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.
00:02 __Ace__ joined 00:06 theorb joined 00:15 njbartlett joined 00:18 theorbtwo left 00:21 penk left 00:23 jedai left 00:28 AzureBOT joined 00:29 LimbicRegion joined 00:31 Limbic_Region left, lisppaste3 left, qmole left, jiing left, cookys left, AzureBOT_ left, c9s left, rgs left, Kattana left, pugs_svn left, exp_evalbot left, buu left, broquaint left, nipotaway left, LimbicRegion is now known as Limbic_Region 00:32 c9s joined, broquaint joined 00:33 buu joined, jiing joined 00:34 Limbic_Region left, Limbic_Region joined, lisppaste3 joined, qmole joined, cookys joined, AzureBOT_ joined, Kattana joined, pugs_svn joined, exp_evalbot joined, nipotaway joined, LimbicRegion joined 00:38 Limbic_Region left, LimbicRegion is now known as Limbic_Region 00:48 AzureBOT_ left
pugs_svn r19279 | ruoso++ | [yap6] More documentation in the TODO file... 00:53
00:55 dduncan joined 00:56 nipotaway is now known as nipotan
pugs_svn r19280 | ruoso++ | [yap6] bool is not a type, false is just a const, everything else is true. 01:01
01:14 penk joined 01:15 lyokato_ joined 01:18 cnhackTNT joined 01:35 justatheory left 01:39 devogon left 01:48 muixirt2 joined 01:49 han__ joined 01:51 han_ left 02:00 skids joined
skids use v6-alpha; async { sleep 1; "a" ~~ m/a/; } sleep 2; yields a segfault on pugs 6.2.13. FWIW. In case there's a big "list o' bugs" somewhere. 02:01
(any attempt to run a .match inside an async. 02:02
02:04 Muixirt left 02:07 amnesiac joined, penk left 02:12 Limbic_Region left, Limbic_Region joined 02:14 `nipra joined 02:17 skids left 02:23 muixirt2 is now known as muixirt 02:25 turrepurre joined, muixirt left 02:52 xinming left 03:00 jferrero left 03:14 amnesiac left 03:23 justatheory joined 03:34 Romney_08 joined 03:35 Romney_08 left 03:52 __Ace__ left 04:05 penk joined 04:20 jjore-m left 04:21 Limbic_Region left 05:17 Auzon left 05:18 spinclad left 05:25 chacha_chaudhry joined, spinclad joined 05:29 poppper joined 06:09 alc joined 06:21 alc left 06:35 amnesiac joined 06:36 araujo left 06:37 jjore-m joined 06:38 jjore-m left 06:44 devogon joined 07:18 justatheory left 07:34 drrho joined 07:47 xinming joined 07:53 iblechbot joined 08:19 drrho left 08:22 dduncan left, lyokato_ left 08:23 lyokato_ joined 08:52 jisom joined 09:01 poppper left 09:04 iblechbot left
nothingmuch arcanux.org/lambdacats/dumb.jpg 09:42
09:58 amnesiac left 09:59 lyokato_ left 10:06 cnhackTNT left 10:23 jedai joined 10:38 Grrrr left 10:43 Grrrr joined 10:46 Muixirt joined 10:58 njbartlett left 11:01 chacha_chaudhry left 11:05 pmurias joined 11:06 fglock_ joined, fglock_ is now known as fglock 11:07 nipotan is now known as nipotaway 11:10 jisom left 11:15 cognominal_ left, cognominal_ joined 11:18 pmurias left 11:30 chris2 joined 11:58 barney joined 12:01 Lorn left, Aankhen`` joined 12:05 agjal joined 12:23 pmurias joined 12:54 Limbic_Region joined 13:05 cognominal_ left, cognominal_ joined 13:07 Lorn joined 13:08 xinming left, xinming joined
pugs_svn r19281 | pmurias++ | fixed typo 13:11
13:14 fglock left 13:16 njbartlett joined 13:18 test left, moritz_ left 13:23 pen1 joined 13:30 xinming left 13:31 agjal left, agjal joined 13:39 Limbic_Region left, Psyche^ joined 13:40 penk left 13:47 fglock_ joined 13:48 fglock_ is now known as fglock 13:51 xinming joined, BinGOs_ joined 13:53 xinming left 13:55 Patterner left, Psyche^ is now known as Patterner, BinGOs left 14:11 han__ left 14:31 drrho joined 14:41 ofer joined 14:44 nipra joined 14:47 gbacon joined 14:54 gongyiliao joined 15:01 `nipra left 15:03 jhorwitz joined 15:07 rindolf joined
ruoso fglock, so... did you take a look in yap6? 15:47
fglock ruoso: not yet, sorry 15:48
but I'm thinking about an alternate implementation
adding kp6-specific ops and objects to PIR
though I'd rather have these things added to plain parrot, instead of having them as extensions 15:52
16:03 Patterner left, alester joined
PerlJam welcome alester 16:03
16:04 pmichaud_ joined, acmoore joined
fglock hi 16:04
the irclog seems broken: irclog.perlgeek.de/perl6/today 16:06
lambdabot Title: IRC log for #perl6, 2007-12-28
16:08 Psyche^ joined, Psyche^ is now known as Patterner 16:10 __Ace__ joined 16:13 araujo joined 16:16 drrho left 16:22 justatheory joined 16:26 kanru joined
pmurias fglock: hi 16:35
16:37 justatheory left, gbacon left, exp_evalbot left, pugs_svn left, qmole left, cookys left, nipotaway left, Kattana left, lisppaste3 left, Patterner left, Lorn left, devogon left, buu left, c9s left, dwave_ left, r0bby left, felipe left, polettix left, f0rth left, aindilis left, takanori0 left, Helios- left, jrockway left, zev left, nothingmuch left, gugod left, Yappo_ left, awwaiid left, pravus left, rafl left, Gothmog_ left, cj left, shachaf left, lambdabot left, chris2 left, obra left, funktio left, Southen left, jedai left, dmq left, SamB left, [particle] left, TimToady left, buubot left, mattz left, kolibrie left, djm left, szbalint left, pasteling left, Daveman left, Aankhen`` left, spinclad left, drbean left, ting left, zamolxes left, integral left, orevdiabl left, bbkr__ left, thepler left, zostay left, araujo left, __Ace__ left, rindolf left, ofer left, cognominal_ left, turrepurre left, AzureBOT left, Caelum left, ruoso left, ead left, kcwu left, cognominal left, rhc left, baest_ left, wolverian left, reltuk left, arguile left, Eidolos left, dalek left, kanru left, acmoore left, jhorwitz left, fglock left, barney left, Grrrr left, broquaint left, bloonix left, pbuetow left, Schwern left, alester left, sri_ left, allbery_b left, charsbar_ left, silug left, nnunley left, Casan_ left, chruck left, GeJ left, avar left, nipra left, gongyiliao left, Tene left, sunnavy left, yath left, yahooooo left, jql left, clkao left, agjal left, pen1 left, hcchien left, marmic left, mtve left, tewk_ left, stevan__ left, Khisanth left, Lunchy left, dvorak left 16:38 kanru joined, justatheory joined, araujo joined, __Ace__ joined, Patterner joined, acmoore joined, alester joined, rindolf joined, jhorwitz joined, gongyiliao joined, gbacon joined, nipra joined, ofer joined, fglock joined, agjal joined, pen1 joined, Lorn joined, cognominal_ joined, Aankhen`` joined, barney joined, chris2 joined, Grrrr joined, jedai joined, devogon joined, spinclad joined, turrepurre joined, nipotaway joined, exp_evalbot joined, pugs_svn joined, Kattana joined, cookys joined, qmole joined, lisppaste3 joined, buu joined, broquaint joined, c9s joined, AzureBOT joined, dwave_ joined, r0bby joined, dmq joined, hcchien joined, marmic joined, bloonix joined, drbean joined, SamB joined, felipe joined, pbuetow joined, Caelum joined, Schwern joined, obra joined, polettix joined, Southen joined, ting joined, mtve joined, f0rth joined, aindilis joined, tewk_ joined, ruoso joined, stevan__ joined, takanori0 joined, Helios- joined, [particle] joined, szbalint joined, Khisanth joined, zamolxes joined, integral joined, sri_ joined, Tene joined, TimToady joined, sunnavy joined, allbery_b joined, ead joined, charsbar_ joined, jrockway joined, silug joined, rhc joined, kcwu joined, nnunley joined, buubot joined, chruck joined, cognominal joined, mattz joined, dvorak joined, Lunchy joined, yath joined, clkao joined, jql joined, yahooooo joined, avar joined, reltuk joined, wolverian joined, Eidolos joined, thepler joined, djm joined, kolibrie joined, pasteling joined, Daveman joined, arguile joined, baest_ joined, dalek joined, zostay joined, GeJ joined, Casan_ joined, bbkr__ joined, orevdiabl joined, lambdabot joined, Gothmog_ joined, gugod joined, cj joined, Yappo_ joined, shachaf joined, nothingmuch joined, zev joined, rafl joined, pravus joined, awwaiid joined, funktio joined
pmurias ruoso: in the TODO 'the should it be implemented as low-level?' part referes to the whole of the string or just the 3 methods? 16:41
16:44 amnesiac joined 16:47 justatheory left 16:53 gongyiliao left
rindolf PerlJam: ping 16:54
avar fglock: hi there 17:00
pasteling "avar" at 212.30.222.56 pasted "RuleConstantFold.pm" (79 lines, 1.9K) at sial.org/pbot/29452 17:01
pmurias ruoso: i think you could have most of yap6 in p6 but the stage1 (or should it be called stage0) language would we very low level (maybe extending p6 in couple of places)
17:01 Psyche^ joined
pasteling "avar" at 212.30.222.56 pasted "fold-rule-constant.pl" (18 lines, 256B) at sial.org/pbot/29453 17:01
fglock avar: hi
avar that's a post-processor and a broken ast walker I made to optimize rule output
fglock looking 17:02
avar basically fold consecutive constants to one so you have /\Gtoken/g not /\Gt/g && /\Go/g && /\Gk/g ....
pmurias avar: the alternative to fixing the grammar? ;) 17:03
fglock I think we'd rather optimize at the ast level :)
PerlJam rindolf: pong
avar fglock: anyway if you could fix it up that would be neat
maybe we should have Visitor/Optimize/*.pm 17:04
rindolf PerlJam: hi. 17:05
PerlJam What's up?
rindolf PerlJam: do you know anyway to reach MJD/yrlnry except IRC and email?
PerlJam: I'm fine.
PerlJam nope, those are the methods I would use.
rindolf PerlJam: I sent a new Perl QOTW, but there seems to be email problems on plover.com 17:06
Is it "there seems to be email problems" or "there seem to be email problems"?
fglock avar: Visitor::Optimize::* is nice 17:07
pmurias avar: shouldn't push @$nodes, $stmt; be also done if $constant ne '' 17:10
?
17:12 jjore-m joined
PerlJam rindolf: I recieved it. What email problem do you think there is? 17:13
rindolf PerlJam: well: 17:14
avar pmurias: it's totally broken:)
rindolf Oh. news.gmane.org/gmane.comp.lang.perl.qotw.discuss 17:15
lambdabot Title: Gmane Loom
avar the main problem is that I didn't call .emit if the node wasn't a constant, it only work at the "first level" as it were
I was hoping someone better in ast optimization would have a go:) 17:16
pmurias avar: i can try to fix it
pasteling "avar" at 212.30.222.56 pasted "ast output" (58 lines, 2.8K) at sial.org/pbot/29454
17:16 Patterner left, Psyche^ is now known as Patterner
avar What it should be doing is merging all consecutive Rule::Constant calls into one 17:17
rindolf PerlJam: it was delayed by 5 hours.
PerlJam: guess I should start working on the solution now.
PerlJam: and my subscription request from my gmail.com address still didn't receive the confirmation.
avar pmurias: great 17:19
amnesiac AVARZ 17:22
pugs_svn r19282 | pmurias++ | [kp6] the still broken avar's regex optimising visitor
fglock you might need to refactor the regex escapes, they currently only work for specific characters - may not recognize more complex strings
pmurias the build sytem dosn't want to compile that file 17:23
printing the compilation output instead of saving to file
avar did you put it in Visitor/Optimize/ ?
17:23 REPLeffect joined, justatheory joined
pmurias yes 17:24
fixed the problem
pugs_svn r19283 | pmurias++ | [kp6] created Optimize/ in the compiled files directory 17:26
pmurias fixing the visitor now ;)
17:26 Ace_ joined, [particle1 joined 17:29 jjore-m left 17:31 barney left, jjore-m joined
pugs_svn r19284 | pmurias++ | [kp6] fixed Visitor::Optimize::RuleConstantFold 17:32
pmurias avar: you can now test it 17:33
17:40 [particle] left 17:45 __Ace__ left 17:50 justatheory left 17:52 jjore-m left
pmichaud_ the articles about Perl 6 on the perlfoundation wiki all seem to be fairly Pugs-centric. Anyone have any suggestions about refactoring these to include perl6 on Parrot? 17:53
araujo which article? 17:54
pmichaud_ www.perlfoundation.org/perl6/index.cgi?perl_6 17:56
lambdabot Title: Perl 6 / Perl 6
18:02 Psyche^ joined
pugs_svn r19285 | pmurias++ | [kp6] fixed another error in RuleConstantFold 18:03
18:14 Patterner left, Psyche^ is now known as Patterner 18:16 jisom joined 18:27 _sprocket_ joined 18:29 pbuetow left 18:30 justatheory joined, qmole left, pugs_svn left, exp_evalbot left, nipotaway left, cookys left, Kattana left, lisppaste3 left, gbacon left 18:31 gbacon joined, lisppaste3 joined, Kattana joined, pugs_svn joined, exp_evalbot joined, nipotaway joined, cookys joined, qmole joined 18:33 Daveman left 18:38 cookys_ joined 18:40 Daveman joined
_sprocket_ could somebody explain (or point out a url that explains) the differences between the pugs/not quite perl/kinda perl 6/and the perl6 in the parrot languages directory? 18:44
18:44 chacha_chaudhry joined
_sprocket_ i'm a little confused as to how they all fit together 18:44
pmurias _sprocket_: they are diffrent implementations 18:45
not quite perl is a limited subset
pmichaud_ pugs is the "most complete" implementation, but development on it seems to have stalled 18:46
languages/perl6 is the official "Perl 6 on Parrot" implementation
it's still primitive but increasing rapidly
"Not Quite Perl" is a tool to help compiler writers create subroutines for Parrot
_sprocket_ so if i were wanting to start helping out on one of the projects, is there one to pick over another? 18:47
18:47 fglock left
pmichaud_ we use NQP to help build compilers; it's not intended to be an end-user programming language 18:47
pmurias _sprocket_: depends what do you want to work with
_sprocket_ pmichaud_: ah ok. do some the perl impelementations leverage NQP then? 18:48
pmichaud_ a big way to help out all of the projects would be to write up something that explains the differences :-)
spinclad: the perl6 implementation is using NQP
sorry, that was meant for _sprocket_ (bad tab completion)
_sprocket_ pmichaud_: that would probably be a helpful document, yes. :)
18:49 orevdiabl is now known as revdiablo
pmichaud_ it could go on the wiki: www.perlfoundation.org/perl6/index.cgi?perl_6 18:49
lambdabot Title: Perl 6 / Perl 6
_sprocket_ pmichaud_: okay, i should have some time later this weekend, i'll try to get something up there 18:50
18:50 cookys left
pmichaud_ actually, I'll start one here: www.perlfoundation.org/perl6/index....mentations 18:50
lambdabot Title: Perl 6 implementations / Perl 6, tinyurl.com/3bl3g7
18:51 justatheory left 18:52 chacha_chaudhry_ joined
_sprocket_ pmichaud_: is there anything that can be brought over from the other perl6 implementations (say, pugs) that can be used in the office perl6 one in parrot? 18:52
ie) could i start porting tests over? 18:53
there doesn't seem to be nearly as many tests in languages/perl6 as are distributed with pugs
[particle1 this speaks to the relative maturity of the two implementations
pmichaud_ _sprocket_: we're going to be refactoring the test suite. The current plan is that the test suite will continue to live in the pugs repo and that the parrot repo will export them from the pugs repo
18:53 [particle1 is now known as [particle]
_sprocket_ pmichaud_: ah, okay 18:54
[particle] we could use help refactoring the test suite
pmichaud_ so, the pugs repo holds "the official test suite", and we have some make targets in perl6 that cause it to import the official tests and run them
[particle] _sprocket_: do you have a pugs commit bit?
pmurias _sprocket_: as for kp6 it can access some parts of the test suit as it lives in the same repo
_sprocket_ [particle]: no, i don't at the moment. i have a very old parrot one, but i've probably forgotten all the info 18:55
pmurias v6/v6-KindaPerl6
18:55 drrho joined
[particle] _sprocket_: send me your email and preferred username, and i'll give you a pugs commit bit now 18:55
grr, commitbit is acting up again 18:56
_sprocket_ [particle]: [email@hidden.address] should work 18:57
[particle] any feather admins here?
Juerd_: ping
19:00 rindolf left
pmichaud_ here's a draft page describing the various implementations: www.perlfoundation.org/perl6/index....mentations 19:02
lambdabot Title: Perl 6 implementations / Perl 6, tinyurl.com/34fg4j
_sprocket_ pmichaud_: cool, ok. i was planning on prodding the various implementations this weekend, so i'll fill it out a bit 19:03
pmichaud_ that would be excellent. 19:05
thanks.
_sprocket_ no problem
pmichaud_ I'll be away from irc most of the weekend, but others can help, and feel free to send me email
_sprocket_ will do, thanks :)
19:05 chacha_chaudhry left 19:07 pbuetow joined 19:10 chris2 left 19:13 acmoore left 19:15 chacha_chaudhry joined 19:17 _sprocket_ left 19:23 Limbic_Region joined 19:26 jjore-m joined
jjore-m This is an utter newbie-ism but what is proper way to read [1..9] three at a time to produce [[1..3],[4..6],[7..9]]? 19:27
19:28 chacha_chaudhry_ left
alester so pmichaud: Got some time? 19:29
pmichaud_ alester: alas, I was about to head off to lunch
I should be back in 60, though
alester OK.
Although, I think that we should blame the delays in Perl 6 on your excessive lunching.
just ping me when yr back.
And we'll hope that I don't have any fires here. 19:30
pmichaud_ (blame excessive lunches) I entirely agree. Lunch is the reason Perl 6 is delayed :-) 19:33
alester scapegoat.perl6.org
PerlJam I thought it was being delayed by all of the people asking when it will be here. 19:34
[particle] all those people are out to lunch 19:35
pmichaud_ www.perlfoundation.org/perl6/index....erl6_tasks 19:36
lambdabot Title: perl6 tasks / Perl 6
19:36 ofer left
alester pmichaud: Is that our scratchpad? 19:36
pmichaud_ I'm going to use it as one, yes. 19:37
19:37 pmichaud_ is now known as pmichaud
alester ok 19:37
pmichaud of course, there's a question as to whether the scratchpad makes more sense in the repo or in the wiki 19:38
PerlJam pm: did I see you say that you're going to be incommunicado over the weekend? 19:39
pm: I'm thinking of some things I want to write and I may email them to you over the weekend. 19:40
pmichaud PerlJam: I should still have email access, but irc time may be limited. I'm headed to CRP for the weekend 19:41
but please do email me
PerlJam pm: heh, tired of the cold?
pmichaud pj: no, I'm attending a wedding 19:43
we fly down tomorrow morning, fly back sunday afternoon
PerlJam ah.
pmichaud afk # lunch 19:45
alester pmichaud: For this first draft, I think wiki will be fine. 19:46
19:53 justatheory joined 19:54 jisom left 19:55 amoore joined 20:05 jferrero joined 20:06 drrho left
TimToady PerlJam: the reason Perl 6 is taking so long is the same reason I can keep going as long as it takes. In my view, Perl 6 is simply the result of a protracted negotiation among everyone who wants to participate, from RFC authors to implementors. 20:09
It really is still the community's rewrite, as first proposed.
I am merely the language designer. 20:10
[particle] hey, we could use your help implementing, too :P 20:11
TimToady see toke.c for a sample of my work. :) 20:13
wolverian oh come on, that was forgiven ages ago..
[particle] okay, we'll ask you again when it comes to optimization
^Wobfuscation 20:14
wolverian so I take it writing the perl6 script in perl6 would be a bad idea at this point..
[particle] i can't parse that
TimToady well, STD.pm is written in Perl 6, but I seriously doubt perl6 can parse it. 20:15
wolverian the simplified cli, from the wiki page. if one writes it in perl6, it'd be hard to launch it.
and the whole idea was to make it easier to launch perl6 :)
well, a shebang line would work, I suppose. 20:16
[particle] nope, perl6 can't parse std yet, but we're getting there 20:17
need to get <alias=rule> working, for one
and, ya know, hashes 20:18
TimToady anyway, the only part of the implementation I'm focusing on right now is the longest-token matcher, which is suitable for my optimization^Wobfuscation style. 20:19
[particle] well said
TimToady basically, I only deal with implementation where it impacts the semantics of syntax 20:20
since that's the part that all the implementations *must* agree on exactly
[particle] so, the closures inside STD.pm are there (instead of in an action grammar) because everyone must do it that way? 20:21
TimToady presumably, though that's certainly negotiable in spots 20:22
and certainly there could be room for "as long as it's equivalent to this"
in which case the correctness is determined by testing instead 20:23
[particle] yes, we have equivalencies in perl6
so we've left the closures out of the rules
TimToady and we all know how well "separate but equal" works in practics :P 20:24
[particle] it seems less cluttered to us that way, but i understand the need for some closures
TimToady *practice
[particle] sure, that's why we have unisex bathrooms everywhere :P 20:25
TimToady well, in that case, usually the same person is doing QA on all the implementations... 20:26
Limbic_Region I feel like I must be missing where all the cool kids hang out 20:28
TimToady anyway, the purpose of STD at this point is to guide convergence where the test suite is inadequate, in some cases necessarily so
Limbic_Region I have seen lots of posts lately talking about a large surge in folks asking where perl 6 is or complaining, etc - but I haven't seen the actual inquiries/complaints/etc 20:29
wolverian perlbuzz, perlmonks mostly
Limbic_Region no, not really
wolverian that is where I have seen them, that is
TimToady well, perlmonks is mostly just Cop
Limbic_Region there is 1 article on perlbuzz
TimToady and perlbuzz is mostly just Andy :)
wolverian true enough :)
alester TimToday: Not at all.
Limbic_Region and the recent stuff on perlmonks is about 5.10.0 - not 6
wolverian but individuals can generate a lot of noise 20:30
TimToady alester: it was a joke</hint>
alester Take a look at what's getting posted, and what's getting replied to.
Limbic_Region wolverian - to be more specific, I see people talking in those places about all the hub bub (as well as use.perl), but I am missing the actual hub bub
alester TimToady: Ok, forgive my diminished joke detector. I'm taking a lot of shit for what I'm trying to get going.
wolverian Limbic_Region, oh, I dunno. echo chamber, perhaps.
[particle] it's a myth. nobody wonders when perl 6 will be done. 20:31
Limbic_Region alester - well, I appreciate what your doing
TimToady but I personally never respond to second-hand worries. I have enough trouble with the first-hand worries.
alester Limbic_Region: Thanks.
Limbic_Region alester - <joke>Besides, even bad PR is good PR</joke> 20:32
[particle] of course people wonder. they just don't ask outloud, because they've been shouted down so many times already.
TimToady comes of growing up in a culture where people frequently said "I'm not offended by that, but I'm worried someone else might be offended by it, so you shouldn't do it."
alester TimToady: Let me ask a crazy question. If you were to put a percentage of how far we are from Day 1 of P6 and putting perl-6.0.0.tar.gz on CPAN, how far along are we?
[particle] <timtoady>well into the second 80%</timtoady> 20:33
Limbic_Region heh
alester [particle]: :-P
Are we closer to 10% or 90%? 20:34
TimToady I refuse to make time estimates; historically, any time estimate I have ever made about anything guarantees that it will be completely wrong.
alester Because at this point, I can't even answer THAT question with any authority.
I'm not asking for a time etsimate.
TimToady people can multiply
alester I'm looking for order of magnitude-level guessing.
TimToady well, you already have my estimate, somewhere between 50% and 100% 20:35
Limbic_Region well, most of the work is out of TimToady's hands once the language specification is done
TimToady and if it takes another 7 years, I won't be any worse than Tolkien
Limbic_Region AFAIK, the last major hurdle in that is longest token matcher 20:36
after that - lots and lots of tests need to be written
alester I'd never heard 50% before.
Limbic_Region and bootstrapping needs to happen
then it is all gravy
TimToady "into the 2nd 80%"
alester I didn't know that was an actual quote. I thought he was kidding.
TimToady assuming, of course, there's no 3rd 80%
alester Of course. 20:37
TimToady well, that's part of the problem, people don't know what we've already said, over and over
[particle] yes, it's an actual quote.
pmurias wolverian: by the perl6 script you mean the perl6 compiler?
wolverian pmurias, the thing you invoke to run the compiler 20:38
alester TimToady: And that's part of what I'm trying to do is to make some sort of "Here's what's happened in the last 7 years"
[particle] answer: promises, promises, promises.
TimToady well, we thought we were working to marry Rachel, and got Leah. :) 20:39
we'll get Rachel before the next 7 years are out, but we'll still have to work another 7 years...
speaking of wasting time, lunch & 20:41
[particle] oh, great, another delay for perl 6
Limbic_Region TimToady - when you come back. Regarding junctions, multi-threading, etc. Is there anything in the language specification dictating it is "ok to parallelize" - not just implied parallelization of junctions but keyword stuff. 20:42
[particle] limbic_region: i think [] meta_circumfix does that 20:43
[+] etc
Limbic_Region hrm
I don't think that's what I mean 20:44
for @links { @page.push($_.fetch_page) } 20:45
something that tells perl it doesn't have to do those in order or sequentially
[particle] ok, like a trait saying this block can be parallelized
Limbic_Region right
[particle] is ordered or somesuch
Limbic_Region just jumping on the "parallelization is the future" bandwagon and wondering if there is something in the language specification to future proof perl 6 20:46
last time I talked to TimToady about it, most of the coroutine, threads, etc was being relegated to implementation (parrot) team 20:47
TimToady you're looking for the "hyper" contextualizer 20:48
Limbic_Region which synopse?
[particle] S06? S03? 20:49
20:49 chacha_chaudhry left
Limbic_Region will find it 20:49
[particle] maybe add them and look at S09
TimToady see S02:3004
Limbic_Region thanks
[particle] ah, i knew eager, but forgot hyper 20:50
TimToady and, of course, hyper is implied by hyperops
gbacon TimToady: an elder from my church who's also having eye problems greatly enjoyed your transplant diary
TimToady (but not by reduce) 20:51
wolverian I remember @page.push(«@links») was considered at some point 20:52
Limbic_Region I realized hyper was implied in hyperops
wolverian or maybe I'm confusing it with junctions again :)
TimToady gbacon: thanks. I should really update it with recent history, which includes two cataract surgeries
Limbic_Region: of course, that's why I said "of course" :P
20:53 iblechbot joined
Limbic_Region sorry, I am still in the "sleep is an illusion" phase of having a newborn home 20:53
TimToady I empathize deeply
four times over
Limbic_Region oh, and I know another example of when 1 + 1 != 2 (not just for sufficiently high values of 1) 20:54
it also doesn't equal 2 when the ones in question are children under the age of 2
20:54 REPLeffect left
gbacon Limbic_Region: when it gets bad, just be thankful you don't have quadruplets :-) 20:55
gbacon shudders
Limbic_Region well, my wife stays at home and does the vast majority of the work
TimToady yes, those particular 1s are really boolean answers to the question "is all my time occupied?"
20:56 REPLeffect joined
[particle] talk about hyper... 20:56
gbacon mine too.. which is a good thing because my wife has a tremendous helping of patience
but still, I feel your pain
Limbic_Region wanders off 20:58
avar avar@oe:~/src/pugs/v6/v6-KindaPerl6$ make;perl script/kp6 --do Emit::Perl5Regex -e 'token meh { proto [ foo | bar ] }' | perltidy -pro=util/perltidyrc | grep '_pos.*/g' 21:03
( ( /\Gp/g && /\Gr/g && /\Go/g && /\Gt/g && /\Go/g && do { my $_pos = pos(); ( ( /\Gf/g && /\Go/g && /\Go/g ) ) || ( ( ( pos() = $_pos ) || 1 ) && ( /\Gb/g && /\Ga/g && /\Gr/g ) ) }
avar@oe:~/src/pugs/v6/v6-KindaPerl6$ make;perl script/kp6 --do Optimize::RuleConstantFold,Emit::Perl5Regex -e 'token meh { proto [ foo | bar ] }' | perltidy -pro=util/perltidyrc | grep '_pos.*/g'
( ( /\Gproto/g && do { my $_pos = pos(); ( (/\Gfoo/g) ) || ( ( ( pos() = $_pos ) || 1 ) && (/\Gbar/g) ) }
yay constant folding :)
TimToady interestingly, I have to do just the opposite to make an efficient longest token matcher :) 21:04
where if all the tokens are constant it degenerates to a trie lookup 21:05
avar does rt have an openid thingy? Wondering if I can use an existing account for the www.perlfoundation.org/perl6/index....erl6_tasks wiki 21:07
lambdabot Title: perl6 tasks / Perl 6
wolverian rt being an openid provider? dunno. I use my lj account. ;) 21:08
[particle] i use bitcard with rt
21:08 aaaa joined
wolverian but the wiki doesn't support bitcard iirc 21:08
21:14 aaaa left 21:28 avar left, iblechbot left
obra RT hasn't ever been an openid provider, but I'd love a plugin and happy to help with comments on the rt side of things 21:29
pmurias TimToady: is it possible to detail the type of a subroutine? something like has Code[Int,Int --> Int] $.addition 21:32
s/detail/further specify/ 21:33
21:36 Lorn left
pmurias S03:2006 seems to suggest it would be Code:(Int,Int --> Int) 21:40
pmichaud according to #parrot it looks like the perl6_tasks page may become an RT query
21:40 jferrero left
TimToady yes, though inside [] is just a parameter list, not a sig, so it'd be more like Code[:sig( :(Int, Int --> Int)] in the parametric type form 21:41
but we can probably allow the Code:(...) form too if it's not to ambiguous
or maybe &:(...) is the short form 21:42
In any case it has to know it's parsing a sig, not an expression 21:43
21:44 gbacon is now known as php-barbie 21:45 php-barbie is now known as gbacon
pmurias thanks, i was think if it was possible to map a subset of perl6 to c 21:45
s/think/wandering/ 21:46
21:50 Aankhen`` left
TimToady lessee, kp6 is p5like, nqp is parrotlike, so you want something like "crap", C Rendered As Perl :) 21:51
pmichaud and here I figured it would be "parc" (Perl Rendered As C) 21:52
oh, "prac"
"drat".
ah, we could just have our own PAC. (Perl as C) 21:53
alester pmichaud: THANKS FOR TELLING ME YOU WERE BACK. 21:58
NOW WE ARE EVEN LATER! 21:59
pmichaud oh, sorry
pmichaud blames pmichaud.
alester NOW YOU MUST PAY
pmichaud I done paid already. :-) 22:00
22:01 pmurias left
alester ok, so www.perlfoundation.org/perl6/index....erl6_tasks 22:03
lambdabot Title: perl6 tasks / Perl 6
alester Part of what we need is not a "start to do X" but just "X"
So "start to refactor test suite" is really "refactor test suite"
pmichaud yes, but "X" is often a much bigger task than "start to do X"
alester Yes, I know. 22:04
But then we can break up X.
pmichaud right, that's what we really need here
alester But let's get the highest, biggest-level rocks.
Things like
pmichaud the tasks list is intended to be "things we can get done relatively quickly" as opposed to "implement Perl 6"
alester "Have the language definition complete", which my understanding is not done.
Then we're going after two things. 22:05
pmichaud "complete language definition" is outside of the realm of perl6
"complete language definition" is a Perl 6 task, but not a perl6 one
alester But does the releas of 6.0.0 rely on it?
does the release of perl6 rely on it?
pmichaud if you're looking for milestones, I put those in the dashboard article
as opposed to perl6_tasks
we're moving perl6_tasks to RT anyway, so I should probably delete that page 22:06
yes, the release of 6.0.0 relies on completing the language definition
alester So even though it's outside the realm of perl6, it's still a requirement.
pmichaud correct
but I don't know who decides when it's complete 22:07
alester Nor I
doesn't matter right now.
pmichaud personally, I don't think that "complete language definition" is a milestone that stands on its own
alester because there are sub milestones? 22:08
pmichaud I really think that what will happen is that the language spec and the language implementation(s) will converge to a point where we decide "okay, this is enough to say that we have 6.0.0"
alester That's fine.
pmichaud i.e., 6.0.0 isn't yet a fixed target
alester And then that milsteone is done.
These don't imply order, either.
pmichaud fair enough, as long as we can make the co-dependency clear 22:09
alester yup, and even THAT is beyond right now. 22:10
Do we need docs?
Is there going to be a perl6doc?
pmichaud what sort of docs?
alester Dunno.
What kind of docs will there be?
pmichaud we have the synopses, and others are working on the perl 6 camel
alester Are those requirements? 22:11
Will there be a perl6doc?
Can I say perl6doc perl6op?
pmichaud let's say yes, there will be a perl6doc
although some would argue that it should just be perldoc
alester ok
pmichaud but yes, we need a perl6 version of perldoc 22:12
alester ok
pmichaud tentatively I'd say that we need "perl6doc perlop" to work for a 6.0.0 release, although we may have lots of workable pre-perl-6.0.0 releases before then 22:13
alester where "perlop" stands for whatever list of pages we need.
pmichaud correct
alester Oh, I'm sure we'll have plenty
ok, how about perlbug?
perl6bug?
pmichaud it's probably worth (me) reviewing the history of how Perl 5 was birthed to see what lessons/guidelines can be learned from that 22:14
perl6bug, yes
however
perl6bug may be difficult to handle given that there are multiple implementations
do we have just one perl6bug for all of the various implementations, or is each implementation simply need to provide its own? 22:15
[particle] it can include $?COMPILER
pmichaud do all of the implementations use RT?
[particle] heck, it can dispatch on $?COMPILER
22:15 njbartlett left
[particle] that seems to be more detail than desired at this point 22:16
alester But it's a good question.
pmichaud anyway, yes, we need a standard mechanism for bug reporting.
any given implementation should provide a bug report mechanism
whether each uses its own or there's a common shared one can be left for later 22:17
alester ok so we're getting somewhere
www.perlfoundation.org/perl6/index...._dashboard
lambdabot Title: Development Dashboard / Perl 6, tinyurl.com/276go6
alester Is the p5 compatibility a requirement for p6.0.0? 22:18
pmichaud historically that's been a "yes"
[particle] does perl 6.0.0 need a cp6an?
alester Will we delay a release of p600 if there isn't one?
pmichaud I don't think cp6an is a requirement. I think it will happen pre-6.0.0 anyway.
alester ok 22:19
[particle] we need a definition of 'core' perl
and a way to make distros
pmichaud for 6.0.0?
my opinion: anything that passes the test suite is 'core' perl 22:20
alester ok
[particle] 'core' definition is for scope
alester same as p5?
[particle] core as in, no modules shipped
alester Pragmata?
[particle] or, 12 modules shipped, whatever
perl 5 regex support
*we need ...
alester ? 22:21
pmichaud I don't think 6.0.0 needs a definition of distros or what consitutes 'core' perl
we may not know what 'core perl' is until well after we have a 6.0.0 release
alester Then what goes in the tarball?
Yes, but to know what the release is, we have to know what the modules are.
Again, it may be a converger.
pmichaud sure, but knowing what modules we put into a tarball is a functionof a distro, not of a definition of 'core perl' 22:22
alester I'm talking about shipping perl-6.0.0.tar.gz
pmichaud correct
alester Is that not "core perl" in your mind?
pmichaud not necessarily
alester ok
I'm trying to focus just on that tarball.
updated www.perlfoundation.org/perl6/index...._dashboard 22:23
lambdabot Title: Development Dashboard / Perl 6, tinyurl.com/276go6
pmichaud I think the way to focus on the tarball is to ignore questions about what is "in the core" versus "not in the core". That's been a rathole discussion on perl6-languages for some time.
alester You said "core" not me. :-)
[particle] ok, then let's call it a distro
i said core
pmichaud actually, particle brought up "core"
and I'm saying "ignore core"
alester OK.
I don't even know what "core" means nayway.
Right now I'm just looking through /usr/src/perl-5.10.0 22:24
x2p?
pmichaud alester: exactly my point :-) And if we try to define it, we'll get caught in all sorts of arguments about core vs. non-core that don't help us get to a distro :-)
alester h2xs?
README.*?
[particle] embedding/extension mechanisms
pmichaud x2p, h2xs: not required
[particle] cross-language functionality 22:25
wolverian www.gnu.org/software/lightning/ woot, new stuff
lambdabot Title: GNU lightning - GNU Project - Free Software Foundation (FSF)
22:25 njbartlett joined
pmichaud cross-language functionality: not required, except to the extent that it's part of the spec 22:25
wolverian six registers, heh. 22:26
alester Embedding headers, like perl.h?
pmichaud not required, except as indicated in spec 22:27
(i.e., I don't think the spec has anything about cross-language functionality or embedded headers)
of course, that might change between now and 6.0.0 :-)
alester Headline of the day: www.linuxjournal.com/node/1005966
lambdabot Title: Perl Puts Out For The First Time In Five Years | Linux Journal
alester pmichaud: Yes, of course, everything might change.
wolverian nice title 22:28
pmichaud actually, I'm a little disappointed in the title
alester emacs dir?
pmichaud: Yeah, me too
pmichaud it makes it sound like nothing has been done in five years
I don't think the 5.8.* releases are "nothing"
[particle] editor highlighting files
alester Yes, but also "puts out" means "allows to have sex with" 22:29
pmichaud editor highlighting: nice to have, not a requirement
wolverian alester, slightly confused an article
pmichaud alester: agreed, I caught that connotation of "puts out" also. I think it was intended to be cute, but it's just crass. 22:30
alester Porting/ dir?
[particle] porting is for the vm 22:31
pmichaud porting: not a requirement. That will tend to be a requirement of the underlying engine or individual implementations
s:2nd/requirement/feature/
as I read the spec, a Perl 6 implementation doesn't require a minimal set of supported platforms to be considered Perl 6 22:32
for the perl6 on Parrot implementation, it will be portable to all platforms that Parrot is portable to 22:33
alester I'm putting this on linuxjournal.com 22:34
I understand that the headline "Perl puts out for the first time in five years" is intended to be cute, but it's very disrespectful to the countless hours of work that have gone into this ubiquitous tool that has been around for twenty years now and shows no sign of stopping.
The Perl community deserves better.
22:34 SCalimlim joined
alester Sound OK? 22:34
pmichaud seems okay to me
I wasn't going to publicly comment, personally 22:35
(other than to the extent that #perl6 is publicly logged :-)
22:35 njbartlett left
alester posted. 22:36
pmichaud: is s:2nd/requirement/feature/ valid syntax? 22:37
pmichaud alester: i tis.
"it is".
alester That's just craziness.
pmichaud see S05. :-)
alester Enjoyable craziness, but craziness nonetheless.
pmichaud one can even use junctions and closures: s:nth(1|2|3|5|8|13|21|34)/foo/bar/ 22:38
s:nth{rand()}/foo/bar/ 22:39
wolverian wait, when is the closure executed?
alester www.perlfoundation.org/perl6/index...._dashboard
lambdabot Title: Development Dashboard / Perl 6, tinyurl.com/276go6
alester now has install targets.
pmichaud perl.h can't be a milestone 22:40
because most of the implementations won't have a perl.h :-)
22:41 njbartlett joined
pmichaud wolverian: I don't know when the closure gets executed for :nth 22:41
wolverian: I need to re-read the spec on that
there's been some issue about some adverbs needing resolution at compile-time while others need resolution at runtime, but I don't recall how that worked out 22:42
alester If we don't have to have a CP6AN (and there could not possibly be a worse name for it), then what about the cpan executable? 22:43
pmichaud "the cpan executable" implies there can be only one, somehow 22:44
alester One that ships with it, at least. 22:46
I8 mean, I know we're being flexible, but at some point someone's gotta start saying "this is in, this is out"
see www.perlfoundation.org/perl6/index...._dashboard for more pain
lambdabot Title: Development Dashboard / Perl 6, tinyurl.com/276go6 22:47
pmichaud I don't think we're at that point.
we already have a queue set up for perl6 bugs (perl6-bug@perl.org) 22:50
we don't have a perl6bug front-end script 22:51
TimToady s:nth{...} is illegal syntax in any case, it'd have to be s:nth({...}), in which case the closure obviously deferred
pmichaud TimToady: ah, thanks.
alester Obviously. :-) 22:52
pmichaud does the regex engine need to support things like s:i($foo) ... ?
TimToady closures are always deferred, until they're not
pmichaud fwiw, S05 has : "So are closures: :nth{.is_fibonacci}" 22:53
TimToady if :i cares it can raise an exception on something that is not compile-time constant
pmichaud okay, raising exception works for me
TimToady that would be a fossil 22:54
pmichaud that's where I grabbed my example from :-)
I can adapt PGE so that it can accept runtime values for adverbs -- I was just curious if it would need it 22:55
TimToady well, least-surprise would make $foo run-time there, I expect. 22:56
pmichaud correct
TimToady easy enough to dup for True/False, but some of the others might not be so easy 22:57
pmichaud currently PGE optimizes assuming that flags are compile-time constants, but we could give flags to PGE that mean "don't optimize this"
and then it could dtrt based on the runtime value of the flag
TimToady :nth() is just a smartmatch, though
pmichaud correct -- that one doesn't pose me any particular difficult 22:58
*y
TimToady unless you want to know when n is too large to match ever again
though in a pinch any($n..*) == $MATCH would do, I suppose. might be slow though 23:01
23:18 jferrero joined 23:19 alester left 23:22 amnesiac left 23:34 LimbicRegion joined
LimbicRegion w 23:50
23:50 Limbic_Region left, LimbicRegion is now known as Limbic_Region

Powered by Cro and the Raku® Programming Language.

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

Thank you!