»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
jnthn oh gah, dynpmcs and dynops build into separate libs :/ 00:04
00:05 wamba left
jnthn orka...I fix it tomorrow. 00:08
00:10 skangas left, skangas_ joined, skangas_ is now known as skangas, lue left, cdarroch left, bloonix left, Reaganomicon left, pothos left, dobie_gillis left, mikemol left 00:13 lue joined, cdarroch joined, bloonix joined, Reaganomicon joined, pothos joined, dobie_gillis joined, mikemol joined, yath joined 00:16 tokuhirom joined 00:57 stepnem left 00:58 stepnem joined 01:00 cdarroch left 01:03 mtk left 01:04 woosley joined 01:05 stepnem left 01:08 stepnem joined 01:20 tokuhirom left, tokuhirom joined 01:33 [Coke] left 01:35 [Coke] joined
thou i'm starting to loath socialtext wiki 01:35
i don't think it's -Ofun 01:36
01:37 bluescreen10 left 01:40 icwiener_ joined 01:42 icwiener_ left 01:43 icwiener left 01:54 whiteknight left, woosley left 01:55 woosley joined 02:10 mberends left 02:15 stepnem left 02:20 stepnem joined
thou anyone here know how to turn off rakudo's "readline" functionality, so I can use rlwrap on it? I could swear I saw a blog post or something about this a while back, but can't find it. 02:21
ah, i can recompile and disable it then 02:24
sorear jnthn: ping 02:25
thou somehow 02:26
for the IRC logs: This should work: cd /path/to/src/rakudo; perl Configure.pl --gen-parrot --parrot-option=--without-readline 02:30
02:31 seeif joined 02:32 envi joined
seeif hey 02:33
it's funny
TimToady haha
thou hehe
02:33 stepnem left 02:37 seeif left 02:39 stepnem joined 02:45 lichtkind left 02:56 benabik joined, _jaldhar_ left 02:57 _jaldhar_ joined
colomon pdcawley just asked me if Perl 6 could recognize this sequence: www.ocf.berkeley.edu/~stoll/answer.html 03:02
sorear colomon: Perl 6 is not an AI 03:05
it doesn't recognize things; I don't know what you mean 03:06
colomon sorear: didn't you hear my proposal to have the sequence operator look up sequences on line? ;)
sorear I FORBID.
colomon it does recognize arithmetic and geometric sequences
sorear this is the last thing RJA Sloane needs 03:08
this is an absolutely terrible idea and I will have nothing to do with it 03:09
I hope it was a joke
colomon sorear: it's a joke.
03:11 spq2 joined
sorear sorry. 03:15
03:19 benabik left 03:20 tokuhirom left 03:22 benabik joined 03:25 LaVolta joined 03:26 Su-Shee left 03:27 benabik left 03:28 Su-Shee joined
colomon rakudo 03:28
;
rakudo: say "ttthh" ~~ /(.)\1*/ 03:29
p6eval rakudo 1fcb14: OUTPUT«===SORRY!===␤Unrecognized backslash sequence at line 22, near "*/"␤»
colomon rakudo: say "ttthh" ~~ /(.)$0*/
p6eval rakudo 1fcb14: OUTPUT«ttt␤»
colomon rakudo: say ~"ttthh".comb(/(.)$0*/)
p6eval rakudo 1fcb14: OUTPUT«ttt hh␤»
03:34 spq2 left, mule joined, mule left 03:37 benabik joined
colomon btw 03:37
rakudo: say ~((1, { $_.comb(/(.)$0*/).map({ $_.chars ~ $_.substr(0, 1) }).join } ... *)[^10]) 03:38
p6eval rakudo 1fcb14: OUTPUT«1 11 21 1211 111221 312211 13112221 1113213211 31131211131221 13211311123113112211␤»
colomon is the www.ocf.berkeley.edu/~stoll/answer.html sequence in p6. (bet someone else can make that cleaner...) 03:39
colomon should be in bed.
TimToady I'm pretty sure I've seen that seq on RC 03:40
sorear niecza: say "31131211131221".subst(/(.)$0*/, { chars($/) ~ $0 }, :g) 03:41
p6eval niecza v7: OUTPUT«13211311123113112211␤»
sorear slightly cleaner
not much
oh, and it can be written using *.subst, no need for a block
TimToady rosettacode.org/wiki/Look-and-say_sequence
LaVolta rakudo: my @lns := ('1', *.comb(/(.)$0*/).map({ .chars ~ .substr(0,1) }).join ... *); .say for @lns[^10]; 03:45
p6eval rakudo 1fcb14: OUTPUT«1␤11␤21␤1211␤111221␤312211␤13112221␤1113213211␤31131211131221␤13211311123113112211␤»
LaVolta but...my repl just hangin up there
:-P
03:48 stephanepayrard_ left 03:49 benabik left 03:51 cognominal joined
colomon LaVolta: are you trying to do the my @Ins := blah part by itself? 03:53
TimToady: did you just add that look-and-say? Or did I accidentally make a slightly uglier duplicate of it? 03:54
03:54 molaf joined
TimToady it's been there a while 03:55
colomon RC++
04:01 JimmyZ joined
JimmyZ wonders why planetsix is not updated 04:01
TimToady it's a stuck first article; scroll down to the next one 04:02
sorear the first article is dated July 6 2011. Nothing newer has been posted. 04:03
04:04 benabik joined
JimmyZ it doesn't fetch pmthium.com/feed/ ? 04:07
04:18 satyavvd joined 04:23 satyavvd left, mberends joined 04:24 kaare_ joined 04:26 satyavvd joined 04:27 sorear_ joined 04:28 barika joined 04:31 xinming joined 04:35 khisanth_ joined 04:37 khisanth_ is now known as Khisanth 04:43 lue left 04:49 sorear_ is now known as sorear
JimmyZ what's difference between space and whitespace? 04:53
Eevee maybe your terminal background is black
(kidding)
JimmyZ I ask it because S02 use them 04:54
benabik Whitespace is a category, space is a member of it.
Eevee I assume "space" means the single character and "whitespace" is any printable blank character
04:55 noganex joined
LaVolta maybe use the word 'space bar' should make this clearer 04:55
Eevee that's a key, not a character
LaVolta sorry...what i mean is the char you type use the bar
my bad 04:56
sorear sleep
04:57 noganex_ left
mberends .u space 04:57
phenny U+0020 SPACE ( )
JimmyZ .u whitespace 05:00
phenny JimmyZ: Sorry, no results for 'whitespace'.
mberends whitespace can also be a sequence of non printing characters. Of course Unicode goes to town of such a topic en.wikipedia.org/wiki/Space_%28punc...in_Unicode 05:11
JimmyZ I see 05:16
phenny: tell LaVolta test 05:17
phenny JimmyZ: I'll pass that on when LaVolta is around.
05:17 koban joined 05:22 birdwindupbird joined 05:23 molaf left
dalek : 83bc2a9 | jimmy++ | docs/zh-cn/syn/S02-bits.pod:
reset author
 05:30
: aed13e3 | jimmy++ | misc/dalek-conf.json:
Merge branch 'master' of github.com:perl6/mu
JimmyZ ops, wrong commit message
should be 'revise wordings' 05:31
05:46 mberends left
moritz nom: class G { ... }; class G { } 05:48
p6eval nom: ( no output )
moritz nom: use Test; eval_lives_ok 'class G { ... }; class G { }', 'can redeclare stub classes';
p6eval nom: OUTPUT«not ok 1 - can redeclare stub classes␤»
moritz nom: use Test; eval_lives_ok 'class G { ... }; class G { }', 'can redeclare stub classes'; say $!
p6eval nom: OUTPUT«not ok 1 - can redeclare stub classes␤Null PMC access in find_method('gist')␤current instr.: 'say' pc 222615 (src/gen/CORE.setting.pir:58683) (:2341)␤»
moritz nom: use Test; eval 'class G { ... }; class G { }', say $! 05:49
p6eval nom: OUTPUT«Null PMC access in find_method('gist')␤current instr.: 'say' pc 222615 (src/gen/CORE.setting.pir:58683) (:2341)␤»
dalek : 772adf2 | jimmy++ | docs/zh-cn/syn/S02-bits.pod:
remove plural from chinese
moritz nom: $_ = "Hellow World"; my $x = .uc; 05:50
p6eval nom: OUTPUT«Rebuild in progress␤»
dalek albot: 2bc5c05 | moritz++ | build-scripts/rebuild-nom.sh:
optimize nom rebuild a bit
 05:54
moritz nom: $_ = "Hellow World"; my $x = .uc;
p6eval nom: OUTPUT«Segmentation fault␤»
moritz nom: $_ = "Hellow World"; my $x = .uc;
p6eval nom: OUTPUT«Segmentation fault␤»
moritz nom: $_ = "Hellow World"; .say 05:55
p6eval nom: OUTPUT«Segmentation fault␤»
JimmyZ didn't update parrot?
moritz seems that automagic access to $_ isn't all that robust yet
06:02 wamba joined 06:07 kfo joined
moritz nom: $_ := "Hellow World"; .say 06:08
p6eval nom: OUTPUT«Cannot use bind operator with this LHS at line 1, near " \"Hellow W"␤current instr.: 'nqp;HLL;Grammar;panic' pc 23508 (src/stage2/gen/NQPHLL.pir:6300)␤»
moritz nom: say List ~~ Positional 06:10
p6eval nom: OUTPUT«No method cache and no find_method method in meta-object␤current instr.: '_block1002' pc 28 ((file unknown):0) (:1)␤»
06:10 kfo_ left 06:12 kfo left
dalek kudo/nom: f4cf99c | moritz++ | / (2 files):
four more passing test files, note a fail in NOMMAP
 06:13
06:15 _jaldhar_ left 06:16 _jaldhar_ joined 06:22 koban left 06:23 arnsholt left 06:25 kanishka joined 06:27 arnsholt joined 06:29 Mowah joined 06:31 wtw joined 06:34 wamba left, wamba joined, daxim joined 06:52 kanishka left 07:08 kanishka joined
dalek : a2cc05a | jimmy++ | docs/zh-cn/syn/S02-bits.pod:
revise omissive wordings
 07:18
07:25 kanishka left 07:37 mj41 joined 07:40 kanishka joined 07:41 arnsholt left 07:46 Chillance left, arnsholt joined 07:51 JimmyZ_ joined 07:55 JimmyZ left, JimmyZ_ is now known as JimmyZ 07:57 daniel-s joined, arnsholt left 08:02 bkolera joined, arnsholt joined 08:04 mberends joined 08:08 bkolera left 08:26 kanishka left 08:29 thou left, koban joined 08:30 birdwindupbird left, bkolera joined 08:31 birdwindupbird joined 08:40 dakkar joined 08:46 sjn is now known as SJN 08:47 cooper left 08:56 tzhs joined
moritz developer.android.com/reference/and...Throwable) 08:58
the wtf, "What a Terrible Failure" log method :-)
09:05 huf joined 09:23 SJN is now known as sjn 09:27 woosley left
JimmyZ what's the wtf? :) 09:41
daniel-s lol
09:41 donri joined 09:44 MayDaniel joined, MayDaniel left, MayDaniel joined 09:49 MayDaniel left
jnthn moritz: ($_) see a few tasks down in nommap :) 09:53
moritz jnthn: ah, I was too eager 09:54
jnthn: I guess it's what makes many of those spectests fail
jnthn moritz: Yeah, I started fixing it last night, but ETOOSLEEPY 09:56
09:57 bkolera left
moritz is really happy with nom progress 10:02
10:03 JimmyZ left 10:05 JimmyZ joined 10:15 LaVolta left 10:20 jesk is now known as JESK 10:21 _jaldhar_ left, _jaldhar_ joined 10:27 timbunce joined 10:28 kaare_ left 10:51 lichtkind joined
flussence rakudo: say 10.fmt('%d %1$x'); 10:55
p6eval rakudo 1fcb14: OUTPUT«Not enough arguments supplied for the given format string␤ in 'Cool::sprintf' at line 2779:src/gen/core.pm␤ in 'sprintf' at line 2813:src/gen/core.pm␤ in 'Cool::fmt' at line 2601:src/gen/core.pm␤ in main program body at line 22:/tmp/m4nyPZtr3l␤»
flussence rakudo: say sprintf('%d %1$x', 10); 10:56
p6eval rakudo 1fcb14: OUTPUT«Not enough arguments supplied for the given format string␤ in 'Cool::sprintf' at line 2779:src/gen/core.pm␤ in 'sprintf' at line 2813:src/gen/core.pm␤ in main program body at line 22:/tmp/_84xx89ynV␤»
flussence hm
.oO( one of these days I'm gonna fix sprintf.t... )
 10:57
pmichaud what should that do, ooc? 10:58
flussence same as perl5/C, 1$ reuses the first arg
pmichaud okay.
flussence S32/Str:420 mentions it 10:59
(I wonder if I can take that grammar and use it for coverage testing...) 11:00
dalek kudo/nom: 2630e16 | pmichaud++ | src/core/ (4 files):
Improve handling of .reify(*) and infinite lists.
 11:11
kudo/nom: b1d9956 | pmichaud++ | src/core/ (4 files):
Implement array slices, including autotrimming on infinite indexes.
kudo/nom: c3d66c7 | pmichaud++ | src/core/ (2 files):
Add WhateverCode slicing to postcircumfix:<[ ]>.
11:12 molaf joined
jnthn \o/ pmichaud++ 11:12
pmichaud nom: say 'building'? 11:13
p6eval nom: OUTPUT«Confused at line 1, near "say 'build"␤current instr.: 'nqp;HLL;Grammar;panic' pc 23508 (src/stage2/gen/NQPHLL.pir:6300)␤»
pmichaud nom: my @a = 1..50; say @a[10..*]; # no longer infinite heat death
p6eval nom: OUTPUT«ListIter<36544736>␤»
pmichaud well, once it updates :)
11:14 tokuhirom joined
pmichaud nom: my @a = 1..50; say @a[10..*]; # no longer infinite heat death yet? 11:15
p6eval nom: OUTPUT«ListIter<49426656>␤»
pmichaud evalbot rebuild nom
p6eval OK (started asyncronously)
11:16 kanishka joined 11:18 icwiener joined, mtk joined 11:19 mtk left, mtk joined 11:22 DarthGandalf left
pmichaud nom: my @a = 1..50; say @a[10..*]; # no longer infinite heat death yet? 11:22
p6eval nom: OUTPUT«11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50␤»
pmichaud \o/
my @a = 0..50; say @a[1,4,7...*] 11:23
nom: my @a = 0..50; say @a[1,4,7...*]
p6eval nom: OUTPUT«1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49␤»
pmichaud nom: my @a = 'a'...'z'; say @a[1,4,7...*] 11:24
p6eval nom: OUTPUT«b e h k n q t w z␤»
pmichaud wfm
oh, and also: 11:25
nom: my @a = 'a'...'z'; say @a[*-2]
p6eval nom: OUTPUT«y␤»
11:25 Katibe left
moritz nom: my @a = 1..*; say @a[2..*].[3] 11:25
p6eval nom: OUTPUT«(timeout)» 11:26
pmichaud well, if you *want* it to blow up :)
moritz :-)
pmichaud array slices aren't lazy (yet?)
moritz plays masak
pmichaud I'm not sure they should be, given that a slice is supposed to return a Parcel
11:26 REPLeffect left
moritz and parcels are supposed to be eager? 11:27
pmichaud well, immutable. the slice should be a parcel of the individual elements sliced, I think.
so that implies some level of eagerness.
11:28 Katibe joined
moritz aren't argument lists passed as parcels (at least notionally)? 11:28
pmichaud sure
a parcel can contain lazy elements
moritz hm, ok
pmichaud I'm just not sure that a slice should have lazy elements in it.
moritz waves his hands vaguely and says "just make it work" :-) 11:29
11:30 REPLeffect joined 11:32 molaf left 11:33 DarthGandalf joined
pmichaud anyway, I can do lazy slices of infinite lists, if we decide that's needed. 11:35
time for a nap -- bbl
11:38 kanishka left, kaare_ joined 11:43 kfo joined 11:47 am0c joined, SHODAN joined 11:51 satyavvd left 11:54 cognominal_ joined 11:58 cognominal left 12:02 jedai_ joined 12:05 jedai left 12:11 orafu left 12:13 orafu joined 12:23 bluescreen10 joined 12:24 Katibe left 12:27 Katibe joined 12:30 dalek left 12:31 Katibe left
lichtkind dalek is dead!!! 12:33
12:34 Katibe joined
Trashlord oh no!!! 12:34
moritz ... long live dalek!
mberends he wasn't dead, he was just resting 12:35
12:36 p6eval left
lichtkind he got exterminated :) 12:36
12:36 Katibe left 12:38 Katibe joined 12:54 geekosaur left, benabik left 12:57 geekosaur joined 13:01 xinming left 13:02 xinming joined 13:04 wolfram_ joined, dalek joined, ChanServ sets mode: +v dalek 13:05 wolfram_ left, Holy_Cow joined 13:06 Holy_Cow left 13:09 dalek left 13:26 dalek joined, ChanServ sets mode: +v dalek 13:29 p6eval joined, ChanServ sets mode: +v p6eval 13:36 woosley joined
JimmyZ \o jnthn 13:37
moritz long live dalek! long live p6eval! 13:38
jnthn o/ JimmyZ 13:44
JimmyZ: Flying in 5 hours or so. :) 13:45
JimmyZ Did Mr Xu contact you? 13:46
jnthn JimmyZ: Yes, and Qiang also. :) 13:47
JimmyZ Oh, OK 13:48
JimmyZ doesn't know Qiang 13:49
moritz nom: sub f { fail 'foo' }; say f.WHAT 13:50
p6eval nom: OUTPUT«Failure()␤»
moritz nom: sub f { fail 'foo' }; say f.^methods
p6eval nom: OUTPUT«new BUILD defined Bool Int Num Str eager flat list elems infinite grep join map min postcircumfix:<[ ]> postcircumfix:<{ }> new ACCEPTS WHERE Bool defined CREATE bless Str Stringy item say gist perl DUMP DUMP-ID isa does dispatch:<var> dispatch:<::> dispatch:<.^> dispatch:<.…
moritz nom: sub f { fail 'foo' }; say f.^methods(:local) 13:51
p6eval nom: OUTPUT«new BUILD defined Bool Int Num Str␤»
jnthn Introspekshun. We has it. :)
moritz nom: say Failure.defined
p6eval nom: OUTPUT«Cannot access attributes in a type object␤current instr.: 'defined' pc 260475 (src/gen/CORE.setting.pir:75972) (:2705)␤»
JimmyZ that's great 13:52
moritz I guess a whole lot of methods actually want :D
on the invocant
I kinda think that should be the default 13:53
jnthn moritz: Note that :D will _not_ call .defined
moritz jnthn: yes, that's kinda the point :-)
jnthn moritz: Thus why we probably chould rename it to :C (concrete)
moritz: oh, I see what you're getting at
moritz jnthn: nonetheless literal Failure.defined shouldn't fail like it does now
jnthn :)
yeah 13:54
moritz nom: say Any.^methods(:local)
p6eval nom: OUTPUT«eager flat list elems infinite grep join map min postcircumfix:<[ ]> postcircumfix:<{ }>␤»
moritz nom: say Mu.^methods(:local)
p6eval nom: OUTPUT«new ACCEPTS WHERE Bool defined CREATE bless Str Stringy item say gist perl DUMP DUMP-ID isa does dispatch:<var> dispatch:<::> dispatch:<.^> dispatch:<.=> dispatch:<.?> dispatch:<.+> dispatch:<.*>␤»
JimmyZ nom: (1..*).eager.say 13:55
p6eval nom: OUTPUT«...␤»
moritz jnthn: should method defined() in Mu be a multi (+ has a proto), and then Failure.defined just adds a Failure:D multi?
jnthn moritz: That'd make sense.
13:55 Moukeddar joined
JimmyZ nom: (1...*).say 13:56
p6eval nom: OUTPUT«... ␤»
JimmyZ nom: (1..*).say
p6eval nom: OUTPUT«1..Inf␤»
moritz it's kinda disappointing that (1...*) does not include the 1
though I see why that happens 13:57
13:57 Trashlord left
JimmyZ I think that's a bug 13:58
jnthn It's better than a hang, I guess... 14:01
14:01 colomon left
jnthn But could be more awesome. 14:02
moritz ouch
my patch makes parrot out-of-mem during compilation
... and I have 4GB on that machine, no ulimit 14:03
ah no
moritz should read error messages
Zero-byte allocation not allowed in mem_sys_allocate_zeroedParrot VM: PANIC: Out of mem!
C file src/gc/alloc_memory.c, line 100
jnthn Whoa.
jnthn wonders how that was triggered.
Patch? 14:04
moritz jnthn: moritz.faui2k3.org/tmp/foobar.patch
mberends passes moritz a zero-length array of zero-filled bytes
14:05 colomon joined
colomon is in Ingy's talk (yapc::na) 14:08
moritz what's the talk about?
colomon Acmeism
moritz some crazy new module? 14:09
daxim worse
colomon "Postmodern Module Packaging"
14:09 PacoLinux joined
jnthn moritz: Oddness, it looks like you add proto method Num(|$) {*} 14:10
twice
where once should be proto method Str
But I'm also confused why it's not complaining about that.
moritz ah, I see my error
colomon moritz: cdent.org/ -- supports Perl 6 but not Perl so far?!!!
moritz colomon: Perl is no. 1 on that last 14:11
colomon moritz: for output
moritz: look at the third paragraph
moritz ah 14:12
probably because p5's OO is so quirky
jnthn: even if I s/Num/Str/ in the second proto, I get the same error
jnthn moritz: Any chance you can catch it under the debugger? 14:13
moritz: And get a (c is fine) stack trace?
moritz tries 14:14
urks
jnthn moritz: Is there anywhere that you have a proto with no multis to go with it?
moritz it runs to an exit()
jnthn though, I thought I fixed taht one. 14:15
moritz so I don't get a bt easily :/
not that I can see
14:18 benabik joined
moritz builds a debug parrot 14:18
jnthn moritz: OK. Then I suspect either that helps, or we're down to trying to golf. 14:20
jnthn packing at the moment, though :)
14:26 odrm left
moritz waits for the nqp build, and for new MoR chapters 14:29
14:33 zby_home_ joined 14:34 Mowah left
moritz jnthn: moritz.faui2k3.org/tmp/backtrace.txt 14:35
that does look like a multi without candidates 14:36
jnthn moritz: yeah, darn, thought I fixed that issue
moritz nom: say Failure ~~ Mu
p6eval nom: OUTPUT«Bool::True␤»
jnthn moritz: From further down the stack trace, it's that somewhere there's a proto for defined with no candidates, it seems 14:37
14:38 JimmyZ_ joined
moritz jnthn: maybe the MOP uses defined somewhere while installing the candidate for defined() ? 14:38
jnthn moritz: Hm, maybe, though doesn't seem too likely 14:39
14:40 JimmyZ left, JimmyZ_ is now known as JimmyZ 14:41 mberends left
flussence
.oO( oh no, I've finally written perl5 code so horrible even vim can't highlight it right... )
 14:42
jnthn moritz: Patching the zero-byte, that'll give a better idea.
moritz fwiw 1) the changes in Failure.pm are irrelevant for triggering the bug 2) it only complains if I apply the changes to .defined. .Bool is fine as proto+multi 14:43
3) compiling the setting on a debug-build parrot is dog slow 14:45
14:45 koban left 14:46 tokuhir__ joined
jnthn Maybe "a debug-build parrot is dog slow" is also generally true :) 14:47
benabik jnthn: No maybes about it.
14:50 tokuhirom left 14:53 kanishka joined 14:54 tokuhir__ left 14:55 tokuhirom joined
pmichaud good morning, #perl6 14:55
14:55 woosley left
benabik o/ pmichaud 14:56
moritz good am, pm
slavik and inbetweens? 14:57
14:58 daxim left
pmichaud ...we could also get defined to work without having to make it a multi. 14:58
...we could also leave it as non-multi in Mu, and just make it multi in Failure
jnthn moritz: Patched the zero-alloc bug at lesat
moritz that's what I'm doing as a workaround right now
dalek kudo/nom: be2aea9 | jonathan++ | src/binder/multidispatch.c:
Avoid zero memory allocations for protos with no multi candidates.
 14:59
moritz (not using a multi at all)
jnthn Would be curious to know what's going on.
Or, what it is that calls defined
dalek kudo/nom: 2c86d40 | moritz++ | src/core/ (2 files):
fix Failure methods on the type object
15:01 thou joined
pmichaud nom: say Mu.Int 15:02
p6eval nom: OUTPUT«Method 'Int' not found for invocant of class 'Mu'␤current instr.: '_block1002' pc 75 ((file unknown):32012038) (:1)␤»
pmichaud nom: say Any.Int
p6eval nom: OUTPUT«Method 'Int' not found for invocant of class 'Any'␤current instr.: '_block1002' pc 75 ((file unknown):70) (:1)␤»
pmichaud nom: say Cool.Int
p6eval nom: OUTPUT«Method 'Int' not found for invocant of class 'Cool'␤current instr.: '_block1002' pc 75 ((file unknown):51672246) (:1)␤»
pmichaud I don't think we need the multi/proto on Failure.Int/Num/Str 15:03
thou morning, #perl6!
moritz nom: Failure.Int
p6eval nom: OUTPUT«Cannot access attributes in a type object␤current instr.: 'Int' pc 260595 (src/gen/CORE.setting.pir:76051) (:2711)␤»
moritz that is an internal error
pmichaud just write
moritz and should never be visible for the user
jnthn: moritz.faui2k3.org/tmp/backtrace2.txt
pmichaud just eliminate "multi" and leave the rest the same. 15:04
moritz ok
pmichaud method Int(Failure:D:) { $!handled ?? 0 !! $!exception.rethrow; }
moritz what about .Bool? same?
rakudo: say Int.Bool
jnthn moritz: ah
p6eval rakudo 1fcb14: OUTPUT«Bool::False␤»
pmichaud Bool has to be defined on all objects
jnthn moritz: Yeah, that explains a lot :)
It does indeed call .defined 15:05
moritz \o/ I was right, even though I have no clue about the actual code :-)
jnthn I was wrong, even though I wrote the actual code. :-P
moritz pmichaud: will do, after commute
pmichaud moritz++
15:07 spq2 joined
jnthn tries a fix to MultiMethodContainer 15:07
So we can make defined a multi if we wish
15:08 Moukeddar left 15:09 tzhs left, dolmen joined
tadzik exam passed! 15:10
benabik \o/ tadzik
jnthn tadzik++ \o/ 15:11
dalek kudo/nom: f6d775d | jonathan++ | src/Perl6/Metamodel/MultiMethodContainer.pm:
Don't rely on .defined in MultiMethodContainer, or we run into fun when adding a multi defined in Mu. :-)
benabik Oh yes. I forgot the ++. tadzik++
pmichaud fwiw, unless we make :D the default, users are going to see "Cannot access attributes in a type object" on a lot of methods, both for core and non-core classes. 15:14
jnthn didn't really consider it an internal error, fwiw 15:15
pmichaud in the core we can add :D to all of the methods... but if we have to do that it starts to feel like :D ought to be the default.
right, I don't think of it as an "internal error" either.
jnthn Is there a way we can phrase it that makes it more helpful?
pmichaud not coming up with one at the moment 15:19
jnthn That error at least feels quite searchable for answers :)
jnthn does final bit of travel prep 15:20
JimmyZ jnthn++ 15:21
pmichaud jnthn++ # yesterday's nom blog posting
JimmyZ will go to sleep
pmichaud jnthn: hope I didn't steal your thunder with my blog post :)
JimmyZ didn't see any jnthn++ and pmichaud++ blog posting from planetsix.perl.org 15:22
15:22 timbunce left
pmichaud JimmyZ: they're there now... just need to scroll past masak's post 15:23
masak++ has discovered a way to keep top billing for a month or more :)
JimmyZ oh 15:24
15:27 leprevost joined 15:32 am0c left, ingyfoo is now known as ingy, zby_home___ joined 15:33 zby_home_ left
ingy colomon: Perl 6 is a much more regular syntax than Perl 5 15:33
colomon: also native OO
colomon ingy: still, it really nice to see Perl 6 used for something cool like that 15:34
nice talk, btw
ingy :)
ta! 15:35
arnsholt wonders what colomon and ingy are talking about
colomon arnsholt: cdent.org/ 15:36
arnsholt Oh right, I remember that coming up in the past 15:39
15:40 Holy_Cow joined 15:45 am0c joined
ingy arnsholt: cdent.org/examples/hello-world/ 15:50
that's pretty much the full functionality at this point... :/
jnthn remembers just in time to replace the fuse in the one and only travel adapter he owns that'll let him plug his laptop in in China. :) 15:51
ingy but I have plans to add assignment, conditional and loops in the next month
dukeleto ingy: i would really like to see that 15:52
ingy dukeleto is everywhere
dukeleto ingy: that is because I can travel ..... without moving. 15:53
ingy well then I could implement CPAN/PyPI/* package
dukeleto ponders writing Spice.pm
ingy dukeleto: do you have pm6 modules yet? 15:54
I forget where they live
dukeleto ingy: i am ashamed to say no. I need to fix that at YAPC::NA
ingy and I actually have some
dukeleto ingy: modules.perl6.org ?
15:54 spq2 left 15:56 JimmyZ left 16:00 colomon left 16:01 tokuhir__ joined 16:05 tokuhirom left 16:08 colomon joined, benabik left 16:09 yves left 16:11 colomon_ joined, colomon left, colomon_ is now known as colomon 16:12 tokuhir__ left, tokuhirom joined 16:15 molaf joined, molaf left 16:16 yves joined 16:18 mj41 left
jnthn airport, bye bye o/ 16:21
colomon \o
pmichaud jnthn: good travels! 16:23
jnthn thanks! back in 15ish hours, provided hotel wifi is good ;) 16:25
pmichaud ...from behind the great firewall? ;-)
this could be interesting. :)
16:27 am0c left 16:31 Trashlord joined 16:40 tokuhir__ joined 16:43 tokuhirom left 16:46 cdarroch joined, cdarroch left, cdarroch joined 16:50 dolmen left 16:52 birdwindupbird left 17:03 ashleydev joined 17:04 wolfman2000 left
sorear good * #perl6 17:05
colomon o/
sorear pmichaud: JimmyZ has no issues with IRC behind "GFW"
17:05 dakkar left 17:09 molaf joined 17:14 masak joined
masak ohi #p6 17:14
colomon \o
17:16 nsh_ joined 17:17 nsh left, DarthGandalf left 17:19 colomon left 17:20 DarthGandalf joined 17:21 benabik joined 17:22 colomon joined
sorear masak: hi 17:22
17:24 tokuhir__ left, tokuhirom joined, bluescreen10 left 17:25 bluescreen10 joined
masak sorear: have you seen Perlude? search.cpan.org/~marcc/Perlude-0.42...Perlude.pm 17:26
it was rewritten last weekend, so that's officially an Old Version, I guess. newest one is here: github.com/eiro/p5-perlude
sorear never heard of it. cute. 17:29
17:29 cooper joined
moritz sorear: 'make test' fails for me in niecza 17:29
17:29 cooper left, cooper joined, cooper left
moritz mono run/Niecza.exe -c test.pl 17:30
Unhandled exception: Unable to resolve method shift in class Str at line 0 (ExitRunloop @ 0) at /home/moritz/p6/niecza/src/NAMOutput.pm6 line 0 (NAMOutput C19_ANON @ 0) at line 0 (KERNEL map @ 2)
... plus lots for of backtres
*backtrace
I get the same error for the spectests too
masak "backtrès" could be a portmonteau for "a lot of backtrace" :) 17:36
colomon moritz: I was getting that too. git clone from scratch and rebuilding cured it. :) 17:37
moritz will do a 'git clean -xdf' instead 17:38
17:44 colomon left 17:48 colomon joined 17:50 Rotwang joined
Rotwang hello 17:50
diakopter howdy
cognominal_ masak++ # but one says portmanteau in English (from the french porte-manteau) or mot-valise in French.
17:50 Holy_Cow left 17:52 cooper joined, cooper left 17:54 cooper joined
masak cognominal_: ah; thanks 17:57
dalek kudo/nom: 0ab6112 | pmichaud++ | src/core/ (3 files):
Add hash slicing, .[*], and .{*}.
 17:59
colomon TimToady: why have the hash produce an index into the packed int value array? wouldn't it be more efficient to just have the hash produce the int value itself? 18:02
TimToady troo
colomon (discussing current YAPC talk, sorry everyone else)
TimToady was thinking of the frozen keys to mutable int array 18:03
colomon you could still have mutable ints as values, no?
just no related int array anywhere
TimToady not if you've calculated a perfect hash function with an immutable key -> int mapping 18:04
colomon (maybe not, I admit it's been a long time since I looked at perfect hash code)
fair enough
TimToady you'd have to recalculate the hash function every time you muted :)
eiro hello world 18:05
benabik Hash shouldn't depend on value, just key.
colomon so, p6 enums could be implemented with a perfect hash and immutable key, no?
TimToady that's how they're defined
eiro is there a keyword associated to =<> in perl6 ?
nine I got the OK from my university to do some work on rakudo for my bachelor paper, so now I need to get to the details of what I'm gonna do. Someone suggested S17 as needing some attention, but I'm not sure what exactly I could do and how to go on about it. 18:06
benabik nine++
18:08 Chillance joined, dukeleto left 18:09 dukeleto joined
colomon eiro: what are you trying to do? 18:10
eiro: or perhaps, do you mean <=> ?
TimToady std: $_ = <>; 18:12
eiro colomon, i know that perl6 have some operators that can be called by function names
p6eval std 37a0cdd: OUTPUT«===SORRY!===␤Unsupported use of <>; in Perl 6 please use lines() to read input,␤ or ('') to represent the null string,␤ or () to represent Nil at /tmp/4vdsDesXjX line 1:␤------> $_ = <⏏>;␤Parse failed␤FAILED 00:01 120m␤»…
TimToady eiro, is that what you mean? ^^
or are you asking about infix:<=> ?
eiro no ...
colomon eiro: all perl6 operators have a function name: for instance, you can call $a + $b using infix:<+>($a, $b) 18:13
eiro sorry guys: i have to introduce the pb:
colomon rakudo: say infix:<+>(4, 5)
p6eval rakudo 1fcb14: OUTPUT«9␤»
eiro we wrote Perlude (some kind of functionnal prelude in perl) and we have implementations of some lazy perl6 operators 18:14
but we can't overload anything: everything is functions
nine Can someone help me getting started with defining the topic for my paper and how to start developing?
eiro we decided that when a function exists in perl6 or haskell, we steal the name 18:15
i'm pretty sure that long time ago, i saw a perl6 function that takes a filename and returns a lazy list
colomon eiro: lines
TimToady except lines doesn't take a filename anymore 18:16
eiro cool! so we made a good choice :)
TimToady just an IO
colomon TimToady: lines($filename) doesn't work?
TimToady not supposed to
eiro TimToady, so do you have a f :: Path -> IO ?
benabik $filename.IO.lines?
TimToady lines($text) splits the text into lines
colomon we need to get stuff sorted so we can actually implement it
rakudo 18:17
;
TimToady nine: someone can help with that
colomon rakudo: say lines("foo")
p6eval rakudo 1fcb14: OUTPUT«Operation not permitted in safe mode␤ in 'Safe::forbidden' at line 2:/tmp/Vj_2VddG3U␤ in 'lines' at line 5932:src/gen/core.pm␤ in main program body at line 22:/tmp/Vj_2VddG3U␤»
TimToady awwaiid just did a talk on similar topics in p5
colomon rakudo: say "foo".lines()
p6eval rakudo 1fcb14: OUTPUT«foo␤»
eiro ooohhh .... so we have a pb there: this would be a big difference between perlude and perl6 lines
colomon back soon
18:17 colomon left
TimToady rakudo: say "foo\nbar\n".lines[1] 18:18
p6eval rakudo 1fcb14: OUTPUT«bar␤»
TimToady rakudo: say lines("foo\nbar\n")[0]
p6eval rakudo 1fcb14: OUTPUT«Operation not permitted in safe mode␤ in 'Safe::forbidden' at line 2:/tmp/T5pL75QG0r␤ in 'lines' at line 5932:src/gen/core.pm␤ in main program body at line 22:/tmp/T5pL75QG0r␤»
TimToady rakudo has .lines right but lines() wrong
pmichaud rakudo is still under the old interpretation, I think.
spectests welcome :)
eiro well ... TimToady this is a method for strings when "our" lines is more than a keyword 18:19
nine TimToady: any idea who that someone might be? :)
eiro tests lines in perl6
TimToady a lot of us are embedded in YAPC::NA at the moment, so a bit distracted 18:20
18:20 TSa joined, colomon joined
TimToady but as I say, awwaiid has been thinking about these things already, and others of us have a few ideas too 18:21
18:21 mberends joined
nine TimToady: ah, didn't recognize this as an answer for me. So I'll just ask him. 18:22
eiro yapc::eu is comming. i'm really excited to give my first talk there
masak \o/ 18:25
eiro \o/ masak
fpw'11 was awesome! the best conference i attended to. ever! 18:26
masak it was very good. 18:29
hard to say why, really :)
pmichaud #phasers in 30 18:30
colomon pmichaud++
pmichaud Keep those cards and letters coming in, folks. :-)
eiro haha .. my guess is: funny people, funny talks, funny ideas, funny hackathlon :) 18:32
masak, are you in contact with kivutar about the panda binding ? 18:33
it would be great to promote perl6 as a game dev. vm
[Coke] can rakudo/nom be built with parrot/master ? 18:35
pmichaud should be able to do so, yes.
perl Configure.pl --gen-parrot=master
18:37 Moukeddar joined
[Coke] Note: checking out '2011.06-7-g342cd53'. 18:38
?
pmichaud that seems wrongish.
pmichaud tries.
[Coke] big warning that I'm now in a detached HEAD state. 18:39
pmichaud that happens frequently, yes.
although it probably shouldn't with --gen-parrot=master
trying it locally
benabik Parrot master is d099f
[Coke] looks like that warning was for NQP - is it checking that out too?
pmichaud NQP is supposed to honor the --gen-parrot flag also. 18:40
[Coke] (and does it need a version of nqp outside of what's in parrot?)
pmichaud nom doesn't use nqp-rx
parrot comes with nqp-rx, not nqp.
[Coke] so, yes? ;)
ok.
pmichaud so, yes.
:)
--gen-parrot=master worked for me 18:42
maybe try removing any parrot/ subdirectory and let it do a fresh checkout 18:43
I'll see if I can reproduce the error your getting
PerlJam [Coke]: did you already have a parrot clone in your rakudo dir? Possibly on a different branch?
[Coke] PerlJam: possibly? I almost never build rakudo that way, but haven't touched this checkout dir in ages. 18:44
PerlJam or a nqp clone I guess if that's what gave the error
18:44 bitpart joined
[Coke] ah, no, becaues it took 30s or more to git clone the thing. 18:44
first time building the nom branch, so definitely not an old nqp hanging about. 18:45
pmichaud nom builds parrot for nqp, it doesn't use nqp's Configure to do it
PerlJam [Coke]: AFAIK, there is no logic to handle the case that there's already a parrot or nqp clone that's not in the right state
18:45 bluescreen10 left
pmichaud yes, there is. 18:45
nom's Configure.pl checks the versions of parrot/nqp to make sure they're correct, _especially_ if =master is given.
18:45 bluescreen10 joined
moritz just svn checkouts of parrot might cause confusion 18:46
pmichaud nom doesn't use svn at all.
moritz but parrot did, back in the days 18:47
pmichaud yeah, --gen-parrot=master appears to be working just fine for me 18:48
even when there's another parrot checkout already in parrot/
[Coke] well, I'm not going to muck with either the parrot or nqp clones in this dir, no worries. 18:49
was just a scary message.
"make test" expected to fail?
PerlJam pmichaud: Hrm. I just did one and it said ... gist.github.com/8645068f1c8d186ad2ed 18:50
tadzik [Coke]: yeah, some of it
[Coke]: try spectest
pmichaud yes, we still have 'make test' fails
PerlJam pmichaud: that last line tells me that something didn't work quite right.
pmichaud (still waiting for the gist to appear) 18:51
[Coke] (rant. I am still not over changes in the summary output of make test from like 5 years ago.)
pmichaud PerlJam: ah yes, there does seem to be a bug with not doing a git pull or fetch there.
[Coke] nopaste? 18:52
PerlJam pmichaud: I've also had a weird behavior when the parrot clone is on another branch, but I don't remember the specifics nor can I reproduce it readily from memory. (I know that doesn't help, but maybe something to watch out for) 18:53
pmichaud I think it's a similar issue. 18:54
okay, I'll work on that.
[Coke] did make spectest got a page long git message about asking to pull without saying what to pull. I assume that's a previous clone issue.
pmichaud yes
likely
you may need to hop into parrot and manually do a git pull
(if you didn't clear parrot/ earlier)
[Coke] (tihs is my home desktop, which I've been unable to ssh into for some months now. things are alittle dusty in here.) 18:55
PerlJam Is there a "clean" target that nukes parrot and nqp from rakudo?
pmichaud rm -rf parrot nqp
PerlJam other than that :)
"make reallyreallyclean" or something
[Coke] ... that spectest run was incredibly faster than I remember.
pmichaud you're only running about 10% of the tests of master
tadzik look how fast nom is! :) 18:56
pmichaud otoh, nom is a lot faster, yes.
(i.e,. the tests run faster)
[Coke] no point in mucking with TODO passes until nom is master, aye?
PerlJam [Coke]: unless you can make them untodo in nom :)
pmichaud I'm thinking we could start mucking, and that we go ahead and peg master to the current spectest suite (and not worry about keeping it up to date)
colomon like we did with alpha and ng. +1 18:57
Moukeddar Hello guys
miss you
tadzik hello Moukeddar
pmichaud I doubt there's going to be any significant spectest-related changing to master from here on out
Moukeddar how's the weather? 18:58
pmichaud Hot.
mberends PerlJam: just checked again, there's currently no Makefile target to do that
tadzik as it should be. Not too sunny, not to rainy
Moukeddar yes, very (46􏿽xB0C) 18:59
tadzik oh, I stole all the laziness available in the universe it seems
Moukeddar i think it's time to move
pmichaud tadzik: I was going to do that, but I was too lazy. :)
Moukeddar tadzik, you forgot how we split it ?
tadzik Moukeddar: oh, I had quite an exhaustive day
plus I have an exam tomorrow
18:59 bitpart left
Moukeddar i finished all my exam 19:00
[Coke] are the commented out test files in t/spec* from the list of passing ones in master?
Moukeddar so i'm litteraly high !)
pmichaud [Coke]: yes.
Moukeddar good luck tadzik
tadzik thanks. I already passed this one once, but I need a few points more 19:01
Moukeddar i passed 16 exam last week 19:02
it was hell 19:03
tadzik oh my
whatever kind of exams that was, it sounds terrifying
Moukeddar in 45􏿽xB0C
tadzik hell on earth indeed
Moukeddar 20 KM from home
but i made load of friends(sweaty ones) :p 19:04
tadzik :)
19:04 wtw left
Moukeddar what subject you'll be tested in ? 19:04
Su-Shee tadzik: my coworker btw knew the movie with the complicated polish name joke within the first three words of a sentence I started to tell him.. ;) 19:05
tadzik :D
Moukeddar: Probabilistic and Statistic Maths
Su-Shee tadzik: I just had to say "someone showed me the joke with the german and the polish name.."
tadzik :)
Moukeddar interesting
we've done some of that
money :p 19:06
tadzik the most interesting thing for me now is "let me go and learn on my own again"
Moukeddar why would you do that ? 19:07
dalek kudo/nom: 4af0ae1 | pmichaud++ | src/core/Mu.pm:
Mu.ACCEPTS() needs to be able to handle non-Any arguments.
kudo/nom: f5c25da | pmichaud++ | / (5 files):
More array-related methods and fixes. This _almost_ gets us

starts working.
tadzik Moukeddar: I feel that I'm wasting time on stupid things instead of putting my effort into interesting things. I had a different views on my studies before I started them, really 19:10
Moukeddar i'm dying to have a decent teacher 19:11
that will not burry my butt in vb6 code or C calculators projects
tadzik oh my
vb6 sounds preety bad
Moukeddar it's the lamest thing ever 19:12
yet, they still teach it like if it's the 90's
masak Moukeddar, hi.
Moukeddar masak, Le Man :) 19:13
hi you
masak back from L'Hexagon :)
tadzik to be honest, I never had a pleasure to work with it, but it sound as a synonym for a ridiculous thing which should be burned and buried. Maybe I'm wrong, maybe it's a pretty good language :)
Moukeddar le Hexagon*
masak really? I thought the H was silent.
Moukeddar in french it is 19:14
masak and I was speaking...? :)
Moukeddar frenglish
masak :P
pmichaud "touche'"
masak indeed.
pmichaud (too lazy to type the accented-e :)
PerlJam tooshay
masak touchy. 19:15
Moukeddar touch􏿽xE9
pmichaud Moukeddar +1
PerlJam masak: no touchy.
masak Moukeddar: either you just made a really funny encoding joke, or you need to change your client to UTF-8 :)
Moukeddar it's Touch􏿽xE9 19:16
masak it's currently ISO-8859-1.
no, it's "touché" :)
flip that utf-8 switch.
Moukeddar yes i should "decapitate" words :)
masak make the world smile ☺ 19:17
Moukeddar does he even smile?
masak Moukeddar: no, I wasn't talking about capital letters. I was talking about encodings.
Moukeddar: do you know about UTF-8? it's OK if you don't.
Moukeddar idk what encoding this client uses
i know UTF-8
just don't know where to change the encoding 19:18
masak it uses latin-1 from what I can tell.
what client?
maybe I can help.
Moukeddar X-chat
19:19 colomon left
flussence bottom of the network settings window 19:19
masak flussence++
19:19 birdwindupbird joined
masak Moukeddar: see the topic: "UTF-8 is our friend!" ☺ 19:19
Moukeddar where?
masak Moukeddar: /topic 19:20
flussence all the way to the right
Moukeddar wait a sec please
still looking for the encoding setting
19:21 colomon joined
flussence hm, I don't actually have xchat installed here.. 19:21
Moukeddar and i didn't find the setting neither 19:22
oh well
touch􏿽xE9 missed with my mind
flussence it *should* be on the thing where you edit a single IRC network's details...
but IIRC it actually defaults to some sort of hand-wavey encoding detection option 19:23
19:23 Mowah joined
Moukeddar can't find it really 19:24
flussence they don't exactly make that window easy to find :) 19:25
Moukeddar i wish i've discovered IRC sooner
earlier*
masak "sooner" is fine. 19:26
19:28 envi left
flussence fwiw, your grammar is better than the average native-lang=en internet user, so don't worry about it :) 19:29
pmichaud nom: my $a = (foo => 'bar'); say "foo"; 19:30
p6eval nom: OUTPUT«␤»
pmichaud cries.
benabik pmichaud: Wha? 19:31
pmichaud I'm thinking something is wrong with the constants table or something like that.
any fatarrow key seems to destroy the use of that string in the remainder of the code.
flussence std: no foo ???
p6eval std 37a0cdd: OUTPUT«ok 00:01 118m␤»
pmichaud nom: my $a = :foo('bar'); say "foo"; 19:33
p6eval nom: OUTPUT«␤»
pmichaud hmm, looks like any compile-time Pair will do it.
19:33 Kivutarrr joined
pmichaud ohhhhhhhh, I see the problem. 19:34
sub make_pair($key_str, $value) {
my $key := $*ST.add_constant('Str', 'str', $key_str);
$key.named('key');
that's causing the constant "key" to always act like a named parameter.
(i.e., it's applying .named() to the constant) 19:35
benabik Ew.
pmichaud okay, I think I can fix that. 19:36
mberends Moukeddar: just looked in my Linux XChat, the sequence is XChat, Network List, FreeNode, Edit, Character Set. 19:37
Moukeddar thank you very much mberends
done:) 19:38
19:43 TSa left 19:44 SHODAN left
masak Moukeddar: write "touché"! 19:47
Moukeddar touch􏿽xE9
when i wrote it at first 19:48
what did you guys see in the screen?
masak it still looks like "touch\351" here... :/
sorear I see touchU+FFFD
benabik I see touché 19:49
Moukeddar hahah, funny
sorear 一 ?
moritz looks fine here too, but my client does latin1/utf8 autoguessing 19:50
Tene I can't find any anything /touch./ that looks like it's rendering wrong to me, fwiw.
dalek kudo/nom: 5bce9dc | pmichaud++ | src/Perl6/Actions.pm:
Using a constant string as a Pair key mustn't mark it as :named('key') for all other uses of the constant string in the code.
 19:51
19:54 wolfman2000 joined 19:59 MayDaniel joined
dalek ast: fb3efa1 | pmichaud++ | S02-builtin_data_types/nil.t:
Update test: Stringifying Nil.WHAT no longer produces 'Nil()'
 20:00
ast: 7fa87bc | pmichaud++ | S02-builtin_data_types/anon_block.t:
Update use of multi to include proto declaration.
ast: 526840c | pmichaud++ | S02-builtin_data_types/array_extending.t:
Remove #?rakudo skip fudge.
lichtkind thou: cheers
thou lichtkind: howdy! 20:01
howdy, pardner
dalek kudo/nom: 6538713 | pmichaud++ | t/spectest.data:
spectest.data update: uncomment more passing tests, add notes about why some are failing.
[Coke] moukeddar's touche looks fine to me. 20:02
(and so does camilla. 20:03
20:03 birdwindupbird left
thou while on the topic of IRC clients and input, can anyone share what client/config allows to write (and nicely format) multi-line text to p6eval? i've seen examples working with :to 'EOF' including newlines that work OK, and i can't make my client (LimeChat on OS X) send that. 20:06
flussence copying and pasting the "␤" works :) 20:07
benabik thou: I use flussence's method. 20:08
[Coke] that, however, is an unrenderable box.
thou hmm, ok. seems that it didn't in my tests, lemme try again
benabik thou: But be careful to only send... 10? at a time
benabik keeps meaning to write an AppleScript to add delay in multilines for Colloquy 20:09
TimToady on gnome ctrl/shift-u 2424 gets ␤ too
diakopter thou: use a gist 20:10
make a gist on github and send the url to p6eval
rakudo: <gist url> 20:11
p6eval rakudo 1fcb14: ( no output )
thou nice
niecza: say (qq :to 'EOF').lines.perl␤foo␤bar␤EOF␤
sorear like flussence and benabik, I tend to copy and paste the unrenderable boxes
p6eval niecza v7: OUTPUT«["foo", "bar"].list␤»
thou mine actually renders as a N/L glyph, where N is in the top left and L is in the bottom right (LucidaGrande on Mac) 20:14
benabik Hm. My 10 at a time was me crossing conversation streams. A script that turned "\n" into that N
L char would be nice
Too
Bah, iPad keyboard.
diakopter a gist works well if you want to try the same script over and over, also for scripts with lots of lines or just a very long script. 20:15
benabik And you can update and retry the gist.
TimToady if something renders as a box, it's probably merely a font problem
benabik And comments, etc etc
diakopter a detriment of gists is, while the change history of the gist is retained, the alignment to p6eval output isn't retained for posterity. 20:16
assuming it's evaluated in #perl6 20:17
benabik diakopter: Can compare change times to log times.
diakopter that's true
heh
I used it heavily when working on rosettacode entries and debugging perlesque
moritz masak: fwiw I've now release WebService::Libris (though p5 :/) to CPAN, now featuring tests :-) 20:19
I wonder if I should tell libris folks about it 20:20
20:20 dukeleto left
masak moritz: might be a good idea. 20:20
moritz: anyway, good job.
20:21 dukeleto joined
dalek ecs: 30347a2 | sorear++ | S32-setting-library/Numeric.pod:
Simplify handling of trig bases

Following discussion on #phasers, we've decided it's not worthwhile to have a second argument on trig functions, nor for methods to look at CALLER. C<use trigbase> has been redefined to still be useful, and is also extensible, if Gradians aren't exotic enough for you.		 20:26
20:28 molaf left
masak sorear++ 20:29
moritz sorear++ indeed 20:30
20:33 jevin_ joined
dukeleto is there something like Module::Starter for perl 6 yet? 20:33
20:33 jevin left
tadzik I wrote one 20:33
it can even create github repos, iirc :)
dukeleto tadzik: woot! What is it called?
tadzik dukeleto: I don't know, let me search through my github
dukeleto: github.com/tadzik/Module-Starter it seems 20:34
don't trust it too much :)
dukeleto tadzik: ingy ridiculed me for not having written a Perl 6 module yet, so now I have to defend my honor 20:35
tadzik I see :)
colomon ingy++
masak recruitment by ridicule! 20:36
TimToady whatever works
masak not really our way around here, but hey, it worked :)
diakopter ridiculous 20:38
dukeleto tadzik: your module-starter has a shebang line of "perl" but it seems to be perl6. Have you aliased perl to perl6 on your system already? ;)
masak oh, that's radical.
tadzik: sewing doubts in the enemy ranks, saying "Perl" but meaning "Perl 6" :) 20:39
er, sowing*
diakopter lol
tadzik dukeleto: a LHF would be to make it use github v3 api
masak my Onglish sycks.
tadzik ouchie, the shebang
colomon the enemy would do well to doubt my sewing
tadzik that's an lhf too, yes :) 20:40
if you don't fix it, I will :)
tomorrow, after The Last of the Exaams
dukeleto tadzik: give me a commit bit and I will fix it :) 20:41
tadzik dukeleto: done
I've spent a while looking for "dukeleto" on github
. o O ( mumble mumble stupid search box mumble mumble ) 20:42
dukeleto tadzik: yeah, that is one of the few places that I am "leto"
masak dukeleto: do you actually own a duchy? 20:43
dukeleto masak: i am saving up
benabik masak++
masak :) 20:44
as a backup plan, you could always move to the Netherlands. 20:45
colomon and here I thought you were just a big John Wayne fan.
though one of my old Star Wars gaming buddies is a Duke in the SCA... 20:46
benabik Need to fight to get those titles though 20:47
20:47 frettled left
colomon my impression is that he is indeed a skilled fighter 20:47
20:47 frettled joined 20:48 MayDaniel left
masak colomon: Society for Creative Anachronism? 20:49
colomon masak: yes
20:49 Moukeddar left
colomon my old college roommate / eventual best man at my wedding got involved with that after college, so I ended up knowing a number of members 20:49
masak nice. 20:50
sorear my parents were involved in that
colomon unsurprisingly, there tends to be a lot of overlap between SCA and role playing games, too.
benabik SCA = more dangerous LARP 20:51
colomon benabik: I dunno, at least they're armored when they start hitting each other. ;)
20:51 benabik left 20:52 benabik joined
diakopter transition from south carolina to north carolina 20:52
benabik colomon: Most LARPs have a "no touch" rule, so even armored is more dangerous.
colomon benabik: I know the rule, but I know some of us aren't very good at upholding the rule in practice. I've only LARPed a tiny bit, though. 20:53
20:53 pmichaud joined
diakopter pmichaud: wb 20:53
masak rakudo: my %h = a => 1, b => 2, c => 3; for %h { .say } 20:54
p6eval rakudo 1fcb14: OUTPUT«a 1␤b 2␤c 3␤»
masak rakudo: my %h = a => 1, b => 2, c => 3; for %h.pairs { .say } 20:55
p6eval rakudo 1fcb14: OUTPUT«a 1␤b 2␤c 3␤»
masak so... in list context, %h means %h.pairs?
sorear Yes.
masak I approve.
:)
dalek kudo/nom: 0516515 | pmichaud++ | / (3 files):
Add Parcel.Numeric and Any.Array, so we now pass parcel.t .
 20:58
kudo/nom: 69398ca | pmichaud++ | src/core/ (2 files):
Add List.reverse and Any.reverse.
sorear 657 unrun test files in Niecza... 20:59
21:01 zby_home___ is now known as zby_home_, bluescreen10 left
colomon rakudo: sub postfix:<degrees>($rad) { $rad / pi * 180 }; say sin(90degrees) 21:02
p6eval rakudo 1fcb14: OUTPUT«-0.954091467472818␤»
21:02 bluescreen10 joined
masak oh, dang. 21:02
rakudo: sub postfix:<degrees>($rad) { $rad * pi / 180 }; say sin(90degrees)
p6eval rakudo 1fcb14: OUTPUT«1␤»
masak \o/
tadzik self.sleep 21:03
masak rakudo: sub postfix:<degrees>($deg) { $deg * pi / 180 }; say sin(90degrees)
p6eval rakudo 1fcb14: OUTPUT«1␤»
21:03 bluescreen10 left
jlaire rakudo: sub postfix:<°>($deg) { $deg * pi / 180 }; say sin(90°) 21:04
p6eval rakudo 1fcb14: OUTPUT«1␤»
masak jlaire++
that's very readable. 21:05
jlaire rakudo++ perl6++ # that's actually awesome
benabik jlaire++
masak it is! it's even more awesome when you picture the parser running over that code and picking up an operator as it goes along...
dalek ecs: 6faacb9 | (Felix Herrmann)++ | S06-routines.pod:
[S06] typo fix: remove lonely ')'
 21:06
21:06 kanishka left
masak suddenly realizes why jnthn is missing 21:06
flussence rakudo: sub postfix:<π>($rad) { $rad }; sub π { pi }; say sin(π/2); # ;) 21:07
p6eval rakudo 1fcb14: OUTPUT«1␤»
moritz tweets the line with postfix:<°> :-)
masak moritz++ 21:08
21:10 dukeleto left
cognominal_ just curious, what about the metamodel and prototype based language like self and javascript? 21:10
21:11 Kivutarrr left, dukeleto joined
masak cognominal_: could you restate that question into something that can be answered? :) 21:11
cognominal_ I thouhgt the metamodel was general enough to model any object oriented language. Is that true for javascript or self? 21:12
colomon flussence: I played around with doing one step better than that -- sort of -- and creating a new Real type that was internally a Rat but whose actual value was the Rat times pi. :) 21:13
masak cognominal_: Perl 6 is already fairly prototype-based.
cognominal_: it just happens to have a 'class' keyword, too :)
cognominal_ there is a chain of inheritance but it goes thru objects not classes.
masak right.
cognominal_: the short answer is that you can code up your metaclass any way you like.
cognominal_ Also methods are in objects because there is no class to store them. 21:14
21:14 impious joined, impious left
cognominal_ including searching methods starting from the object? 21:15
the last line was an answer to masak. 21:16
masak right.
cognominal_: as far as I can tell, the metamodel all the way down to KnowHOWs lets you change a fair number of things. prototype-based MROs should definitely be among the possible things. 21:17
how well it can be made to play nice with the rest of Perl 6 is another question.
dukeleto it seems that nothing that starts with perl6-* in projects.list in ecosystem.git shows up on modules.perl6.org 21:21
am I missing something?
for instance, tadzik's Module::Starter doesn't show up on modules.perl6.org, but is in projects.list
sorear KnowHOW *is* a prototype system
masak sorear: oh, indeed. but with Perl 6's class/role system built on top. 21:22
colomon dukeleto: huh. perl6-Testing shows up, but as Testing
masak rakudo: my @a = [0,0,0], [1,1,1], [2,2,2]; say @a.flat.elems 21:24
p6eval rakudo 1fcb14: OUTPUT«3␤»
masak ah.
rakudo: my @a = [0,0,0], [1,1,1], [2,2,2]; say [+] @a>>.elems
21:24 dobie_gillis left
p6eval rakudo 1fcb14: OUTPUT«9␤» 21:24
21:25 colomon left 21:27 colomon joined
cognominal_ sorear, indeed KnowHOSs are objects. 21:30
21:34 molaf joined
colomon tadzik: raw.github.com/perl6/ecosystem/master/META.list seems to lack module-starter? 21:35
how is that list generated?
dalek kudo/nom: 2f4210b | pmichaud++ | src/core/Range.pm:
Fix Range.perl for non-numeric ranges.
 21:36
kudo/nom: 726436a | pmichaud++ | / (9 files):
Add infix:<===> and stub in some core types and their roles. This
21:37 dukeleto left, dukeleto joined 21:38 george_z0rwell joined 21:39 Reaganomicon left
dukeleto starts writing his first perl6 module whilst sitting next to colomon++ and watching lightning talks at YAPC::NA 21:40
PerlJam dukeleto: did he infect you? 21:41
colomon tadzik: why does ecosystem have projects.list and META.list? They're not consistent atm... 21:42
21:42 REPLeffect left, Psyche^ joined 21:46 Patterner left, Psyche^ is now known as Patterner
sorear wha? roast contains a test that "moose".bytes == 5 21:49
masak blog post! strangelyconsistent.org/blog/june-2...structures
PerlJam masak: aren't you about a week behind?
masak sorear: TimToady has tried to explain the thing with .bytes to me. AFAIU, literal strings can double as Bufs.
PerlJam: yes. see previous post. 21:50
sorear niecza thinks "moose".bytes == 10
PerlJam aye
sorear ooc, how much attention is my little project getting at yapc?
masak how much Perl 6 is there at YAPC::NA? 21:51
colomon sorear: TimToady mentioned it in the keynote
I think my talk is the only p6 talk
masak and stop calling it YAPC like it's the only one... :P
PerlJam masak: one of the semi-CS majors that I have as a student worker mentioned earlier today that, to learn a new programming language, he always trys to write a text adventure game in it.
sorear masak: it's the current one
there is at least one other #perl6 but I never need to qualify 21:52
dukeleto will mention perl 6 in his parrot talk, but there aren't many perl6-specific talks
PerlJam masak: so, your choice for days 29 & 30 is quite good I think :)
masak PerlJam: :)
colomon sorear: I certainly plan to mention niecza a good bit in my talk
dukeleto seemingly the perl 6 gradient decreases as you go westward over the atlantic
masak PerlJam: that was basically the original idea: to build up to an adventure game.
colomon it's great to have two distinct implementations with different strengths and weakness to talk about
masak PerlJam: having games every fifth day as well turned out to be a good idea, though. 21:53
colomon dukeleto: sure it's not a north/south thing? There was more last year in ohio... ;)
dukeleto colomon: this is my first YAPC. Obvioiusly, interpolation with a single data point is LTA :)
colomon And I'm only interpolating with two data points... 21:54
21:55 molaf left
sorear uhm 21:55
my $r = \$x
say $$r
21:55 REPLeffect joined
sorear is this correct at all currently? 21:55
substr.t is trying to use it
PerlJam sorear: I thought it had to be say $($r)
masak sorear: that looks like Pugs-era syntax. 21:57
PerlJam: no, those two are equivalent.
felher masak++ for makeing me a 'more harmonious human being' ;)
masak but I don't see people doing that kind of scalar dereferencing much. or at all.
felher: :)
21:58 kaare_ left
masak You're welcome. ;-) 21:58
PerlJam masak: for some reason I thought all dereferencers need parens
dukeleto what is the recommended way to run my tests for a perl6 module ?
masak PerlJam: they do in Rakudo.
dukeleto currently I am doing: PERL6LIB=lib prove -e perl6 -rv t/
masak dukeleto: 'ufo && make test' :) 21:59
PerlJam needs to reread the spec with fresh eyes
(shouldn't be too hard since I haven't really read them in a good long while)
lichtkind TimToady: is there any chance to get from $*IN via getc just a char without hiting enter?
22:00 whiteknight joined
masak lichtkind: heh, I asked that the other day :) 22:00
dukeleto masak: not using ufo nor do I have a makefile :)
masak dukeleto: well, you asked... :)
dukeleto should I be pushing './lib' onto @*INC in my test? 22:01
lichtkind masak: i like the idea
masak dukeleto: I've seen some people do that.
lichtkind: yes, me too.
lichtkind it superb for [Y/N] and alike
masak lichtkind: I want to write a Nethack-like game in Rakudo, so I need this. 22:02
lichtkind maybe call it not getc because as i read syn theres a problem with getc
the way we split out roll
i mean $larry did :)
22:04 leprevost left
dalek ecza: 3cbb32e | sorear++ | / (2 files):
Pass 9 new S32-str test files
 22:04
ast: e640168 | sorear++ | S32-str/ (6 files):
Remove named-arg tests from substr, fudge rest of S32-str for niecza
lichtkind masak: do you remeber who wanted to wirte a wiki in perl 6? 22:08
masak lichtkind: besides me and viklund, you mean? 22:09
if so, then no. maybe the backlog remembers. 22:10
22:11 Moukeddar joined
Moukeddar masak, how's your sourdough bread ? 22:11
lichtkind masak: yes i mean the guy who talked here some day ago of the idea to make a git based wiki in perl 6 so we can wave decent wiki and nive diff reports for the chatbot 22:12
colomon dukeleto: I'll talk to you about @*INC when we get a chance 22:15
masak Moukeddar: it looks tasty :) 22:19
Moukeddar the Q is : does it taste as it looks?
masak lichtkind: if that's all you want, you could have that today through gist.github.com :)
Moukeddar: I'll get back to you on that.
Moukeddar sure,
lichtkind masak: what is gist? 22:21
benabik lichtkind: gist.github.com IIRC 22:22
dalek kudo/nom: e8d75f6 | pmichaud++ | / (3 files):
Move scalar assignment decontainerization a bit lower down in
 22:23
Tene masak: I started a roguelike in rakudo quite a while ago; you can see what I got together here: shorl.com/sisebedrerito
August 2009, looks like 22:24
lichtkind masak: i need 4 features, decent wiki syntax, make change labels/notes, spamproof, and migration path for the data
masak Tene: I remember it.
Tene Unfortunately, it relied on cross-language-loading, which has been broken since shortly after that time.
use Curses:from<parrot>;
masak ooh.
does that still work?
Tene 16:24 < Tene> Unfortunately, it relied on cross-language-loading, which has been broken since shortly after that time. 22:25
As far as I know, it's still broken.
masak I suspected that. 22:26
Tene If so, you can get it working anyway with some Q:PIR
load_bytecode 'Curses.pbc'; screw around with some namespaces;
^ literal PIR code, totally not a lie 22:27
masak sounds promising.
lichtkind Moukeddar: how serous are you in creating a wiki? 22:28
Moukeddar i think you mean masak :)
sorear Tene: note that rakudo nom no longer uses parrot namespaces
masak lichtkind: gist has Markdown, I don't know what you mean by labels/notes, only you can edit your gists, and it should be fairly easy to migrate.
pmichaud nom's ability to do anything with parrot libraries is likely to be severely limited for a while. 22:29
Tene Given the number of people who have expressed interest in picking up my language interop work in the years since my burn-out, I'm a little disappointed that nobody actually has.
pmichaud the object systems are just too different/separate
Tene pmichaud: you don't need much of an object system to call curses functions with strings and integers.
lichtkind masak: i ment that i attach a text, a commit message to a change
benabik pmichaud: More reason to get 6model into Parrot core. 22:30
Tene masak: So, I recomment you write write MyCurses.pm, which loads the parrot curses library and wraps its functions appropriately into a few Rakudo classes.
masak lichtkind: gist doesn't have that. 22:31
Tene: aye.
pmichaud another approach might be to figure out how to get a zavolaj equivalent on nom, and use that to wrap the curses library :) 22:33
Tene pmichaud: the Parrot curses library is just dlopen, repeated dlsym, approximately.
So, it does about the same thing. 22:34
22:35 Moukeddar left 22:39 colomon left, dorlamm joined 22:40 lue joined 22:43 benabik_ joined 22:44 colomon joined 22:46 benabik left 22:47 benabik_ is now known as benabik
masak 'night, #perl6 22:48
22:48 masak left 22:50 benabik_ joined, icwiener left, icwiener joined, benabik__ joined 22:51 benabik left, benabik__ is now known as benabik 22:52 cotto joined 22:54 benabik_ left 22:57 dukeleto left 22:58 dukeleto joined
dalek kudo/nom: 6031e3a | pmichaud++ | / (6 files):
Fix some comparison operators, pass S03-operators/comparison-simple.t .
 22:58
22:58 dukeleto left 22:59 dukeleto joined, dukeleto left, dukeleto joined 23:00 dukeleto left, dukeleto joined 23:07 kanishka joined 23:15 dukeleto left, dukeleto joined, drbean left 23:17 wamba left 23:22 pernatiy left, Mowah left 23:24 bluescreen10 joined 23:27 Zapelius left 23:37 icwiener left
dalek kudo/nom: e02a620 | pmichaud++ | / (5 files):
More pir:: -> nqp:: conversion, spectest.data updates.
 23:41
23:43 mberends left
dalek ecza: 5348797 | sorear++ | lib/CORE.setting:
Add 4-argument substr, $$capture
 23:47
pmichaud sorear: where could I find sources for the regex engine you're currently using (in niecza or otherwise)? I need to update nqp's engine for ltm and the like. 23:49
sorear pmichaud: src/RxOp.pm6 with support routines in lib/Cursor.cs 23:51
23:52 stepnem left
sorear the general setup is a lot like nqp's regex compiler; the parser generates regex nodes, and then each regex node generates some code 23:52
23:52 alester left
sorear the LTM stuff is handled by the lad() methods on regex nodes 23:53
which produces a simple data structure representing the declarative prefix of the node
the actual LTM automaton is generated by C# code in Cursor.cs, mostly by LADNode and it subclasses, Lexer, and LexerNode 23:54
23:56 dorlamm left
pmichaud sorear: okay, thanks. I may have some questions at some point, Is what you have much different from whatever gimme5 and the like are using? (Do we even still have a gimme5? ;-) 23:57
sorear gimme5 is gone, I rewrote it because I couldn't understand it ;-) 23:59
it is based to some extent on what viv and STD's Cursor use

Powered by Cro and the Raku® Programming Language.

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

Thank you!