»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:00 whiteknight left 00:25 thou_ left 00:28 arlinius joined 00:30 thou joined, tokuhiro_ joined 00:43 simcop2387 left 00:46 simcop2387 joined 00:49 Chillance left 01:13 hypolin joined 01:15 Targen left 01:22 thou left 01:29 thou joined 01:48 anuby joined 02:04 kst left 02:08 thou left 02:09 rafl joined 02:11 kst joined
diakopter short backlog is 02:24
jnthn morning o/ 02:30
02:43 popl left
[Coke] jnthn: ... almost. 02:43
jnthn [Coke]: In Japan :P 02:46
[Coke] ah, say hi to my cousin. :) 02:47
"surely you'll run into him while you're there!" 02:48
jnthn Yeah, I mean, there's only 30 million people in Tokyo :P 02:49
02:58 _jaldhar_ joined
dalek kudo/nom: dca2ab5 | jonathan++ | src/Perl6/ (2 files):
Make a real pad etc. for a signature literal, to prevent various things exploding in nasty ways.
 03:01
kudo/nom: 61cbec6 | jonathan++ | src/ (3 files):
Give signatures a backlink to the code object they are associated with.
kudo/nom: 6cd3432 | jonathan++ | src/ (3 files):
Implement capture ~~ signature smart-matchng.
kudo/nom: 8f5fc4d | jonathan++ | t/spectest.data:
Turn on S03-smartmatch/capture-signature.t.
jnthn uh...where's the 5th commit? :) 03:02
huh, it shows up on github...
[Coke] jnthn: he's... about 90 min SW of tokyo by car. if I were a guessing man. 03:03
lazy dalek.
jnthn: oddly, shows up in #parrot 03:06
jnthn oddness 03:09
03:12 fgomez left
jnthn lunch & 03:14
03:25 fgomez joined 03:34 arlinius left 04:05 thou joined 04:10 tokuhiro_ left 04:22 telex left 04:26 telex joined 04:33 telex left 04:40 telex joined 04:49 skids left
moritz \o 05:12
jnthn++ # smart-matching against signature literals
dalek kudo/nom: 9af8f20 | moritz++ | docs/ChangeLog:
update ChangeLog
 05:13
05:24 wk left 05:34 kaleem joined 05:38 birdwindupbird joined 05:41 Maddingue left, Maddingue joined 05:42 arnsholt left, arnsholt joined 06:04 wtw joined 06:10 Pleiades` left 06:19 stepnem joined 06:23 Pleiades` joined 06:38 wk joined 06:45 preflex left 07:00 thou left 07:06 fhelmberger joined 07:09 brrt joined 07:11 ServerSage joined 07:14 Yappocall left, Yappocall joined 07:20 kurahaupo left 07:21 kresike joined
kresike good morning all you happy perl6 people 07:21
07:22 flightrecorder joined 07:33 Psyche^ joined, Patterner left, Psyche^ is now known as Patterner
moritz \o kresike 07:41
sorear o/ 07:45
07:49 FROGGS joined 08:05 fgomez left
brrt \o kresike 08:05
08:12 fgomez joined 08:17 fgomez left 08:19 fgomez joined 08:24 mucker joined 08:26 daniel-s_ left 08:27 daxim joined 08:30 fgomez left 08:31 cognominal joined, fgomez joined 08:32 dakkar joined 08:40 cognominal left 08:48 cognominal joined 08:56 rhr_ left 08:57 eternaleye left 08:58 wamba joined 08:59 cognominal left 09:03 rhr joined 09:06 cognominal joined
jnthn back 09:07
arnsholt o/
Did you find the problem in Zavolaj with the struct tests? 09:08
jnthn Now to finish my talk :)
arnsholt: Not had chance yet.
arnsholt Cool
I've not had much time the last month or so either
09:11 markov1 left 09:12 Circlepuller_ joined 09:14 Circlepuller left 09:15 cognominal left 09:16 wamba left 09:19 cognominal joined 09:30 cognominal left, mucker left
jnthn Hm, after fixing Test::Mock I'm now realizing I probably have too much material in my talk to include it without being crazy tight on time... 09:33
09:33 c1sung left 09:34 fgomez left, cognominal joined 09:35 leont joined 09:36 grondilu joined, fgomez joined
grondilu Is there a shorter way to write map &f, @a ? 09:36
09:38 taniavs666 joined
grondilu with metaoperators or something? 09:38
lumi__ r: sub f($x) { $x + 1 }; my @a = 0..2; say @a>>.&f
p6eval rakudo 9af8f2: OUTPUT«1 2 3␤»
tadzik even shorter? :)
jnthn Note that if you use >> then it doesn't promise order, whereas map will evaluate things in order 09:39
grondilu seems cool. I'm surprised it works though, since f is not a method.
jnthn Note the & 09:40
09:40 c1sung joined
jnthn $foo.&f really means f($foo) 09:40
grondilu I had no clue. Good to know.
09:42 cognominal left 09:43 mathw_ is now known as mathw 09:48 cognominal joined 09:56 cognominal left 10:01 cognominal joined 10:05 hypolin left
pmichaud TimToady: Happy Birthday! 10:07
moritz ooh 10:08
TimToady++
grondilu Say I have two functions f and g with an identical signature. How do I call f inside g with the exact same arguments, without tediously copying the arguments?
leont Happy birthday!
moritz r: sub f($a, $b) { say ::MY.keys }; f 1, 2 10:09
tadzik Happy Birthday TimToady! \o/
p6eval rakudo 9af8f2: OUTPUT«0␤»
moritz r: sub f($a, $b) { say MY.keys }; f 1, 2
p6eval rakudo 9af8f2: OUTPUT«0␤»
jnthn grondilu: Capture syntax
moritz meh, how did that work?
jnthn moritz: MY::.keys
moritz r: sub f(|c) { g(|c) }; sub g($a, $b) { say "$a|$b" }; f(1, 2) 10:10
jnthn grondilu: sub g(|c) { f(|c) }
p6eval rakudo 9af8f2: OUTPUT«1|2␤»
moritz r: sub f($a, $b) { say MY::.keys }; f 1, 2
p6eval rakudo 9af8f2: OUTPUT«$a $b $_ $/ $! call_sig $*DISPATCHER &?ROUTINE RETURN␤»
moritz r: sub f($a, $b) { g(|::('call_sig')) }; f 1, 2; sub g($a, $b) { say "$a|$b" }
p6eval rakudo 9af8f2: OUTPUT«No such method 'FLATTENABLE_LIST' for invocant of type 'CallContext'␤ in sub f at /tmp/aoC8uy882T:1␤ in block at /tmp/aoC8uy882T:1␤␤»
moritz nice try though :-)
jnthn Uh, don't mess with call_sig :)
We should really magically hide that or something... :) 10:11
moritz r: sub f(|c ($a, $b)) { g(|c) }; sub g($a, $b) { say "$a|$b" }; f(1, 2)
p6eval rakudo 9af8f2: OUTPUT«1|2␤»
grondilu do I really have to explicitely write |c ($a, $b)? Isn't there a default name for the capture of the current procedure signature? 10:13
10:15 cognominal left
pmichaud r: sub f($a, $b) { g( |callframe.args ); }; sub g($a, $b) { say "$a|$b" }; f(1,2); 10:15
p6eval rakudo 9af8f2: OUTPUT«No such method 'args' for invocant of type 'CallFrame'␤ in sub f at /tmp/dFkwryI0ic:1␤ in block at /tmp/dFkwryI0ic:1␤␤»
pmichaud r: sub f($a, $b) { &g.callsame }; sub g($a, $b) { say "$a|$b" }; f(1,2); 10:16
p6eval rakudo 9af8f2: OUTPUT«No such method 'callsame' for invocant of type 'Sub'␤ in sub f at /tmp/MA91S9rIaM:1␤ in block at /tmp/MA91S9rIaM:1␤␤»
pmichaud looks like NYI 10:17
might even be EASY :-)
grondilu "callsame" would be nice 10:18
Timbus wouldnt a wrapper function work. or even just using slurpy or @_ params?
pmichaud .callsame is mentioned in S06
S06:2859 10:19
10:20 wamba joined
jnthn how on earth would &g.callsame work? 10:20
10:20 anuby left
pmichaud it calls &g with the arguments of the current Routine 10:20
(i.e., it doesn't have to go through a dispatcher, if &g is single dispatch) 10:21
jnthn I really don't like to overload the meanings there
10:21 kurahaupo joined
jnthn callsame is always about a currently ongoing dispatch so far. 10:21
As in, iterating a candidate list.
pmichaud S06 says that's "callsame without an explicit destination" 10:22
10:22 cognominal joined
pmichaud i.e., callsame (the function) uses the currently ongoing dispatch, while .callsame (the method) starts a new dispatch 10:22
jnthn You can say sub f(|c, $a, $b) { } fwiw 10:23
pmichaud also, S06 seems to have a fair number of |$ fossils. 10:25
&thermo.wrap( -> |$args { callwith(|$args) * 2 } );
10:26 domidumont joined, kurahaupo1 joined, kurahaupo left
felher TimToady: terjbe.ui ! 10:27
10:34 kurahaupo1 left 10:48 grondilu left, taniavs666 left
moritz r: sub f(|c, $a, $b) { } 10:49
p6eval rakudo 9af8f2: OUTPUT«===SORRY!===␤Cannot put required parameter $a after variadic parameters␤at /tmp/DCfUbFDaFD:1␤»
moritz r: sub f(|c, ($a, $b)) { }
p6eval rakudo 9af8f2: OUTPUT«===SORRY!===␤Cannot put required parameter after variadic parameters␤at /tmp/81xyMGmwAS:1␤»
moritz r: sub f(|c ($a, $b)) { }
p6eval rakudo 9af8f2: ( no output )
jnthn Hmm 10:52
std: sub f(|c, $a, $b) { }
p6eval std 77327a4: OUTPUT«Potential difficulties:␤ $b is declared but not used at /tmp/HDIlbmLdvn line 1:␤------> sub f(|c, $a, ⏏$b) { }␤ $a is declared but not used at /tmp/HDIlbmLdvn line 1:␤------> sub f(|c, ⏏$a, $b) { }␤ok 00:00 43m␤»
11:00 leont left 11:01 am0c left 11:02 gootle left
jnthn argh, prompt is broken? 11:04
Did we release Rakudo with that?
It needs Ctrl+Z here to actually react to the input :/ 11:05
11:05 Su-Shee_ joined
FROGGS I heard ppl talking about the issue, so it really looks like it is in the release 11:06
11:06 Circlepuller joined
jnthn ah crap 11:06
Did it come from the Parrot IO changes?
I'd mistakenly understood it to be dealt with already and hoped that at worst it was Windows only.
FROGGS ohh, I thought it was windows only 11:07
jnthn Oh.
TimToady problem on linux too
11:07 Circlepuller_ left
jnthn arrrgh 11:07
FROGGS well, release early, release often 11:08
jnthn
.oO( revert early, revert often? )
FROGGS hehe
11:09 Su-Shee left
jnthn tries the lastest Parrot to make sure it isn't actually already fixed 11:15
11:16 Su-Shee_ is now known as Su-Shee
jnthn Actually, just saw the version history and don't see any applicable patches. 11:18
The REPL calls readline_interactive, not readline. Maybe only the first was fixed. 11:24
11:28 Vlavv_ left
jnthn No, switched .get to use readline_interactive doesn't help either :/ 11:36
Did Rakudo actually get any value out of the Parrot IO changes? What if I shove us back to a Parrot version before the changes? 11:38
phenny: tell pmichaud I think we can't put out a Star release until the prompt issue is resolved one way or another. It'll break a load of stuff. Debugger is broken, for example. 11:39
phenny jnthn: I'll pass that on when pmichaud is around.
jnthn phenny: tell pmichaud If our solution is "tie ourselves to a Parrot that wasn't busted" that's fine by me too. 11:40
phenny jnthn: I'll pass that on when pmichaud is around.
jnthn goes to find something to snack on
moritz jnthn: well, some small socket changes would need to be reverted 11:41
FROGGS we are talking about the release 2012-08, right? 11:44
moritz aye 11:45
11:46 cognominal_ joined 11:47 cognominal left
FROGGS is this commit related to it? github.com/parrot/parrot/commit/ea...6f1d0e3439 11:48
ohh, there are more commits about that 11:49
moritz there was a whole big branch merge, no?
hm, or maybe that was already done earlier 11:50
FROGGS still reading commits... there are quiet some ;o)
ya, thera was a io_cleanup1 merge on aug 27th
11:52 domidumont left, domidumont joined 11:55 skids joined
dalek rlito: ff4958d | (Flavio S. Glock)++ | t5/01-perlito/25-syntax-defined-or.t:
Perlito5 - parser - add tests for defined-or vs. match
 11:56
jnthn I'm curious why the REPL works 12:03
12:03 cognominal_ left 12:05 wamba left 12:06 taniavs666 joined 12:12 cognominal_ joined
jnthn What's really odd is that a simple say prompt '>'; doesn't seem to have the issue now 12:18
moritz define "now" 12:19
rakudo 2012.09?
jnthn I dunno, I have a local patch that s/readline/readline_interactive/
I need to undo it to be sure 12:20
but it doesn't fix things in the debugger
dalek rlito: 7cdd110 | (Flavio S. Glock)++ | / (2 files):
Perlito5 - parser - fix defined-or vs. match
rlito: 8a25831 | (Flavio S. Glock)++ | html/perlito5.js:
Perlito5 - rebuild
12:20 fgomez left 12:25 cognominal_ left 12:26 flightrecorder left, FROGGS left, cognominal_ joined, FROGGS_ joined, flightre1 joined 12:27 wamba joined 12:28 Ulti left, cognominal_ left, cognominal joined
jnthn ah, for now I just write my own non-busted prompt in the debugger so I can use it tomorrow 12:32
12:39 wamba left 12:43 cognominal left 12:49 cognominal joined 12:57 cognominal left 12:58 PacoAir joined 13:01 telex left 13:02 telex joined 13:06 cognominal joined 13:08 skids left 13:10 mikemol joined 13:15 cognominal left 13:16 wamba joined 13:21 mdupont left, mdupont joined 13:22 GlitchMr joined
GlitchMr Why almost every Twitter Bootstrap page looks so similar... that is annoying 13:23
For example, perhaps I'm wrong, but Perl 6 Maven looks like something made with Twitter Bootstrap 13:25
(but it's probably worse with things like Octopress - who changes default layout anyways?) 13:26
13:26 grondilu joined
grondilu perl6: .say for map &sin, 0, pi/2 13:28
p6eval niecza v22-6-g9e5350d: OUTPUT«0␤1␤»
..rakudo 9af8f2: OUTPUT«This type cannot unbox to a native integer␤ in method reify at src/gen/CORE.setting:5239␤ in method reify at src/gen/CORE.setting:5141␤ in method reify at src/gen/CORE.setting:5141␤ in method reify at src/gen/CORE.setting:5231␤ in method reify at src/gen/CORE.…
GlitchMr I have yet to see Octopress page which doesn't use default layout without changes and that isn't Octopress page itself
perl6: &sin.arity
p6eval rakudo 9af8f2, niecza v22-6-g9e5350d: ( no output )
GlitchMr perl6: say &sin.arity
p6eval niecza v22-6-g9e5350d: OUTPUT«1␤»
..rakudo 9af8f2: OUTPUT«0␤»
GlitchMr grondilu: it's because &sin has 0 arity 13:29
It still wasn't fixed
grondilu perl6: .say for map { sin($_) }, 0, pi/2 13:30
p6eval rakudo 9af8f2, niecza v22-6-g9e5350d: OUTPUT«0␤1␤»
grondilu .say for map { sin($_) }, 0, 0.2 ... 1
perl6: .say for map { sin($_) }, 0, 0.2 ... 1
p6eval rakudo 9af8f2: OUTPUT«Cannot call 'map'; none of these signatures match:␤:(&code, *@values)␤␤ in block at /tmp/OHjy8Ig5Ux:1␤␤»
..niecza v22-6-g9e5350d: ( no output )
13:31 birdwindupbird left
GlitchMr 0.2 ... 1? That wouldn't ever work 13:32
It's sequence of numbers... 0.2, 1.2, 2.2, 3.2, 4.2, 5.2
It never touches 1, so it's infinite
grondilu r: say 0, 0.2 ... 1
p6eval rakudo 9af8f2: OUTPUT«0 0.2 0.4 0.6 0.8 1␤»
GlitchMr Oh, I forgot about 0
Sorry
grondilu r: .1 13:33
p6eval rakudo 9af8f2: ( no output )
grondilu r: say .1
p6eval rakudo 9af8f2: OUTPUT«0.1␤»
GlitchMr grondilu: well... I think it's because subroutine itself also is part of sequence 13:34
perl6: .say for map { sin($_) }, (0, 0.2 ... 1)
p6eval rakudo 9af8f2: OUTPUT«0␤0.198669330795061␤0.389418342308651␤0.564642473395035␤0.717356090899523␤0.841470984807897␤»
..niecza v22-6-g9e5350d: OUTPUT«0␤0.19866933079506122␤0.38941834230865052␤0.56464247339503537␤0.71735609089952279␤0.8414709848078965␤»
grondilu ok
GlitchMr I've feeling that ... is too clever for its good 13:35
But sequences in Excel are even more clever, so I don't know
I have no idea how Excel can continue sequences like 1, 52, 29, 421.521, 43, 3.14, 582 13:36
grondilu Is there an actual logic in this one?
GlitchMr I have no idea 13:37
It does something, but I don't know what - I'm currently on Linux
13:38 xinming_ joined
GlitchMr LibreOffice appears to copy every sequence + 1 when the sequence isn't simple 1, 2, 3, 4, 5 13:38
But Excel does something way more complex 13:39
In LibreOffice, sequences like 1, 2, 4 are expanded with 2, 3, 5, 3, 4, 6, 4, 5, 7
13:40 colomon joined 13:41 replore joined 13:42 benabik left, xinming left
colomon 's MBP is fixed! 13:42
13:42 benabik joined
grondilu I could compile rakudo today. Thanks for your work, guys :) 13:43
GlitchMr > (1, 2, 4, sub ($a, $, $) { $a + 1 } ... *)[^10]
1 2 4 2 3 5 3 4 6 4
13:44 cognominal joined
grondilu "..." should not accept a code as a first argument, imho. 13:44
colomon grondilu: why not? 13:45
moritz grondilu: it's very versatile
grondilu so we could write map &f, 1, 2 ... 4
colomon n: say ({ rand } ... * )[^10]
p6eval niecza v22-6-g9e5350d: OUTPUT«0.11837241990415957 0.83748304882900926 0.85082873322573893 0.73417876369048785 0.10460571344225002 0.52821420018012366 0.43011551230685574 0.57213686852349754 0.85243803023008535 0.68109084324962033␤»
GlitchMr grondilu: just no...
grondilu colomon: indeed, you're right. 13:46
GlitchMr That would mean varying level of precedence
That couldn't be determined until runtime
moritz grondilu: that's a question of precedence, not of whether ... accepts a code object or not
grondilu ok ok
perl6: .say for map *+1, 0, 1 ... 3 13:47
p6eval niecza v22-6-g9e5350d: ( no output )
..rakudo 9af8f2: OUTPUT«Cannot call 'map'; none of these signatures match:␤:(&code, *@values)␤␤ in block at /tmp/tYvczcuaW7:1␤␤»
grondilu I guess we do need parens indeed. 13:48
GlitchMr But... subs with arguments should be forbidden by ...
The question is... why this even works
grondilu perl6: .say for map *+1, (0, 1 ... 3)
p6eval rakudo 9af8f2, niecza v22-6-g9e5350d: OUTPUT«1␤2␤3␤4␤»
GlitchMr Your example with { rand } makes sense 13:49
But when function has arity, it simply doesn't make sense
13:51 raiph joined
grondilu yeah, it should accept only 0 arity on first position, 1 arity on second and so on. Shouldn't it? 13:52
13:53 thou joined
colomon Why? taking too much arity too early will definitely fail, but it's very reasonable to have "too little" 13:54
n: say (0, 1, 2, { rand } ... *) [^10]
GlitchMr He probably meant "at most"
p6eval niecza v22-6-g9e5350d: OUTPUT«===SORRY!===␤␤Unable to parse bracketed infix at /tmp/5l97LSZph9 line 1:␤------> say (0, 1, 2, { rand } ... *) [⏏^10]␤Couldn't find final ']'; gave up at /tmp/5l97LSZph9 line 1:␤------> say (0, 1, 2, { rand } ... *)…
colomon n: say (0, 1, 2, { rand } ... *)[^10]
p6eval niecza v22-6-g9e5350d: OUTPUT«0 1 2 0.91410435639047272 0.24678441521096248 0.16036482162790597 0.33466130650353682 0.31822322696364636 0.77577368904639676 0.59004591106904947␤»
13:54 mucker joined
moritz r: say <== map *+1 <== 1, 2, 4 ... 64 13:54
p6eval rakudo 9af8f2: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix <== instead␤at /tmp/Xlm2PejEAo:1␤»
GlitchMr perl6: say() <== map *+1 <== 1, 2, 4 ... 64 13:55
p6eval niecza v22-6-g9e5350d: OUTPUT«␤Unhandled exception: Feed ops NYI␤ at /home/p6eval/niecza/lib/CORE.setting line 1437 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3353 (infix:<<==> @ 4) ␤ at /tmp/fdJEgE4Ii3 line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting li…
..rakudo 9af8f2: OUTPUT«2 3 5 9 17 33 65␤»
grondilu (0, 1, 2, { rand } ... *) could be interpreted as list concatenation of (0, 1, 2) and ({ rand } ... * )
GlitchMr Actually, I've interpreted it like that at first 13:56
colomon grondilu: trying to come up with interpretations which make the parser work harder is a bad idea
GlitchMr But, it's probably best syntax that would work
Unless you have better idea
colomon n: say (0, 1, 2, *.rand ... *)[^10]
p6eval niecza v22-6-g9e5350d: OUTPUT«0 1 2 1.2861544635548043 0.20919612994498515 0.19472512659656913 0.089665246528384132 0.067629248614021337 0.03793104336496745 0.02617297972594048␤»
GlitchMr You don't have to use sequences if you don't like them
But they are a nice shortcut 13:57
We could just program in assembly and use JMP everywhere
But we don't. 13:58
At least, most of time.
(I know that you would usually use conditional jumps in assembly) 13:59
grondilu just looked at the code for "..." 14:01
raiph jnthn: note irclog.perlgeek.de/perl6/2012-09-19#i_6002246 thru irclog.perlgeek.de/perl6/2012-09-19#i_6002337
GlitchMr I find it interesting that in Brainfuck you have highlevelish while(){} loop 14:02
14:02 flightre1 left, FROGGS_ left
jnthn raiph: thanks 14:04
raiph jnthn. yw. especially last entry
GlitchMr Wait... why $*IN.get would be different too $*IN.get()? 14:05
jnthn I...dunno.
GlitchMr That is the only difference I see 14:06
raiph and variation even on same OS: irclog.perlgeek.de/perl6/2012-09-19#i_6002353
GlitchMr While it doesn't happen on my machine, I'm going to try that on minefield 14:07
(that would be VM)
It doesn't happen on my VM 14:08
14:09 flightrecorder joined
dalek rlito: 06276a4 | (Flavio S. Glock)++ | src5/lib/Perlito5/Javascript3/CORE.pm:
Perlito5 - js3 - fix substr()
 14:09
jnthn GlitchMr: Does the debugger work for you, ooc?
GlitchMr What is "ooc"?
And yes, it works
'out of context'? 14:10
14:10 wtw left
tadzik curiosity 14:11
14:11 stopbit joined
raiph GlitchMr: ooc = out of curiosity 14:13
14:13 mucker left
moritz
.oO( out of countries )
 14:14
14:14 mucker joined 14:15 FROGGS_ joined
[Coke] I'd rather more pages looked like twitter bootstrap than like the normal default web page. 14:17
14:20 wk left
jnthn GlitchMr: out of curiosity :) 14:23
GlitchMr: Did you hit enter to single-step? 14:24
GlitchMr yes
jnthn ok
that's broken for me :(
well, I have a hack now locally...
14:26 cognominal left
jnthn sleep & 14:28
14:30 fgomez joined
FROGGS_ sleep well 14:30
tadzik heh, I'm plenty of timezones away and I wouldn't mind a nap either 14:32
FROGGS_ dito 14:33
14:33 cognominal joined 14:34 replore left, skids joined 14:36 kaare_ joined 14:42 spider-mario joined
[Coke] if there's a way we can write an automated test for this sort of thing, that'd be awesome. 14:48
14:51 thou left
[Coke] sees resistances to enums as strings because of == ... but if you expect == to work, then you expect them to be integers, which means you should expect 0 to be false. 14:56
moritz well, old code would break through that change 14:58
but if that's something we really want, it shouldn't stop us
15:00 taniavs666 left 15:06 colomon left 15:07 thou joined 15:08 taniavs666 joined 15:09 brrt left, grondilu left 15:11 birdwindupbird joined 15:14 taniavs666 left 15:24 Targen joined 15:32 colomon joined 15:34 hoelzro joined, wamba left 15:41 orafu joined 15:48 hoelzro is now known as hoelzro|away 15:49 wamba joined 15:52 preflex joined 15:55 araujo left, araujo joined 15:56 birdwindupbird left 15:57 kaleem left 15:58 Targen left 16:00 flightrecorder left, FROGGS_ left, cognominal left 16:02 tokuhiro_ joined 16:03 wk joined 16:05 fhelmberger left 16:08 literal_ is now known as literal 16:09 MayDaniel joined 16:13 colomon left
kresike bye folks 16:14
16:14 kresike left, wamba left 16:18 colomon joined 16:28 wamba joined 16:30 sjohnson left, sjohnson joined 16:33 sivoais left 16:36 dakkar left 16:39 sivoais joined 16:44 zby_home joined, Vlavv_ joined 16:47 sjohnson left, sjohnson joined 16:48 bruges left 16:49 bruges joined 16:52 bruges_ joined 16:53 fgomez left 16:54 bruges left 16:56 Chillance joined 16:58 bruges_ left, preflex left 17:00 zby_home__ joined, zby_home left, kaare__ joined, kaare_ left 17:02 DarthGandalf left 17:04 DarthGandalf joined
sorear good * #perl6 17:07
colomon o/
17:09 bruges joined 17:10 jlaire joined
raiph hi sorear 17:18
17:18 cognominal joined
pmichaud can someone bring me up to speed on the status of the readline / prompt issue? I hadn't seen any bug reports about it. 17:20
phenny pmichaud: 11:39Z <jnthn> tell pmichaud I think we can't put out a Star release until the prompt issue is resolved one way or another. It'll break a load of stuff. Debugger is broken, for example.
pmichaud: 11:40Z <jnthn> tell pmichaud If our solution is "tie ourselves to a Parrot that wasn't busted" that's fine by me too.
pmichaud I'm happy with "tie ourselves to a non-busted Parrot", although I think that involves a fair bit of reverting. 17:21
We can also basically release Star using 2012.08 compiler (and corresponding nqp/parrot)
which would mean that 2012.08 would be almost identical to 2012.09 . 17:22
17:22 cognominal_ joined, replore joined
[Coke] we could also try to get a patched parrot out the door if it's an obvious fix. 17:24
17:24 cognominal left
pmichaud testing rc1 under windows now 17:33
17:33 replore left
pmichaud repl command history seems to work for me under windows 17:47
I'll try the debugger
17:49 benabik left
pmichaud command history seems to work there also, although I still don't have proper screen colors/formatting 17:55
raiph pmichaud: i think felher suggested his (the?) problem was older than 2012.08 17:56
pmichaud raiph: that may be so, although jnthn++ was citing some issues earlier today 17:57
felher raiph: no. I merely said my problem occured when switching to a new machine, so it doesn't have to be a problem with up-to-date rakudo :)
pmichaud ah, jnthn++ was listing "prompt" as a problem. I'll try that 17:58
irclog.perlgeek.de/perl6/2012-09-27#i_6021962
felher raiph: or this was at least what i thought i've said :)
pmichaud Yes, "prompt" definitely fails for me under windows. 17:59
....and under linux. 18:00
18:00 cognominal_ left 18:03 cognominal joined
sorear o/ felher, raiph 18:03
felher o/ sorear :)
raiph felher: ah. ok. do you think the problem you had looks like it might well be the same as the one jnthn found? 18:05
pmichaud so, we definitely cannot use 2012.09 compiler for a star release. (We can create a 2012.09.1, or we can go back and use 2012.08.)
[Coke] does prompt work in git-latest? 18:06
pmichaud I'll look.
I'll have to build it.
[Coke] votes "ever onward" where possible.
pmichaud same here.
[Coke] pmichaud: I'll test it out on OS X. 18:07
pmichaud I have a boringly simple test program.
(nopasting)
felher raiph: might well be :) I can not check at the moment, 'caue the machine i had the problem on is down for some time.
18:08 cognominal_ joined, cognominal left
pmichaud gist.github.com/3795470 # test program 18:09
felher Hm, my problem occured with $*IN.get. I didn't even need prompt, so this might be a different problem. :) 18:10
18:11 MikeFair_ left
pmichaud jnthn's problem appears to have been with prompt. 18:12
regardless, it's not something we can have a regression on.
felher right 18:14
skids Was irclog.perlgeek.de/perl6/2012-09-26#i_6020795 replicatible by anyone else? 18:17
18:18 kurahaupo joined 18:20 benabik joined
[Coke] pmichaud: still an issue on os x with git latest. (have to hit ^D, not just enter) 18:30
skids: I can duplicate it, aye 18:31
pmichaud [Coke]: was that git latest with parrot, ooc? 18:35
18:36 cognominal joined 18:37 thou left, cognominal_ left
pmichaud i.e., you were using parrot latest? 18:39
18:41 thou joined
pmichaud ah. Rakudo's HEAD seems to "fix" the problem in github.com/rakudo/rakudo/commit/b7...230bdc0405 18:42
it adds an explicit flush to $*OUT. Previously Parrot would autoflush stdout when reading input from stdin.
[Coke] pmichaud: I grabbed rakudo nom, git clean -xdf'd, and then rebuilt nqp & parrot 18:43
pmichaud ...and it failed on osx? weird.
oh, wait.
[Coke] I am pretty sure I have that commit
pmichaud let me re-test.
[Coke] I think the problem is IN, not OUT, no?
pmichaud one of the problems with "prompt" in 2012.09 is that my $abc = prompt "abc: "; doesn't flush $*OUT prior to waiting for the input 18:44
[Coke] I have that commit, true.
18:44 cognominal left
[Coke] oh. I thought the problem was that hitting enter didn't stop prompt. 18:44
pmichaud thus the user never sees the prompt
that may be a problem also, I need to check that.
I was simply looking for the prompt, not seeing it, and saying "oh, that's a regression" 18:45
[Coke] I see the prompt.
pmichaud on my system, though, rakudo head doesn't seem to require the ^D
rurban pmichaud: Did you ever checked for an empty cpu_type? github.com/parrot/parrot/issues/846 I changed it now to "unknown"
I don't think you care about cpu types.
18:45 kurahaupo left
pmichaud cpu type hasn't ever popped up on my radar, no. 18:46
[Coke] I see no reference to cpu_type in rakudo/src
rurban good
[Coke] (only in parrot specific stuff in the install)
18:47 cognominal joined 18:55 cognominal left 18:56 fgomez joined 18:58 arlinius joined
dalek rlito: 8b919d2 | (Flavio S. Glock)++ | / (5 files):
Perlito5 - js3 - use FETCH() method
 18:58
pmichaud felher: the problem you're encountering is similar to the one that Coke++ and jnthn++ are reporting. 19:01
the fix in b72c4346 isn't sufficient to resolve the regression. 19:04
gist.github.com/3795822 # difference from 2012.08 to current Rakudo HEAD 19:08
I have to get some lunch and pick up kids, I'll start filing tickets when I get done with those activities. Others are welcome to file tickets and report problems also. 19:09
dalek rlito: 25e21d4 | (Flavio S. Glock)++ | / (2 files):
Perlito5 - js3 - use assign()
19:12 leprevost joined 19:17 sirrobert joined
sirrobert hi p6 =) 19:17
phenny sirrobert: 26 Sep 05:35Z <GlitchMr> tell sirrobert or * + * for the anonymous function you want
19:25 pernatiy joined 19:26 sjohnson left 19:27 sjohnson joined
[Coke] rurban pointed out this bug: github.com/parrot/parrot/issues/834 19:27
19:28 mtk left 19:30 mtk joined 19:31 zby_home__ left 19:32 sjohnson left, sjohnson joined 19:33 daxim left 19:39 birdwindupbird joined 19:41 fgomez left
masak TimToady: happy 0x3a-th birthday. 19:52
PerlJam masak++ 19:53
TimToady: Happy Birthday!
sorear happy birthday! 19:54
19:55 arlinius left
flussence same day as google too, according to their logo scribble... 19:55
[Coke] r: say 0x3a 19:56
p6eval rakudo 9af8f2: OUTPUT«58␤»
[Coke] r: say 2012-1954
p6eval rakudo 9af8f2: OUTPUT«58␤»
PerlJam Coke: nah, they're celebrating Larry's birthday too :)
masak r: say 2012 - 1954
p6eval rakudo 9af8f2: OUTPUT«58␤»
masak g'ah, becursed train lag. 19:57
[Coke] (just checking, as '54 is my mom's)
masak r: say 0b111010
p6eval rakudo 9af8f2: OUTPUT«58␤»
19:57 birdwindupbird left
masak smiles at still being able to convert directly from hex to bin 19:58
not that it's particularly hard.
my uncle taught it to me when I was 0b100.
19:59 cognominal joined, birdwindupbird joined
sjohnson happy brithday perl man! 20:00
20:01 GlitchMr left
masak imagines a Superman suit with a big letter P on the front 20:01
sorear has he worn that? 20:02
sirrobert r: "one two three" ~~ m/one \s* two \s* three/; 20:12
p6eval rakudo 9af8f2: ( no output )
sirrobert r: say "one two three" ~~ m/one \s* two \s* three/;
p6eval rakudo 9af8f2: OUTPUT«｢one two three｣␤␤»
sirrobert is there a matching mode that automagically ignores whitespace differences?
to make something like this work: "one two three" ~~ m/one two three/ 20:13
?
sorear r: say "one two three" ~~ mm/one two three/;
p6eval rakudo 9af8f2: OUTPUT«===SORRY!===␤Confused␤at /tmp/Tr8vVFdqIS:1␤»
sorear r: say "one two three" ~~ ms/one two three/;
p6eval rakudo 9af8f2: OUTPUT«｢one two three｣␤␤»
sirrobert hmm cool, thanks
sorear r: say "one two three" ~~ m:s/one two three/;
p6eval rakudo 9af8f2: OUTPUT«｢one two three｣␤␤»
sorear r: say "one two three" ~~ m/:s one two three/;
p6eval rakudo 9af8f2: OUTPUT«｢one two three｣␤␤»
sirrobert hrm 20:14
20:14 kaare__ left
sirrobert is there any difference? (between ms, m:s, m/:s) 20:14
sorear no
sirrobert ok, cool
thanks
sorear under the scope of :s (short for :sigspace), " " means <.ws> 20:15
ws can be overridden in a grammar; its default definition is \s* *but* if between two \w, at least one space is needed
masak sorear: not to my knowledge. 20:16
(re "has he worn that?")
[Coke] 16:10 <@pmichaud> we almost certainly won't be able to issue the star release today... too many IO regressions.
sorear r: say "one+two" ~~ m:s /one "+" two/ # no space needed here
[Coke] 16:10 <@pmichaud> I'm trying to get them all catalogged... _and_ I have other (non-rakudo) events taking place today :-|
p6eval rakudo 9af8f2: OUTPUT«｢one+two｣␤␤»
[Coke] (from #parrot)
sorear r: say "oneplustwo" ~~ m:s /one "plus" two/ # but the space IS mandatory here
p6eval rakudo 9af8f2: OUTPUT«#<failed match>␤»
20:17 skids left 20:18 cognominal left 20:25 Targen joined
pmichaud (from #parrot) github.com/parrot/parrot/issues/847 # one of the .readline issues 20:25
20:26 birdwindupbird left
raiph masak: hobbified: "[JS] modules don't install themselves into a global namespace." chromatic: "Perl 6 sort of edged toward that idea, but I don't remember if it ever made it quite far enough." 20:49
from www.reddit.com/r/perl/comments/10j9...rl/c6e1w22
20:52 flightrecorder joined 20:57 hoelzro|away is now known as hoelzro
[Coke] if I have an old commit checked out, and rather than going all the way to HEAD, I want to just progress one more commit in that direction, is there shortcut for that? 21:02
21:03 spider-mario left
doy no, because there's not really any way to tell what "one more commit" actually means in that context (in the presence of branches, for instance) 21:04
21:16 raiph left 21:17 Circlepuller_ joined 21:19 Circlepuller left 21:20 arlinius joined 21:21 Circlepuller joined 21:22 Circlepuller_ left 21:25 benabik left 21:44 araujo left, mucker left 21:45 robinsmidsrod joined 21:54 fgomez joined 21:56 skids joined 21:57 flightrecorder left, araujo joined
TimToady I think it's useful for the first enum to boolify to False by default; many enums want a "non-value" 22:04
I see little use case for making all enums True by default
22:06 thou left 22:08 felher joined
rurban pmichaud: parrot readline #847 fixed in github.com/parrot/parrot/commit/1fe3b78b36 22:08
22:09 mtk left
rurban but on windows we still return \r\n ending strings. I'm not sure 100% if that's a regression. 22:09
22:12 thou joined, wamba left 22:15 preflex joined 22:16 stopbit left, Circlepuller_ joined 22:17 Circlepuller left 22:18 hoelzro is now known as hoelzro|away 22:22 mtk joined 22:28 Moukeddar joined 22:29 Moukeddar left, Moukeddar joined 22:30 Moukeddar left, Moukeddar joined, Moukeddar left 22:47 replore_ joined 22:51 ServerSage left 22:53 Targen left 22:54 yarp joined 22:58 replore_ left 23:02 ServerSage joined 23:05 replore_ joined 23:06 replore_ left, PacoAir left 23:07 _jaldhar_ left, _jaldhar_ joined 23:11 whiteknight joined, ServerSage left 23:12 kid51 joined 23:16 kurahaupo joined 23:17 ServerSage joined
japhb Hippo Birdie Two Ewes, TimToady! 23:24
23:31 Chillance left 23:35 _jaldhar_ left 23:43 yarp left 23:44 MayDaniel left 23:53 dayangkun joined, dayangkun left, popl joined
diakopter pmichaud: ping 23:54
23:55 blez joined
diakopter phenny: tell pmichaud when used as a queue (push/shift or unshift/pop), afaict qrpa's storage will grow without bounds as time goes on. I recommend converting it to a ring buffer.. 23:58
phenny diakopter: I'll pass that on when pmichaud is around.
blez does perl5-6 use parrot currently?
diakopter one implementation of Perl 6 does 23:59
blez what about python? wasn't parrot going to unify perl and python's libraries
(something I've read years ago)

Powered by Cro and the Raku® Programming Language.

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

Thank you!