not deploy 23:25
well, anyways i already submitted a note via this form www.programmersheaven.com/other/contact_us.htm
jnthn peepsalot: Thanks. Since I'm lead dev there now, I'm sure I'll find out about it on Monday. ;-) 23:27
23:27 rwt left
jnthn Thought it was a page markup bug at first, but the renderer is certainly fubar'd. 23:28
23:31 jiing left
spinclad .eat($food).sleep(:hour(10)) 23:31
&
mncharity spinclad: good night. thanks for the interesting discussion.
spinclad glad to contribute! 23:32
peepsalot pugs: my @a = 1,2,3; @a,4;
exp_evalbot RESULT[([1, 2, 3], 4)]
23:34 jferrero left 23:35 jiing joined, IllvilJa joined
peepsalot how does the reverse function work here? www.programmersheaven.com/2/Perl6-F...nes#slurpy 23:36
TimToady that does not look correct to me
peepsalot seems like it would return an array like [[[[[[1],2],3],4],5] or something 23:37
TimToady should just flatten @a into the list
peepsalot hmm, what's the difference between a list and an array?
TimToady, you mean the pugs result is not right? 23:38
or the example in the link?
TimToady yes, that's what I mean
peepsalot ok
jnthn is still confused about what is not right 23:39
The link or the Pugs output?
TimToady pugs is wrong, I think
jnthn Ah, OK.
The one that I'm not to blame for writing. PHew. :-)
jnthn is busy enough creating bugs to get blamed for in Rakudo 23:40
peepsalot TimToady, how does perl decide to make it into one big list instead of nested ones?
TimToady pugs: my @a = 1,2,3; (@a,4).join(':')
exp_evalbot RESULT["1:2:3:4"]
Juerd peepsalot: Context!
peepsalot what if I wanted a nested list?
TimToady it's probably just not in any valid context
Juerd peepsalot: Then you don't use something with a @ or () in list context.
peepsalot pugs: my @a = 1,2,3; (@a,4); 23:41
Juerd But something with \@, $ or []
exp_evalbot RESULT[([1, 2, 3], 4)]
Juerd Or you put (,) in list context, which results in an array, and which I still think is a bad idea because it's utterly confusing and leads to questions like yours :) 23:42
TimToady pugs: my $a = (1,2,3); ($a,4).join(':')
exp_evalbot RESULT["1 2 3:4"]
TimToady that one's still nested
Juerd afk
Actually, do we really need the list/array distinction? 23:43
peepsalot where can i learn more about this context you speak of
Juerd afk
peepsalot this faq I am reading does not explain much about lists and arrays AFAIK 23:44
jnthn peepsalot: The one on PH? 23:47
peepsalot yes 23:48
jnthn Yeah, agree it could say much more.
Will see if I can get some time to work on it. 23:49
peepsalot cool
jnthn I wrote the FAQ originally, BTW. 23:50
PH funded me to work on it.
Think they'll be happy to fund me to update it some now too.
There's been some design changes in bits of Perl 6 since it was written. 23:51
23:52 jferrero joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!