Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2009-11-20

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 estrai_ joined #perl6
00:00 zaslon joined #perl6
00:00 pointme joined #perl6
00:00 sjohnson joined #perl6
00:00 estrai joined #perl6
00:16 facsimile joined #perl6
00:16 * facsimile hugs Larry Wall (at least in spirit)
00:17 * facsimile (was reading nasty blog posts about perl 6)
00:17 facsimile hm everyone not just Larry
00:17 spinclad hugme: hug everyone
00:17 * hugme hugs everyone
00:17 spinclad hugme++
00:18 * spinclad hugs everyone, too
00:19 spinclad /me is off for a holiday week-and-a-half, have -Ofun, y'all!
00:19 facsimile I had no idea people would get angry about it but probably a topic that's come up lots of times so no more to say about it
00:19 facsimile -funroll-loops!
00:21 spinclad facsimile: i'm glad to see a small spake of bloggings this week trying to ease people's antipathies. masak++ mst++ etc++
00:22 * perigrin tried to be fair and balanced.
00:22 spinclad ~Y~
00:23 perigrin emoticon fail
00:23 perigrin are you blinking with a migraine?
00:48 Wolfman2000 joined #perl6
00:49 Wolfman2000 evening
00:50 gfx joined #perl6
00:53 sjohnson hi
00:59 envi^office joined #perl6
01:01 agentzh joined #perl6
01:02 frew_ joined #perl6
01:20 cognominal_ joined #perl6
01:26 quietfanatic uh oh
01:26 quietfanatic rakudo: say 0 <=> -1/2
01:26 p6eval rakudo 7347ec: 0␤
01:26 quietfanatic -1/2 is not 0!
01:30 quietfanatic (bug submitted)
01:32 kst joined #perl6
01:33 Wolfman2000 rakudo: say -1 / 2
01:33 p6eval rakudo 7347ec: -0.5␤
01:34 quietfanatic rakudo: say 0 < -1/2
01:34 quietfanatic rakudo: say 0 > -1/2
01:34 p6eval rakudo 7347ec: 0␤
01:34 p6eval rakudo 7347ec: 1␤
01:34 diakopter rakudo: say 0/4 <=> -1/2
01:34 p6eval rakudo 7347ec: Multiple Dispatch: No suitable candidate found for 'cmp_num', with signature 'PP->I'␤in Main (file <unknown>, line <unknown>)␤
01:34 quietfanatic rakudo: say 0 == -1/2
01:34 p6eval rakudo 7347ec: 0␤
01:35 quietfanatic wait PP->I is a parrot signature.
01:37 stephenlb joined #perl6
01:47 justatheory joined #perl6
01:54 facsimile joined #perl6
01:56 diakopter rakudo: say 0 <=> 1/
01:56 p6eval rakudo 7347ec: Confused at line 2, near "/"␤in Main (file <unknown>, line <unknown>)␤
01:56 diakopter rakudo: say 0 <=> 1/1
01:56 p6eval rakudo 7347ec: -1␤
01:57 diakopter rakudo: say 0 <=> 2/1
01:57 p6eval rakudo 7347ec: -1␤
01:57 diakopter rakudo: say 0 <=> -2/1
01:57 p6eval rakudo 7347ec: 1␤
01:57 diakopter rakudo: say 0 <=> 1/2
01:57 p6eval rakudo 7347ec: 0␤
01:57 diakopter rakudo: say 0 <=> -1/2
01:57 p6eval rakudo 7347ec: 0␤
02:00 ihrd left #perl6
02:00 diakopter ng: say Inf ** (1/Inf)
02:00 p6eval ng 2a3149: 1␤
02:00 diakopter ng: say Inf ** (-Inf)
02:00 p6eval ng 2a3149: 0␤
02:01 cotto_work rakudo's math is not impressive
02:03 diakopter ng: say &(Inf)
02:03 p6eval ng 2a3149: Method 'item' not found for invocant of class 'Float'␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
02:09 Tene [particle]: re rewrite: remember that there's also the 3600 lines of setting that's mostly not being rewritten.
02:13 kst joined #perl6
02:14 TimToady yeah, they're only replacing the brain, simple...
02:17 * diakopter must've missed something about a rewrite
02:18 diakopter Tene: what rewrite?
02:18 Tene diakopter: ng
02:19 diakopter rakudo: say Inf ** (-Inf)
02:19 p6eval rakudo 7347ec: 0␤
02:27 sjohnson snuggle
02:27 sjohnson wrong chan sorry
02:30 TimToady hugme: snuggle sjohnson
02:30 Tene sjohnson: np, I do that all the time.
02:31 sjohnson i might have to roll up my sleeves and implement that hugme functionality
02:31 sjohnson i can't quite snuggle myself
02:31 sjohnson well, without grossing out anyone
02:31 diakopter too late.
02:32 Tene You're jus tt not trying hard i was gonna say, you're just not trying hard enough.
02:32 Tene Argh, lag. :(
02:33 sjohnson <--- going to watch a chick flick, and have a glass of wine for 1 by a warm fire
02:35 colomon diakopter: Are you writing new spectests for those math issues you're finding?
02:40 pugs_svn r29134 | lwall++ | [S02,3,4] attempt to rename void context to sink context and see if it clogs
02:42 Exodist joined #perl6
02:43 facsimile joined #perl6
02:46 JimmyZ joined #perl6
02:47 pugs_svn r29135 | lwall++ | [STD] put constant into the same category as subset and enum
02:53 kst joined #perl6
03:04 jaldhar joined #perl6
03:10 pugs_svn r29136 | lwall++ | [constant.t] tweaks for change of constant category to be more typish
03:10 am0c joined #perl6
03:18 kst joined #perl6
03:30 Wolf2k_Pinged joined #perl6
03:38 pugs_svn r29137 | lwall++ | [STD] improve the clarity of borg messages
03:40 diakopter colomon: if I knew they were issues, I might
03:41 colomon surely 0 <=> 1/2  --> 0 is wrong?
03:42 justatheory joined #perl6
03:47 colomon pmichaud: ping?
03:57 pmichaud colomon: pong
03:57 colomon pmichaud: I decided I really wanted to have a very basic set of range tests we can pass now for refactoring it.
03:58 pmichaud sounds like a good idea
03:58 colomon I've created t/02-basic/01-range.t here and hooked it up.
03:58 pmichaud okay
03:58 colomon is that a good name / approach?  (hooked it up to make test, I mean.)
03:59 pmichaud works for me.  Is there any reason why we can't have basic range tests in t/spec, though?
03:59 colomon Something like t/spec/blah/range-basic.t?
03:59 pmichaud sure
03:59 colomon works for me, I just thought you'd suggested the t/02-basic approach.
03:59 pmichaud this is partially why I've carped from time to time about some of the spectests relying on too many advanced p6 features
04:00 pmichaud I did suggest the t/02-basic approach, but I'd prefer that as a fallback position rather than a primary one
04:00 colomon okay, I'll move what I've got over to the spectest.
04:01 colomon (and what I've got is really, really slight so far.  :)
04:01 pmichaud that just seems easier overall, and given that we hope other implementations of Perl 6 will be coming up, it would be useful for them also.
04:01 colomon though now that I've said it, I may have to hack on master a bit to get even what I've got now working there.  :)
04:03 pmichaud well, it doesn't have to go into spectest.data in master :)
04:03 colomon BTW, the problem here is just that the range.t tests are way too complex.  The first one is "is ~(3..6), "3 4 5 6", which requires the operator to work, the range to be constructed properly, and stringification to act in a certain way which I'm not sure is actually in the spec.
04:04 colomon pmichaud: good point!  :)
04:05 justatheory_ joined #perl6
04:06 pmichaud afk for a while
04:08 benabik joined #perl6
04:16 pugs_svn r29138 | colomon++ | [t/spec] Start a set of very basic tests for Range.
04:19 dj_goku joined #perl6
04:25 pugs_svn r29139 | jafelds++ | If using XHTML (...why?), put ' /' where needed.
04:25 Wolfman2000 ...does pugs_svn not recognize the directory I'm uploading from or something/
04:26 * Wolfman2000 also wonders why the feather index page isn't updated with his changes yet.
04:31 pugs_svn r29140 | lwall++ | [S12] start process of distingishing abstract ~ from concrete .Str
04:31 pugs_svn r29140 | (likewise for + vs .Num)
04:35 pugs_svn r29141 | lwall++ | [S06,S09] rename 'is ref' to 'is parcel' and make it synonymous with \
04:46 am0c joined #perl6
04:47 justatheory joined #perl6
04:59 astrojp left #perl6
05:15 vamped joined #perl6
05:22 gfx joined #perl6
05:41 synth joined #perl6
05:52 synth joined #perl6
05:52 synth joined #perl6
05:59 QtPlaty[HireMe] joined #perl6
06:04 TiMBuS joined #perl6
06:12 pugs_svn r29142 | lwall++ | [Spec] slaughter various confusing overloadings of the term 'context'
06:12 pugs_svn r29142 | contextual variables are now dynamic variables
06:12 pugs_svn r29142 | specific dynamic contexts are now just call frames
06:12 pugs_svn r29142 | (for now we've left alone 'context' used as in 'context switching')
06:16 JimmyZ_ joined #perl6
06:20 yahooooo joined #perl6
06:30 pnate2 joined #perl6
06:36 meppl joined #perl6
06:45 diakopter std: ~ ~ ~ ~ ~ ~ 7
06:45 p6eval std 29142: ok 00:01 108m␤
06:45 diakopter std: - - - - - - - - - 7
06:45 p6eval std 29142: ok 00:01 108m␤
06:46 diakopter std: ? ! ? ! ? ! ? ! ? ! ? ! 7
06:46 p6eval std 29142: ok 00:01 108m␤
06:46 iblechbot joined #perl6
06:53 simcop2387 joined #perl6
06:58 quietfanatic joined #perl6
07:04 diakopter quietfanatic: hi
07:05 quietfanatic hi
07:05 rfordinal joined #perl6
07:06 diakopter rakudo: say [].PARROT
07:06 p6eval rakudo 7347ec: Perl6Array␤
07:06 diakopter rakudo: say [].WHAT
07:06 p6eval rakudo 7347ec: Array()␤
07:07 diakopter rakudo: say []()
07:07 p6eval rakudo 7347ec: invoke() not implemented in class 'ResizablePMCArray'␤in Main (file <unknown>, line <unknown>)␤
07:08 kaare joined #perl6
07:09 JimmyZ_ rakudo: \.PARROT
07:09 p6eval rakudo 7347ec:  ( no output )
07:10 JimmyZ_ rakudo: say \.PARROT
07:10 p6eval rakudo 7347ec: Failure␤
07:10 pugs_svn r29143 | lwall++ | [S03] tweaks from TheDamian++
07:10 pugs_svn r29143 | Various coercion clarifications
07:14 [particle]1 joined #perl6
07:23 Su-Shee joined #perl6
07:23 Su-Shee good morning.
07:26 meppl good morning, Su-Shee
07:26 meppl sushi ,)
07:28 cognominal_ rakudo-ng:    '' ~~ / { say "hello" } /
07:28 cognominal_ hum
07:29 perigrin I think ti's just ng:
07:29 cognominal_ ng:    '' ~~ / { say "hello" } /
07:29 p6eval ng 2a3149: Confused at line 1, near "'' ~~ / { "␤current instr.: 'perl6;HLL;Grammar;panic' pc 448 (src/stage0/HLL-s0.pir:300)␤
07:29 perigrin ng: /me often has a Grammar:panic too
07:29 p6eval ng 2a3149: Confused at line 1, near "/me often "␤current instr.: 'perl6;HLL;Grammar;panic' pc 448 (src/stage0/HLL-s0.pir:300)␤
07:30 cognominal_ apparently closure is not yet supported
07:30 perigrin hmm and apparently an irssi one
07:30 perigrin cognominal_: nor irc ACTION commands
07:30 perigrin :/
07:33 cognominal_ pmichaud has updated the STATUS file for nqp-rx  five days ago
07:34 cognominal_ A lot of thing already there...
07:34 cognominal_ pmichaud++
07:34 cognominal_ jnthn++
07:42 _eMaX_ joined #perl6
07:43 cognominal_ Competition yoga,  Comme dirait le dude à walter : You make everything a fucking  travesty!    http://www.nytimes.com/2009/11/​19/fashion/19fitness.html?_r=1
07:44 _eMaX_1 joined #perl6
07:48 cognominal_ oops wrong window
07:49 justatheory joined #perl6
07:49 TimToady it was entertaining nonetheless
07:50 moritz_ good morning
07:55 diakopter good morning to you
07:56 barney joined #perl6
08:16 quietfanatic There.  It's not near where I need for Gamebase, but here's Events for Perl 6
08:16 quietfanatic http://github.com/quietfanatic/Events
08:17 moritz_ cool
08:17 moritz_ quietfanatic: you should add that to the proto project list
08:18 quietfanatic moritz_: how?  Also I don't have a Makefile...
08:18 moritz_ quietfanatic: see their PIONEERS file, there's not much you have to do
08:18 moritz_ hugme: add quietfanatic to proto
08:18 * hugme hugs quietfanatic. Welcome to proto!
08:18 quietfanatic oh cool
08:19 moritz_ quietfanatic: and now you can edit projects.list :-)
08:19 quietfanatic Heh
08:20 moritz_ resistance is futile :-)
08:22 rfordinal joined #perl6
08:22 quietfanatic Hmm, how do I make it install things without learning how to write Makefiles?
08:22 moritz_ quietfanatic: as far as I understand the installed-module branch provides a reasonable default installing system
08:23 orafu joined #perl6
08:23 moritz_ quietfanatic: for example my SVG::Plot has no Makefile at all, and proto installs it just fine
08:23 quietfanatic moritz_: meaning copy the contents of lib/ into the user's library?
08:23 quietfanatic I'll have to rearrange a bit then
08:23 moritz_ right
08:23 JimmyZ joined #perl6
08:23 moritz_ yep, standard layout helps
08:25 ejs joined #perl6
08:28 quietfanatic moritz_: Okay, I think I did it right.  I have no way of knowing if I just broke your proto or not. :)
08:28 quietfanatic (until I try it for myself...)
08:34 moritz_ quietfanatic: I'm sure somebody will tell you if you broke it :-)
08:36 KatrinaTheLamia joined #perl6
08:39 pugs_svn r29144 | lwall++ | [Specs] more constant cleanups
08:39 pugs_svn r29144 | add 'anon' declarator in place of 'my'/'our' misuse
08:39 ejs joined #perl6
08:47 rfordinal left #perl6
08:59 eternaleye joined #perl6
09:05 eternaleye joined #perl6
09:07 xenoterracide joined #perl6
09:07 jrockway joined #perl6
09:26 jnthn oh hai
09:29 moritz_ oh morning, oh jnthn
09:30 masak joined #perl6
09:30 masak \o/
09:30 jnthn welcome to the wi-fi internets, masak \o/
09:31 sjohnson yo masak
09:31 masak man, this feels good.
09:31 sjohnson wireless internet?
09:31 masak you don't know how nice wifi/laptop battery is until it is taken away from you. :)
09:31 sjohnson yeah i just woke up from a nap and still typing in bed
09:31 jnthn lol i stole his laptop ;-)
09:32 masak :P
09:32 masak jnthn is joking.
09:32 * moritz_ knows quite many mac users who forget their charger :-)
09:32 masak yeah, learning the hard way what a new one costs...
09:33 moritz_ ~100€ I think?
09:33 masak aye.
09:34 moritz_ ou.ch
09:34 sjohnson jnthn: diakopter recommended console2 to me yesterday.  it does the whole resize thing with the mouse, both x and y.  quite nice if you ask me :)
09:35 sjohnson just thought i'd pass that along
09:35 moritz_ $somebody contacted me and asked if I wanted to place ads for programming jobs on my blog or website
09:35 masak do you?
09:35 masak of all kinds of ads, ads for programming jobs are the least despicable, I think.
09:36 moritz_ well, I'm toying with the idea
09:36 moritz_ and I want to get opinions from my readers
09:37 masak I don't get annoyed by ads, since I filter them out. :)
09:37 moritz_ if it were text-only somewhere in the left side bar... would that annoy you?
09:37 masak oh, text.
09:37 jnthn Text ads dont't annoy me. :-)
09:38 masak moritz_: probably not, no.
09:38 jnthn Unless they use the <OMGBLINK> tag.
09:38 masak they annoy me if they blink. or are marquee.
09:38 moritz_ jnthn: I plan to continue to use valid xhtml, and no javascript :-)
09:38 jnthn Oh yes, or marquee.
09:38 masak if they do both, I usually want to stab someone.
09:38 masak which is unusual.
09:39 jnthn Please can we all make a concerted effort not to send masak links to sites that use marquee and blink this weekend? kplzthnx.
09:39 * masak sticks out tongue
09:40 * moritz_ remebers the xcd layout on occasion of shutting down geocities
09:40 masak :)
09:47 masak quietfanatic++ # http://github.com/quietfanatic/Events
09:49 moritz_ use Event; # Haack!
09:50 masak quietfanatic++ # event sub aliens_attack { ... }
09:50 masak quietfanatic: both jnthn and I were surprised one can even do that in Rakudo!
09:50 * moritz_ too
09:51 moritz_ it took a while to understand that quietfanatic++ didn't add a scope declarator :-)
09:51 jnthn yes!
09:51 masak he's kinda worked around it :)
09:52 moritz_ rakudo: sub a(&foo) { say foo() }; a sub space { 'rocks' }
09:52 p6eval rakudo 7347ec: rocks␤
09:52 masak question is, why does he need both 'events sub' and declare the sub name in the Event::Type constructor call?
09:52 moritz_ ng: sub a(&foo) { say foo() }; a sub space { 'rocks' }
09:52 p6eval ng 2a3149: rocks␤
09:52 masak ng++ # rocks
09:52 moritz_ perl6: sub a(&foo) { say foo() }; a sub space { 'rocks' }
09:52 p6eval rakudo 7347ec: rocks␤
09:52 p6eval ..pugs: *** ␤    Unexpected "space"␤    expecting "=", subroutine parameters, trait or block␤    at /tmp/EvHH6JQVa6 line 1, column 34␤
09:52 p6eval ..elf 29144: /home/p6eval/pugs/misc/STD_red/match.rb:141:in `block in to_dump0': undefined method `to_dump0' for true:TrueClass (NoMethodError)␤   from /home/p6eval/pugs/misc/STD_red/match.rb:140:in `each'␤      from /home/p6eval/pugs/misc/STD_red/match.rb:140:in `map'␤        from
09:52 p6eval ../home/p6eval/pugs/misc/STD_red/ma…
09:53 moritz_ std: sub a(&foo) { say foo() }; a sub space { 'rocks' }
09:53 masak seems rakudo-specific :)
09:53 p6eval std 29144: ok 00:01 112m␤
09:53 moritz_ not quite :-)
09:55 eternaleye joined #perl6
10:00 masak Hacker News covers 'Fuck Perl 6', and more or less completely mis-groks it: http://news.ycombinator.com/item?id=950215
10:00 masak jjames at the bottom actually does the research.
10:01 moritz_ then upvote him!
10:02 masak (Dave Cross)++ # http://perlhacks.com/2009/11/my-sekrit-project.php
10:02 masak as you see, I'm catching up on news here. :)
10:03 moritz_ wow
10:03 moritz_ that's exactly what I had in mind of perlblog.org
10:05 synth joined #perl6
10:35 zaslon lolmasakhazblogged! masak++ 'November 19 2009 -- a feeling of power, a feeling of release': http://use.perl.org/~masak/journal/39921?from=rss
10:37 am0c joined #perl6
10:38 sjohnson i would like to give suggestions and the like for this new Perl 6 pdf book
10:38 masak sjohnson: go ahead; either here or on #perl6book.
10:38 sjohnson thanks masak
10:39 sjohnson finally, something i could contribute to perl6
10:39 sjohnson ( `ー´)
10:39 sjohnson is it p michaud who runs rakudo's site?
10:42 sjohnson pmichaud: grep downloadthe on http://rakudo.org/node/58
10:50 ihrd joined #perl6
10:50 ihrd left #perl6
10:55 masak food &
11:02 [particle] joined #perl6
11:02 drbean_ joined #perl6
11:03 athaba joined #perl6
11:12 justatheory joined #perl6
11:15 moritz_ sjohnson: typo corrected
11:19 sjohnson *thumbs up*
11:20 cognominal joined #perl6
11:27 dakkar joined #perl6
11:56 simcop2387 joined #perl6
12:05 ejs joined #perl6
12:15 envi^home joined #perl6
12:17 payload joined #perl6
12:20 simcop2387 joined #perl6
12:34 colomon joined #perl6
12:35 colomon o/
12:36 sjohnson hi colomon
12:58 pmichaud good morning, perl6
13:00 jnthn morning, pmichaud
13:01 colomon \o
13:03 pmichaud today's goal:  eval.  and maybe repl.
13:03 moritz_ a worthy goal.
13:04 colomon \o/
13:04 moritz_ and being able to call eval_{lives,dies}_ok will bring us back a few tests, I think
13:04 masak joined #perl6
13:04 jnthn lolitsmasak
13:05 jnthn moritz_: Yes, I think so.
13:05 colomon and being able to call eval will bring us back a few more.
13:05 jnthn I've had tests that passed all until they hit a dies_ok.
13:05 moritz_ dies_ok should work
13:05 moritz_ just the eval_ versions not
13:05 jnthn uh, eval_does_ok I ment.
13:05 masak TimToady: in S32/Containers, the documentation for the .pick method on Set looks copy/pasted from that for .pick in Bag, and not altogether applicable.
13:05 jnthn *ie
13:06 moritz_ pmichaud: do you plan to ever merge ng into master, or simply replace it?
13:06 pmichaud replace
13:06 masak TimToady: I'll probably change it right away, but might want to look over the result, to see that I didn't miss anything, or introduce new infelicities.
13:06 pmichaud likely:  rename master to something else, rename ng to master
13:07 moritz_ pmichaud: any objections to cherry-picking some useful commits from master ng?
13:07 pmichaud no objections
13:10 colomon moritz_: do you have something in mind?
13:11 moritz_ colomon: lots of things actually
13:11 moritz_ the doc changes from the release
13:11 takadonet morning all
13:11 moritz_ KyleHa's patch to parallelize tools/update-passing-test-files.pl
13:12 colomon sounds good to me.
13:13 moritz_ I'm just a bit afraid of losing overview, and forgetting some important things
13:13 colomon rakudo: say "hello".perl
13:13 p6eval rakudo 7347ec: "hello"␤
13:14 colomon rakudo: say '$range'.perl
13:14 p6eval rakudo 7347ec: "\$range"␤
13:14 colomon ng: say '$range'.perl
13:14 p6eval ng 2a3149: sh: ./perl6: No such file or directory␤
13:15 pugs_svn r29145 | masak++ | [S32/Containers] fixed copy/paste error
13:15 moritz_ it's rebuild time
13:16 masak I feel this line is non-sensical, or misleading. http://github.com/rakudo/rakudo/blob​/master/src/setting/Any-list.pm#L53
13:16 masak can I remove it? (if I make spectests, and nothing breaks)
13:17 colomon wfm
13:17 moritz_ I'd rather remove the Int type constraint
13:17 masak maybe that'd be better, yes.
13:17 masak I'll do that.
13:17 bluescreen joined #perl6
13:17 pmichaud +1
13:18 masak also, I'm not really sure I like the current behavior of %h.pick
13:18 colomon the idea that .pick(1.5) would be valid kind of scares me
13:18 masak but I've not quite decided yet.
13:18 iblechbot joined #perl6
13:19 colomon would .Int be more appropriate than .floor anyway?
13:20 masak colomon: yes, that sounds like a good idea.
13:20 moritz_ +1
13:20 masak intent rather than procedure.
13:21 * masak writes `$num .= Int`, which reads quite well
13:21 * moritz_ wants $num as Int
13:21 cognominal I thought I had seen in S05 a convention for bracketted construct with the brackets before the rule for bracketted stuff. Am I wrong?
13:22 masak moritz_: nod.
13:22 masak cognominal: A ~ C B
13:22 moritz_ cognominal: the "goal matching" thing, yes
13:22 moritz_ cognominal: '(' ~ ')' [ stuff ]
13:22 cognominal thx
13:23 masak std: my $a; $a .= Int
13:23 p6eval std 29145: ok 00:02 109m␤
13:23 masak rakudo: my $a = 5.5; $a .= Int
13:23 p6eval rakudo 7347ec: .= must have a call on the right hand side at line 2, near " .= Int"␤in Main (file src/gen_setting.pm, line 2751)␤
13:23 * masak submits rakuodbug
13:24 masak ng: my $a = 5.5; $a .= Int
13:24 p6eval ng 2a3149:  ( no output )
13:24 cognominal '(' ~ ')'  <term> ** \,   # so that makes sense?
13:24 masak ng++
13:25 masak cognominal: yes.
13:25 moritz_ cognominal: you might need brackets around the <term> ** \,
13:25 moritz_ I'm not sure about the relative precdence of ~ and **
13:25 moritz_ and rakudo used not to be sure too
13:26 cognominal So I suppose it would fail to work as desired if <term> may match a closing parenthesis
13:27 masak that sounds reasonable, yes.
13:29 pugs_svn r29146 | colomon++ | [t/spec] More very basic Range tests.
13:30 cognominal ng:   "ab" ~~  / a ~ b <?> /
13:30 p6eval ng 2a3149: Confused at line 1, near "\"ab\" ~~  /"␤current instr.: 'perl6;HLL;Grammar;panic' pc 448 (src/stage0/HLL-s0.pir:300)␤
13:30 moritz_ cognominal: ng doesn't have regex literals yet
13:31 cognominal ng:   "ab" ~~  / 'a' ~ 'b' <?> /
13:31 p6eval ng 2a3149: Confused at line 1, near "\"ab\" ~~  /"␤current instr.: 'perl6;HLL;Grammar;panic' pc 448 (src/stage0/HLL-s0.pir:300)␤
13:31 masak moritz_: which is kinda ironic.
13:31 moritz_ i know :-)
13:31 moritz_ masak: tell pmichaud (after he's done eval)
13:32 masak I'm not going to imply to pmichaud that I want a schedule change. :)
13:32 masak I want ng to be as good as possible, that's all.
13:32 cognominal rakudo:   "ab" ~~  / 'a' ~ 'b' <?> /
13:32 * moritz_ wants it to be better than possible :-)
13:32 p6eval rakudo 7347ec:  ( no output )
13:32 cognominal rakudo:   say "ab" ~~  / 'a' ~ 'b' <?> /
13:32 p6eval rakudo 7347ec: ab␤
13:33 cognominal how can he test anything without regex literals?
13:33 moritz_ grammar Foo { token TOP { regex here } }; Foo.parse('str')
13:34 cognominal ok :)
13:42 cognominal sukria, ça peut t'intéresser
13:42 cognominal oops
13:43 masak ça certainement ne peut m'intéresser!
13:43 masak pfui! :)
13:44 moritz_ colomon: you coul add range-basic.t to spectest.data
13:44 moritz_ masak: 'ne'... always requires 'pas' or 'plus' or something like that in the end
13:44 colomon moritz_: I did in ng
13:44 masak moritz_: dang. :/
13:44 colomon don't know if it works in master
13:44 masak it's been too long.
13:44 masak ça certainement ne peut PAS m'intéresser!
13:44 moritz_ colomon: if so you didn't push
13:45 moritz_ $ grep range t/spectest.data | grep -vE \# | wc -l
13:45 moritz_ 0
13:45 ng_feed rakudo-ng: colomon++
13:45 ng_feed rakudo-ng: Add S03-operators/range-basic.t to the tested spectest files for ng.
13:47 colomon moritz_: apparently I am not always fully functional right before going to bed.
13:48 colomon s/right before going to bed//
13:48 ruoso joined #perl6
13:49 moritz_ nobody is always fully functional :-)
13:50 colomon I think this explains why ng_feed never reported my commit last night.  :)
13:50 moritz_ hugme: hug ng_feed
13:50 * hugme hugs ng_feed
13:51 moritz_ up to 279 spectests now
14:00 sjohnson masak: tu parles francais aussi? :)
14:00 masak sjohnson: non. je ne peux parler français.
14:00 masak sjohnson: non. je ne peux PAS parler français.
14:00 masak dang again. :/
14:01 sjohnson heh
14:01 moritz_ :-)
14:01 sjohnson ou as-tu appris le francais?
14:01 * moritz_ smatters français :-)
14:02 masak sjohnson: j'ai l'appris, mais je ne peux pas parler.
14:02 jnthn Je suis une merdique en francais.
14:03 moritz_ jnthn is female in french? :-)
14:03 masak I'm afraid so. :P
14:03 sjohnson est-ce qu'on peut parler en francais sans trouble si on le fais a 06h du matin?
14:03 jnthn Ooh la la...
14:03 cognominal masak: non, ton français est merdique  :)      don't confuse the adjective with the noun
14:03 masak cognominal: don't confuse me with jnthn  :)
14:04 cognominal anyway, we pardon you french  :)
14:04 jnthn We pardon you English too.
14:04 cognominal oops
14:04 masak touché :)
14:05 masak .oO( your father was a hamster and your mother smelled of elderberries )
14:05 * moritz_ has heard worse name calling
14:06 masak moritz_: you... kniggit!
14:08 cognominal "touché", "c'est la vie"  and some others are French expressions rarely used except by anglosaxons  :)
14:08 masak cognominal: "double entendre" :)
14:08 PerlJam good $localtime all
14:08 cognominal indeed
14:09 sjohnson hi
14:10 cognominal we would say "double sens"
14:10 moritz_ hi PerlJam
14:10 PerlJam cognominal: but which consonants would you not pronounce?  :)
14:10 masak cognominal: why is the adjective before the noun in that phrase?
14:11 PerlJam TimToady++  (is parcel)
14:12 cognominal I don't know any definitve rule for the place of the ajectives which come usually last.
14:12 PerlJam TimToady++ (s/contextual/dynamic/  Isn't that were we started?  :)
14:12 masak cognominal: when I learned French, I was taught that they all come last except for a select few.
14:13 cognominal Sometimes,  the meaning varies with the postiion : "Un  homme grand" is a tall guy,  but "un grand homme" is an very important man. De Gaulle was both
14:14 cognominal masak: yes, that the general rule.
14:14 masak cognominal: nodnod. that makes very much sense, because the most meaning-carrying word gets to go first.
14:15 cognominal more funny is the sexism,  "un homme public" is a well-known man,  "une femme publique" is a prostitute
14:15 masak all too common in languages.
14:15 moritz_ we have that (most important first) in Perl 6 too
14:15 masak a wizard is something cool. a witch is something bad.
14:15 moritz_ that's why you can write statement-modifying 'if', 'for', 'while'
14:16 astrojp joined #perl6
14:16 moritz_ masak: witch isn't negative to me, but that might be because I read all 7 Harry Potter books :-)
14:16 moritz_ where it's used neturally
14:16 masak moritz_: right. but look at children's stories in general.
14:17 masak moritz_: I'm sure I could think up more pairs than witch/wizard, too.
14:17 moritz_ yes, sure
14:17 PerlJam I don't think I've heard tell of a "nice witch" or a "kind witch" or a "beautiful witch"  (Except in Oz)
14:17 PerlJam usually  they are evil and/or ugly
14:18 colomon PerlJam: Think of Oz.  Two bad witches, two nice witches.
14:18 PerlJam as I said, except in Oz
14:18 colomon ah, yes.  :)
14:18 colomon Oz and Harry Potter.
14:19 PerlJam Aesop certainly had a very negative view of witches
14:19 PerlJam And the brothers Grimm as well (but they seemed to be very negative towards everybody)
14:20 moritz_ I guess 'sorceress' isn't exactly very positive either :-)
14:20 colomon in my head, if I strip away modern stuff, wizard is pretty dodgy as well.
14:21 colomon maybe not always out and out evil, but dark and mysterious and arrogant.
14:21 moritz_ colomon: I guess you count Tolkien as "modern"?
14:21 colomon moritz_: he certainly is next to Aesop and Grimm.
14:23 moritz_ because Gandalf is one of the most likeable characters in LOTR and Little Hobbit, IMHO
14:23 colomon and even in Tolkien, dark and mysterious and arrogant fits Saruman to a T, and really isn't that far from Gandalf.
14:24 colomon we just see Gandalf's lighter side, too.
14:24 moritz_ he has one :-)
14:28 elmex_ joined #perl6
14:29 * masak is looking forward to the adaptation of The Hobbit
14:30 masak too bad the canceled "Ender's Game", though. :/
14:30 * colomon is actually pretty happy with the existing adaptation of The Hobbit.
14:31 PerlJam I was looking forward to a proper treatment of LotR, but was disappointed that a good chunk of Return of the King was left out :(
14:31 masak colomon: hm, haven't seen that one.
14:32 PerlJam colomon: are you talking about the animated ones?
14:32 moritz_ PerlJam: I was especially disappointed that the return to the shire was so different from the book
14:32 colomon masak: It might not have the same effect on people who weren't 9 when they saw it the first time.  :)
14:32 PerlJam moritz_: indeed
14:32 colomon PerlJam: yes
14:32 moritz_ OTOH it was way too long for a single movie anyway :-)
14:33 moritz_ PerlJam: I found that the LOTR movies are good if you treat them as disconnected from the book
14:33 PerlJam moritz_: yes, I suppose that's true.
14:33 PerlJam But I've read the books so many times, I *know* the story and the movies aren't quite it
14:34 colomon I'm guess I'm something of a heretic -- IMO the movies strip away a lot of cruft from the books.  They're definitely not perfect, but at this point I'd be more inclined to rewatch the movies than to reread the books.
14:34 PerlJam I can understand leaving out things like their encounter with Tom Bombadil but other things are less forgiving.
14:35 moritz_ PerlJam: I think Tom Bombadil is not essential for the story, but still essential for the spirit of the book
14:35 PerlJam s/forgiving/forgivable/
14:35 colomon (Though I fear I won't react so well to the new Hobbit movie... that book is much tighter.)
14:37 athaba joined #perl6
14:38 * colomon would be vastly more likely to reread some Stevermer or Wolfe anyway...
14:39 synth joined #perl6
14:44 KyleHa joined #perl6
14:48 pmurias joined #perl6
14:49 pmurias ruoso: hi
14:50 ruoso hi pmurias
14:51 pmurias 7 mildew tests are remaining for mildew-js :)
14:51 * ruoso envy pmurias, because he can actually code...
14:51 pmurias ruoso: still doing the budget planning?
14:52 ruoso now planning the details on the budget execution
14:53 masak .oO( budget executioner )
14:55 moritz_ .oO( execution executioner )
14:57 pmurias ruoso: i think having a .handled on exceptions is incorrect as the spec seems to imply that the exceptions are rethrown at the end of the block if you don't exit the block with a when
14:58 athaba joined #perl6
15:00 pmurias joined #perl6
15:01 perlygatekeeper joined #perl6
15:08 ejs joined #perl6
15:09 ruoso pmurias, the concept of halded exceptions exist anyway
15:09 ruoso s/halded/handled/
15:10 ruoso even if that is not the definition for the rethrow
15:10 ruoso because that applies to $!
15:10 Psyche^ joined #perl6
15:12 PacoLinux joined #perl6
15:12 * ruoso lunch &
15:13 payload joined #perl6
15:20 pmichaud jnthn: ping
15:20 KyleHa moritz_: How about we rename $impl in t/harness in ng to 'ng' and then fudge stuff in the suite for it so we can run more files sooner?  When ng is about ready to go master, we switch $impl back to 'rakudo' and take out the fudge.  Eh?
15:22 jnthn pmichaud: pong
15:22 KyleHa colomon: I don't think you need done_testing; if your plan has an explicit number (I'm looking at range-basic.t).
15:22 pmichaud interesting conversation in #parrot last night about compilation performance; thread starts at http://irclog.perlgeek.de/p​arrot/2009-11-20#i_1751189
15:22 colomon KyleHa: You're probably right.  :)
15:22 pmichaud takeaway:   mark and sweep is killing us
15:23 KyleHa colomon: Also, I think the best way to test the result of .perl is to eval() it since there might be multiple legitimate (but not string eq) ways to represent what's getting serialized.
15:24 KyleHa colomon: And thanks for range-basic.t !  8-)
15:24 colomon KyleHa: Agreed, but that relies on eval actually working.  And it doesn't.
15:24 pmichaud I should have eval shortly.
15:24 pmichaud as in, today.
15:24 [particle] KyleHa: i would call it rakudo-ng to be more descriptive to the casual observer
15:24 KyleHa pmichaud++  # eval!
15:24 KyleHa [particle]: Good idea.
15:24 colomon pmichaud++  #indeed!
15:25 jnthn pmichaud: That's more than I'd have expected.
15:25 jnthn profiling++
15:25 pmichaud well, let's think about it this way
15:25 cognominal joined #perl6
15:26 pmichaud when compiling (nqp) Actions.pm, we end up creating 23,994 Capture objects
15:26 pmichaud that sounds about rightish to me, given that every parse, PAST, and POST node is a Capture object
15:26 masak o.O
15:26 pmichaud (they're trees, after all)
15:26 pmichaud furthermore, they're long lived
15:27 pmichaud so, as our source programs get bigger, we end up (repeatedly) mark+sweep on that entire tree
15:28 KyleHa How often does Parrot collect garbage?
15:28 jnthn Yes, thus the need for a more generational scheme.
15:29 pmichaud in that particular run, there were 92 gc runs
15:29 pmichaud resulting in 500,000 marks of Capture objects
15:29 rfordinal joined #perl6
15:30 masak o.O
15:30 jnthn Given that so far, two GSOC projects iirc have tried to improve GC and not ended up getting merged back (granted, the code likely got a bunch cleaner), I'm not optimistic of that changing too soon... :-/
15:30 masak that sounds computationally intensive. half a million marks.
15:31 pmichaud that's half a million marks for 24,000 PMCs
15:31 pmichaud so each one of those PMCs ended up being marked (on average) about 20 times
15:31 jnthn That's seriously painful.
15:31 pmichaud note that by way of comparison, close to 3.5 million PMCs were created for PCC
15:32 jnthn Well yes.
15:32 pmichaud but they resulted in far fewer gc marks
15:32 jnthn Right, 'cus they die very young.
15:32 pmichaud exactly
15:32 jnthn Almost alwys.
15:32 pmichaud 1.8 million Context and CallSig PMCs resulted in only 30,000 marks.
15:33 jnthn They're probably for the most part found unreachable and so we need not mark them.
15:33 Juerd Hmm, 3.5 cookies :)
15:33 Juerd +million
15:33 jnthn .oO( 3.5 million stroopwaffels )
15:33 Juerd s/ff/f/
15:34 pmichaud jnthn: yeah.  So, we were concerned about contexts-as-pmcs increasing GC pressure.... but that tends to not happen much
15:34 jnthn Well, yes and no.
15:34 pmichaud relatively speaking it doesn't happen much, because what matters is not how many context pmcs get created, but how many are live when gc runs occur
15:34 jnthn We still pay an allocation/collection cost.
15:34 alester joined #perl6
15:34 pmichaud well, we were getting allocation/collection cost even when they weren't PMCs
15:34 jnthn But they don't stay around and need marking.
15:34 jnthn Yes, true.
15:35 pmichaud (maybe not as much, but we did have to allocate/collect contexts)
15:35 pmichaud anyway, I'm wondering if we could use this information to get another parrot hackathon to fix gc :)
15:35 pmichaud similar to the way we got people working on pccupdate
15:35 jnthn :-S
15:36 jnthn Dunno. It's non-trivial to design and implement and debug.
15:36 pmichaud true, but whiteknight already has a good start on it
15:36 pmichaud and pcc was also non-trivial to design and implement and debug
15:37 pmichaud anyway, I think I may post these numbers to parrot-dev and see how people respond
15:37 jnthn Yes, it's worth making the bottleneck more widely known.
15:40 explorer joined #perl6
15:42 pugs_svn r29147 | masak++ | [spec/S04-declarations/constant.t] unbroke singly-quoted string w/ apostrophe
15:43 [particle] does it make rakudo unacceptably slow?
15:43 [particle] or, noticibly slower than expected?
15:43 pmichaud it makes compiling slow
15:44 pmichaud the benchmark was a straight compile, no execution
15:44 [particle] well, sure, but that's what the benchmark is...
15:44 dalek rakudo: 6bf0179 | masak++ | src/setting/Any-list.pm:
15:44 dalek rakudo: [Any-list.pm] removed Int constraint on $num in .pick
15:44 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/6​bf017930e6a020312229abdce7c93bf5ff7f857
15:45 [particle] ooc, is compiling rakudo-ng today slower than compiling rakudo two or three months ago?
15:45 pmurias rakudo: my @foo;say @foo.push(7);
15:45 p6eval rakudo 7347ec: 7␤
15:45 pmurias rakudo: my @foo;say @foo.push(7,4);
15:45 p6eval rakudo 7347ec: 74␤
15:45 [particle] i'm wondering what slow compilation is benchmarked against.
15:46 * [particle] runs off to a meeting &
15:46 pmurias the return value of Array.push is not specced
15:47 pmichaud [particle]: (slow compilation) -- the fact that we end up compiling around only 50 lines of source code per second?
15:51 KyleHa I wonder how that compares to compilers that took their input from punched cards.
15:52 masak :)
15:52 masak KyleHa: it's kinda tragic when that comparison even comes to mind.
15:53 pugs_svn r29148 | lwall++ | [S03] fix fossil found by edwin.steiner++
15:53 pugs_svn r29148 | nail down that constants are still constants
15:53 KyleHa I remember my dad telling me that an important innovation during his college years was that you could hand your box of punched cards to the tech in the computer room and then GET THEM BACK a few minutes later instead of having to wait until the job had run.
15:54 KyleHa (They copied them and gave them back.
15:55 TimToady pmurias: "Add to the end of C<@array>, all of the subsequent arguments and returns the modified array." S32-Containers
16:02 pmurias TimToady: thanks, not sure how i missed that
16:04 zaslon lolmasakhazblogged! masak++ 'November 20 2009 -- big battles, small commits': http://use.perl.org/~masak/journal/39922?from=rss
16:06 masak std: Container
16:06 pugs_svn r29149 | lwall++ | [S32/Containers] minor typos
16:06 p6eval std 29147: Undeclared name:␤     'Container' used at line 1␤ok 00:01 107m␤
16:06 masak TimToady: 'Container' is in S32 but not in STD.
16:07 TimToady feel free to add it to CORE.setting :)
16:07 masak not remove it from S32/Containers? :)
16:07 TimToady either is fine by me until I get my coffee...
16:10 TimToady okay, got my coffee, and it doesn't help...guess I'll have to drink some of it
16:10 allbery_b IV?
16:11 TimToady they took out my PIC line when I left the hospital, alas
16:12 masak allbery_b: problem is, after going IV there's really no level above that. you've kind of run out of boosters.
16:13 allbery_b high-saturation caffeine ringers drip
16:14 TimToady I asked for the genetic mod to manufacture my own caffeine, but they said I was a poor risk, since my head might explode.
16:16 masak I can only concur in that assessment.
16:16 masak about the risk, I mean.
16:16 masak furthermore, I don't think anyone here at #perl6 would want TimToady to run headless.
16:19 pugs_svn r29150 | lwall++ | [S03] typo
16:21 pugs_svn r29151 | lwall++ | [S32/Containers] typo
16:22 TimToady obviously I shouldn't type either in the evening or in the morning
16:23 TimToady rakudo: say Container.WHAT
16:23 p6eval rakudo 7347ec: Could not find non-existent sub Container␤in Main (file src/gen_setting.pm, line 324)␤
16:23 TimToady hmm, can anyone envision a use for the Container role?
16:23 TimToady what methods does it have?
16:24 masak TimToady: its primary purpose seems to be to tie together List and Hash.
16:25 masak which in itself is kinda dis-levelly.
16:25 masak List and Associative I'd understand, or Array and Hash.
16:25 masak erm.
16:25 masak List and Map.
16:26 TimToady maybe I will add it to CORE.setting when someone writes a test that uses it.  :)
16:27 TimToady (a test that tests its functionality, not its existence)
16:27 TimToady ((if it has any functionality))
16:27 masak I think it's bogus.
16:27 jnthn FETCH and STORE?
16:29 ejs joined #perl6
16:29 TimToady mebbe, have to think about it when my thinker is thinkier.
16:29 jnthn :-)
16:30 TimToady ng: say Container.WHAT
16:30 p6eval ng c8de8d: Could not find non-existent sub &Container␤current instr.: '_block14' pc 29 (EVAL_1:0)␤
16:30 TimToady then there's that datapoint...
16:30 jnthn Well, yeah.
16:30 jnthn We don't have a Container role
16:30 jnthn We do have FETCH and STORE methods on all our containers, iirc.
16:31 TimToady but with differing signatures, I'll wager
16:31 jnthn mebbe, pmichaud can probably answe that right off since he wrote them. :-)
16:31 TimToady meaning you can't write generic Container code in any case
16:32 jnthn In that case, yes, the container role is less useful.
16:32 TimToady unless someone wants ~~ Container for some reason that I can't fathom at this depth
16:33 TimToady early in the morning my right brain loves to free associate while my left brain isn't looking
16:33 masak :)
16:34 masak TimToady.rightBrain ~~ Associative
16:34 masak o/
16:34 jnthn masak: I needz beer1
16:34 jnthn oh yay...we can go to the bar! \o/
16:34 jnthn o/
16:35 TimToady too early!!!
16:36 explorer__ joined #perl6
16:47 TimToady hmm, both "bridge building efforts"; we may have two bridges for sale, if they don't meet in the middle :)
16:50 lambdabot joined #perl6
16:52 payload joined #perl6
17:18 rfordinal joined #perl6
17:20 diakopter re "Ender's Game" film canceled - I suspect that tricking a child into effecting genocide/obliteration of an entire planet is a notion that, when portrayed on the big screen, would stir up too-strong emotions... so it's a very risky project.
17:21 diakopter even if the alternative was destruction of humanity...
17:22 colomon isn't that already the plot of the Last Starfighter?  ;)
17:22 TimToady Without knowing a thing about it, I suspect Scott Card of putting the kabosh on it for getting out of hand
17:23 TimToady I doubt he would have signed over all his editorial rights.
17:24 diakopter ... I'm sure that's true.. producers wanted to tame/alter the psychology
17:24 diakopter I'm guessing.
17:24 TimToady are you always sure something is true when you're guessing? :P
17:25 diakopter 'that' referred to what you said... "I'm guessing" referred to "producers..."
17:26 diakopter but, to answer your question, unfortunately, yes
17:26 pugs_svn r29152 | lwall++ | [fudge] allow a name like rakudo-ng
17:26 diakopter it could've been rakudong
17:26 TimToady rofl
17:27 diakopter I can't find the referent of "bridge building efforts"
17:28 alester diakopter: Oh good, ruin the book for me.
17:29 TimToady diakopter: http://irclog.perlgeek.de/​perl6/2009-11-20#i_1751739
17:30 TimToady then follow the linked link
17:30 qp_pq joined #perl6
17:31 qp_pq hi and sorry to bother you
17:31 qp_pq moritz_: my salutations
17:31 qp_pq moritz_: I very much appreciate you for taking the initiative to write a book
17:31 qp_pq I think the effort is very well spent
17:31 rfordinal left #perl6
17:32 diakopter alester: :(  I'm sorry.  it's an excellent book, even after the first time.
17:32 qp_pq but I do have an OBJECTION
17:33 diakopter qp_pq: objection overruled
17:33 diakopter (kidding) - what is it?
17:33 qp_pq since the syntax of p6 hasn't yet come to a stable form(prove me wrong?) try to keep the code in teh book pseudo-code for now...or you can just link to it outside the book...
17:33 qp_pq like, a book is very good to make the language known and such...
17:33 qp_pq but if the language is not finished
17:33 qp_pq can you expect someone coming(who doesn't know what to expect)
17:34 qp_pq he comes and writes a few lines of Perl6 seems something is broken
17:34 qp_pq (very likely I presume)
17:34 qp_pq he will get the bad idea
17:34 diakopter qp_pq: that's a good idea (each code example has a tinyurl to its updated form online), but the current version should still be printed
17:34 qp_pq so first finish the language, then you can publish a big book on it and everyone will be happy
17:34 qp_pq (including me)
17:34 Tene ew ew tinyurl
17:34 diakopter at least some id.
17:35 Tene and what happens when tinyurl dies?  or if they expire old links?
17:35 diakopter I don't mean _the_ tinyurl
17:35 diakopter I mean the book's site
17:36 diakopter or github. or whatever.
17:36 diakopter a smallish url[-portion]
17:36 Tene qp_pq: Are you saying that nobody should do any work on a book until the language is "finished"?
17:37 Tene qp_pq: also, what language do you know that is "finished" and allows no additional changes?
17:37 TimToady if you print out a book for an alpha-ish chunk of code, it should warn you at the front, and that should be sufficient warning for an attentive reader
17:38 rgrau joined #perl6
17:38 TimToady but part of the bootstrap of Perl 6 is to get people to start using preliminary implementations to flush out the problems
17:38 Tene qp_pq: The book is likely going to discuss Perl 6 as it is when Rakudo Star is released, is my understanding.  The majority of the language implemented so far hasn't changed much, and the changes aren't made for no reason... they're in response to real problems people are encountering.
17:40 TimToady we'll probably go into a design "slush" when rakudo * comes out so that we can converge on a solution for 6.0.0
17:41 TimToady we'll still tweak things after that, but we'll start marking more issues for "post 6.0.0"
17:41 PerlJam convergence++
17:41 diakopter like a slushee machine, but with lava instead of ice
17:41 TimToady right now we're still trying not to fall into the trap of "we already have 11 users, so we can't change anything"
17:41 Tene diakopter++
17:42 alester But once we hit 14 users, we're etching it all in stone!
17:42 diakopter TimToady: but that's 11 in base-999999999
17:43 pugs_svn r29153 | lwall++ | [S03] add infix:<else> to match infix:<also>, suggested by TheDamian++
17:52 stephenlb joined #perl6
17:53 quietfanatic masak: the reason for both the event sub and the Event::Type declaration is that I expect module authors to define and call the types and users of their module to write the handlers.
17:54 quietfanatic Wait he's not here!
17:55 quietfanatic phenny: tell masak The reason for having both the event sub and the Event::Type declaration is that I expect module authors to define and call the types, and users of their module to write the handlers.
17:55 phenny quietfanatic: I'll pass that on when masak is around.
17:56 quietfanatic phenny: tell masak Also the Event::Types will have many more options in the future that control just how handlers are dealt with.
17:56 phenny quietfanatic: I'll pass that on when masak is around.
17:56 pugs_svn r29154 | lwall++ | [STD] implement infix:<else> to see if it causes any problems
18:01 diakopter what's infix:<else>
18:01 TimToady opposite of infix:<also>
18:01 TimToady short-circuit junctional
18:02 quietfanatic Does that make "A if B else C" work as expected?
18:03 TimToady depends on what's expected
18:03 diakopter is "A if B else C" the same as B ?? A !! C
18:03 [particle] A if B | C
18:04 cdarroch joined #perl6
18:04 diakopter o
18:04 [particle] i'm sure it'll cause synaptic dissonance
18:05 TimToady well, we could also s/also/then/ :)
18:05 [particle] oy.
18:05 [particle] save the headaches for monday, will you please?
18:06 TimToady 'mkay
18:06 diakopter 'during' might be more time-appropriate for junction &  than 'then'
18:07 TimToady I don't see how that works at all
18:07 TimToady implies short-circuit the other direction to me
18:07 [particle] meetoo
18:07 quietfanatic like 'while'
18:07 diakopter I must not comprehend "short-circuit junctional"
18:08 TimToady well, here's another play, if we want both &/| and &&/|| semantics, it's &&&/|||
18:08 TimToady sorta like anding 1 and 2  :)
18:08 TimToady er, oring
18:08 TimToady or, erring
18:09 TimToady .oO(I have an O-ring for an earring)
18:09 * [particle] oohs and adds
18:09 chromatic joined #perl6
18:09 [particle] sigh, dvorakian slip
18:09 TimToady doohdadds
18:12 TimToady and then A ??? B !!! C would obviously be junctional too
18:14 ProperNoun joined #perl6
18:14 quietfanatic Let's not claim every possible combination of ascii characters as operators.
18:14 * diakopter missed the boat on the wrong side of the bed this morning
18:17 TimToady I'm with you lock, stock, and sinker... er, hook, line, and barrel....
18:18 TimToady the whole nine yards for small values of 9
18:19 * TimToady thinks he is funny, and is occasionally correct, but not the way he imagines...
18:19 astrojp_ joined #perl6
18:20 diakopter who's on zeroeth
18:21 eiro_ joined #perl6
18:21 TimToady I don't know doesn't know
18:21 TimToady and the manager has sent me to the showers... &
18:24 jan___ joined #perl6
18:30 Su-Shee_ joined #perl6
18:33 KyleHa joined #perl6
18:39 KyleHa TimToady++  # [fudge] allow a name like rakudo-ng
18:51 payload joined #perl6
19:05 TimToady wow, the shower really helped, and I know what to do with short-circuit junctions now.
19:05 perigrin shower them?
19:06 perigrin they probably won't short after they're clean and dry
19:16 FullMetalHarlot joined #perl6
19:19 TimToady to make a long story short-circuited, there's going to be an S metaoperator that suppress
19:19 TimToady suppresses any out-of-order execution
19:19 TimToady either explicit, with junctions or hypers
19:19 TimToady or implicit because the optimizer thought it could get away with it
19:19 KyleHa Nice.
19:20 TimToady S for sequential/serial/synchronous/short-circuit
19:20 TimToady and with relationship to sink/sync
19:20 TimToady not to mention single-threaded
19:23 pmurias TimToady: what's the point of suppresing out-of-order execution made by the optimizer?
19:26 TimToady if you know that the optimizer is going to make a suboptimal decision in the larger scheme of things, like using up a bunch of threads to do something fast that is not going to be used soon.
19:26 TimToady things the optimizer can't know
19:27 TimToady but it also gives us S& and S| for free, not to mention S^ and S«op»
19:27 TimToady the optimizer can't know how safe some evalutions are, because figuring that out might be a halting-ish problem
19:28 TimToady so S is also for "safe"
19:28 KyleHa S for slow?
19:29 TimToady that too :)
19:35 diakopter but how do you distinguish halting-ish problems from halting problems?
19:36 chromatic Check the exit status.
19:37 [particle] S for suboptimal!
19:37 TimToady diakopter: obviously, you check their isomorphic status bits
19:39 diakopter :( it keeps hanging on that check.
19:55 TimToady you forgot to mark the variables as volatile!!!
20:06 takadonet ahh volatile
20:09 TimToady explosive variables make it really easy to determine when it will halt.  It will halt NOW!
20:13 pugs_svn r29155 | lwall++ | [S03] destroy two low-powered, obscure, short-circuiting operators
20:13 pugs_svn r29155 | in favor of one high-powered, obscure-but-transparent, S metaoperator
20:16 pugs_svn r29156 | kyle++ | [t/spec] Test for RT 70664: Try <=> with rationals
20:17 dalek nqp-rx: b387ccf | pmichaud++ |  (2 files):
20:17 dalek nqp-rx: Allow bareword GLOBAL:: package identifier.
20:17 dalek nqp-rx: review: http://github.com/perl6/nqp-rx/commit/b​387ccfc432da46e854f30faa4a6feb41ee17292
20:17 dalek nqp-rx: 2443dcc | pmichaud++ | src/stage0/ (3 files):
20:17 dalek nqp-rx: Update bootstraps.
20:17 dalek nqp-rx: review: http://github.com/perl6/nqp-rx/commit/2​443dcc1d14f30463d21b279d630240d94268550
20:17 qp_pq Tene: latin
20:19 qp_pq TimToady: the people that will be using it will all be testers... an average Perl6 user might require a good baggage of Perl5 and some serious functional programming experience...
20:19 pugs_svn r29157 | lwall++ | [S03] tweaks, add conjectural RSR case
20:20 qp_pq Tene: yes, but about 3 books on Perl6 have already been rendered obsolete so far...
20:21 TimToady and all of them have said, "This is all subject to change".  or should have...
20:23 cognominal joined #perl6
20:23 TimToady the optimal strategy for coming up with a 100-year language does not involve freezing designs that will come back to haunt us later.
20:24 TimToady the process must be convergent, but forcing convergence to rapidly is as suboptimal longterm as forcing it too slowly.
20:24 TimToady *too
20:29 japhb TimToady, What exactly does 'fiddly' mean in the context of infix operators?
20:30 Exodist joined #perl6
20:39 TimToady it is precisely those operators to show up marked as "fiddly" when the metaoperator decides whether to allow it or not.
20:39 pugs_svn r29158 | kyle++ | [t/spec] more rational relational tests
20:39 TimToady most fiddly markings come straight from the precedence table
20:39 TimToady but an operator can declare itself to be fiddly or non-fiddly as well
20:40 astrojp_ left #perl6
20:41 astrojp joined #perl6
20:41 TimToady mostly, it means, "has strange side-effects" such as control flow or modifying its arguments
20:42 TimToady in a similar vein, the % operator declares itself to be 'iffy' so that !% works
20:44 qp_pq can a language like Perl6 have a person saying
20:44 qp_pq "hey, let's get a f*cking plan together and finish this thing ?"
20:44 qp_pq or is this too hostile ?
20:44 qp_pq why doesn't this happen ? does someone always come up with a new idea
20:44 TimToady eh, people do that all the time, only politely
20:44 qp_pq and then effort is dispersed in that direction ?
20:46 TimToady when you say "why can't we all go this direction", I immediately get a picture of the respected doctor in the original Posaidon Adventure
20:46 TimToady who confidently led a large group of people to the bow of the ship, where they all died.  :)
20:47 TimToady *Poseidon
20:50 TimToady so I always listen very hard to the little boys who think the emperor has no clothes, because sometimes they're right.
20:51 KyleHa Let's all go this direction:  http://www.code-muse.com/blog/wp-con​tent/uploads/2007/11/df20021001.jpg
20:51 TimToady if that gives the appearance of herding cats, then so be it
20:51 qp_pq TimToady: just watched the trailer to poseidon adventur
20:51 TimToady cats is all we've got
20:51 qp_pq TimToady: funny as hell http://www.youtube.com/watch?v=dd03qev59Jo
20:51 qp_pq TimToady: thanks for a good laugh
20:52 sjohnson qp_pq: the new one or old one?
20:52 TimToady dunno if the new one matches the old one in plat though, never saw the new one
20:52 QtPlaty[HireMe] TimToady: Sometimes the little boys who say the emperor has no cloaths are boys who cry wolf.  To compleatly mangle a metaphore.
20:52 TimToady *plot
20:52 qp_pq sjohnson: I only saw one trailer(the '70s one)
20:52 sjohnson new one is technologically "newer", ie, cell phones, portable game systems that kids use, etc
20:52 sjohnson and the tables aren't nailed to the floor as in the old one
20:53 sjohnson except for the piano
20:53 sjohnson my uncle when he was a child was given the novel, and he misread it as "the positron adventure".  he was probably a bit disappointed when he started reading the book
20:53 TimToady wow, Blue Angels-ish critters flying overhead
20:54 TimToady it's one of those films I classify as heavily existentialist, along with One Flew Over the Cuckoo's Nest
20:55 qp_pq TimToady: I saw 2012
20:55 qp_pq TimToady: and today 30 days of night
20:55 qp_pq they were both incredible crap
20:56 KyleHa Is 2012 a sequel to 2010?
20:56 TimToady yeah, well, I saw Star Wars when it was just *one* movie, and it was pretty okay. :)
20:57 TimToady pity GL learned the wrong lessons at several points
21:00 perigrin Star Wars before it had a subtitle.
21:01 takadonet left #perl6
21:03 IllvilJa joined #perl6
21:05 sjohnson anyone here like Donnie Darko?
21:05 qp_pq me
21:05 qp_pq it was a very confusated movie
21:06 justatheory joined #perl6
21:06 qp_pq I can honestly say I did not understand anything after watching it
21:06 qp_pq but it was interesting to watch
21:06 sjohnson one reason why i enjoyed it so much
21:09 pmurias qp_pq: freezing the spec would speed up implementing Perl 6
21:09 pmurias * wouldn't
21:10 TimToady you see--we're all conflicted about it :)
21:11 KyleHa std: sub foo($opt?, $req) { say "1 $opt, 2, $req" }
21:11 p6eval std 29158: [31m===[0mSORRY![31m===[0m␤Can't put required parameter after optional parameters at /tmp/uOpdCBaIqM line 1:␤------> [32msub foo($opt?, $req[33m⏏[31m) { say "1 $opt, 2, $req" }[0m␤    expecting any of:␤   constraint␤       parameter␤        trait␤  whitespace␤FAILED 00:01 109m␤
21:14 pugs_svn r29159 | lwall++ | [STD] add S metaoperator, delete also/else
21:14 diakopter std: say 3 [S+] 4
21:14 p6eval std 29158: [31m===[0mSORRY![31m===[0m␤Two terms in a row at /tmp/c7s8Vbc32u line 1:␤------> [32msay 3 [[33m⏏[31mS+] 4[0m␤    expecting any of:␤ infix or meta-infix␤      infix stopper␤    standard stopper␤ terminator␤FAILED 00:01 106m␤
21:15 pmurias TimToady: why do we allow the optimiser choice in marking a routine as hard or soft? won't it lead to portability problems?
21:16 pugs_svn r29160 | lwall++ | [also.t] rewrite to S&, too lazy to rename file
21:16 pmurias with code that uses a routine as a soft one without declaring that breaking using a different (version/configuration) of the optimiser
21:17 pugs_svn r29161 | kyle++ | [t/spec] Test for RT 66822: required param after optional param
21:17 TimToady pmurias: I don't know what you mean by hard/soft
21:18 [particle] hrmm, i wonder how often i'll use the Sand operator
21:19 diakopter Surely more often than you'll use Sbut
21:19 pmurias TimToady: hard - unwrappable
21:20 pmurias TimToady: hard - HardRoutine, soft - SoftRoutine
21:20 [particle] i wonder if we'll talk about operations being Sandy or Sorry
21:22 [particle] (S metaop)++
21:23 TimToady pmurias: it is necessary to give optimizers some room to cheat, or they can never optimize anything, and you either end up with an unoptimizable language, or a B&D language
21:24 TimToady we're looking for a happy medium
21:24 TimToady because neither extreme is happy.
21:25 diakopter std: say 3 [S+] 4
21:25 p6eval std 29161: ok 00:01 106m␤
21:26 TimToady when it comes to optimizers, Don't Care is as important as Yes and No.
21:26 TimToady it's yet another way of writing generic code
21:27 diakopter std: say 3 [SRSRSRSRSRSRSRSRSR+] 4
21:27 p6eval std 29161: ok 00:01 106m␤
21:27 [particle] S+ is what R is based on.
21:27 zamolxes joined #perl6
21:27 [particle] but that's a horse of a different color.
21:36 payload joined #perl6
21:41 diakopter std: [[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[Sa​nd]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]] 4,5,6,7,8
21:41 p6eval std 29161: ok 00:01 111m␤
21:47 diakopter std: *&!*%$$*_!=$_&&$_
21:47 p6eval std 29161: ok 00:01 110m␤
21:48 diakopter rakudo: my $$a = 5; say $$a;
21:48 p6eval rakudo 7347ec: 5␤
21:49 diakopter rakudo: my $$$$$a = 5; say $$$$$a;
21:49 p6eval rakudo 7347ec: 5␤
21:49 diakopter O_O
21:51 diakopter someone, help?
21:51 hcchien joined #perl6
21:52 Su-Shee left #perl6
22:06 sjohnson sdssswqrakudo: my $$$$$a = 5; say $$$$a;
22:06 sjohnson rakudo: my $$$$$a = 5; say $$$$a;
22:06 p6eval rakudo 7347ec: Symbol '$$$$a' not predeclared in <anonymous> (/tmp/cpfJw6xJdz:2)␤in Main (file <unknown>, line <unknown>)␤
22:08 sjohnson curiosity killed the compiler
22:10 meppl joined #perl6
22:11 astrojp left #perl6
22:30 quietfanatic joined #perl6
22:30 diakopter quietfanatic: help
22:30 diakopter rakudo: my $$$$$a = 5; say $$$$$a;
22:31 p6eval rakudo 7347ec: 5␤
22:31 quietfanatic what's up diakopter?
22:31 quietfanatic I'm not terribly surprised
22:31 diakopter std doesn't complain
22:31 diakopter ng: my $$$$$a = 5; say $$$$$a;
22:31 p6eval ng c8de8d: Malformed my at line 1, near "$$$$$a = 5"␤current instr.: 'perl6;HLL;Grammar;panic' pc 448 (src/stage0/HLL-s0.pir:300)␤
22:31 diakopter oh
22:32 diakopter 'tis good
22:32 quietfanatic Apparently variable names can have as many sigils as they want :|
22:32 diakopter fixed in ng I guess
22:32 quietfanatic At least currently
22:48 zaphar_ps joined #perl6
22:59 perigrin_ joined #perl6
23:08 TiMBuS joined #perl6
23:16 solarion joined #perl6
23:34 payload joined #perl6
23:40 payload joined #perl6
23:48 zamolxes joined #perl6
23:50 bluescreen joined #perl6

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

Perl 6 | Reference Documentation | Rakudo