»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
jnthn No, I'm not sure where hits it 00:00
timotimo i'll let the branch rest for a little longer then and drive home for now
jnthn Can you add a debugging .WHAT on the thing that decont gives back?
timotimo can do
00:00 dmol joined
timotimo oh, i'm still building for two backends 00:03
diakopter jnthn: won't the .WHAT box it automatically?
timotimo this could take a few seconds longer ... 00:04
diakopter jnthn: oh nm, I suppose that's fine :)
timotimo i wrote it in the parrot block instead of the !parrot block >_> 00:05
jnthn diakopter: Oh, I just meant the .WHAT to help us debug it 00:07
Mouq std: say 1 Δ 2; BEGIN sub infix:<Δ>(\a, \b) { &[-](b,a) } # Should this work?
camelia std 8adbc60: OUTPUT«===SORRY!===␤Two terms in a row at /tmp/3Ir3SuQy3Q line 1:␤------> say 1 ⏏Δ 2; BEGIN sub infix:<Δ>(\a, \b) { &[-](␤ expecting any of:␤ feed_separator␤ infix or meta-infix␤ infixed function␤ statement modif…»
Mouq r-j: say 1 Δ 2; BEGIN sub infix:<Δ>(\a, \b) { &[-](b,a) } # Should this work?
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
Mouq r-p: say 1 Δ 2; BEGIN sub infix:<Δ>(\a, \b) { &[-](b,a) } # Should this work?
camelia rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/UqH02Sj4s8␤Two terms in a row␤at /tmp/UqH02Sj4s8:1␤------> say 1 ⏏Δ 2; BEGIN sub infix:<Δ>(\a, \b) { &[-](␤ expecting any of:␤ postfix␤ infix st…»
jnthn That looks like the most pointless ues of BEGIN ever... :P
It doesn't work 'cus BEGIN takes a blorst and statementlist contextualizes %*LANG, iirc
00:08 raiph left
jnthn oh, no, it's a statement 00:08
hmm
oh..
But of course
Mouq jnthn: I ask because it has implications for something like this:
jnthn You still don't define the operator
Mouq std: sub infix:["\x227c"](\a, \b) { 1 + 2 }; say 1 ≼ 2;
camelia std 8adbc60: OUTPUT«===SORRY!===␤Confused at /tmp/W0Jjey5g97 line 1:␤------> fix:["\x227c"](\a, \b) { 1 + 2 }; say 1 ⏏≼ 2;␤ expecting any of:␤ feed_separator␤ infix or meta-infix␤ infixed function␤ statement modifier loop␤P…»
jnthn I mean, the BEGIN is pointless because it's a sub declaration which is already compile time
The Δ still needs to be declared so the parser knows about it 00:09
Mouq I already knew it wouldn't work because of how the parser works
jnthn I think the latest one you just mentioned should work, though
I mean, "\x227c" is constant at compile time. 00:10
I don't see how the two relate, fwiw :)
The first is about one-pass parsing, the second is about what counts as constant, afaict.
timotimo running into some trouble with debugging
00:12 Psyche^_ joined
Mouq jnthn: Alright. So then something like: sub infix:["{chr(4*2207)}"](\a, \b) { 1 + 2 }; say 1 ≼ 2; shouldn't work 00:13
jnthn Mouq: I think that counts as pushing your luck... :) 00:14
Mouq: but TimToady will probably come and tell us, "yeah, that should work" :P
Mouq r: eval "sub infix:[{chr(4*2207)}](\a, \b) { 1 + 2 }; say 1 ≼ 2;" # :D 00:15
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling eval_0␤Preceding context expects a term, but found infix ≼ instead␤at eval_0:1␤------> sub infix:[≼⏏](, ) 3; say 1 ≼ 2;␤»
Mouq r: eval "sub infix:<{chr(4*2207)}>(\a, \b) { 1 + 2 }; say 1 ≼ 2;" # :D
r: eval "sub infix:<{chr(4*2207)}>(\\a, \\b) { 1 + 2 }; say 1 ≼ 2;" # :D
Bluh
r: eval "sub infix:<{chr(4*2207)}>(\\a, \\b) \{ 1 + 2 \}; say 1 ≼ 2;" # :D 00:16
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling eval_0␤Malformed parameter␤at eval_0:1␤------> sub infix:<≼>(⏏, ) 3; say 1 ≼ 2;␤ expecting any of:␤ colon pair␤ quote words␤»
rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling eval_0␤Missing block␤at eval_0:1␤------> sub infix:<≼>(\a, \b) ⏏3; say 1 ≼ 2;␤ expecting any of:␤ new term to be defined␤ constraint␤»
rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«3␤»
timotimo drives home now 00:19
00:21 risou_awy is now known as risou
Mouq Is anyone going to look at Arcterus' PR for Rakudo? github.com/rakudo/rakudo/pull/217 His stuff is kind of getting ignored for whatever reason 00:26
jnthn Mouq: It's meant to be on my plate but my plate is overflowing as usual :/ 00:28
Last time I reviewed the stuff it looked overall sane 00:29
pippo r: my @array; for ^10_000 {@array[$_]=$_}; say @array 00:30
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 …»
pippo r: my @array; for ^10_000 {@array[$_]=$_}; say @array 00:31
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 …»
pippo my @array; for ^1_000 {@array[$_]=$_}; my $file = @array.join(";"); my @array_clone=$file.split(";"); say @array_clone 00:35
r: my @array; for ^1_000 {@array[$_]=$_}; my $file = @array.join(";"); my @array_clone=$file.split(";"); say @array_clone
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 …»
pippo r: my @array; for ^10_000 {@array[$_]=$_}; my $file = @array.join(";"); my @array_clone=$file.split(";"); say @array_clone 00:36
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 …»
pippo r: my @array; for ^10_000 {@array[$_]=$_}; my $file = @array.join(";"); my @array_clone=$file.split(";"); say @array_clone 00:37
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 00:38
..rakudo-parrot 3cef56: OUTPUT«0 1 2 3 4 5 6 7 8 9 10 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 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 …»
00:38 pippo left 00:39 risou is now known as risou_awy 00:42 dmol left
timotimo i'm trying to add debug stuff to the creation of the buildplan, but that's kinda exploding 00:48
nqp-j: say(nqp:isnull(nqp:null_s())); 00:49
camelia nqp-jvm: OUTPUT«Confused at line 2, near "say(nqp:is"␤ in panic (gen/jvm/stage2/NQPHLL.nqp:377)␤ in comp_unit (gen/jvm/stage2/NQP.nqp:922)␤ in TOP (gen/jvm/stage2/NQP.nqp:820)␤ in parse (gen/jvm/stage2/QRegex.nqp:1247)␤ in parse (gen/jvm/stage2/NQPHLL.nqp:1376)␤ in…»
timotimo nqp-j: say(nqp::isnull(nqp::null_s())); 00:50
camelia nqp-jvm: OUTPUT«0␤»
BenGoldberg r-p: my @a is default(42); say @a[5] 00:51
camelia rakudo-parrot 3cef56: OUTPUT«42␤»
BenGoldberg r-p: my @a is default (42); say @a[5]
camelia rakudo-parrot 3cef56: OUTPUT«True␤»
BenGoldberg is confused
Why does that space change it to True?
r-j: 0 00:54
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
timotimo a mere say in the create buildplan gives me a null pointer exception :|
ah, but an nqp::say seems to work 00:55
Mouq rn: say "a"~~token a { <sym> }
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«Useless declaration of a has-scoped method in mainline␤｢｣␤ sym => ｢｣␤␤»
..niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤Methods must be used in some kind of package at /tmp/Ortg2lMSTS line 1:␤------> say "a"~~token a ⏏{ <sym> }␤␤<sym> is only valid in multiregexes at /tmp/Ortg2lMSTS line 1:␤------> sa…»
BenGoldberg r-j: my $t = token { <sym> }; say "a" ~~ $t;
r-p: my $t = token { <sym> }; say "a" ~~ $t; 00:56
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
rakudo-parrot 3cef56: OUTPUT«｢｣␤ sym => ｢｣␤␤»
00:56 colomon joined
BenGoldberg r-p: say "a" ~~ token { <sym> }; 00:56
camelia rakudo-parrot 3cef56: OUTPUT«｢｣␤ sym => ｢｣␤␤»
BenGoldberg r-p: say "a" ~~ token a { <sym> };
camelia rakudo-parrot 3cef56: OUTPUT«Useless declaration of a has-scoped method in mainline␤｢｣␤ sym => ｢｣␤␤»
Mouq n: grammar A {token a { <sym> }}; A.parse("a")
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤<sym> is only valid in multiregexes at /tmp/JGRdZPIluw line 1:␤------> grammar A {token a { <sym⏏> }}; A.parse("a")␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib…»
Mouq Niecza gets a cookie :) 00:57
BenGoldberg n: my $t = token { <sym> }; say "a" ~~ $t;
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤<sym> is only valid in multiregexes at /tmp/0wgJbrdMmz line 1:␤------> my $t = token { <sym⏏> }; say "a" ~~ $t;␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/COR…»
BenGoldberg n: my $t = rule { <sym> }; say "a" ~~ $t;
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤<sym> is only valid in multiregexes at /tmp/csPYjiURBm line 1:␤------> my $t = rule { <sym⏏> }; say "a" ~~ $t;␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE…» 00:58
00:59 woosley joined
Mouq std: grammar A {token a { <sym> }}; 01:00
camelia std 8adbc60: OUTPUT«ok 00:01 124m␤»
Mouq n: grammar A {token a { <sym> }; token sym { "This shouldn't blow up" } }; 01:01
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤<sym> is only valid in multiregexes at /tmp/5oK41JSCcq line 1:␤------> grammar A {token a { <sym⏏> }; token sym { "This shouldn't blow up␤␤Unhandled exception: Check failed␤␤ at /home/…»
Mouq Niecza gets half a cookie
Hmm. Maybe 01:02
diakopter I just realized.. the more you spam #perl6, Programming Perl 6 is delayed by 1 more second... 01:07
<WINK>
timotimo wtf, this makes no sense 01:23
unless of course there is some kind of recursion that i'm not seeing 01:24
doesn't seem to be the case
i've sprinkled the whole method with debug says, and it seems to not happen inside the while loop at all, even though that's what i see in the backtrace 01:25
diakopter ETOOMANYSPRINKLES 01:29
timotimo so ... apparently the posat that gets the code is the culprit? but ... whaaat? 01:33
so, the result of 4 + $typespec is not unboxable to a native int? 01:37
r-j: say nqp::objprimspec(my int $x).WHAT;
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
timotimo bluh 01:38
benabik r-j: # timeout 01:39
timotimo using nqp::add_i instead of + seems to fix that little bit, great! >_>
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
benabik rakudo:
Mouq benabik: camelia doesn't bother if you don't give it a program 01:40
timotimo jnthn: Cannot unbox a type object is the result of nqp::say(nqp::decont(%attrinit{$key_name}).WHAT) 01:42
oh, i may need to just-say that
benabik Mouq: I was just trying to find the minimal program that times out. I guess less than just a comment isn't going to happen. 01:44
Mouq r-j: (timeout) 01:45
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
Mouq rnp: say (timeout)
rnp: say "(timeout)" # dummy 01:46
Do we no longer have rnp mapped?
rn: say "(timeout)" # dummy
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56, niecza v24-98-g473bd20: OUTPUT«(timeout)␤»
timotimo great, i got it to work on rakudo-jvm 01:47
the only thing wrong with the code was the 4 + $typecode
benabik Finally! I have nqp-[pmj] and rakudo-[pj] installed. 01:49
Mouq nqp: grammar A { token TOP { <a>* }; proto token a {*}; token a:!sym { <sym> } }; say(A.parse("000000")) #heh
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«000000␤»
timotimo oh you silly :)
Mouq can't figure out where, but a:!b becomes a:b<0> and a:b becomes a:b<1> somewhere in the parse stage of NQP 01:51
Or... somewhere 01:52
timotimo easy 01:53
method colonpair in src/NQP/Actions.nqp
it takes the second branch and puts the IVal there with !$<not>
Mouq Okay, but how does that value end up in %*RX<name>? 01:54
Oh
It actually does take the .ast
Mouq is dumb
timotimo where in the spectests do i put tests for the buildplan?
ah, easy 01:55
S12-construction seems to be the place 01:56
01:57 PZt joined 02:03 SmokeMachine joined 02:06 pmurias left 02:08 SmokeMachine left 02:09 SmokeMachine joined
timotimo i should hit the hay and continue this tomorrow. 02:10
02:13 SmokeMachine left 02:22 risou_awy is now known as risou 02:25 wayland-mobile joined
dalek p: 1931744 | Mouq++ | src/ (3 files):
Allow multiregexes of the form category:usage

Before, we always had to use "category:sym<usage>" even if "usage" was entirely <ident>		 02:29
Mouq nqp: grammar A { rule TOP { <fruit> + }; proto token fruit {*}; token fruit:sym<banana> { <sym> s? };}; say(A.parse("banana bananas banana")) 02:34
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«banana bananas banana␤»
Mouq fruit:sym<banana> can now just be fruit:banana, is all that does. Of course, it's not bootstrapped and won't work in Rakudo Perl, so the only place you can really use it ATM is the Rakudo NQP source 02:35
02:39 risou is now known as risou_awy
lue I still get those failures. :( 03:01
which means my test modifications didn't get pulled. Considering the returned value is correct, I say the spectest passed successfully, but I'd like further input before moving to modules-test 03:02
Mouq lue: "[S05-mass] Fix punctuation tests."? 03:04
lue yes. 03:05
BenGoldberg r: print "(timeout)" 03:07
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«(timeout)»
03:07 wayland-mobile left
BenGoldberg :) 03:07
r: 1
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 03:08
( no output )
03:24 dansamo joined
lue Spectest has two failures, one's my fault, and the other can't be changed unless someone moves tags around in rakudo to get a rakudo that pulls the updated specs. I'll go through the rest of Step 10 and then wait for someone to respond. 03:26
03:30 benabik left
lue Welp, DBIish failed. (That's the only broken one by test) 03:40
✘ ⑩ (Release PAUSED until further notice) 03:41
jnthn, moritz, other release pros: what should I do about gist.github.com/lue/7188705 (ignore the access failure, I have a noatime mount) and the failed DBIish module test? I did fix the failing charset test (¶ and § *are* punctuation), but it appears R* isn't pulling the fixed version. 03:46
04:17 dansamo left 04:22 risou_awy is now known as risou 04:39 risou is now known as risou_awy 04:43 fridim_ left
TimToady would like to point out to various and sundry that infix:<{...}> or infix:<xdeaf> will never interpolate inside <> 04:46
we've seen that braino a couple times in the last several days now
lue infix:«{...}» works though, right? 04:47
04:48 BenGoldberg left
TimToady BenGoldberg: re irclog.perlgeek.de/perl6/2013-10-28#i_7774302 a trait is just sugar for an colonpair, so without an argument defaults to True; dunno what the (42) is doing other than maybe poking a subsig pattern into the signature 04:48
lue: presumably it oughta do that, unless people think it should work like {} in shell quoting 04:50
lue "it" being «» or <> ?
TimToady {} 04:51
lue (2nd it)
TimToady «» is assumed already
lue assumed when? infix:<X> ? inside {}?
TimToady you already assumed «» in your question! 04:52
you asked if «{...}» should work 04:53
r: say « a { 40 + 2 } z »
lue gets a hand on his caffeine before continuing... his brain must be slowing down :) 04:54
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«a42z␤»
TimToady rp++
lue r-j: say "Hi TimToady o/"
TimToady so it looks liek it works in rp anyhoo
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
lue suspects the startup time is too much for camelia
TimToady rj also has problems from time to time with recursive errors, but you'd think that'd show up in the message if so 04:55
j: say 42
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
lue j: True 04:56
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
JimmyZ j: say (timeout)
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
JimmyZ look, it works 04:57
lue that's not good; it doesn't even parse the syntax!
std: say (timeout)
TimToady consistency is overrated in this case
04:57 preflex_ joined, ChanServ sets mode: +v preflex_
camelia std 8adbc60: OUTPUT«===SORRY!===␤Undeclared routine:␤ 'timeout' used at line 1␤Check failed␤FAILED 00:01 123m␤» 04:57
TimToady j: 1 2
Mouq rn: sub infix:«\x227c»(\a, \b) { 1 + 2 }; say 1 ≼ 2;
04:57 preflex left, preflex_ is now known as preflex
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 04:58
..rakudo-parrot 3cef56, niecza v24-98-g473bd20: OUTPUT«3␤»
TimToady j apparently can't even get the parser going long enough to detect a syntax error 04:59
TimToady wonders if it's fighting for cpu at the moment
lue would like CORE.setting to not b0rk up unicode parsing, so all those ugly infix:["x2A03"]'s can be transformed to infix:<⨃>
TimToady the limit is wall clock time, isn't it?
lue You should slow down your clock to give it a better chance then :P 05:01
moritz \o 05:02
lue o/
moritz: I had some spectest failures on ⑩ earlier, if you're backlogging :) 05:04
moritz lue: I am; IMHO you should proceed anyway 05:10
lue did you see that DBIish failed its tests too?
moritz lue: since the compiler release tested successfully for several people, we can discount the failures as platform specific for now
lue: no, I did not
lue: have a link to the failures? 05:11
lue updates the gist to include broken module
gist.github.com/lue/7188705 05:12
moritz: ^^^ (I know the noatime is my fault, and I personally think rakudo is correcter than the specific spectest revision it pulls on punctuation. So only the module concerns me.)
moritz lue: DBIish seems to be caused by the rakudo regression regarding whatever-currying of *[0] etc 05:13
and work-aroundable
so I'll do that soon
but first, breakfast for $daughter 05:14
lue for @family { &breakfast; POST { query "proceed to ⑪ or pause, then?" } } # :) 05:15
05:17 kaleem joined 05:21 [Sno] left
lue (in any case, I suspect 12-15 at least will happen tomorrow) 05:27
05:43 xenoterracide left
lue ♘ all o/ 05:45
dalek Iish: b0c5d75 | moritz++ | lib/DBDish/Pg.pm6:
work around rakudo regression

  lue++ for noticing it during the R* release process		 05:53
05:53 dansamo joined
dalek kudo/nom: a7f7522 | benabik++ | tools/build/ (2 files):
Teach j-install about DESTDIR
 05:55
06:14 sqirrel joined 06:16 FROGGS joined 06:19 xenoterracide joined 06:23 risou_awy is now known as risou 06:42 risou is now known as risou_awy 06:56 sqirrel left 06:58 risou_awy is now known as risou 07:01 denis_boyun joined 07:02 [Sno] joined 07:10 ssutch left 07:21 denis_boyun left 07:34 dmol joined 08:01 zakharyas joined 08:07 FROGGS[mobile] joined 08:19 PacoAir left
masak moarning, #perl6 08:24
moritz \o masak
sjn o/ 08:25
tadzik masak! asdfmovie7 is out :D 08:27
masak .oO( the latest instalment in the asdf franchise ) 08:29
FROGGS[mobile] Morning
masak good morning, FROGGS in a mobile.
masak has a mental image of rubbing a mobile phone with a piece of cloth, and FROGGS pops out all ghost-y and grants him three wishes 08:30
hoelzro tadzik: what. 08:31
hoelzro is off to Youtube
tadzik hah hah
it has colours
hoelzro mwahaha 08:33
Desmond the Moon Bear reprise!
FROGGS[mobile] masak: what what do I wear in your phantasy? 08:34
how was she called, beloved ginny? 08:35
but maybe I would be more like Myrte 08:40
masak :)
FROGGS[mobile]: you're in a long stylish bathrobe which kinda morphs into mobile phone blue smoke.
FROGGS[mobile] I am fine with that 08:41
.oO( of course the smoke is blue, it is digital smoke)
08:42
08:47 fhelmberger joined
masak of course. 08:56
diakopter masak: without looking at the source, I'm guessing that the current macro implementation has to do hoop-jumping to save ASTs..? 09:19
masak: I'm actually thinking about wicked things like macros working in string evals after they've been declared.. .oO( does that already work,ish? ) 09:22
dalek kudo/nom: 27fcd70 | (Timo Paulssen)++ | src/ (2 files):
WIP: creating buildplan thingies for native attrs.
 09:24
kudo/nom: 2e5e1e3 | (Timo Paulssen)++ | src/core/Mu.pm:
WIP
kudo/nom: 6c853ea | (Timo Paulssen)++ | src/Perl6/Metamodel/BUILDPLAN.nqp:
make native attribute buildplan work on jvm, too.
kudo/nom: 17ecfcc | (Tobias Leich)++ | src/ (2 files):
Merge pull request #224 from timo/nativeattr_buildplan

buildplan operations for native attributes
09:25 SamuraiJack_ joined 09:26 xinming left 09:27 xinming joined
masak diakopter: I promise to backlog yesterday properly and ask followup questions when I have the cycles to process your insight. 09:30
diakopter: hoop-jumping -- well, a Quasi needs to save the QAST as a data structure, because it has (unquote) holes in it that need to be filled. 09:31
diakopter: I can easily believe that you have found a better way to do that.
diakopter masak: perhaps.. it's pretty tenuous/shaky/amorphous 09:32
but it may be more elegant, but then again might not be
the "easy" way to do it would be to create a closure for each placeholder that runs a routine generated from the expression passed as the argument, and then for each "invocation" of the macro, it creates another closure with those arguments curried in 09:35
"easy" in quotes because it's not super-efficient.. 09:36
09:36 daxim joined
diakopter (as it could be if the code were duplicated/inlined 09:36
)
masak diakopter: I also want to make sure we don't trade away AST introspection. 09:37
09:37 sqirrel joined
diakopter how is that used? 09:37
the only "problem" with the way I described it is it creates intermediate frames that the user might not have imagined 09:38
moritz r: class A { has $!x = 42; method get() { ::('$!x') } }; say A.new.get 09:41
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«No such symbol '$!x'␤ in method gist at gen/parrot/CORE.setting:11916␤ in method gist at gen/parrot/CORE.setting:974␤ in sub say at gen/parrot/CORE.setting:12816␤ in block at /tmp/RqeHydpurZ:1␤ in any at /tmp/RqeHydpurZ:1␤ in any at …»
09:41 dakkar joined, spider-mario joined
moritz should that work? 09:41
r: class A { has $!x = 42; }; my $obj = A.new; say A.^attributes[0].get_value($obj); 09:42
masak moritz: my first thought is "I don't see why it shouldn't".
camelia rakudo-jvm 882e33: OUTPUT«(timeout)» 09:43
..rakudo-parrot 3cef56: OUTPUT«42␤»
moritz masak: my first thought was "rakudobug"
j: say 1
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
diakopter moritz: it's been timing out on nearly everything since you went asleep
masak omg, it's powered by moritz awakitude!
quick, moritz never go to sleep 09:44
p: class A { has $!x = 42; method get() { ::('$!x') } }; say A.new.get 09:45
camelia rakudo-parrot 3cef56: OUTPUT«No such symbol '$!x'␤ in method gist at gen/parrot/CORE.setting:11916␤ in method gist at gen/parrot/CORE.setting:974␤ in sub say at gen/parrot/CORE.setting:12816␤ in block at /tmp/fkp8grFoPq:1␤ in any at /tmp/fkp8grFoPq:1␤ in any at …»
masak submits rakudobug 09:46
diakopter masak: but what if it's not private? 09:48
masak diakopter: there's always a $!x container, whether it was declared as $!x or $.x 09:49
r: class A { has $.x; method foo { say $!x } }; A.new(:x(42)).foo 09:50
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«42␤»
masak r: class A { has $.x; method foo { say ::('$!x') } }; A.new(:x(42)).foo
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
..rakudo-parrot 3cef56: OUTPUT«No such symbol '$!x'␤ in method gist at gen/parrot/CORE.setting:11916␤ in method gist at gen/parrot/CORE.setting:974␤ in sub say at gen/parrot/CORE.setting:12816␤ in method foo at /tmp/UATA_w4Lf0:1␤ in block at /tmp/UATA_w4Lf0:1␤ in an…»
diakopter hm
moritz rj: 1 09:52
camelia rakudo-jvm 882e33: OUTPUT«(timeout)»
moritz rj: 1
camelia rakudo-jvm 882e33: OUTPUT«Error while reading '/home/p6eval_eval/p6eval-token': Permission denied at /home/p6eval/jvm-rakudo/eval-client.pl line 10.␤»
masak does anyone want to field finanalyst's FAQ on p6l?
moritz rj: 1 09:53
camelia ( no output )
moritz \o/
masak it's about the "help, I can't fit a bunch of Xs into an Array of X!" again...
moritz rj: say 42
camelia rakudo-jvm 882e33: OUTPUT«42␤»
masak moritz++
moritz masak: no, that's not the frequently asked question
the FAQ is why sub bla(Int @a) { ... }; my @b = 1, 2, 3; bla(@b) doesn't work 09:54
but he actually declared my Int @b = ...
so I don't know off-hand want's wrong
or did I misread the question? 09:55
masak hmmm 09:56
does it have anything to do with the gather, perchance?
r: my Int @b = gather for 1, 2, 3 { take $_ }; say @b
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«1 2 3␤»
masak guess not.
moritz ok, bug_4.p6 is exactly the FAQ
masak: note that the assignment doesn't bomb, but the call 09:57
masak *nod* 09:58
dalek p: 0d34a3e | (Tobias Leich)++ | / (2 files):
mapped getlexrelcaller
masak if it were up to me, we should never give the error message "expected 'Array[A]' but got 'Array[A]'" :)
moritz yes, the error reporting is LTA 09:59
no, LTA doesn't cover it
it sucks.
masak LTA does cover it, even when it's an understatement.
LTA was the term I was looking for.
arnsholt Oh, this is the "class + mixin !~~ class + same mixin" problem, isn't it? 10:11
moritz probably
masak when it's described like that, I'm not sure it's bug rather than a feature. 10:12
r: role R {}; say (Int but R) === (Int but R)
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True␤»
masak r: role R {}; say (Int but R) ~~ (Int but R)
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«False␤»
masak hrm.
yeah, that's weird. 10:13
10:14 pdurbin left
arnsholt IIRC jnthn has said it's a bug, but also that it'll be a bit complicated to fix properly 10:15
masak can't recall whether it's in RT already 10:18
FROGGS I'd think you would have to check if A consists of the same parts as B, rather than to check if they have the same WHICH 10:19
masak that sounds... risky.
FROGGS hmmm, why?
the problem is that you create a new thing in a signature, so it is always a new thing and will never match others 10:20
masak by "same parts", do you mean "same mixins"? 10:21
because you also have to check the order in which they were applied, and that order *may or may not matter*.
FROGGS identical mixins 10:22
hmmm
well, if the order matters it must be checked, yes :o)
basically you would have to compare how both things where composed 10:23
so, do we record that?
like the build-plan
the blueprint 10:24
10:24 darutoko joined 10:31 donaldh joined 10:37 colomon left 10:38 colomon joined 10:41 wsri joined
moritz the merge base for Arcterus' pull request is from July 10:43
no wonder I get lots of conflicts while merging 10:44
masak maybe rebase it locally? 10:46
moritz problem is, I don't know much about the code he is touching 10:47
that makes it much harder for me to resolve conflicts
masak *nod* 10:48
11:01 FROGGS[mobile] left 11:08 denisboyun joined 11:10 risou is now known as risou_awy 11:12 fhelmberger left 11:16 fhelmberger joined 11:19 lizmat joined
diakopter Beyond Less Than Awesome 11:25
11:27 lue joined, rindolf joined
masak well, don't forget that "Less Than Awesome" was always meant as an understatement, implicitly meaning "this sucks". 11:28
it's only because it's now become so ingrained that we feel it's not enough.
11:31 pernatiy joined
dalek ast: 92e4192 | (Elizabeth Mattijsen)++ | S02-types/whatever.t:
Unfudge now passing tests, timotimo++
 11:44
lizmat did we recently add § and ¶ as punctuation chars ? 11:46
sorry, remove as punctuation chars ?
t/spec/S05-mass/charsets.t is failing test #10 because of that, at least on parakudo 11:47
or did lue forget to add a "2" to the todo 11:49
11:53 pochi_ joined, jnthn_ joined 11:54 mathw_ joined, [particle]1 joined
dalek p: ffcebf9 | (Tobias Leich)++ | src/vm/ (24 files):
make nqp::backendconfig available for all backends
 11:57
11:58 jferrero joined
dalek ast: 3e12598 | (Elizabeth Mattijsen)++ | S02-types/native.t:
Added tests for #102416
 11:59
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=102416
FROGGS nqp: say($_.key) for nqp::backendconfig() # this should work in a few(?) minutes
camelia nqp-moarvm: OUTPUT«ccoptiflags␤uvrule␤tomobjects␤ldshared␤sharule␤ccinstflags␤ldlibs␤ldimp␤shaobjects␤hasreadline␤mtclean␤ccdefflags␤obj␤ldflags␤lddir␤name␤noreturnattribute␤dllexport␤defs[0]␤defs[1]␤dcclean␤moar␤ldrpath␤dcrule␤ldusr…»
..nqp-parrot: OUTPUT«Error while compiling block : Error while compiling op for (source text: "say($_.key) for nqp::backendconfig() # this should work in a few(?) minutes"): Error while compiling op backendconfig (source text: "nqp::backendconfig()"): No registered operation handler f…»
..nqp-jvm: OUTPUT«No registered operation handler for 'backendconfig'␤ in compile_op (gen/jvm/stage2/QAST.nqp:231)␤ in as_jast (gen/jvm/stage2/QAST.nqp:3696)␤ in as_jast (gen/jvm/stage2/QAST.nqp:2971)␤ in (gen/jvm/stage2/QAST.nqp:1207)␤ in (gen/jvm/stage2/QAST.nqp:228…»
11:59 denisboyun left
lizmat n: my @a = 0..4; @a.=[1, 2]; 12:02
camelia ( no output )
lizmat n: my @a = 0..4; @a.=[1, 2]; say @a
camelia niecza v24-98-g473bd20: OUTPUT«1 2␤»
lizmat n: my @a = 0..4; @a.=[1, 2]; say @a.perl
camelia niecza v24-98-g473bd20: OUTPUT«[1, 2].list␤»
lizmat n: my @a = 0..4; @a=[1, 2]; say @a.perl 12:03
camelia niecza v24-98-g473bd20: OUTPUT«[[1, 2]].list␤»
lizmat n: my @a = 0..4; @a=(1, 2); say @a.perl
camelia niecza v24-98-g473bd20: OUTPUT«[1, 2].list␤»
12:03 kaleem left
lizmat n: my @a = 0..4; my $b=[1,2]; @a.=$b; say @a.perl 12:04
camelia niecza v24-98-g473bd20: OUTPUT«Unhandled exception: Unable to resolve method postcircumfix:<( )> in type Array␤ at /tmp/L7VIX4riYw line 1 (mainline @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4583 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4584 (modul…»
lizmat r: my @a = 0..4; my $b=[1,2]; @a.=$b; say @a.perl
camelia rakudo-jvm 882e33: OUTPUT«No such method 'postcircumfix:<( )>' for invocant of type 'Array'␤ in any at gen/jvm/BOOTSTRAP.nqp:1658␤ in method dispatch:<.=> at gen/jvm/CORE.setting:1129␤ in block at /tmp/qb9GOnVq4z:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in …»
..rakudo-parrot 3cef56: OUTPUT«No such method 'postcircumfix:<( )>' for invocant of type 'Array'␤ in any at gen/parrot/BOOTSTRAP.nqp:1687␤ in any at gen/parrot/BOOTSTRAP.nqp:1669␤ in method dispatch:<.=> at gen/parrot/CORE.setting:1138␤ in block at /tmp/_vg0NCwkvs:1␤ …»
lizmat masak: apart from it being legal in STD, what does the .=[] syntax bring exactly? 12:05
FROGGS r: my @a; my $b = 'perl'; say @a.=$b
camelia rakudo-parrot 3cef56: OUTPUT«No such method 'postcircumfix:<( )>' for invocant of type 'Str'␤ in any at gen/parrot/BOOTSTRAP.nqp:1687␤ in any at gen/parrot/BOOTSTRAP.nqp:1669␤ in method dispatch:<.=> at gen/parrot/CORE.setting:1138␤ in block at /tmp/eA_xRF628z:1␤ i…»
..rakudo-jvm 882e33: OUTPUT«No such method 'postcircumfix:<( )>' for invocant of type 'Str'␤ in any at gen/jvm/BOOTSTRAP.nqp:1658␤ in method dispatch:<.=> at gen/jvm/CORE.setting:1129␤ in block at /tmp/j8PRltvuN6:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in an…»
lizmat this wrt #120383
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=120383
masak lizmat: it allows me to reassign an array to a slice of that array. 12:06
lizmat: @a.=[@slice] is a short form for @a = @a[@slice]
lizmat: beyond that, I don't quite understand the question :)
lizmat ah, ok, better use different values in examples then, as in this case the indexes where the same as the values 12:07
which sort of clouds the issue :-)
masak ah; sorry, didn't mean to use a confusing example... :)
lizmat writes a test
masak lizmat++
lizmat this looks indeed as fallout from the [] {} method -> sub migration, wouldn't you say, masak ? 12:10
masak that was my suspicion, but I wasn't sure. 12:14
lunch &
dalek p: c91640d | (Tobias Leich)++ | docs/ops.markdown:
document getlex_n<[inso]> and bindlex_n<[inso]>
 12:31
[Coke] ~~ and heads off to work.
arnsholt Does anyone happen to know who created most of the tags in the NQP issues queue? 12:35
FROGGS arnsholt: you mean the seven labels? 12:36
arnsholt Yeah
FROGGS no, is that important to know?
arnsholt Or more specifically, the all uppercase ones (TESTNEEDED, PARROT)
Just curious. I'd prefer them to be cased normally (and maybe even a space, "Test needed"), but I'm not sure if it's intended on jnthn's part or some such 12:37
FROGGS irclog.perlgeek.de/perl6/2013-10-21#i_7740656
arnsholt Oooh, cheers!
I suspect the all uppercase may have happened because initially we had two labels: JVM and LHF 12:38
FROGGS yeah
but I'd prefer "Parrot" too
arnsholt renames 12:39
There
dalek ast: cc153df | (Elizabeth Mattijsen)++ | S09-subscript/slice.t:
Added tests for #120383
 12:44
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=120383
dalek p: dcfdfe2 | (Tobias Leich)++ | docs/ops.markdown:
fix getlex and bindlex
 12:49
13:03 SamuraiJack_ left
tadzik synopsebot, u so clever 13:05
13:10 jnap joined, risou_awy is now known as risou 13:11 fhelmberger_ joined, rurban joined 13:14 fhelmberger left 13:18 _ilbot left, _ilbot joined 13:21 fhelmberger joined 13:22 fhelmberger_ left 13:25 sqirrel left 13:26 risou is now known as risou_awy 13:29 rurban left 13:34 colomon left 13:38 colomon joined
dalek p: 61e97a9 | (Tobias Leich)++ | docs/ops.markdown:
documented getlex[outer|caller|rel|reldyn|...]
 13:40
colomon JSON::Tiny is failing two tests? 13:41
the smoker is completely borked because it cannot rebuild panda right now...
13:42 bluescreen10 joined
FROGGS colomon: can you already paste something wrt panda? 13:42
colomon ==> Testing JSON::Tiny 13:43
t/01-parse.t ...... Failed 2/90 subtests
is the relevant bit, I think
13:43 benabik joined 13:46 fridim_ joined, ajr joined 13:47 ajr is now known as Guest21062, Guest21062 is now known as ajr_
lizmat didn't moritz recently committed some sigwhitespace related fixes for JSON::Tiny ? 13:48
*commit
13:49 rurban joined 14:10 denis_boyun joined 14:15 PacoAir joined 14:22 kaare__ joined 14:28 sqirrel joined
masak r: my &oct = { :8(~$_) }; my &dec = { :10(~$_) }; say (oct 31) == (dec 25) 14:30
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True␤» 14:31
masak I couldn't get :8(~*) to work; nor do I know if it should...
lizmat so does that mean Christmas is 3 days from now? 14:34
14:36 fridim_ left
PerlJam On #perl6, it's always Christmas :) 14:39
moritz lizmat: I did 14:40
14:43 rurban_ joined, denis_boyun left
lizmat re: #120383, the error is generated in src/Perl6/Actions, line 3753 14:44
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=120383
lizmat but I think the cause is deeper, since the past.op *not* be callmethod at all in the case of .=[]
*should 14:45
rurban_ pointme: dalek
lizmat but rather, it should be 'call'
masak PerlJam: Christmas is already here, it's just unevenly distributed ;) 14:46
rurban_: hm, 'pointme' sounds familiar -- what does/did it do? 14:47
r: my @a = <foo bar baz>; say @a.[1]
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«bar␤»
masak r: my @a = <foo bar baz>; @a.=[1]; say @a
PerlJam masak: exactly!
camelia rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/KPgpK0fADi␤Cannot use .= on a non-identifier method call␤at /tmp/KPgpK0fADi:1␤------> my @a = <foo bar baz>; @a.=[1]⏏; say @a␤»
..rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/2UMo_8JCmd␤Cannot use .= on a non-identifier method call␤at /tmp/2UMo_8JCmd:1␤------> my @a = <foo bar baz>; @a.=[1]⏏; say @a␤»
masak moritz: do the above two outputs differ? oh! the temp files :/ 14:48
lizmat: seems to me that whatever mechanism @a.[1] uses, @a.=[1] should use, too.
lizmat moritz: I guess the tempfile names should be normalized in the comparison ? 14:51
rurban_ masak: It expanded the src url for the bots and more 14:52
I couldn't find dalek on the web
FROGGS dalek: help 14:53
diakopter dalek itself just echos lines from a file # sorear++ 14:56
15:01 dmol left 15:04 darutoko left 15:11 risou_awy is now known as risou 15:14 lowpor03 joined, tipdbmp joined 15:19 sqirrel left 15:21 [Sno] left
dalek kudo/moar-support: 753af50 | (Tobias Leich)++ | src/Perl6/ (2 files):
added QAST::VM variants for moar idential to jvm
 15:21
kudo/moar-support: 6e556ae | (Tobias Leich)++ | src/vm/moar/ (2 files):
added p6capturelex, which really needs reviewing
 15:23
lizmat seems the storm has died down here a bit, so we're out cycling&
FROGGS lizmat: keep away from trees please :/
lizmat we'll trie :-) 15:24
FROGGS :P
masak we have a nice fresh wind going here too. it's rattling against the window panes. 15:25
TimToady diakopter++ should also keep away from trees right now
15:28 lowpor03 left, risou is now known as risou_awy 15:30 lowpor03 joined 15:31 lowpor03 left
diakopter TimToady++ should also tree away from keeps 15:32
TimToady might get more writing done in a keep... 15:33
killing the trees comes later
15:33 raiph joined
timotimo moritz: you'll be glad to know, that the workaround you put into DBIish today for rakudo star can already be removed again for rakudo nom :) 15:33
diakopter ..but them your knowledge of Perl 6 will become outdated.. oh wait.... 15:34
TimToady diakopter: maybe you should pull down one of those oxygen masks 15:35
diakopter can't reach it from lying flat 15:36
..out
er, *laying
er
TimToady either way you're hosed in English
15:37 bluescreen10 left
TimToady but I'm less likely to mistake you for a hen than for a liar :) 15:37
15:37 bluescreen10 joined
TimToady so I'd go with "laying" even if the proscriptivists proscribe it 15:37
dalek ast: 9b8425a | (Timo Paulssen)++ | S12-construction/construction.t:
tests for native attribute BUILDPLAN.
 15:39
15:42 atroxaper joined
atroxaper Hello #perl6 15:42
diakopter pregramming is hard; let's go scribing 15:43
tadzik scrabbling
masak scrobbling 15:44
TimToady atroxaper: er, hello
atroxaper Who can tell me how i can get something via network. Some http get request. I have tried to install LWP::Simple but panda said that tests fails.
FROGGS atroxaper: I'd propose to fix LWP::Simple 15:46
15:46 ajr_ left
FROGGS (together) 15:47
tadzik that'd be awesome
as far as I remember from porting it to JVM, the chunked transfers seem to behave a bit strangely
15:47 ajr joined
FROGGS tadzik: is that a known issue? 15:47
tadzik yes :(
FROGGS :/
tadzik maybe we need LWP::Simplier
diakopter "patches compelled.."
benabik Heh. My terminal says: ./perl6-p CORE.setting has stages start, parse, syntaxcheck, ast, optimize, jast, classfile, jar
15:48 ajr is now known as Guest17946
benabik (The lovely confusion of parallel make.) 15:48
timotimo %)
FROGGS atroxaper: so, you are on the jvm?
tadzik or look how panda fetches files over http
(prepare a paperbag)
diakopter .. and a caldera 15:49
tadzik github.com/tadzik/panda/blob/maste...tem.pm#L54 ;)
atroxaper FROGGS: no. I have win R* "This is perl6 version 2013.05 built on parrot 5.3.0 revision 0"
FROGGS atrodo: ahh, a bit outdated though...
tadzik I remember hacking Curl binding with bbkr++ on PLPW, but I don't remember if we got it to semi-work 15:50
timotimo yeah, that's quite old
FROGGS atroxaper: I think I got a 2013.09 on a windows box...
diakopter FROGGS: ah well, I'll just transliterate the parrot chain, since I can sorta tell what it does, but don't understand why
timotimo windows, though ...
tadzik github.com/bbkr/cURL may be worth a try
FROGGS starts his windows 15:51
timotimo gentlemen, start your windows 15:52
atroxaper FROGGS: i'm installing 2013.09 R* now 15:53
FROGGS k
15:56 Guest17946 left, ajr_ joined 15:57 sqirrel joined 15:58 _ilbot left 15:59 _ilbot joined
atroxaper R* 2013.09 is even stranger... gist.github.com/atroxaper/7199471 15:59
16:00 btyler joined 16:01 bluescreen10 left 16:02 Rotwang joined, Rotwang left, Rotwang joined
FROGGS ahh, I know that one... 16:02
right after we made the 2013.09, we discovered that unlink does not delete protected files... 16:03
it was fixed a short while after the release though
timotimo that seems to happen often :P
that kind of thing
FROGGS yeah
timotimo just like all the unexpected fallout from the postcircumfix changes 16:04
atroxaper FROGGS: what i should do with that?) 16:05
FROGGS atroxaper: I am currently thinking about creating star 2013.10
does somebody know if lue got through it? 16:06
timotimo not yet
atroxaper It would be nice)
timotimo 06:27 < lue> (in any case, I suspect 12-15 at least will happen tomorrow)
FROGGS k 16:07
timotimo did we get a fix for .=[] and such in? 16:08
otherwise i could probably look at that now
FROGGS timotimo: well, 2013.10 happened like more than a week ago
timotimo er, i mean into nom 16:09
today
16:10 bluescreen10 joined
FROGGS I'd still would build a star for 2013.10 16:11
atroxaper FROGGS: So, should i wait 2013.10? Ans as i understand LWP:Simple is corrupted now?
FROGGS atroxaper: I dont know what is wrong with LWP::Simple@rakudo-parrot 16:12
you can always build rakudo HEAD, and then fetch panda on your own
if you have git and MSVC and activeperl, and ... 16:13
atroxaper FROGGS: ok. I'l try. It's will a usefull experiens i think ) 16:14
FROGGS atroxaper: it is definitely better than waiting a month for a fix :o)
and yes, you only an learn
atroxaper Do you recomend to use Rakudo.jvm already as well? 16:15
FROGGS depends
timotimo kind of a sad state, but understandable because our brightest engineers are currently working on moar rather than rakudo itself
FROGGS if you have long running stuff, then yes
for super-short scripts parrot is better 16:16
16:16 xinming left 16:17 xinming joined
atroxaper FROGGS: i read that there is a way to call java statements from Rakudo.jvm now. Is there way to do vice versa stuff? 16:18
FROGGS: once we started to talk ) 16:19
FROGGS atroxaper: I dont think so, but my knowledge about that is rather limited
atroxaper FROGGS: ok. Thank you. Have a good time 16:20
FROGGS atroxaper: you might want to watch jnthn++'s talk videos about jvm interop
timotimo perl6 level subs are exposed as something that implements Callable to java, so you can at least have callbacks into perl6 code
FROGGS :o)
atroxaper timotimo: thank you too ) 16:21
16:21 xinming left
atroxaper FROGGS: where is the video? 16:21
timotimo you can find it on youtube, i believe 16:22
16:22 xinming joined
atroxaper timotimo: yes. I found i think) Thank you all. Going to learn... 16:23
timotimo enjoy! :)
perl6 is cool as long as you can dance around the bugs that inevitably pop up everywhere :P
timotimo is just a bit frustrated from yesterday's experience 16:24
FROGGS atroxaper: have fun! :o)
atroxaper o/
TimToady \o 16:25
16:25 atroxaper left
FROGGS timotimo: yeah :/ 16:25
timotimo: but keep in mind, it is all TimToady's fault :o) 16:27
timotimo at least i got a few good commits out of the experience, so that's good! :)
16:27 iSlug joined
dalek p: c119b90 | (Tobias Leich)++ | src/vm/moar/QAST/QASTOperationsMAST.nqp:
stole set_[core|hll]_op_result_type, attach_result_type from jvm
 16:33
16:33 SamuraiJack_ joined 16:37 spider-mario left 16:38 zakharyas left
TimToady Rule 1: It's all TimToady's fault. Rule 2: Even if TimToady repents, it's still his fault for not repenting sooner. 16:40
FROGGS well, you could have invented PHP and then we would not have these types of problems :o) 16:42
TimToady no, I ended up with a problem of types instead
japhb__ That sounds like a good title: "A Problem of Types" 16:44
I'm not sure if that should be a paper, a book, or an art film.
FROGGS TimToady: do you have an opinion about this? irclog.perlgeek.de/perl6/2013-10-28#i_7775529 (10:11 to 10:24) 16:45
timotimo took the time to read and fill out the CLA
japhb__ timotimo: Did you find the source of irclog.perlgeek.de/perl6/2013-10-27#i_7773084 ?
timotimo: Does perl6-bench need unbitrotting? 16:46
(That's in regards to irclog.perlgeek.de/perl6/2013-10-27#i_7773538 _ 16:47
s/_/)/
timotimo japhb__: did not find the source, sorry :(
japhb__ timotimo: re: irclog.perlgeek.de/perl6/2013-10-27#i_7773747 , there is some sample code, but no one has added it to rakudo-jvm yet, IIRC 16:48
timotimo the second one: run seems to no longer give the changes to the environment to the processes it runs
japhb__ Gah.
But shell does? 16:49
timotimo i think so
japhb__ Rather than work around in perl6-bench, I'd like to just fix the regression in Rakudo.
timotimo that would be good, yes 16:50
there's one regression we hit in perl6-bench that i've fixed already =)
japhb__ Oh?
PR?
timotimo already merged
into rakudo nom, that is
japhb__ Oh, gotcha
timotimo the *.[] and *.{} thing
japhb__ And thank you for that. :-) 16:51
masak r: my @a = <foo bar baz>; @a.=[1]; say @a
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/zDOY2XLsZ9␤Cannot use .= on a non-identifier method call␤at /tmp/zDOY2XLsZ9:1␤------> my @a = <foo bar baz>; @a.=[1]⏏; say @a␤»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/OCC19FtPba␤Cannot use .= on a non-identifier method call␤at /tmp/OCC19FtPba:1␤------> my @a = <foo bar baz>; @a.=[1]⏏; say @a␤»
timotimo oh wow, i scanned that PDF in super-high-res
masak timotimo: you're free to look at @a.=[1], though. it's in a recent RT ticket, too. 16:52
timotimo: lizmat looked at it in today's backlog.
timotimo i seem to recall that
masak moritz: how hard would it be to get camelia to use the same /tmp/* filename, so that rakudo-jvm and rakudo-parrot yield the same output when possible?
timotimo she didn't attempt a fix?
japhb__ contemplates whether direct Rakudo hacking to fix problems, or creating a list of the top most painful Rakudo issues with hopes of others working on them, is a more effective use of my tuits today. 16:53
japhb__ notes that in camelia's last response, the two rakudos had different sha1s. 16:54
16:55 fhelmberger left
timotimo lizmat said the error comes from dotty:sym<.*>, but the error message doesn't match up with what i get from (my @a = 1, 2, 3).=[1] 16:56
oh, because that still tries to find a method, righto
oh, curious! there's different error cases from .=[1,2] and .=[1] 16:58
japhb__ timotimo: Looked through my notes. The rp open sketch was at: gist.github.com/donaldh/6901992
[Coke] I actually slightly preferred the all caps tags, but am fine with the change. 17:00
dalek rl6-roast-data: 189a03a | coke++ | / (5 files):
today (automated commit)
 17:01
rl6-roast-data: 173d2ee | coke++ | p (2 files):
today (automated commit)
17:02 kaare__ is now known as kaare_
timotimo japhb__: i don't understand why from and to have the same implementation o_O 17:02
[Coke] preflex: seen au?
preflex au was last seen on #perl6 246 days, 22 hours, 4 minutes and 28 seconds ago, saying: pity that Unicode terminals doesn't support strike-through very well :)
17:02 [Sno] joined
[Coke] raw.github.com/coke/perl6-roast-da...ummary.out - well, that's disappointing. 17:03
17:05 rurban left
masak .oO( au has been implementing strike-through in Unicode terminals for 246 days!? ) 17:07
17:07 kaare_ left
dalek p: 353787c | (Tobias Leich)++ | src/vm/moar/QAST/QASTCompilerMAST.nqp:
added sub want, needed for coercion
 17:08
17:09 ajr_ left
timotimo Use of uninitialized value $thirdpartyjars in join or string at tools/build/create-jvm-runner.pl line 42. 17:09
what's this all about?
probably just need to reConfigure.pl 17:10
17:12 risou_awy is now known as risou
[Coke] geekosaur: hey. You're a haskell person! 17:14
geekosaur I wonder if we can get away with upgrading pugs to work with modern stuff instead of complaining about all the deprecated stuff 17:15
in fact, some of it we're going to have to fix; as of the next ghc version, manual Typeable instances are forbidden 17:16
deriving (Typeable) is the only permitted way to do it
TimToady someone should just write a Haskell backend for rakudo now :)
FROGGS ohh noes /o\
TimToady how hard can it be?
masak .oO( GHC-hard ) 17:17
[Coke] so easy, even TimToady can do it. :P
geekosaur might find that easier than trying to understand au++ --- clearly he is not nearly as smart...
17:17 k1lldash9 joined
[Coke] geekosaur: updates to pugs that don't fail more spectests always welcome. 17:18
how can I find out why gist.github.com/coke/7200634 has the locale error it does? is there a way to see what locales are installed on the box I'm testing?
geekosaur look in /usr/share/locale usually 17:19
k1lldash9 locale should show you
17:19 BenGoldberg joined
geekosaur I was starting to say that over in #haskell when you pinged me over here :) 17:19
masak .oO( the peoples of #perl6 and #haskell have always had friendly relations with each other, despite their obvious evolutionary differences ) 17:21
TimToady
.oO(whitespace as syntax?!?)
geekosaur someone alredy pointed at default(42) vs. default (42) last night, and IIRC there is similar with fucntion calls; we already have whitespace as syntax 17:22
17:22 jnthn_ is now known as jnthn
TimToady
.oO(indentation as syntax?!?)
 17:22
17:23 dansamo left
masak TimToady: heredocs. 17:23
TimToady there's a huge difference between "Whitespace, yes or no?" and "Whitespace, how much?"
17:23 kaare_ joined
TimToady masak: the exception that proves the rule :P 17:24
masak :P
masak .oO( a *true* whitespace Scotsman...! )
BenGoldberg I understand the difference between with-whiteapce and without-whitespace for function calls.
But not default(42) and default (42)
TimToady again, it's the difference between :default(42) and :default (42) 17:25
we don't allow ws inside colonpairs either
BenGoldberg r: my @a is default (42); say @a[5] 17:26
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True␤»
BenGoldberg r: my @a is default (42); say @a
TimToady r: say :default
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«␤»
TimToady r: say (:default)
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«"default" => Bool::True␤»
BenGoldberg r: my @a is default; say @a
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«␤»
BenGoldberg What is 'is default' with no argument supposed mean, anyway? 17:27
TimToady exactly the same as default(True)
I keep sayin'
BenGoldberg Ahh, ok, I get it
r: my @a is default; say @a[1..5] 17:28
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True True True True True␤»
BenGoldberg r: my @a isn't default; say @a[1..5]
camelia rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/C79jY_FJvu␤Confused␤at /tmp/C79jY_FJvu:1␤------> my @a is⏏n't default; say @a[1..5]␤ expecting any of:␤ whitespace␤»
..rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/s3W1vSBvXq␤Confused␤at /tmp/s3W1vSBvXq:1␤------> my @a is⏏n't default; say @a[1..5]␤ expecting any of:␤ whitespace␤»
BenGoldberg :)
TimToady thing is, the parser doesn't catch the ttiar because siggies are parsed differently
std: my @a is default (42);
camelia std 8adbc60: OUTPUT«ok 00:01 124m␤»
BenGoldberg r: say True(42) 17:29
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&True' called (lines 1, 1)␤»
TimToady r: my @a is default (42); @a = 43;
camelia ( no output )
BenGoldberg Identical error messages! A first! :)
TimToady is probably just ignoring the subsig
BenGoldberg r: my @a isn't default(False); say @a[1..5] 17:30
camelia rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/GAZSiNU7VK␤Confused␤at /tmp/GAZSiNU7VK:1␤------> my @a is⏏n't default(False); say @a[1..5]␤ expecting any of:␤ whitespace␤»
..rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/9JdOxhsRwM␤Confused␤at /tmp/9JdOxhsRwM:1␤------> my @a is⏏n't default(False); say @a[1..5]␤ expecting any of:␤ whitespace␤»
BenGoldberg r: my @a is default(False); say @a[1..5]
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«False False False False False␤»
BenGoldberg r: my @a is default(False) (42); say @a[1..5]
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«False False False False False␤»
17:30 risou is now known as risou_awy
TimToady r: my @a is default ($b,$c); @a = 43; 17:30
camelia ( no output )
TimToady r: my @a ($b,$c); @a = 43;
BenGoldberg r: my @a is default(False) (now see here); say @a[1..5]
camelia ( no output )
rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/GjuRKuV2I7␤Invalid typename 'now' in parameter declaration.␤at /tmp/GjuRKuV2I7:1␤------> my @a is default(False) (now⏏ see here); say @a[1..5]␤»
..rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/uV0NSK4UIg␤Invalid typename 'now' in parameter declaration.␤at /tmp/uV0NSK4UIg:1␤------> my @a is default(False) (now⏏ see here); say @a[1..5]␤»
TimToady r: my @a ($b,$c); @a := 43; 17:31
camelia rakudo-parrot 3cef56: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Int'␤ in block at /tmp/68zqDAlxk7:1␤ in any at /tmp/68zqDAlxk7:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ in any evalfil…»
..rakudo-jvm 882e33: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Int'␤ in block at /tmp/GrkMuoPulC:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.nq…»
TimToady r: my $a ($b,$c); @a := 43;
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/cEkXHZ80_Y␤Variable '@a' is not declared. Did you mean '$a'?␤at /tmp/cEkXHZ80_Y:1␤------> my $a ($b,$c); @a⏏ := 43;␤ expecting any of:␤ postfix␤»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/6xKUvOGmMM␤Variable '@a' is not declared. Did you mean '$a'?␤at /tmp/6xKUvOGmMM:1␤------> my $a ($b,$c); @a⏏ := 43;␤ expecting any of:␤ postfix␤»
BenGoldberg r: my @a is default(False) ($do-i-need-to-declare-this); say @a[1..5]
TimToady r: my $a ($b,$c); $a := 43;
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«False False False False False␤»
( no output )
BenGoldberg r: my @a is default(False) ($do-i-need-to-declare-this, $how-bizzarre-how-bizzarre); say @a[1..5]; # :)
TimToady r: my $a ($b,$c) = 43;
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«False False False False False␤»
( no output )
TimToady yeah, seems to ignore the subsig entirely 17:32
r: my $a ($b,$c); $b = 43; say $b
camelia rakudo-jvm 882e33: OUTPUT«java.lang.NullPointerException␤ in block at /tmp/GlPGERLxuQ:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:1194␤ in any command_eval at s…»
..rakudo-parrot 3cef56: OUTPUT«(signal SEGV)»
TimToady wahoo
arnsholt jnthn: I've been pondering a refactor of NQP/QRegex so that HLL::Grammar.O takes actual named arguments, instead of parsing a string. Nay or yay? 17:33
BenGoldberg n: my @a is default(False) ($do-i-need-to-declare-this); say @a[1..5]
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤Trait default not available on variables at /tmp/2PlGtcgkDI line 1:␤------> ault(False) ($do-i-need-to-declare-this)⏏; say @a[1..5]␤␤Postconstraints, and shapes on variable declarators NYI …»
BenGoldberg n: my $a ($b, $c); $a = 42; say $a
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤Postconstraints, and shapes on variable declarators NYI at /tmp/URPGqNsiwj line 1:␤------> my $a ($b, $c)⏏; $a = 42; say $a␤␤Potential difficulties:␤ $b is declared but not used at /tmp/…»
BenGoldberg std: my $a ($b, $c); $a = 42; say $a
camelia std 8adbc60: OUTPUT«ok 00:01 127m␤»
BenGoldberg r: my $a ($b, $c); $b = 42; say $b 17:34
camelia rakudo-parrot 3cef56: OUTPUT«(signal SEGV)»
..rakudo-jvm 882e33: OUTPUT«java.lang.NullPointerException␤ in block at /tmp/eAZa2puHK4:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:1194␤ in any command_eval at s…»
masak whoa.
masak submits rakudobug
TimToady FROGGS: re mixin identities, structural equivalence tends to be a world of hurt for everyone involved, especially implementors, so I'm inclined to go with name equivalence here, i.e. you want the same mixin, then name it. 17:36
17:36 ssutch joined
TimToady that being said, if the optimizer wants to combine identical mixins to save space, we might want to rethink that a bit 17:36
masak feels weird to rely on though, if it's an optimization. 17:37
so conservatism seems to favor "no for now"
jnthn What on earth does "my $a ($b, $c)" even mean?
TimToady is inclined to think that, since classes are mutable, they should not automatically try to share 17:38
otherwise we might end up with rubyesque difficulties 17:39
benabik jnthn: $b := a[0]; $c := $a[1] # ?
TimToady well, presumably the same as in sub foo ($a ($b, $c)) {...}, only different :) 17:40
jnthn But it needs an extra pair of aprens if it's getting parsed as a signature, no? 17:42
my ($a (4b, $c)) # doesn't surprise
masak r: my $a ($b, $c); say $b
TimToady 4b or not 4b, that is the question
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«(Mu)␤»
masak r: my $a ($b, $c); $b = 42 17:43
camelia rakudo-jvm 882e33: OUTPUT«java.lang.NullPointerException␤ in block at /tmp/f7mLLTPKnG:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:1194␤ in any command_eval at s…»
..rakudo-parrot 3cef56: OUTPUT«(signal SEGV)»
jnthn Too much lag to spot my typos in time :P
I'm surprised it even parses; I must be misremembering something in the grammar...
TimToady it's just coming in as a postconstraint
17:43 iSlug left
jnthn oh... 17:44
I suspect we don't try to handle that code path at all :)
So heck knows what comes of the $b and $c... :)
TimToady r: my $a where 42; $a = 43
camelia ( no output ) 17:45
TimToady yeah
r: subset Even of Int where * %% 2; my Even $x; $x = 3; 17:48
camelia rakudo-parrot 3cef56: OUTPUT«Type check failed in assignment to '$x'; expected 'Even' but got 'Int'␤ in block at /tmp/V229oUIPh3:1␤ in any at /tmp/V229oUIPh3:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ in any ev…»
..rakudo-jvm 882e33: OUTPUT«Type check failed in assignment to '$x'; expected 'Even' but got 'Int'␤ in block at /tmp/NJS1GDNiDG:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPH…»
TimToady I guess we're not splitting that test out into post_constraint space as we do in siggies 17:49
r: subset Even of Int where * %% 2; my $x where Even; $x = 3;
camelia ( no output )
BenGoldberg r: subset Even of Int where * %% 2; my Even $x; if( rand > 1 ) { $x = 3 } 17:50
TimToady wonders if the assignment to typed object could be simpler on the nominal end if we did split that out
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!===␤Word 'if' interpreted as 'if()' function call; please use whitespace instead of parens␤at /tmp/d3fhiDyzel:1␤------> ven of Int where * %% 2; my Even $x; if⏏( rand > 1 ) { $x = 3 }␤Unexpected block…»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!===␤Word 'if' interpreted as 'if()' function call; please use whitespace instead of parens␤at /tmp/qNpxAu_Av5:1␤------> ven of Int where * %% 2; my Even $x; if⏏( rand > 1 ) { $x = 3 }␤Unexpected bl…»
BenGoldberg r: subset Even of Int where * %% 2; my Even $x; if ( rand > 1 ) { $x = 3 }
camelia ( no output )
moritz masak: (re camelia patch) no idea 17:51
BenGoldberg r: subset Even of Int where * %% 2; my Even $x; if ( rand > 1 ) { $x = "string" }
camelia ( no output )
TimToady BenGoldberg: "instead of" :P
BenGoldberg How would I get that to die at compile time, due to "string" not being of type Even? 17:52
moritz by patching the optimizer, somhow... 17:53
jnthn By patching Rakudo to do such things, yeah :)
17:53 sqirrel left
BenGoldberg p: subset Even of Int where * %% 2; my Even $x; if ( rand > 1 ) { $x = "string" } 17:53
camelia ( no output )
BenGoldberg P: subset Even of Int where * %% 2; my Even $x; if ( rand > 1 ) { $x = "string" }
camelia pugs: OUTPUT«*** ␤ Unexpected "% 2"␤ expecting "::"␤ at /tmp/wPVmDJINhC line 1, column 29␤»
TimToady no %% op
BenGoldberg P: my Int $x; if ( rand > 1 ) { $x = "string" } 17:54
camelia ( no output )
TimToady implement type inference?
or range inference if want rand > 1 to turn into False :) 17:55
[Coke] diakopter: the pugs tests really want "en_US.ISO-8859-1" available on the testing machine. Any chance we can get that?
-> $dayjob 17:56
TimToady US English doesn't include Latin :P
BenGoldberg In perl5 (and in Java!), an if( False ) { code } will remove { code } from the parse tree, for appropriate values of False
Does perl6's optimizer do that yet?
TimToady yes, I know--I implement the p5 part of that 17:57
*mented
diakopter [Coke]: it has that, you just need to set the env vars
TimToady these days should probably ISO-8859-15 or thereabouts to include the euro...
arnsholt BenGoldberg: Good question. If you feel up for it, Rakudo's src/Perl6/Optimizer.nqp is the place to look. It might not be too hard to implement 17:58
TimToady otoh, I would also not mind ignoring locales out of existence
if False { use locale; } 18:00
18:00 dakkar left
BenGoldberg Shirely, even in perl6, 'use' gets performed the moment the statement it's in is fully evaluated? 18:01
jnthn BenGoldberg: yes
BenGoldberg So that would need to be BEGIN { if False { require locale } }, or somesuch 18:02
TimToady was a joke 18:03
BenGoldberg :)
TimToady now, it'd be *interesting* if the compiler could figure out that the block was going away even before it was parsed, and disable any BEGIN-stuff inside it 18:04
but not even p5 tries to do that 18:05
(that I know of...)
18:06 SamuraiJack_ left
BenGoldberg That would be pretty deep magic :) 18:06
18:10 sqirrel joined
timotimo i can't get rid of Use of uninitialized value $thirdpartyjars in join or string at tools/build/create-jvm-runner.pl line 42. 18:10
dalek p: 8ea4967 | dwarring++ | examples/rubyish/ (2 files):
general rubyish sigspace tidies. added :s<hs> for horizontal white-space
 18:23
timotimo and also open: No such file or directory at tools/build/create-jvm-runner.pl line 35. 18:28
Creating '/home/timo/build/rakudo/install/home/timo/build/rakudo/install/bin/perl6-j' 18:29
well, this doesn't make sense
FROGGS ohh 18:30
destdir or prefix confusion
timotimo i've temporarily reverted the recent commit that seems to have touched it 18:31
18:33 sqirrel left 18:35 btyler left
timotimo that makes it work again 18:41
who here is Brian Gernhardt? 18:43
retupmoca timotimo: irclog.perlgeek.de/perl6/2013-10-28#i_7774849 ? 18:46
FROGGS timotimo: benabik
timotimo yes
benabik timotimo: Hello? 18:47
timotimo hi there :)
can you help me with the jvm runner installer?
benabik Probably. I tried not to break it too badly. ;-)
timotimo Creating '/home/timo/build/rakudo/install/home/timo/build/rakudo/install/bin/perl6-j' - this is obviously wrong, but i'm not sure how to fix it
this is using Configure.pl --backends=jvm,parrot --gen-parrot --gen-nqp=master 18:48
benabik Erm
I think I forgot quotes in Makefile-JVM.in
Tene timotimo: Try again with HOME=/home/timo/home/timo/home/tene/home/timo
timotimo Tene: :3
benabik Setting DESTDIR=/ might be a workaround. 18:49
timotimo i'd prefer a fix. this doesn't stop me from working, mind you :)
benabik Doublequotes around $(DESTDIR) on the create-jvm-runner lines in Makefile-JVM.pl 18:50
I can submit a PR in a moment.
timotimo ah, for some reason i thouht you had a commitbit
timotimo applied for one today :)
benabik I have one for Moar, NQP, and Parrot but not Rakudo. 18:51
lue hello world o/ 18:53
dalek p: e13e2f4 | dwarring++ | examples/rubyish/ (2 files):
updated rubyish README etc. nqp => nqp-p
benabik trying to build without DESTDIR before submitting the PR. 18:54
lue so, can I move on from 10? :)
FROGGS lue: only if unicode chars are involved :o) 18:56
lue :) I ask if it's OK to go on with that DBIish fail and the one/two minor spectest fails. gist.github.com/lue/7188705 18:57
FROGGS lue: no idea 18:58
lue moritz: do you know if I should go ahead? (Or did you e.g. fix DBIish and I need to re-do the tarball?) 18:59
19:00 ajr joined, ajr is now known as Guest90674 19:01 Guest90674 is now known as ajr_ 19:02 rindolf left 19:03 sqirrel joined 19:04 dmol joined
arnsholt moritz++ # New build system is still awesome =) 19:04
FROGGS it is, yeah 19:05
timotimo moritz++ # what arnsholt said!
FROGGS moritz++ # :o)
benabik moritz made it awesome, I just broke it. :-D
19:06 jeffreykegler joined
timotimo fwiw, i don't actually know how to fix the .=[...] stuff 19:07
benabik If someone with a bit could merge github.com/rakudo/rakudo/pull/225 that'd be great.
arnsholt I like the body of the message =D
benabik :-D
19:10 pecastro joined 19:13 risou_awy is now known as risou
FROGGS timotimo: the PR solved your problem? 19:14
19:15 jaffa4 joined
jaffa4 hi all 19:15
I cannot see the module I added to Panda.
timotimo did not try yet 19:16
jaffa4 Is it not approved yet?
lue moritz, jnthn, other release pros: if a DBIish fix isn't there/imminent, can I continue to Step 11?
timotimo DBIish did get the fix 19:18
or at least a fix 19:19
lue time to go back and spend a few more testing hours then :) 19:20
19:22 pernatiy left
lue jaffa4: I can see your Path::Util on modules.perl6.org/, so perhaps you just need to update your panda's module list (forget the command to do that exactly, though there should be some --help that tells you) 19:25
timotimo i think it's panda update 19:27
jaffa4 yes, that was the problem.
dalek p: 724a9b1 | (Arne Skjærholt)++ | src/NQP/ (2 files):
Make 'return' a term taking an optional argument.

This way, 'return;' won't be silently ignored in a function. Closes #127.		 19:28
19:29 Rotwang left
masak arnsholt++! \o/ 19:29
arnsholt: you are my hero tonight.
19:29 risou is now known as risou_awy
arnsholt Oh, yay! \o/ 19:29
masak arnsholt: next beer is on me :)
lue ✔ ⑤ Update modules (⑥ and ⑦ still ✔) 19:30
19:30 Mouq left
dalek ar: e5501b9 | lue++ | modules/DBIish:
Pull DBIish update.
 19:30
lue ✔ ⑧ push un-pushed changes
19:31 sftp joined
lue ✔ ⑨ Create tarball 19:32
19:32 xdbr joined
arnsholt masak: Anyways, happy to be of help. It's been on my hitlist of NQP bugs for a while, but last time I tried to fix it I couldn't quite get there 19:32
19:32 Shozan joined
arnsholt You coming to Oslo in the near future, BTW? 19:33
19:33 sqirrel left 19:34 daxim left
masak arnsholt: nothing on the schedule, though I had a Bergen a few days ago, and will have an Ålesund next week. 19:35
FROGGS Å is pronounced like an O, right? 19:36
arnsholt Right. Both are a bit far away for just a beer =) 19:37
FROGGS: Yeah
lue thinks of å as a "scandinavian o sound" :P 19:39
arnsholt Yeah, the exact phonetic value is probably going to be different from German/English
Close enough though 19:40
FROGGS cool..., we have a software at $work call Blå, which is the color blue as I remember
dalek kudo/nom: 9952073 | benabik++ | tools/build/Makefile-JVM.in:
Fix j-install for no DESTDIR
lue for some reason my mind is saying å ~~ german u ... I'll check the IPAs
19:41 denis_boyun joined
geekosaur front-a, isn't it? 19:41
19:41 sqirrel joined
masak ah, the sensual dance of the Lexical and the Dynamic. 19:41
moritz lue: å is pronounce more like the German o
oh, FROGGS++ already said that in the backlog :-) 19:42
GlitchMr perl6: .say for (('' xx 2, 'Fizz') xx * Z~ ('' xx 4, 'Buzz') xx *) Z|| 1 .. 100;
lue yeah.
camelia niecza v24-98-g473bd20: OUTPUT«===SORRY!===␤␤This macro cannot be used as a function at /tmp/1aB2oDIVIb line 1:␤------> zz') xx * Z~ ('' xx 4, 'Buzz') xx *) Z||⏏ 1 .. 100;␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niec…»
..rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«1␤2␤Fizz␤4␤Buzz␤Fizz␤7␤8␤Fizz␤Buzz␤11␤Fizz␤13␤14␤FizzBuzz␤16␤17␤Fizz␤19␤Buzz␤Fizz␤22␤23␤Fizz␤Buzz␤26␤Fizz␤28␤29␤FizzBuzz␤31␤32␤Fizz␤34␤Buzz␤Fizz␤37␤38␤Fizz␤B…»
arnsholt lue: German u is more similar to Norwegian o
FROGGS my colleagues tend to say it like Blä >.<
masak moritz: at least in Swedish, we screwed up the letter 'o' so that it (most of the time) has more of an /u/ sound.
FROGGS: heh :)
FROGGS masak: and in germany the e gets lost :o) 19:43
19:43 Rotwang joined
arnsholt Same in Norwegian. It's mostly /u/, but occasionally /o/ just to keep foreigners on their toes =) 19:43
lue really needs more practice in German :P
FROGGS for example most ppl say 'nein' like 'nain'
so, at some point we can scratch that e 19:44
lue All I had to do was say "Ostwestfälisch" to myself and I could confirm å == o :)
arnsholt lue: Which is your first language?
FROGGS .oO( Perl )
19:45 jeffreykegler left
lue arnsholt: de & en, though the de part's been lacking since I was 5 :) 19:45
arnsholt Ah, right
masak .oO( Leben, das Universum und der ganze Rest )
lue is pretty sure some of his ancestors are rolling in their graves because il sait français :) 19:47
moritz masak, arnsholt: which means in the theory there are 4 ways to pronounce "Oslo", though in practice I have heard only three of the variants 19:48
:-)
lue .oO[ Then again, they should know my opinions on East Prussia and Alsace-Löttrigen :) ] 19:49
masak moritz: only 4? I think you're discounting both vowel length and stress. :) 19:50
moritz: not to mention speech tone.
arnsholt moritz: The inherited pronounciation is probably /ushlu/ 19:51
moritz masak: they all sound the same to me :-)
masak count all those, and you could have 64 different pronunciations of "Oslo".
GlitchMr r: .say for ('' xx 2, 'Fizz') xx * Z~ ('' xx 4, 'Buzz') xx * Z|| 1 .. 100;
arnsholt And not /uslu/
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/ba8RO5tbZZ␤Bogus statement␤at /tmp/ba8RO5tbZZ:1␤------> Z~ ('' xx 4, 'Buzz') xx * Z|| 1 .. 100;⏏␤ expecting any of:␤ postfix␤ prefix or term␤ …»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/5G5nmaqRBw␤Bogus statement␤at /tmp/5G5nmaqRBw:1␤------> Z~ ('' xx 4, 'Buzz') xx * Z|| 1 .. 100;⏏␤ expecting any of:␤ postfix␤ prefix or term␤ …»
GlitchMr r triggers both Rakudos? Why? 19:52
masak ah, yes, the /s/ or /sh/ distinction is another bifurcation. so 128 :)
GlitchMr: because... they're both rakudos?
moritz GlitchMr: becase rakudo-moarvm doesn't work yet well enough
masak :)
19:52 Mouq joined
lue
.oO(Oslo -> Ahzzz-low)
 19:52
masak slaps lue with a doormat labeled "WELCOME TO THE USA" 19:53
moritz
.oO( maybe 'rs' should launch all rakudos and 'r' a random rakudo )
masak rakoulette!
FROGGS randudo! 19:54
moritz masak: sounds tasty :)
masak rak-u-tell-me
FROGGS o-rak-el?
lue R-A-K-U-D-O-!: say "hello"
masak .oO( because what we need from camelia is... more indeterminism...!? )
moritz and a proper source of entropy, please 19:55
lue sm: -> surprise me, any compiler camelia has available could be used :)
moritz bdsm: -> be a dear and surprise me :-) 19:56
oh sorry, we are trying to be family friendly to all time zones
masak .oO( we crossed a line? where? ) :P 19:57
moritz we didn't? good :-)
masak far as I'm concerned, unexpanded acronyms are SFW in-jokes. 19:58
moritz woah, rakudo-j setting build just segfaulted the JVM
19:59 BenGoldberg left
arnsholt Whoops! 20:00
lizmat is back from cycling and was not hurt by any tree parts 20:03
TimToady just whole trees? 20:04
lizmat timotimo: wrt to .=[], I tried to figure out where to apply a fix, but could not find the place, nor how to fix it
Mouq Uh, re: #120383: There's actually two different rules trying to handle what you're trying to do
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=120383
lizmat hmmm... isn;t a whole tree also a tree part ?
Mouq r: my @a = (4,5); @a.=[1]; say @a
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/Jy_2TLX7ui␤Cannot use .= on a non-identifier method call␤at /tmp/Jy_2TLX7ui:1␤------> my @a = (4,5); @a.=[1]⏏; say @a␤»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/UOqokADIlO␤Cannot use .= on a non-identifier method call␤at /tmp/UOqokADIlO:1␤------> my @a = (4,5); @a.=[1]⏏; say @a␤»
Mouq r: my @a = (4,5); @a .= [1]; say @a
camelia rakudo-jvm 882e33: OUTPUT«No such method '&postcircumfix:<[ ]>' for invocant of type 'Array'␤ in method dispatch:<.=> at gen/jvm/CORE.setting:1129␤ in block at /tmp/gYPsuez15t:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nq…»
..rakudo-parrot 3cef56: OUTPUT«No such method '&postcircumfix:<[ ]>' for invocant of type 'Array'␤ in method dispatch:<.=> at gen/parrot/CORE.setting:1138␤ in block at /tmp/vaHxV07r4l:1␤ in any at /tmp/vaHxV07r4l:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any…» 20:05
lizmat
.oO( feels a set or category theory remark coming up )
Mouq :P We have an infix <.=> and a term <.*> that includes .=
Mouq was messing with this stuff when he made dotty postfixes work 20:06
timotimo lizmat: yup, me either :|
20:06 btyler joined
lizmat Mouq: indeed, and that's when I started to lose track 20:06
timotimo at least list and hash access are going to get faster ... i guess
lue Just to clear this up ahead of time (starting spectests now): if the only failures I get are those two from the spectest (noatime and :Punctuation), can I proceed to 11? 20:07
masak lizmat, Mouq: despite my example being confusing, I think it's still perfectly legitimate even after the postcircumfix method->sub refactor.
Mouq masak: I agree
lizmat masak: agree
masak \o/
lizmat it's an implementation detail
masak it's little things like @a.=[@indices] that makes Perl 6 a joy to use. 20:08
lue
.oO(Perhaps @a[] is a sub and @a.[] is a method? :P)
timotimo it's not supposed to be :|
masak inner dialogue goes "should this work? it *looks* like it should work. yay, it works!"
well, modulo all the bugs. :)
but I'm submitting those :P
Mouq I think we should "just" make postcircumfixes pretend to be methods everywhere and be done with it :P 20:09
timotimo wouldn't that be a step back?
lue Now that I said it, I actually like the method/sub thing I mentioned. After all, you have foo(@bar) and @bar.foo, this would be the same thing
TimToady timotimo: that was a joke
lue but with funny syntax for the operator's names 20:10
timotimo ah, of course
you fave foo(@bar) and @bar.&foo, that's a difference 20:11
(er, or am i misunderstanding?)
masak lue: subs and methods are different.
r: class A { method foo { say "method" }; method bar { foo() }; sub foo { say "sub" } }; A.foo; A.bar
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«method␤sub␤»
lue sub postcircumfix:<[ ]>(@a, *@b) { @a.[@b] } is what I was thinking, like how you have sub foo($a, $b) { $a.foo($b) } 20:12
20:12 iSlug joined
masak lue: under the current semantics, that would probably regress infinitely :) 20:13
lue :)
arnsholt r: grammar Foo { token foo { :temp $foo = 3; foo } }; say "alive"
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/qoyLzsr3tl␤Unrecognized regex modifier :temp␤at /tmp/qoyLzsr3tl:1␤------> grammar Foo { token foo { :temp⏏ $foo = 3; foo } }; say "alive"␤»
..rakudo-parrot 3cef56: OUTPUT«===SORRY!=== Error while compiling /tmp/iZUqBPlpCW␤Unrecognized regex modifier :temp␤at /tmp/iZUqBPlpCW:1␤------> grammar Foo { token foo { :temp⏏ $foo = 3; foo } }; say "alive"␤»
hoelzro ahoy perl6ers 20:14
timotimo hey rob :) 20:15
jaffa4 panda does not work for me, it cannot delete a file.
timotimo what are you up to these days?
jaffa4: the new release is going to fix that, lue is preparing it at this moment
arnsholt hoelzro o/
hoelzro timotimo: today, converting a Perl 5 script I wrote to Perl 6
lue Ooh! I got it! sub postcircumfix:<.[ ]>(@a, *@b) { @a[@b] } :D 20:16
masak: ^^^
jaffa4 Are you sure?
hoelzro tomorrow, looking at taking advantage of the free time I'm finding in my schedule to work on stuff =)
Grrrr 26
lue (I can, in fact, proceed to step 11 if I get just the atime and punct fails, right?) 20:17
hoelzro so I'm having an issue with grammars/actions: gist.github.com/hoelzro/7203812
isn't make at the TOP method supposed to determine what parse returns? 20:18
timotimo you need to .ast on that
hoelzro G.parse(...).ast?
timotimo i think so
hoelzro ah ha!
timotimo++
TimToady lue: there's no bosses here; you'll have to use your judgement :)
hoelzro thank you!
timotimo if all things were so easy to solve :)
hoelzro indeed =) 20:19
lue TimToady: My judgement says yes; my inexperience says "since this is my first time releasing..." :)
TimToady you could put a "known issues" into the release notes 20:20
lizmat
.oO( .ast should really be called .made, as it is the result of the inner make() )
arnsholt Well, that makes good punny sense, but sometimes practical names make sense too =) 20:21
masak lue: +1 on TimToady's "known issues"
lue I could; but the atime failure is if you have a noatime drive, and the punct failure is because I can't get 2013.10 rakudo to use the newer tests (which I had fixed because I deemed the tests wrong)
timotimo hoelzro: github still doesn't use the perl6 highlighter, at least not for nqp files :(
hoelzro timotimo: I know =( so frustrating!
lue (I will probably put a "known issues" in the announcement though.) 20:22
hoelzro hmm
masak lue: will users who download the new R* see test failures?
lizmat arnsholt: I would think that hoelzro would have thought about a .made sooner than about an .ast
hoelzro I *may* have forgotten to put it in linguist...
Mouq lizmat: I may have figured it out
lue If they run spectest, yes.
lizmat Mouq++
(for persevering) 20:23
timotimo lue: if you put a known issues section there, you can include *<>, *{} and *[] not currying correctly
lue The specific punct failure, by the way, is :Punctuation delivering § and ¶ (category Po in Unicode), and the old tests not having those two characters in the expected.
moritz submitted the JVM bug at icedtea.classpath.org/bugzilla/show...gi?id=1590
lue hopes updating the announcement doesn't require an update of the tarball
it does :/ 20:24
arnsholt One more squashed NQP bug and we're down to 25 open tickets
moritz arnsholt++ 20:25
lue Plan: finish spectest, update announcement, new tarball, go through the testing procedure again (but SKIP the spectest; that shouldn't change), 11.
moritz lue: sounds fine
lue (I could probably just stop the spectest know, because I don't see how I'll randomly get new failures.)
s/know/now/
hoelzro why is it that action methods must explicitly take $/ as an argument? 20:28
you'd think that would be implied
timotimo nah, how would you tell an action class apart from a non-action class?
moritz scoping
timotimo and you can have non-action methods in an action class, too
hoelzro I see
moritz $/ is lexically scoped
lue method action_rule($not-the-slash) { $not-the-slash<hash> }
^ works too 20:29
Mouq Oh
moritz also if you want to do a regex match, you'd better not name the argument $/ :-)
Mouq ".=" in <.*> was there so that we could do @list».=meth
lizmat moritz: so would $*/ not work then, as it is marked as dynamic, afaik ?
I mean CALLER::$/ ? 20:30
r: say $/.VAR.dynamic
camelia rakudo-parrot 3cef56: OUTPUT«True␤»
..rakudo-jvm 882e33: OUTPUT«Unhandled exception: java.lang.OutOfMemoryError: GC overhead limit exceeded␤ in (gen/jvm/stage2/QAST.nqp:288)␤ in map_classlib_core_op (gen/jvm/stage2/QAST.nqp:287)␤ in (gen/jvm/stage2/QAST.nqp:1889)␤ in (gen/jvm/stage2/QAST.nqp)␤ in (src…»
moritz $CALLER::</> maybe
what, $/ is dynamic? 20:31
lizmat the thing is, that $/ is usually only needed for error messages
yup, as is $_ and $!
r: say $_.VAR.dynamic
moritz
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True␤»
lizmat r: say $!.VAR.dynamic
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«True␤»
moritz 's world view collides
erm, collapses
it collapses so hard, I can'T even type anymore
lizmat sorry 20:32
:-)
lizmat starts picking up the pieces
lue r: my $::<###!@#@###!!$$> = 42;
camelia rakudo-parrot 3cef56: OUTPUT«===SORRY!===␤ResizablePMCArray: index out of bounds!␤»
..rakudo-jvm 882e33: OUTPUT«===SORRY!===␤VMArray: Index out of bounds␤»
hoelzro hmm
lue std: my $::<###!@#@###!!$$> = 42;
camelia std 8adbc60: OUTPUT«ok 00:01 122m␤»
moritz how to make the JVM not segfault, 101: enable core dumps
hoelzro feels like there *must* be shorthand for :foo($<foo>) 20:33
moritz hoelzro: $/.hash<foo>:p
or just $<foo>:p if that works
hoelzro =O
lue what does the p adverb do? :P
moritz if not, $/<foo>:p
lue: return a Pair, not just the value
hoelzro bwa 20:34
blown away =)
masak moritz: ISTR $_, $/, $! are "lexical if <mumble>, otherwise dynamic".
moritz meh, I want intuitive semantics *and* clear, simple rules! :-) 20:35
and more wows than WATs 20:36
and an OOM killer that takes that fat, ever-growing memory eater instead of my window manager 20:37
raiph felixge.de/2013/03/11/the-pull-request-hack.html # the wisdom of liberal commit bits 20:38
masak raiph: au++ would be proud :) 20:39
raiph: she was 8 years ahead of her time.
moritz raiph: we do that for projects under the perl6/ organization
(at least I do)
for rakudo, there are some non-trivial legal aspects to consider
raiph yeah, I've seen it increasingly regularly over the last 2 years
I've always wanted more of au's thinking applied 20:40
masak and indeed, HN traces the idea back to au++: news.ycombinator.com/item?id=5358645
moritz au++ was even more "agressive", handing out commit bits to people who might possibly want to commit
masak "trust the anarchy" ♡
moritz I remember seeing Guido van Rossum in the list of pugs commit bitters 20:41
masak ...because he badmouthed Perl 6? :P
masak .oO( "I don't know who this Guido guy is, but he sounds like he wants to improve things" ) 20:42
20:42 Rotwang left
raiph do p5 contributors have to sign a CLA? 20:42
hoelzro hmm
I can do [:r ...] to lexically enable ratcheting, right?
moritz hoelzro: yes 20:43
hoelzro so how can I disable ratcheting lexically?
moritz [:!r ... ]
hoelzro ah ha
that makes sense =)
moritz++
moritz don't we also have modifiers for quantifiers that enable backtracking? 20:44
moritz hasn't read through S05 in a long time
masak moritz: yes.
moritz is that the * suffix? 20:45
masak the "long forms" are *! (greedy) *? (frugal) and *: (ratchet). 20:46
and similarly for all the other quantifiers. 20:47
depending on the context you're in, you can use just *
moritz ah, !
masak yeah.
which works really well for "greedy" :)
ISTR *+ is a p5-compatible STD-spec'd synonym for *? 20:48
jnthn back from dinner :) 20:49
masak jnthn! \o/
hoelzro moritz: I have a goal of summarizing it, someday...
moritz hoelzro: I have a goal of understanding it, someday 20:50
hoelzro heh
moritz I think I have come pretty far, but there are still areas that scare me 20:51
masak moritz: pro tip: implement a grammar engine :>
moritz masak: like, one that does transitive LTM, and bactracking into the NFA? 20:52
moritz shudders
masak moritz: well, one must start somewhere. 20:53
moritz: those are likely the last few features one implements.
moritz last, because after that you give up programming due to constant frustration :-) 20:54
masak yeah, but it's *good* frustration!
20:54 |newbie| joined
moritz I'm picking up good frustration ♫ 20:57
20:57 jaffa4 left, jferrero left, jferrero joined 20:58 colomon left 20:59 colomon joined
masak moritz: now *that* should be the programmer's mantra, instead of "aaaargh don't interrupt me while I'm thinking very complicated thoughts" :) 21:00
moritz: programming, at its very heart, is frustration.
jnthn moritz: You don't "backtrack into the NFA", you just throw away cursors ;)
lue mantra: I will PERSONALLY torch AMD's factories if you don't work THIS TIME! 21:01
FROGGS .oO( Please consider the environment before throwing away cursors! ) 21:02
lizmat I thought cursors were made of pure recycled electrons only ? 21:03
FROGGS lizmat: used electrons... it is the same like with batteries :|
moritz jnthn: so ther NFA enerates a passing cursor each time it is in an accepting state? 21:04
FROGGS: the grow weaker over time?
21:04 Rotwang joined
FROGGS yeah... positrons would grow weaker back into time though 21:04
21:04 kaare_ left
jnthn moritz: Not quite. In alternation mode it just queues up the branches on the bstack in reverse order and immediately backtracks to hit the first one. 21:04
moritz: In protoregex mode it returns an ordering 21:05
moritz: And we just call the candidate rules in that order
Mouq r: say (my@=^10).=&postcircumfix:<[ ]>(2) 21:06
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«2␤»
moritz jnthn: and rakudo's bstack roughly coresponds to a list of cursors in the model advocated by S05, right?
FROGGS humm, I read "parrot 3cef56" ad "parrot sketch" 21:07
jnthn moritz: Yeah, the bstack is like, "Cursors ain't that cheap, let's go integering" :)
moritz sounds very integer :-) 21:08
lue is there a windows equivalent of "noatime" ? 21:10
FROGGS lue: serverfault.com/questions/33932/how...fs-windows 21:11
(google #1)
lizmat depends on the file system no?
lue :)
[this is for the announcement's known issues section]
moritz lue: then don't mention the windows stuff 21:12
FROGGS yeah, I guess it is not that common to do that on windows
lue Oh, I just wanted to mention the non-unix equivalent of noatime, just in case. 21:13
21:13 risou_awy is now known as risou
lue Should I mention the noatime issue in the first place, in case someone runs the spectest? 21:14
moritz I wouldn't give the whole thing a high priority, since the test is a few months old iirc, and so far we didn't have any complaints 21:19
dalek ar: 344b246 | lue++ | docs/announce/2013.10.md:
Update Release Announcement
 21:23
Mouq r: say (my@=^10) .= &postcircumfix:<[ ]>((2,3,4)) #fine
camelia rakudo-parrot 3cef56, rakudo-jvm 882e33: OUTPUT«2 3 4␤»
21:23 benabik left
Mouq r: say (my@=^10) .= '&postcircumfix:<[ ]>'((2,3,4)) #less so 21:23
lue I would love some review to tell me what issues should be added/removed/reworded
camelia rakudo-jvm 882e33: OUTPUT«No such method '' for invocant of type 'Array'␤ in method dispatch:<.=> at gen/jvm/CORE.setting:1129␤ in block at /tmp/ZUohq1LKAG:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any co…»
..rakudo-parrot 3cef56: OUTPUT«No such method '' for invocant of type 'Array'␤ in method dispatch:<.=> at gen/parrot/CORE.setting:1138␤ in block at /tmp/Ws_a7vFgTl:1␤ in any at /tmp/Ws_a7vFgTl:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/…»
Mouq Err, what am I doing?
r: say (my@=^10).'&postcircumfix:<[ ]>'((2,3,4)) #less so
camelia rakudo-jvm 882e33: OUTPUT«No such method '&postcircumfix:<[ ]>' for invocant of type 'Array'␤ in block at /tmp/0qQfzZIAu9:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.n…» 21:24
..rakudo-parrot 3cef56: OUTPUT«No such method '&postcircumfix:<[ ]>' for invocant of type 'Array'␤ in block at /tmp/YMtr9_IGhS:1␤ in any at /tmp/YMtr9_IGhS:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ in any evalfi…»
Mouq lue: The :sigspace thing is more like "the only place whitespace will be significant is after atoms that match something" 21:27
Although your description is technically correct 21:28
lue Mouq: that's what it already says "leading whitespace NOT" ~~ "non-leadng whitespace IS"
21:28 iSlug left
lue (I didn't write that bit though; it's mostly left as-is. Although I do agree, I did have initial trouble reading it.) 21:29
I'd like someone to tell me the notability of L81 of the announcement.
Mouq Oh, I understand what it's saying now 21:30
lue & --until="top of the hour"
21:30 risou is now known as risou_awy 21:37 logie joined 21:41 logie left 21:50 sqirrel left 21:51 donaldh_ joined
lue All the modules pass, so unless anyone has anything else to say about the announcement (I'll wait a few minutes), the current version is what will come in the tarball. 21:52
lizmat argh, "is deprecated" should read "is DEPRECATED" :-( 21:55
well, not so important 21:59
lue lizmat: got it.
dalek ar: bec08fa | lue++ | docs/announce/2013.10.md:
Spelling fix in Announcement.
 22:01
lue will now tarball and test once more, sans spectest 22:02
22:02 rurban joined
timotimo Greatly reduce object creation during Regex parsing. <- hm, but this is only during code parsing, no? well, and anything that uses ident 22:03
japhb__ Which is a lot of things. 22:04
lue timotimo: I assume all regexes are affected, though I could add a "especially for Perl 6 itself" or something to that effect.
japhb__ timotimo: Do you have local changes to perl6-bench?
timotimo japhb__: i've been trying to change components.json to use the new configure.pl system 22:05
and also try to get nqp-moarvm updated
japhb__ Are you timotimo on github? 22:06
timotimo no, i'm timo 22:07
japhb__ Ah, OK
timotimo i got in early enough to grab that short name :)
22:08 rurban left
lue ✔ ⑨ create tarball 22:08
lizmat lue++
japhb__ timotimo: In honor of au++, you now have a perl6-bench commitbit.
lue (I didn't add the "esp. Perl 6" though, I don't think it's necessary)
timotimo \o/
japhb__: i wish i could try out the changes to components.json, but i'll have to massively downgrade rakudo or something >_> 22:09
japhb__ If you have any ideas on dealing with components with changing build instructions, I'm all ears. Right now, I'm imagining specifying multiple sets of instructions per component, each with a "minimum git revision" that set of instructions should apply to. 22:11
Downgrade rakudo? Why?
timotimo oh
to get a working "run" again
japhb__ Oh, right.
timotimo i don't think we are at a point where we need that feature, though :)
japhb__ Yeah, that's more of something to care about when we're watching for regressions. 22:12
timotimo yup
japhb__ Still, never hurts to vaguely plan ahead. ;-)
OK, run bugs I'm currently aware of: 1) JVM getting boolean value of run result wrong, 2) Run not passing on updated %*ENV, even though shell does. Any others? 22:14
timotimo calling run often will cause segfaults
or "memory corruption" of any sort
22:15 bluescreen10 left
FROGGS on parrot only, right? 22:15
timotimo yes
couldn't test it on moar yet ;)
FROGGS *g*
we're working on it :o)
btw, if someone would help me, we could fix run()
we only need an implementation (in C) for spawn 22:16
22:16 iSlug joined
japhb__ OK, so segfaulting run on Parrot makes 3. 22:16
FROGGS and only for linux... since the windows version is done
japhb__ Are there any known shell() bugs?
22:16 stevan_ joined
FROGGS not that I know 22:16
japhb__ OK, how many of the shell fixes can we steal for run?
22:16 denis_boyun left
FROGGS all 22:17
timotimo oh, JVM doesn't have open pipe yet
japhb__ Does donaldh's pipe sketch work on JVM 22:18
?
timotimo didn't try, my rakudo is now going to be too outdated :)
FROGGS timotimo: btw, I believe we can fix issue #3 by replacing "$state but $state" by Process::State as specced 22:19
donaldh_ japhb_: yes :-) 22:20
timotimo: JVM has nqp::openpipe
FROGGS this would return the return code, but boolifies to return_code == 0
japhb__ Then by gosh, let's make it happen!
timotimo doesn't that require the same sort of thing that we need for IntStr and friends?
like, do we have to derive it from Str and Bool? 22:21
japhb__ Ewww.
timotimo r: class Process::State is Int is Bool { } 22:22
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!===␤java.lang.ClassFormatError: Duplicate method name&signature in class file __P6opaque__95␤»
..rakudo-parrot 995207: OUTPUT«===SORRY!===␤Duplicate box_target for native int␤»
22:22 stevan_ left
timotimo not Str and Bool, fwiw 22:22
jnthn I sure hope Process::State ain't gonna derive from the two of those :P
timotimo r: class Process::State is Int is Bool is repr(Int) { }
22:22 stevan_ joined
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!===␤Cannot unbox a type object␤» 22:22
..rakudo-parrot 995207: OUTPUT«===SORRY!===␤Cannot unbox a type object as a native str␤»
jnthn timotimo: The problem there is that Int and Bool both want to own the slot that unboxes as a native int. 22:23
japhb__ class Process::State is Int { method Bool { self == 0 } }; Process::State $a = 0; say so $a; Process::State $b = 1; say so $b;
jnthn japhb__: that's more like it :)
timotimo yeah, and i have no idea how to specify that kind of thing
japhb__ r: class Process::State is Int { method Bool { self == 0 } }; Process::State $a = 0; say so $a; Process::State $b = 1; say so $b;
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/FbD5VcuWGZ␤Two terms in a row␤at /tmp/FbD5VcuWGZ:1␤------> od Bool { self == 0 } }; Process::State ⏏$a = 0; say so $a; Process::State $b = 1␤ expecting any of:␤ …»
..rakudo-parrot 995207: OUTPUT«===SORRY!=== Error while compiling /tmp/Ga62JEBT0B␤Two terms in a row␤at /tmp/Ga62JEBT0B:1␤------> od Bool { self == 0 } }; Process::State ⏏$a = 0; say so $a; Process::State $b = 1␤ expecting any of:␤ …»
FROGGS r: class P::S { method Numeric() { 42 }; method Bool { False } }; my $state = P::S.new; say ?$state
timotimo needs a my
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«False␤»
FROGGS r: class P::S { method Numeric() { 42 }; method Bool { False } }; my $state = P::S.new; say +$state
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«42␤»
japhb__ r: class Process::State is Int { method Bool { self == 0 } }; my Process::State $a = 0; say so $a; my Process::State $b = 1; say so $b; 22:24
FROGGS r: class P::S { method Numeric() { 42 }; method Bool { False } }; my $state = P::S.new; say $state unless $state
japhb__ r: class Process::State is Int { method Bool { self == 0 } }; my Process::State $a = 0; say so $a; my Process::State $b = 1; say so $b;
camelia rakudo-parrot 995207: OUTPUT«Type check failed in assignment to '$a'; expected 'Process::State' but got 'Int'␤ in block at /tmp/gg2Ef2SOnK:1␤ in any at /tmp/gg2Ef2SOnK:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ …»
..rakudo-jvm 882e33: OUTPUT«Type check failed in assignment to '$a'; expected 'Process::State' but got 'Int'␤ in block at /tmp/yrUOSGMNhh:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/s…»
rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«P::S.new()␤»
rakudo-jvm 882e33: OUTPUT«Type check failed in assignment to '$a'; expected 'Process::State' but got 'Int'␤ in block at /tmp/cc8rDhDBg6:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/s…»
..rakudo-parrot 995207: OUTPUT«Type check failed in assignment to '$a'; expected 'Process::State' but got 'Int'␤ in block at /tmp/7BmIjsmxiP:1␤ in any at /tmp/7BmIjsmxiP:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ …»
22:24 spider-mario joined
japhb__ r: class Process::State is Int { method Bool { self == 0 } }; my Process::State $a .= new(0); say so $a; my Process::State $b .= new(1); say so $b; 22:25
camelia rakudo-jvm 882e33: OUTPUT«Default constructor for 'Process::State' only takes named arguments␤ in method dispatch:<.=> at gen/jvm/CORE.setting:1171␤ in block at /tmp/u5rewtmoLz:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.n…»
..rakudo-parrot 995207: OUTPUT«Default constructor for 'Process::State' only takes named arguments␤ in method new at gen/parrot/CORE.setting:809␤ in method new at gen/parrot/CORE.setting:804␤ in method dispatch:<.=> at gen/parrot/CORE.setting:1187␤ in block at /tmp/NkyTRG…»
FROGGS japhb: it is not specced as being an Int
timotimo oh?
FROGGS it can give you the PID too
but numifies/gistifies to state, and boolifies, well, you know
timotimo does it also do the +> 8 thing? 22:26
FROGGS I dont think so
japhb__ r: class Process::State is Int { has int $.state is box_target; has int $.pid; method Bool { self == 0 } }; my Process::State $a .= new(:state(0), :pid(42)); say so $a; my Process::State $b .= new(:state(1), :pid(23)); say so $b; 22:27
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!===␤java.lang.ClassFormatError: Duplicate method name&signature in class file __P6opaque__96␤»
..rakudo-parrot 995207: OUTPUT«===SORRY!===␤Duplicate box_target for native int␤»
japhb__ r: class Process::State { has int $.state is box_target; has int $.pid; method Bool { self == 0 } }; my Process::State $a .= new(:state(0), :pid(42)); say so $a; my Process::State $b .= new(:state(1), :pid(23)); say so $b;
camelia rakudo-jvm 882e33: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: Mu *%_)␤ in any at gen/jvm/BOOTSTRAP.nqp:1201␤ in sub infix:<==> at gen/jvm/CORE.setting:3978␤ in method Bool at /tmp/peodoN2WvE:1␤ in block at /tmp/peodoN2WvE:1␤ in any ev…»
..rakudo-parrot 995207: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: Mu *%_)␤ in any at gen/parrot/BOOTSTRAP.nqp:1218␤ in any at gen/parrot/BOOTSTRAP.nqp:1209␤ in method Numeric at gen/parrot/CORE.setting:994␤ in sub infix:<==> at gen/parrot…»
japhb__ r: class Process::State is Cool { has int $.state is box_target; has int $.pid; method Bool { self == 0 } }; my Process::State $a .= new(:state(0), :pid(42)); say so $a; my Process::State $b .= new(:state(1), :pid(23)); say so $b; 22:28
camelia rakudo-jvm 882e33: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: Mu *%_)␤ in any at gen/jvm/BOOTSTRAP.nqp:1201␤ in sub infix:<==> at gen/jvm/CORE.setting:3978␤ in method Bool at /tmp/zJHeuS3ylq:1␤ in block at /tmp/zJHeuS3ylq:1␤ in any ev…»
..rakudo-parrot 995207: OUTPUT«Cannot call 'Numeric'; none of these signatures match:␤:(Mu:U \v: Mu *%_)␤ in any at gen/parrot/BOOTSTRAP.nqp:1218␤ in any at gen/parrot/BOOTSTRAP.nqp:1209␤ in method Numeric at gen/parrot/CORE.setting:994␤ in sub infix:<==> at gen/parrot…»
japhb__ Harumph.
FROGGS ETOOMUCHMAGIC 22:29
japhb__ r: class Process::State is Cool { has int $.state is box_target; has int $.pid; method Bool { $.state == 0 }; method Numeric { $.state }; }; (my Process::State $a .= new(:state(0), :pid(42)); say so $a; my Process::State $b .= new(:state(1), :pid(23)); say so $b;
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/_iZuWb2oFR␤Unable to parse expression in parenthesized expression; couldn't find final ')' ␤at /tmp/_iZuWb2oFR:1␤------> .= new(:state(1), :pid(23)); say so $b;⏏<EOL>␤ …»
..rakudo-parrot 995207: OUTPUT«===SORRY!=== Error while compiling /tmp/Y3JivOpK_9␤Unable to parse expression in parenthesized expression; couldn't find final ')' ␤at /tmp/Y3JivOpK_9:1␤------> .= new(:state(1), :pid(23)); say so $b;⏏<EOL>…»
japhb__ r: class Process::State is Cool { has int $.state is box_target; has int $.pid; method Bool { $.state == 0 }; method Numeric { $.state }; }; my Process::State $a .= new(:state(0), :pid(42)); say so $a; my Process::State $b .= new(:state(1), :pid(23)); say so $b;
lizmat gnight, #perl6!
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«True␤False␤»
japhb__ o/ lizmat
22:29 Rotwang left
japhb__ FINALLY. SHEEESH. 22:29
Sorry for the bot spam, folks. 22:30
timotimo japhb__: did you notice how you could use a native int, but not have to make a custom BUILD? :)
japhb__ You made that work? 22:31
timotimo++
|newbie| timotimo: is it done?
timotimo yes, just today :D
well, between yesterday and today i guess
|newbie| timotimo: is it done?
timotimo is what exactly done? 22:32
|newbie| the panda change
timotimo i'm not sure i follow
22:33 ajr_ left
|newbie| panda has a bug, it cannot delete files 22:33
timotimo ah, yes
FROGGS |newbie|: this was fixed right after the 2013.09 release
|newbie| so it is not in any relase 22:34
FROGGS |newbie|: not yet, but possible today or tomorrow
well, it is in the 2013.10 compiler release fwiw
|newbie| because of the new release? 22:35
22:35 ajr joined
FROGGS lue is working on a new star release, contain this and a lot of other things 22:35
22:36 ajr is now known as Guest55742
lue I suspect 12 and the rakudo.org part of 13, specifically the availability of its permission-holders, are going to be the only blockers to 2013.10 at this point. 22:36
FROGGS containing*
22:36 Guest55742 is now known as ajr_
|newbie| what is the bug? Permission of folders, right? 22:37
timotimo unlink didn't persist if permissions didn't allow deletion 22:38
now it deletes files even if it's not permitted to, basically
well, that's worded poorly
|newbie| who is the possible?
how
FROGGS this is possible on windows :o) 22:40
so, unlink behaves now exactly like Perl 5's unlink...
|newbie| what are the permissiion incorrect, anyway?
SHould not be those be fixed?
what ->why
FROGGS we decided that P5's behaviour is okay, and it fixed the problem 22:42
|newbie| why did you choose this way? 22:43
FROGGS I said that right now
|newbie| you found one solution, fixing the unlink 22:44
did you consider the other way as well?
22:45 dmol left
FROGGS yes, but it would be rather annoying if you had to check+change the flags of hundreds of files just to delete them afterwards 22:45
so, if you can gain control over a file, you sort of have already control over it
so it does not make much sense to forbid deleting it 22:46
22:46 xinming left
|newbie| ok bye 22:47
22:47 |newbie| left 22:48 xinming joined 22:49 donaldh_ left
Mouq lizmat: I need to → bed, but basically the problem is that it ends up being called as $var."&postcircumfix<[ ]>"(|whatev), which, beacuse of the quotes, will currently always call as a method. Adding a dispatch:<quote> method and using it at Actions L3863 should work, but I'm not sure if that's really the best way to do this 22:49
timotimo man, rakudo's ram usage ... 22:50
Mouq Maybe there could be two dispatch methods for <.=> 22:51
Mouq out
FROGGS gnight Mouq
jnthn Wait, what's that trying to fix? 22:52
timotimo the @a.=[1,2,3] thing 22:53
FROGGS I believe some splice mutator thing
yeah
jnthn hmmm
Well, i fit's gonna work, it should be a general solution for postfixes too 22:54
timotimo not all postcircumfix are subs :\
22:55 Mouq left
jnthn yeah, postcircumfix:<( )> is the odd one out now :( 22:57
22:59 raiph left 23:01 xinming left
lue so, unless someone insists that I run the exact same spectest on the exact same rakudo... 23:01
✔ ⑩ Verify tarball
timotimo desists 23:02
23:02 xinming joined 23:04 raiph joined, BenGoldberg joined
lue ✔ ⑪ Tag Star "2013.10" 23:04
Unsurprisingly, I have not the permissions for 12. 23:05
BenGoldberg preflex: seen yoleaux 23:06
preflex yoleaux was last seen on #perl6 3 days, 3 hours, 44 minutes and 13 seconds ago, saying: 18:23Z <diakopter> Mouq: awesome work on stdsigspace Mouq++
lue pmichaud, jnthn, masak, PerlJam, tadzik, moritz : if one of you could scp the R* tarball, that'd be great :) the md5sum of the tarball, so you can verify generation went smoothly, is ef8d1e47d11c5d5ad663405897fb4ee0. 23:07
(if you could also put the announcement on rakudo.org that'd be super)
jnthn lue: Sadly, my key for that is sat on my machine at home, which is a long way away. 23:08
lue :(
you couldn't, say... scp the key, could you? :)
FROGGS win *cough* dows 23:09
jnthn The machine isn't even running, given I'm away for over a week.
lue you couldn't, say... scp a power-on command, could you? :P 23:10
lue knows moritz will be around late tonight from his POV, at least.
"his" being me, that is. 23:12
23:14 risou_awy is now known as risou 23:17 raiph left 23:32 risou is now known as risou_awy 23:38 iSlug left 23:39 rurban joined 23:41 spider-mario left
dalek p: 1161844 | dwarring++ | examples/rubyish/ (3 files):
rubyish - implement compound-statement/multi-line interpolation
 23:43
23:43 benabik joined 23:45 Celelibi joined 23:50 btyler left

Powered by Cro and the Raku® Programming Language.

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

Thank you!