pugscode.org/ planetsix.perl.org/ feather.perl6.nl/syn/ perlcabal.org/svn/pugs/log github.com/rakudo/rakudo | nopaste: sial.org/pbot/perl6 | evalbot: "perl6: say 3;" | irclog: irc.pugscode.org/ | Rakudo.org www.perlfoundation.org/perl6/index.cgi?smop perl.net.au/wiki/Elf |~300 days 'til Xmas
Set by mncharity1 on 25 February 2009.
00:01 mberends left 00:04 Ariens_Hyperion left 00:38 alester joined 00:41 meteorjay joined, nihiliad left 00:46 alester left 00:59 Eevee joined
pugs_svn r25634 | putter++ | [elfparse] Starting to sketch in more efficient rx ratcheting support. 01:00
01:05 DemoFreak left 01:40 nihiliad joined
frooh rakudo: sqrt(9) 01:46
p6eval rakudo d3f407: RESULT«3»
frooh rakudo: sqrt(5)
p6eval rakudo d3f407: RESULT«2.23606797749979»
01:46 Whiteknight left
frooh rakudo: my $bar = sub ($foo) { $foo }; 01:49
p6eval rakudo d3f407: RESULT«{ ... }»
frooh rakudo: my $bar = sub ($foo) { $foo }; $bar->(1);
p6eval rakudo d3f407: OUTPUT«Statement not terminated properly at line 1, near "->(1);"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh rakudo: my $bar = sub ($foo) { $foo }; $bar(1);
p6eval rakudo d3f407: RESULT«1»
frooh rakudo: for my $f 1..4 { say $f } 01:52
p6eval rakudo d3f407: OUTPUT«Redeclaration of variable $f␤Statement not terminated properly at line 1, near "1..4 { say"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh rakudo: for my $f, 1..4 { say $f }
p6eval rakudo d3f407: OUTPUT«Use of uninitialized value␤␤Use of uninitialized value␤␤Use of uninitialized value␤␤Use of uninitialized value␤␤Use of uninitialized value␤␤»
pmichaud bacek: I know that chain implies &&. I'm saying I don't like the current factoring. 01:55
frooh how do I do a for loop in perl 6? 01:57
(or where in the fine manual can I find out :-)
TimToady S04
rakudo: for 1..4 -> $f { say $f }
p6eval rakudo d3f407: OUTPUT«1␤2␤3␤4␤» 01:58
TimToady $f is a formal parameter to the block in p6
frooh ah
I was searching for =item for
rakudo: '5' =~ /\d+$/ 02:02
p6eval rakudo d3f407: OUTPUT«Unable to set lvalue on PAST::Val node␤current instr.: 'parrot;PAST;Val;lvalue' pc 566 (src/PAST/Node.pir:161)␤»
frooh rakudo: '5' ~~ /\d+$/
p6eval rakudo d3f407: RESULT«Method 'perl' not found for invocant of class 'Match'␤current instr.: 'parrot;P6metaclass;dispatch' pc 320 (src/classes/ClassHOW.pir:161)␤»
frooh rakudo: if '5' ~~ /\d+$/ { say 'matches' }
p6eval rakudo d3f407: OUTPUT«matches␤»
frooh rakudo: if 'frew' ~~ /\d+$/ { say 'matches' }
p6eval rakudo d3f407: RESULT«Method 'perl' not found for invocant of class 'Match'␤current instr.: 'parrot;P6metaclass;dispatch' pc 320 (src/classes/ClassHOW.pir:161)␤»
frooh rakudo: if my $f = 'frew' { say $f } 02:03
p6eval rakudo d3f407: OUTPUT«frew␤»
02:03 xinming_ joined
frooh when rakudo tells me "No applicable methods" is there any way I can find out what line it's talking about? 02:11
ok, actually the issue is coming from doing this: 02:15
if my $f = %odt{$token} { $f(...) }
I tried replacing $ with & 02:16
still no good
02:17 xinming left
frooh oh weird 02:18
so I had an issue in the function
I had this:
push(@stack, pop(@stack) + pop(@stack)); 02:19
I know that's no idiomatic perl6
but is it wrong/ 02:20
02:31 mikehh left 02:32 mikehh joined
TimToady std: push(@stack, pop(@stack) + pop(@stack)); 02:33
p6eval std 25634: OUTPUT«Potential difficulties:␤ Variable @stack is not predeclared at /tmp/mLVas1R668 line 1:␤------> push(@stack, pop(@stack) + pop(@stack));␤ Variable @stack is not predeclared at /tmp/mLVas1R668 line 1:␤------> push(@stack, pop(@stack) + pop(@stack));␤
.....
TimToady std: my @stack; push(@stack, pop(@stack) + pop(@stack)); 02:34
p6eval std 25634: OUTPUT«ok 00:02 34m␤»
frooh rakudo: my @stack = [1, 2]; push(@stack, pop(@stack) + pop(@stack)); 02:35
p6eval rakudo d3f407: OUTPUT«No applicable methods.␤␤current instr.: '_block14' pc 141 (EVAL_16:52)␤»
TimToady rakudo: my @stack := [1,2]; push(@stack, pop(@stack) + pop(@stack)); 02:36
p6eval rakudo d3f407: OUTPUT«No applicable methods.␤␤current instr.: '_block14' pc 127 (EVAL_16:50)␤»
TimToady rakudo: my @stack := [1,2]; push(@stack, @stack.pop + @stack.pop); 02:37
p6eval rakudo d3f407: RESULT«1»
TimToady rakudo: my @stack := [1,2]; push(@stack, @stack.pop + @stack.pop); say @stack[0]
p6eval rakudo d3f407: OUTPUT«3␤»
TimToady you either have to use := [] or = () 02:38
frooh I replaced all of them with @foo.bar, but it's still a bug right?
well, @stack is defined by push'es
TimToady in theory pop(@stack) ought to work
frooh ok
just making sure
TimToady the push as a function seems to work
frooh it works fine when I use the method syntax, which I prefer anyway, I was just curious 02:39
TimToady rakudo: my @stack = (1,2); say pop(@stack)
p6eval rakudo d3f407: OUTPUT«2␤»
frooh also, what exactly is the diff between := = [] and () ?
TimToady seems to work in isolation
binding vs assignment
[] creates a single scalar containing an array
frooh so = value, := reference?
TimToady if you assign that, you end up with only one element that happens to be an array 02:40
yes, basically
frooh interesting
Tene_ rakudo: my $a = 5; my $b := $a $b++; say $a
p6eval rakudo d3f407: OUTPUT«Statement not terminated properly at line 1, near "$b++; say "␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
Tene_ rakudo: my $a = 5; my $b := $a; $b++; say $a 02:41
p6eval rakudo d3f407: OUTPUT«6␤»
TimToady decommuting &
Tene_ decommuting too!
In 9 minutes.
frooh my $f = [1,2,3]; 02:42
rakduo: my $f = [1,2,3]; $f
rakudo: my $f = [1,2,3]; $f
p6eval rakudo d3f407: RESULT«[1, 2, 3]»
frooh rakudo: my $f := [1,2,3]; $f
p6eval rakudo d3f407: RESULT«[1, 2, 3]»
frooh has trouble seeing the difference. 02:43
rakudo: my $f = (1,2,3); $f
p6eval rakudo d3f407: RESULT«[1, 2, 3]»
frooh rakudo: my @f = (1,2,3); $f
p6eval rakudo d3f407: OUTPUT«Scope not found for PAST::Var '$f' in ␤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)␤»
frooh rakudo: my @f = (1,2,3); @f
p6eval rakudo d3f407: RESULT«[1, 2, 3]»
frooh rakudo: my @f := (1,2,3); @f
p6eval rakudo d3f407: RESULT«[1, 2, 3]»
02:48 Chillance left
frooh pugs: 1 - 2 02:54
p6eval pugs: RESULT«-1»
frooh pugs: 1 R- 2
p6eval pugs: OUTPUT«*** ␤ Unexpected "R"␤ expecting operator␤ at /tmp/XSe9w7YqD2 line 1, column 3␤»
frooh pugs: 1 R/ 2 02:55
p6eval pugs: OUTPUT«*** ␤ Unexpected "R"␤ expecting operator␤ at /tmp/DgcfqjC3D4 line 1, column 3␤»
frooh pugs: 1 R== 2
p6eval pugs: OUTPUT«*** ␤ Unexpected "R"␤ expecting operator␤ at /tmp/Z5RZqbEJBq line 1, column 3␤»
02:56 Sepheebear left, edenc left, broquaint left 02:58 broquaint joined, edenc joined 03:15 Sepheebear joined, Sepheebear left 03:16 Sepheebear joined 03:37 Sepheebear left 03:42 aindilis` joined
frooh sub f(Int $i) { say $i }; f(2); f('2'); 03:49
rakudo: sub f(Int $i) { say $i }; f(2); f('2');
p6eval rakudo d3f407: OUTPUT«2␤Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh rakudo: sub f(Int $i) { say $i }; f(2); f(+'2');
p6eval rakudo d3f407: OUTPUT«2␤Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh rakudo: sub f(Int $i) { say $i }; f(2); f((+'2')); 03:50
p6eval rakudo d3f407: OUTPUT«2␤Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh rakudo: sub f(Int $i) { say $i }; f(2);
p6eval rakudo d3f407: OUTPUT«2␤»
frooh rakudo: 5
p6eval rakudo d3f407: RESULT«5»
frooh rakudo: '5'
p6eval rakudo d3f407: RESULT«"5"»
frooh rakudo: +'5'
p6eval rakudo d3f407: RESULT«5»
frooh hmmm
rakudo: ~5
p6eval rakudo d3f407: RESULT«"5"»
frooh rakudo: sub f(Int $i) { say $i }; f(+'2'); 03:51
p6eval rakudo d3f407: OUTPUT«Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh rakudo: sub f(Int $i) { say $i }; my $foo = '2'; f(+$foo);
p6eval rakudo d3f407: OUTPUT«Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh rakudo: sub f(Int $i) { say $i }; my $foo = +'2'; f($foo);
p6eval rakudo d3f407: OUTPUT«Parameter type check failed for $i in call to f␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh :-(
meppl good night 03:55
03:56 aindilis left
frooh also: for @*ARGS -> $arg doesn't seem to work 03:56
03:58 japhb joined 03:59 meppl left
pmichaud frooh: Rakudo currently sees +'2' as being a Num, not an Int. 04:01
I'm not exactly certain how to resolve that one yet.
frooh huh 04:02
well, I guess that makes sense
but can't it say, "dude, this Num really *is* an Int!"
(I presume it can't or it would :-)
any idea about the @*ARGS thing?
pmichaud as I said, I'm not exactly certain how to resolve that one yet. Currently our prefix:<+> returns a Num. 04:03
I guess we could work on getting prefix:<+> to return an Int if the string doesn't include a decimal point.
frooh yeah, I'm just joking, I think returning a num makes more sense anyway
pmichaud it's a bit complicated on the parrot internals side for that.
frooh maybe I could do (+'f').int 04:04
er
'5'
pmichaud looking at @*ARGS now
looks like it works to me: 04:06
$ cat x.pl
for @*ARGS -> $arg { $arg.say }
$ ./perl6 x.pl a b c
a
b
c
frooh weird...
lemme look again
unless there is a really recent fix for rakudo (I built tongiht) it's busted here still 04:07
exact same code
pmichaud can you pastebin/nopaste what you're testing?
what output are you getting? 04:08
frooh Statement not terminated properly at line 10, near "-> $arg { "
pmichaud it's not even parsing? that's odd.
frooh lemme comment out the rest of the code
pmichaud oh, so it's not the _exact_ same code :_)
I suspect the error is whatever follows the for loop body
frooh weird
well 04:09
it's BEFORE the for loop
it's a sub
pmichaud oh
yes, that's a known bug in Rakudo.
you need a semi after that sub.
frooh really?
what is it exactly?
pmichaud otherwise Rakudo sees it as "sub xyz { } for ... "
frooh oh
hih
pmichaud and thinks that the 'for' is a statement modifier.
I guess I better fix that tomorrow.
frooh so do sub xyz semi { ... }? 04:10
it's not a huge deal
pmichaud (even if it ends up being an ugly-ish fix.
it needs to be: sub xyz { ... }; for @*ARGS -> $arg { ... }
with the semicolon after the closing brace of the sub.
frooh ahhh
git it
pmichaud that's a problem that trips up a lot of folks, and we really need to fix it somehow. 04:11
frooh got*
pmichaud so, it's on my list for tomorrow.
rakudo: sub foo(Int $x) { $x.say }; foo('5'.int);
p6eval rakudo d3f407: OUTPUT«5␤»
pmichaud no + needed.
frooh got it 04:12
04:12 Tene joined
frooh so would it be worth it for me to learn pir to try to help or would it be better to try to help with the setting? 04:12
pmichaud I think helping with setting is what most people will find most useful.
or most enjoyable.
frooh cool
well, that's definately true
but I don't mind doing stuff that no one else wants to do 04:13
anything to make perl6 come sooner :-)
pmichaud the PIR really is intended to be reserved for things that can't easily be done in Perl 6.
frooh what exactly is the PGE that people keep talking about? 04:14
pmichaud it's Rakudo's (and Parrot's) regular expression engine.
It's the thing that handles regex matching, as well as the parsing of Perl 6 code itself.
frooh got it
pmichaud there should be an entry for it in docs/glossary.pod
frooh ahh 04:15
pmichaud oh, there's not! That's a bug.
frooh haha
niec
nice*
dalek kudo: c1f3976 | pmichaud++ | docs/glossary.pod:
Add entry for PGE to glossary.pod (frooh++)
 04:18
04:19 tony_ joined
frooh reads glossary.pod 04:19
so does rakudo actually use STD.pm? or is it just a validator/ 04:21
pmichaud rakudo doesn't use STD.pm yet. 04:22
frooh but it will at some point?
pmichaud PGE isn't sophisticated enough yet to parse STD.pm
frooh ah, got it
pmichaud the goal is to have Rakudo and other implementations converge on STD.pm, yes.
frooh yeah, I didn't know if that meant use STD.pm itself or just something like it
pmichaud whatever we end up with undoubtedly be very close to STD.pm, if not STD.pm itself. 04:23
it might end up being STD.pm with a couple of patches.
frooh also: how about the Setting? is it the goal that implementations can reuse that too?
04:23 idemal joined
pmichaud I don't know to what extent Perl 6 will have a "standard setting" (more) 04:23
there are some functions (primitives) that likely cannot be expressed in Perl 6 04:24
(more)
04:24 Tene_ left
pmichaud and things that are primitives on one platform might be fairly complex on other platforms 04:24
frooh that makes sense
pmichaud so, we may have a standard setting library that says "this is a reference for implementations", but individual implementations may want or need to use a slightly different internal implementation (that gives the same semantics). 04:25
frooh yeah. But the ability to change out settings (or add them) is part of the spec right? 04:26
TimToady std: sub foo {...} for 1..10 -> {...}
pmichaud but yes, I'd like for Rakudo to be providing some settings that other implementations can re-use.
p6eval std 25634: OUTPUT«############# PARSE FAILED #############␤Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?) at /tmp/3BssYYhSx3 line 1:␤------> sub foo {...} for 1..10 -> {...}␤ expecting any of:␤ infix or meta-infix␤ infix
..stopper␤ ...
pmichaud TimToady: you may have answered this already, but should prefix:<+> be smart enough to return an Int for things like +'5' and a Num for +'5.0' ? 04:27
TimToady thinking... 04:29
pmichaud or is there a "generic num/intify" that we should be doing on Str ?
TimToady perhaps it returns a Rat :) 04:30
pmichaud if the answer isn't yet specced, I don't need it right now; was just curious if I had overlooked something (I'll have to leave very shortly) 04:31
TimToady not specced yet
pmichaud but I know that sub foo(Int $x) { ... }; foo(+'5') trips up a lot of folks.
in some sense I think it might be good if we say that such cases require an explicit .int on the argument 04:32
frooh rakudo: +'5/7'
p6eval rakudo c1f397: RESULT«5»
frooh rakudo: +'5+7i'
TimToady on the other hand, if Int does the Num role, it's fine to return an Int for a Num return type
p6eval rakudo c1f397: RESULT«5» 04:33
frooh rakudo: +'5.7'
pmichaud but if Int does Num, then foo(2.3) would seem to match?
p6eval rakudo c1f397: RESULT«5.7»
pmichaud or is my too-tired brain thinking of that backwards again? 04:34
TimToady but it matches Num more closely
frooh pmichaud: I agree, foo(2.3) should at least give a warning, right?
2.3 isn't an Int, and foo only takes Int's... 04:35
TimToady maybe they both do Numeric
or maybe + just doesn't commit to a specific return type
pmichaud well, I was headed that way, but there's also the case of
frooh want's + to be * for numeric coersions :-) 04:36
pmichaud foo('3' * 2)
so it's not just prefix:<+> that we have to consider.
TimToady well, AnyAny would convert to IntInt in that case 04:37
which seems okay
pmichaud I don't understand that last part
TimToady infix:<*>:(Any,Any) converts via +$x * +$y 04:38
which would make two integers and call infix:<*>:(Int,Int)
pmichaud okay.
TimToady unless we short circuit the strings with extra candidates, but it doesn't seem necessary 04:39
pmichaud right.
so then prefix:<+> does become the generic "make me a number" operator.
TimToady basically, if people rely on the Any cases they get slowed down, so there some motivation to cast explicitly
pmichaud I wasn't looking for a guaranteed return type for prefix:<+>; I was more asking if it was possible that prefix:<+>(Str) could choose to return a Num or an Int based on the conversion. 04:40
TimToady yes, + can just make the most specific type it can, I suspect
frooh hurray!
TimToady even something to be said for Rat in some cases 04:41
pmichaud agreed.
frooh what about Complex?
or is that just going too far?
;-)
pmichaud it's not too far, it's just too complex. :-) :-)
TimToady esp if the string is '355/113'
well, if the string is 1+.2i, what're you gonna do, truncate it? 04:42
pmichaud where do we draw the line between numify and eval? ;-)
TimToady we guarantee to recognize radix numbers in strings too... 04:43
frooh user defined types?
pmichaud yes. Somehow I see it as being "numify a term" as opposed to "numify an expression"
TimToady starting to sound like a longest-token problem
parse this string with term:* and ... 04:44
pmichaud so far Rakudo's string-to-num converter follows the parsing for <number>
TimToady but can other types add term: rules, I wonder... 04:45
pmichaud (which includes the radix numbers)
TimToady or maybe we need a number: category
pmichaud anyway, you've given me the short-term answer for prefix:<+>, so I'm good with that for now.
TimToady wouldn't be difficult to arrange
pmichaud number: might be very nice
could be much cleaner for dealing with the complexes and rats 04:46
rakudo: say +( ~ ( 3+.1i ) );
p6eval rakudo c1f397: OUTPUT«3␤»
pmichaud rakudo: say ~(3+.1i) # just checking 04:47
p6eval rakudo c1f397: OUTPUT«3+0.1i␤»
pmichaud might also help with parsing things like "Inf" and "NaN" 04:48
TimToady indeed
I love it when a plan comes together :)
pmichaud token number:NaN { <sym> >> } 04:50
token number:Inf { <sym> >> }
helps to resolve the "Inf" versus "Info" issue, too.
TimToady assuming we require $ too
pmichaud rakudo: say 2 + (Inf)i; # just curious 04:52
p6eval rakudo c1f397: OUTPUT«NaNInfi␤»
TimToady probably we want to short-circuit normal ints and nums, and only revert to parse on fancy stuff
but that depends on the overhead of regex setup 04:53
frooh is the type for a function Sub?
pmichaud depends on the function :-)
frooh haha
anonymous subroutine?
TimToady with or without the "sub"? 04:54
they're different
frooh good question..
I don't know...
let's say without for now
Block?
TimToady pugs: (sub {}).WHAT
p6eval pugs: RESULT«::Sub»
TimToady pugs: ({}).WHAT
p6eval pugs: RESULT«::Hash»
frooh haha
TimToady pugs: ({;}).WHAT
p6eval pugs: RESULT«::Block» 04:55
frooh is there something that would work for both?
TimToady Code
frooh awesome
that's what I would guess
TimToady rakudo: {;} ~~ Code 04:56
p6eval rakudo c1f397: RESULT«1»
TimToady rakudo: sub {;} ~~ Code
p6eval rakudo c1f397: RESULT«1»
pmichaud rakudo: {} ~~ Code
p6eval rakudo c1f397: RESULT«0»
pmichaud ...although why that returns "0" and not "" or "False" is beyond me. 04:57
TimToady because it's a Hash
pmichaud there's a lot of places where people keep using 0/1 instead of True/False
TimToady oh, wait
pmichaud (in the rakudo code base)
TimToady the value of False is 0
pmichaud rakudo: 3 == 4 04:58
TimToady current spec only gets the name via .perl or .name
p6eval rakudo c1f397: RESULT«Bool::False»
pmichaud so, infix:<==> is returning "false"
frooh rakudo: my $foo = "{3==4}";
p6eval rakudo c1f397: RESULT«"0"»
pmichaud I suspect that Code.ACCEPTS() in Rakudo is returning 0/1
I'll have to crack some more whips over those coders :-) 04:59
but right now my wife is cracking a whip at me, so I best go. :-)
frooh pmichaud: thanks for the help on bug avoidance :-)
TimToady so we should probably generally only see 0 or 1 on output, unless it's .perled 05:00
last week it was stringifying to False, but I decided that was inconsistent with the native type behavior 05:01
False should never turn into '' though 05:02
that's a p5ism
frooh is excited about perl6
is there a way to get the signature from a function?
like, &sub.signature?
TimToady try it 05:03
05:03 nihiliad left
frooh rakudo: sub (Int $f, Code $b) { say $f }.signature 05:03
p6eval rakudo c1f397: OUTPUT«Statement not terminated properly at line 1, near ".signature"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh rakudo: sub (Int $f, Code $b) { say $f; }.signature
p6eval rakudo c1f397: OUTPUT«Statement not terminated properly at line 1, near ".signature"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh rakudo: sub foo (Int $f, Code $b) { say $f; }.signature
p6eval rakudo c1f397: OUTPUT«Statement not terminated properly at line 1, near ".signature"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh rakudo: sub foo (Int $f, Code $b) { say $f; }.sig
p6eval rakudo c1f397: OUTPUT«Statement not terminated properly at line 1, near ".sig"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
frooh no? 05:04
05:04 NoirSoldats joined
TimToady not sure rakudo returns a value from a named declaration 05:04
frooh the first few were anon
TimToady rakudo: sub (Int $f, Code $b) { say $f; }.sig
p6eval rakudo c1f397: OUTPUT«Statement not terminated properly at line 1, near ".sig"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
TimToady rakudo: &push.signature
p6eval rakudo c1f397: OUTPUT«Null PMC access in find_method()␤current instr.: '_block14' pc 74 (EVAL_10:41)␤»
TimToady rakudo: &push.sig 05:05
p6eval rakudo c1f397: OUTPUT«Null PMC access in find_method()␤current instr.: '_block14' pc 74 (EVAL_10:41)␤»
TimToady ah well
frooh ok
TimToady pugs: &push.sig
p6eval pugs: OUTPUT«*** No such method in class Sub: "&sig"␤ at /tmp/SRL4Uz7B7t line 1, column 1 - line 2, column 1␤»
frooh so it doesn't work for now
TimToady pugs: &push.signature
p6eval pugs: RESULT«:($?1 is rw is ref, @?1 is copy)»
TimToady there you go
frooh but could I theoretically do this:
sub (&foo.sig) {...} ?
TimToady no 05:06
frooh er
signature*
TimToady you can't eval expressions as parameters
frooh is there a way I could do that?
because it would be cool
TimToady what are you expecting it to mean? 05:07
frooh I want to make a function from another function that has the same signature
I could of course use a slurpy thing, but that loses accuracy
TimToady what would be the use of that?
frooh to make a generalized cache closure thing... 05:08
frooh is going through Higher-Order Perl with perl6
TimToady you can do smart matching on signatures
frooh although I guess with perl6 you could just do &foo.wrap
TimToady so there's probably a way to do it
frooh I'll just use a slurp for now since that works 05:09
05:18 tony_ is now known as FIRERILL 05:19 FIRERILL is now known as firedrillll 05:21 alester joined 05:25 bacek left
frooh will perl6 support prototype based OO? 05:26
skids You mean multi-dispatch? 05:29
frooh uhh
no, like, classless OO 05:30
sorry, I should have been more clear
TimToady yes, smop is proto-based 05:32
frooh smop?
super mega object prototypes?
TimToady the thing ruoso and pmurias have been working on for most of the last year
Simple Matter Of Programming :) 05:33
frooh but regular class based OO is also supported right?
TimToady yes 05:34
frooh can we use both at the same time? like make a class that inherits from a...smop object?
TimToady either can be modeled on the other
frooh awesome... 05:35
TimToady all depends on how you set up the responder interfaces
frooh I don't know what that means exactly, but I am excited anyway?
TimToady this is all swept under the carpet of the object's .HOW metaobject 05:36
frooh rakudo: my $f = 2; $f.HOW;
TimToady which decides how to be an object
p6eval rakudo c1f397: OUTPUT«get_bool() not implemented in class 'P6metaclass'␤current instr.: 'parrot;PCT;HLLCompiler;command_line' pc 1502 (src/PCT/HLLCompiler.pir:790)␤»
TimToady strange error, but you see it's in the metaclass 05:37
frooh yeah
rakudo: my @f = (1,2); @f.HOW;
p6eval rakudo c1f397: OUTPUT«get_bool() not implemented in class 'P6metaclass'␤current instr.: 'parrot;PCT;HLLCompiler;command_line' pc 1502 (src/PCT/HLLCompiler.pir:790)␤»
frooh eh, same, but still
TimToady rakudo: [].HOW.methods 05:38
p6eval rakudo c1f397: OUTPUT«Could not locate a method 'methods' to invoke on class 'P6metaclass'.␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
frooh TimToady: I love how perl lets me do things that are crazy 05:39
thanks again
and perl6 will be even better
TimToady we surely hope so
alester ok, I think rakudo.org is kinda looking decent. 05:42
frooh alester: if you want any input I think the *'s next to username and password might be a little redundant 05:43
alester frooh: That is waaaaaay far down on my list of things to think, but thank you. 05:44
frooh alester: yeah, I don't blame you.
alester hey, the nav links only show up if I log in as admin 05:45
frooh yeah, all I have is Recent posts
skids rakudo: sub foo ($a, *@@s) { $a.perl.say; @@s.perl.say; }; foo(1) <== 1,2,3; 05:54
p6eval rakudo c1f397: OUTPUT«Unable to parse multisig; couldn't find final ')' at line 1, near "@s) { $a.p"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
skids pugs: sub foo ($a, *@@s) { $a.perl.say; @@s.perl.say; }; foo(1) <== 1,2,3;
p6eval pugs: OUTPUT«*** No compatible multi variant found: "&foo"␤ at /tmp/zm9P4pVytW line 1, column 52-68␤»
skids Is that supposed to work eventually? Or does *@@slurp preclude other params? 05:57
alester ok how about now frooh 05:58
frooh nope
hust Search and Recent Posts 05:59
just*
alester nope what?
what were you expecting?
frooh I figured you had more links
sorry
alester you don't have blog posts in the right column?
syndicate?
search?
frooh oh, yeah, those are there 06:00
you said nav links before...
so I thought you meant the Navigation panel on the left
oh
sorry
yeah
that's blank
Syndicate has only an RSS icon
Search is there
alester ok
frooh Recent blog posts is there 06:01
TimToady std: sub foo ($a, *@@s) { $a.perl.say; @@s.perl.say; }; foo(1) <== 1,2,3; 06:02
p6eval std 25634: OUTPUT«ok 00:02 35m␤»
pugs_svn r25635 | lwall++ | [STD] fix ruoso++'s little nested package bug 06:17
skids std: multi sub foo ($a, *@s where { .elems > 1 }) {...} 06:20
p6eval std 25635: OUTPUT«ok 00:02 34m␤»
alester wowI think I have found the theem I love
TimToady std: module Foo { our sub bar($arg) { say $arg; }; module Bla { our sub bar() { say "ok 1"; } } }; say "1..2"; Foo::Bla::bar(); Foo::bar("ok 2\n"); 06:23
p6eval std 25635: OUTPUT«Undeclared names:␤ Foo::Bla::bar used at 1 ␤ Foo::bar used at 1 ␤ok 00:04 35m␤»
TimToady becomes professional thumb twiddler 06:24
06:25 alester left 06:27 mberends joined
TimToady std: module Foo { our sub bar($arg) { say $arg; }; module Bla { our sub bar() { say "ok 1"; } } }; say "1..2"; Foo::Bla::bar(); Foo::bar("ok 2\n"); 06:28
p6eval std 25635: OUTPUT«Undeclared names:␤ Foo::Bla::bar used at 1 ␤ Foo::bar used at 1 ␤ok 00:04 35m␤»
TimToady I don't understand why std changes its revision number out of sync with clobbering lex... 06:29
std: module Foo { our sub bar($arg) { say $arg; }; module Bla { our sub bar() { say "ok 1"; } } }; say "1..2"; Foo::Bla::bar(); Foo::bar("ok 2\n"); 06:31
p6eval std 25635: OUTPUT«Undeclared names:␤ Foo::Bla::bar used at 1 ␤ Foo::bar used at 1 ␤ok 00:04 35m␤»
TimToady std: module Foo { our sub bar($arg) { say $arg; }; module Bla { our sub bar() { say "ok 1"; } } }; say "1..2"; Foo::Bla::bar(); Foo::bar("ok 2\n"); 06:36
p6eval std 25635: OUTPUT«Undeclared names:␤ Foo::Bla::bar used at 1 ␤ Foo::bar used at 1 ␤ok 00:06 39m␤»
06:50 Maghnus_ joined 06:54 bacek joined
TimToady std: module Foo { our sub bar($arg) { say $arg; }; module Bla { our sub bar() { say "ok 1"; } } }; say "1..2"; Foo::Bla::bar(); Foo::bar("ok 2\n"); 06:56
p6eval std 25635: OUTPUT«ok 00:04 35m␤»
TimToady 'bout time...
07:02 abra joined 07:07 Maghnus left, Maghnus_ is now known as Maghnus
mberends TimToady: and what changed? clobbered lex without bumping std revision? 07:08
lambdabot mberends: You have 1 new message. '/msg lambdabot @messages' to read it.
07:18 aindilis` left 07:20 szabgab left, mikehh left, mikehh joined 07:22 kanru joined
mberends @tell masak your further changes to the proto script look very good - I am very pleased with the result 07:24
lambdabot Consider it noted.
TimToady I imagine some cron job woke up and rm'd lex/ 07:25
mberends :)
dalek kudo: 9e24621 | (Stephen Weeks)++ | src/parser/actions.pm:
Fix exception handlers to not catch control exceptions.
 07:38
pugs_svn r25636 | lwall++ | [STD] added number category so we can use <number> for prefix:<+> extensibility 07:50
TimToady @tell ruoso addition of number category to STD may or may not break mildew 07:51
lambdabot Consider it noted.
TimToady @tell pmurias addition of number category to STD may or may not break mildew 07:52
lambdabot Consider it noted.
08:11 firedrillll left 08:16 justatheory left 08:27 DemoFreak joined 08:35 masak joined
mberends masak: a very fine morning to you 08:35
masak mberends: and an excellent on to you, sir.
lambdabot masak: You have 1 new message. '/msg lambdabot @messages' to read it.
masak @massage
lambdabot mberends said 1h 11m 44s ago: your further changes to the proto script look very good - I am very pleased with the result
mberends masak++ 08:36
masak mberends: but wait -- there's more.
mberends ooh
masak I have hacked up a configuration-file thing.
not quite ready yet, but give me 5-10 minutes, and I'll push it.
I was offline when I wrote it, so I need to test it first.
mberends ok 08:37
masak digs in
heh, there ought to be a way to identify 'checkpoints' in all the massive output we're suppressing, and turn those into little dots of progress in the proto output. :) 08:56
that sounds too good not to attempt.
...later.
mberends likes dotty progress bars 08:57
08:58 szabgab joined
masak and we'll take the chance to vertically align the 'completed' messages. 08:58
pretty++
mberends not only pretty, problems will stand out visibly 09:00
masak how so? 09:01
masak gets a build failure when removing the 'rakudo dir' conf pair from the config file 09:03
hm.
mberends generally, if verbose output is scrappy, we do not read it closely, but if it is tidy, the unexpected attracts the eye. a bit like the rationale for consistent source code indentation.
09:03 c9s_ joined, c9s left
mberends RAKUDO_DIR points to the Test.pm file 09:03
masak mberends: ah. but the only info you get from the dots is how quickly each step is going, not what the steps are.
09:04 c9s_ left
masak mberends: that did not compute. :) 09:04
mberends: oh, well. I'll push what I have, since it works in the default case.
09:04 c9s joined
masak and then I'll fix the problem I just found. 09:04
mberends ok. I'll pull that before spending most of the day offline 09:05
masak pushed. 09:06
please deconnect your $PARROT_DIR and try it.
mberends sorry, that RAKUDO_DIR remark concerned Configure and Makefile. pulled. 09:09
masak I sure hope it works for you. I find the new setup very nice to work with. 09:10
things are a bit inconsistent in 'proto' and 'installer', however.
consider it 'release early'.
mberends must get YAML::Syck, not a desirable dependency for a bootstrapper. isn't the required functionality simple enough to inline in proto? 09:12
masak I'm of two minds about that.
one the one hand, I agree fully about the dep.
no deps at all would be swell. 09:13
on the other hand, I don't think we should be reinventing configuration formats.
or attempting our own YAML parsing.
09:13 lisppaste3 left
mberends is up for the challenge of YAML parsing (lite) 09:14
masak mberends: give it a go then. 09:15
maybe we should aim for a very limited subset of it.
right now we're only serializing a hash, after all.
mberends agreed, will do during offline time
masak excellent. 09:16
please take into account all kinds of escaping that might occur.
colons, newlines, quotes, possibly other things.
mberends right. but we mainly have to read just what we previously wrote. 09:17
masak aye.
still, we don't want to be caught misparsing correct YAML. 09:18
better to throw an error on things we recognize but explicitly don't handle.
mberends aye
masak tries to install Parrot-inside-Rakudo 09:25
09:29 abra left 09:33 duke_leto joined 09:35 mikehh left
mberends parrot-inside-rakudo works well for me, and jnthn++ and pmichaud++ seem to have nailed the double-free bug in the fake executable too 09:37
masak aye. 09:38
'faktecutable' should probably be a setting value too. 09:39
perhaps even the default.
mberends masak: an important question similar to one I asked eric256 several months ago: do you approve of proto/projects/lib and proto/projects/bin directories, to act as unification nodes for diverse projects? I would deeply appreciate a 'yes'. 09:40
masak I've fixed the 'Rakudo dir not defined' issue, so if the Rakudo dir is not set, it now sets it and re-saves the config file. I'll push in a minute.
mberends: unh. 09:41
mberends: so you're talking about copying stuff out of their native git directories?
mberends yes. each project's Makefile.in says what to copy in the 'install' target 09:42
masak at present, I'm very reluctant to such an idea.
mberends ok 09:43
masak proto's goal, in my eyes, is to keep the stricture of the projects, while still alleviating much of the pain of having many lib/ dirs
basically, your PERL6LIB changes around a lot when you install new Perl 6 projects, but proto keeps the tabs and tells you what to set it to. 09:45
mberends the PERL6LIB growth and maintenance is what I don't like 09:46
masak that's what proto'll take care of for you.
but szabgab has expressed similar misgivings, so I'm not all made up on this point. 09:47
by the way, one reason I added the config.proto and the 'Proto projects directory' setting, is so that I'll go dogfooding very soon.
I'll also want the 'update' subcommand, so I'll have a one-stop way to update all my installed Perl 6 projects. :) 09:48
mberends and even update rakudo
masak mberends: yes, but remember that the default will be 'released rakudo' 09:49
hm, I'll add that setting too.
...which means that we'll leave git as our primary way to get Rakudo.
mberends there is lots of recent debate about what 'release' etc means. To me, it has to mean 'not broken according to *all* testing' 09:50
masak aye.
that's what we want.
and if it;s proto's default, hopefully Perl 6 projects will have to align to that default. :) 09:51
specifically November, which uses many bleeding-edge features.
mberends I like the fact that the debate is beginning to look at expanding testing to beyond the spectests into some modules or projects. Some of our stuff would be good candidates. 09:52
masak indeed.
I know pmichaud and others have (at times) been wary not to break November when doing great merges.
09:53 protorom joined
masak that has helped a lot, I believe. 09:53
at least that's the impression I get from the few occasion where such care has not been taken. :P
mberends we shall need 100+ november scale test cases eventually
masak November has a fair amount of tests.
mberends plans to expand Pod::Parser and its tests soon 09:54
masak but the fact of the matter is, that as deficiencies in the spectest suite are discovered through November, they usually get assimilated into the spectest suite.
masak will keep expanding SVG though TDD
mberends pmichaud wants spectest to remain as concise (and fast) as possible, but seemed interested in broader testing. When announcing Perl 6, TimToady said CPAN would be the ultimate test suite ;) 09:57
masak mberends: one thing I'm lacking in the config.proto file right now is comments. I'd like to at least be able to give the allowed values for ever setting. so I think custom generation of the config.proto file is a good idea.
mberends yes. good feature. 09:58
masak adds the settings 'Parrot revision', 'Rakudo revision' (both with possible values 'bleeding' and 'release'), 'Test when building', and 'Test failure policy' 10:00
10:00 mikehh joined
masak currently the defaults are 'bleeding', 'bleeding', 'no', and 'die', respectively. 10:00
but the first two should eventually be 'release, and 'release'. 10:01
hm, config.proto file versioning -- good idea or bad? 10:02
maybe duck typing is enough? :)
masak tentatively goes with date versioning
ISO 8601. 10:03
please stop me if I'm overdesigning.
10:03 szabgab left 10:04 ejs joined
mberends goes with dates too, assuming ISO 8601 means YMD ordering 10:04
masak 2009-02-28. 10:05
mberends my favo[u]rite!
masak :)
YAML::Syck ASCII sorts the keys of the serialized hash. 10:06
I'm not sure if I like that.
would be nicer if they came in insertion order, methinks.
10:06 szabgab joined
protorom hi guys, just trying to install parrot-0.9.0 on OS-X and Configure.pl seems to be failing - no parrot_config gets produced :-( 10:06
during the tests I get the error: dyld: lazy symbol binding failed: Symbol not found: _rl_get_keymap 10:07
masak protorom: could you nopaste the output you're seeing?
protorom nopaste?
masak sial.org/pbot/perl6
:)
see above URL. 10:08
mberends Some hashes do preserve order, despite being hashes. We can probably use arrays and grep/map them instead.
protorom coming up...
masak mberends: aye, as long as it's easy to add settings when constructing the $config_info variable, I'm fine with it. 10:09
protorom heres the output: sial.org/pbot/35283
szabgab masak, and he is using the released tarball 10:10
masak szabgab: aye.
protorom: I'm on OS X too. I'll try the tarball and see if I get the same. 10:11
I think it has worked before, here.
masak needs to go soon
szabgab masak, I am getting him to download the svn
protorom that was the tarball :-( I'm just checking out the svn repo now
masak goodie. 10:12
but the tarball should work, too.
mberends masak: I need to go soon too. Can you send out a draft config.proto file? 10:13
masak just a minute. 10:15
pushed. 10:16
it's all yours.
mberends: during the day, I might sketch out an 'update' subcommand, if that's ok. 10:17
I'll be mostly mucking around in 'installer'.
(and in a branch)
mberends very ok. pulled. hopefully see you later 10:18
masak not today, I fear.
but very possibly tomorrow.
mberends ok 10:19
masak thanks for great work on proto so far -- and see you tomorrow.
protorom masak: still getting the same problem. do you know which version worked for you?
masak protorom: the tarball, several times, I think.
10:19 mberends left
masak protorom: now might be a good time to file a Parrot Trac ticket. 10:20
protorom: or to ask on #parrot on irc.perl.org
masak waves
10:20 masak left
protorom ok - thanks 10:21
szabgab rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv -> $k, $v { say "$k, $v" } 10:26
p6eval rakudo 9e2462: OUTPUT«a, 2␤b, 1␤»
szabgab rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort(:by keys) -> $k, $v { say "$k, $v" } 10:27
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near "(:by keys)"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
szabgab how do I sort that ?
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort :by keys -> $k, $v { say "$k, $v" } 10:31
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near ":by keys -"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort -> $k, $v { say "$k, $v" } 10:34
p6eval rakudo 9e2462: OUTPUT«a, b␤1, 2␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort { $b <=> $a } -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near "-> $k, $v "␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort({ $b <=> $a }) -> $k, $v { say "$k, $v" } 10:35
p6eval rakudo 9e2462: OUTPUT«Scope not found for PAST::Var '$b' in ␤current instr.: 'parrot;PCT;HLLCompiler;panic' pc 146 (src/PCT/HLLCompiler.pir:102)␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort({ $^b <=> $^a }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«2, 1␤a, b␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.kv.sort({ $^b.key <=> $^a.key }) -> $k, $v { say "$k, $v" } 10:38
p6eval rakudo 9e2462: OUTPUT«Method 'key' not found for invocant of class 'Str'␤current instr.: 'parrot;P6metaclass;dispatch' pc 320 (src/classes/ClassHOW.pir:161)␤»
szabgab wayland76, thanks
wayland76 But I haven't answered the question yet :)
10:39 duke_leto left
szabgab oh you have an answer :-) 10:39
cool
wayland76 I don't have one yet, but I'm getting closer :)
szabgab great
10:40 pmurias joined
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.p.sort({ $^b.key <=> $^a.key }) -> $k, $v { say "$k, $v" } 10:41
p6eval rakudo 9e2462: OUTPUT«Method 'p' not found for invocant of class 'Perl6Hash'␤current instr.: 'parrot;P6metaclass;dispatch' pc 320 (src/classes/ClassHOW.pir:161)␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.pairs.sort({ $^b.key <=> $^a.key }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«a 2, b 1␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.pairs.sort({ $^b.key cmp $^a.key }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«b 1, a 2␤»
wayland76 Ok, that's the answer 10:42
Does that work for you?
pmurias TimToady: "echo 123 > file;perl viv file" prints --- '123' and it doesn't work for other simple programs too 10:43
lambdabot pmurias: You have 1 new message. '/msg lambdabot @messages' to read it.
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.pairs.sort({ $^b.value cmp $^a.value }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«a 2, b 1␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.pairs.sort({ .key }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«a 2, b 1␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.pairs.sort({ .value }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«b 1, a 2␤»
wayland76 Cool :)
szabgab: That might be a shorter answer :) 10:44
szabgab wayland76++ 10:45
will it be wayland77 now ?
wayland76 Well, that would mean that I was younger, and that could be useful :)
After all, youth is wasted on the young :) 10:46
(I was born in '76)
rakudo: my %h = ("a" => 2, "b" => 1); for %h.sort({ .key }) -> $k, $v { say "$k, $v" }
p6eval rakudo 9e2462: OUTPUT«a 2, b 1␤»
wayland76 rakudo: my %h = ("a" => 2, "b" => 1); for %h.sort({ .value }) -> $k, $v { say "$k, $v" } 10:47
p6eval rakudo 9e2462: OUTPUT«b 1, a 2␤»
wayland76 ...and shorter :)
10:47 misfotto joined
wayland76 szabgab: Btw, I have the specs on my computer, and I searched them with ack (which is grep but better), and first I searched for "sort", and read that section 10:48
...and then I searched for .kv, and read the bits in S02. Of course, I already knew I wanted to ensure I was getting 2 Pairs instead of 4 Scalars, so that helped :) 10:50
Anyway, I have work to do. &
10:56 misfotto left, kidd joined
kidd hi guys. I'm trying to do my first perl6 scripts, and I can't find a way to execute external processes with system nor ``. I haven't found any docs about that. Pointers to documentation are really welcome 11:11
11:16 pmurias left
wayland76 perlcabal.org/syn/ 11:26
That link documents how Perl6 is *supposed* to work
That way the guys writing code know what to write
It seems to imply that `` works as always 11:27
and that an alternative could be qqx/$cmd $arg1/
Or, if you don't want interpolation, qx{cmd arg} 11:28
kidd oh, then I must be doing something wrong... I'll keep reading. thanks for the pointers wayland76
wayland76 The relevant document is the one called S02
szabgab kidd, go over this szabgab.com/perl6.html and let me anything that is not clear
let me "know" ...
wayland76 Which you get to from the "Synopsis" link next to "2 Bits and Pieces"
I hate it when I a word out :) 11:29
11:29 protorom left
wayland76 rakudo: print `echo hi` 11:29
p6eval rakudo 9e2462: OUTPUT«print requires an argument at line 1, near " `echo hi`"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: print qx/echo hi/ 11:30
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near "/"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: print qx/echo hi/;
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near "/;"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: print qx"echo hi";
p6eval rakudo 9e2462: OUTPUT«Statement not terminated properly at line 1, near "\"echo hi\";"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤»
wayland76 rakudo: print qq"echo hi";
p6eval rakudo 9e2462: OUTPUT«echo hi»
wayland76 Well, I guess the sandboxing must work :) 11:31
Anyway, bedtime for me &
kidd gnight
bacek @tell pmichaud I've made single version of REDUCEMETAOP. github.com/bacek/rakudo/blob/84c89c...assign.pir But I don't like it... It's too complex 11:34
lambdabot Consider it noted.
bacek @tell pmichaud and chained version is not short-circuiting anymore... 11:37
lambdabot Consider it noted.
11:37 szabgab left 11:40 rindolf joined 11:55 hercynium joined 11:57 AzureStone left 12:09 AzureStone joined 12:13 rindolf left 12:14 ejs left 12:28 meppl joined 12:41 c9s left, c9s joined 12:53 riffraff joined 13:36 rindolf joined 14:03 ZuLuuuuuu joined 14:08 ZuLuuuuuu left 14:10 Whiteknight joined, aindilis joined 14:34 jan_ left 14:35 jan_ joined 14:52 hercynium left 14:58 brunoV joined 15:04 pmurias joined 15:08 jhorwitz joined 15:11 rindolf left 15:24 kanru left 15:35 kanru joined 15:37 kidd left 15:40 rindolf joined 15:43 kanru2 joined 15:55 cibs joined 16:00 hercynium joined 16:08 kanru left 16:11 Ariens_Hyperion joined 16:13 SamB left 16:16 abra joined 16:17 SamB joined 16:22 Kimtaro joined 16:28 nihiliad joined
dalek kudo: 57a0cf4 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 315 files, 7084 passing, 0 failing
 16:35
kudo: f7a7bf4 | pmichaud++ | tools/test_summary.pl:
We no longer need the "[rakudo]:" part of the commit message.
16:40 justatheory joined, Kimtaro_ joined, Whiteknight left 16:41 Kimtaro_ left 16:42 aindilis` joined 16:53 riffraff left 16:56 aindilis left 16:58 Kimtaro left 17:03 Tene_ joined 17:05 duke_leto joined 17:07 duke_leto left 17:13 rindolf left 17:16 Tene left
pugs_svn r25637 | putter++ | [elfparse] Sketch in more ratcheting (subrules). 17:21
17:26 mikehh left 17:29 pmurias left 17:38 meppl left 17:42 justatheory left 17:51 cognominal left 17:53 meppl joined 17:56 Kimtaro joined, simcop2387 left 17:58 cognominal joined 18:05 mikehh joined 18:08 protorom joined 18:14 simcop2387 joined 18:15 Southen_ joined 18:20 Psyche^ joined
protorom rakudo: help 18:20
p6eval rakudo f7a7bf: OUTPUT«Could not find non-existent sub help␤current instr.: '_block14' pc 53 (EVAL_16:38)␤»
18:22 rindolf joined 18:30 Patterner left, Psyche^ is now known as Patterner 18:32 rindolf left 18:33 Southen left 18:45 abra left 18:49 protorom_ joined, rindolf joined 19:07 protorom left 19:12 mberends joined 19:14 mberends left, mberends joined, mberends_ joined, mberends_ left, mberends left 19:28 protorom_ left
dalek kudo: d4c266b | pmichaud++ | src/parser/actions.pm:
Revert "Fix exception handlers to not catch control exceptions."

which was causing numerous spectests to fail.		 19:47
19:49 protorom joined 19:50 aindilis` left
skids point of confusion -- when we get %h{...}:exists, will we still have %h.exists(...)? 19:50
pmichaud I think the .exists(...) form goes away.
lambdabot pmichaud: You have 3 new messages. '/msg lambdabot @messages' to read them.
skids I'm looking for a definite answer, since that might have been people overinterpreting the demise of sub exists/exists(%h{...}). 19:52
pmichaud neither ".exists" nor ".delete" appear in the synopses. 19:53
nor is there any mention of a subcall form. To me, that'd be pretty good evidence that it's only :exists and :delete. 19:54
skids Where is the notation of "method :name" documented? 19:55
pmichaud I think that's pseudo-notation.
skids Ahh, ahh, aHH no pseudo-notation allowed :-) 19:56
pmichaud i.e., I think that's a bit more made-up, unless it's meant to be a longname form somehow.
I haven't been following the syntactic details on that very closely.
otoh, .exists does appear in the spectests 19:57
so.... I dunno.
pmichaud checks the svn revision history.
skids Yeah my personal hypothesis until I find a definitive source is that the idea that .exists and .delete are going away as methods is a snowballing rumor. 19:58
pmichaud argggh, I can't seem to review the log that far back. 19:59
the message at groups.google.com/group/perl.perl6....d89f7115f1 says 20:03
Note also that existence and deletion are now simply adverbs on subscript
operators.
20:04 pmurias joined
pmichaud and S12 used to have a "%hash.:exists{...}" form, which is now gone. 20:04
(as shown in that same message) 20:05
so, I'm pretty sure the method forms are no more -- they're just adverbs on subscripts. 20:07
skids Well, that explains where the idea of "method :name" might have come from.
pmichaud one thing that does concern me about S29/S32 is that they still show .grep, .join, etc as being methods on List and not Any 20:08
skids To me that seems maybe reading a bit too much into it (and besides, when we implement something like them will be needed at least internal to the class, because we don't want a flag switch run every time in postcircumfix:<{ }>.
pmichaud flag switch? That's what mmd is for. 20:09
skids The only reason to ditch the method form IMO would be if there's no way to elegantly reconcile .delete(), .delete, and zen slices.
Sorry I was mentally bactracking :-) 20:12
pmichaud anyway, a message to p6l is probably in order then. 20:16
skids I'm not subbed there, I just can't handle mls these days. 20:17
20:21 mikehh left 20:23 pmurias left
skids By the way since you're around -- is there supposed to be a way to add multis on top of stuff that's written in PIR yet? Doing so naively didn't work for me. 20:24
20:24 protorom left
pmichaud there's not a way to do that yet, no. 20:26
it's one of the reasons where moving things from PIR into settings/
s/where/we're/
skids In that case, maybe an example of a q:PIR in settings by someone who really knows what they are doing might help encourage things. 20:27
pmichaud sure -- I just haven't had a chance to write that yet. I spent most of the week working on getting our first release out.
skids s/settings/setting/
pmichaud (our first non-Parrot release, that is.)
skids :-) I know we take and take and take... :-)
20:30 viklund joined 20:32 justatheory joined
skids Speaking of MMD, I was looking into what future multidim stuff might look like to optimize common cases and having a few convenience methods for where {...} might make things less ugly. like .dims instead of .shape.elems which might do too much math, and something that quickly checks whether all slices are .elems == 1 to pick out single value accesses. I guess that's far down the road though. 20:42
20:45 mikehh joined
diakopter hm, ./viv doesn't seem to work the same as it once did 20:48
pmichaud std: if(3) { say 'hello'; } 20:50
p6eval std 25637: OUTPUT«############# PARSE FAILED #############␤if() interpreted as function call at line 1 ; please use whitespace instead of parens␤Unexpected block in infix position (two terms in a row) at /tmp/eUeZz8RwQD line 1:␤------> if(3) { say 'hello'; }␤ expecting any of:␤
..in...
diakopter if (3) { say 'hello' } 20:51
std: if (3) { say 'hello' }
p6eval std 25637: OUTPUT«ok 00:02 34m␤»
diakopter ./viv doesn't give an ast anymore ?
20:52 protorom joined 20:53 PZt left
diakopter std: sub if($a) { $a }; if (3) { say if(3) }; 21:03
p6eval std 25637: OUTPUT«ok 00:02 36m␤»
pmichaud std: say 'hello' if(3); 21:07
p6eval std 25637: OUTPUT«############# PARSE FAILED #############␤Syntax error (two terms in a row?) at /tmp/6pLyiRgJ2t line 1:␤------> say 'hello' if(3);␤ expecting nofun␤FAILED 00:02 34m␤»
diakopter TimToady: around? I removed ->item from line 67 of viv and it spits out the ast again, although it doesn't read from stdin 21:16
21:21 Whiteknight joined 21:27 brunoV left
dalek kudo: 0bb4da5 | pmichaud++ | src/parser/grammar.pg:
Slight STD convergence -- switch "rule statement" to "token statement"
 21:30
kudo: dd9b61c | pmichaud++ | src/parser/grammar.pg:
Some more parsing fixes.

  * change some {{-tests to use <?{{...}}> instead.
21:32 rindolf left 21:40 wayland76 left 21:44 SamB left 21:46 protorom left, SamB joined
skids rakudo: multi sub f (%h where { .{'a'} == 1 }) { say "OH HAI" }; multi sub f (%h where { .{'a'} != 1 }) { say "O NOES" }; my %c = <a 1 b 2>; f(%c); 21:57
p6eval rakudo dd9b61: OUTPUT«Method 'postcircumfix:{ }' not found for invocant of class 'Str'␤current instr.: 'postcircumfix:{ }' pc 3797 (src/classes/Associative.pir:77)␤»
pmichaud that's kinda weird. 21:59
skids rakudo: multi sub f (%h where { %^j{'a'} == 1 }) { say "OH HAI" }; multi sub f (%h where { %^j{'a'} != 1 }) { say "O NOES" }; my %c = <a 1 b 2>; f(%c);
p6eval rakudo dd9b61: OUTPUT«Non-Associative argument for %j in call to _block30␤current instr.: 'die' pc 16347 (src/builtins/control.pir:204)␤»
pmichaud rakudo: multi sub f (%h where { .PARROT.say }) { say "OH HAI" }; my %c = <a 1 b 2>; f(%c); 22:00
p6eval rakudo dd9b61: OUTPUT«OH HAI␤»
skids Actually the first one differs from what I get at the cl.
pmichaud rakudo: multi sub f (%h where { .PARROT.say }) { say "OH HAI" }; my %c = <a 1 b 2>; %c.PARROT.say; f(%c);
p6eval rakudo dd9b61: OUTPUT«Perl6Hash␤OH HAI␤»
pmichaud rakudo: multi sub f (%h is rw where { .PARROT.say }) { say "OH HAI" }; my %c = <a 1 b 2>; %c.PARROT.say; f(%c); 22:01
p6eval rakudo dd9b61: OUTPUT«Perl6Hash␤OH HAI␤»
skids There I get "Could not find non-existent sub a" when using .{}
pmichaud rakudo: multi sub f (%h is rw where { .PARROT.say }) { say %h.perl }; my %c = <a 1 b 2>; f(%c);
p6eval rakudo dd9b61: OUTPUT«{"a" => "1", "b" => "2"}␤»
pmichaud rakudo: multi sub f (%h where { .PARROT.say }) { say %h.PARROT }; my %c = <a 1 b 2>; f(%c);
p6eval rakudo dd9b61: OUTPUT«ObjectRef->Perl6Hash␤»
pmichaud I suspect a problem with the where invocation. 22:02
rakudo: multi sub f (%h where { +$_ == 2 }) { say %h.PARROT }; my %c = <a 1 b 2>; f(%c);
p6eval rakudo dd9b61: OUTPUT«ObjectRef->Perl6Hash␤»
pmichaud rakudo: multi sub f (%h where { +$_ == 3 }) { say %h.PARROT }; my %c = <a 1 b 2>; f(%c);
p6eval rakudo dd9b61: OUTPUT«ObjectRef->Perl6Hash␤»
pmichaud yes, the where clause looks like it's having trouble there. 22:03
22:06 szabgab joined 22:23 eternaleye joined
pmichaud std: sub f() { say 'foo' } if; 22:25
p6eval std 25637: OUTPUT«Undeclared routine:␤ if used at 1 ␤ok 00:02 34m␤»
pmichaud std: sub f() { say 'foo' } if 22:26
p6eval std 25637: OUTPUT«Undeclared routine:␤ if used at 1 ␤ok 00:02 34m␤»
22:26 jhorwitz left
skids So what is the class that is "something that is not a List" for MMD-on-return-value purposes? 22:29
22:43 viklund left, ZuLuuuuuu joined 23:01 aindilis joined 23:09 aindilis left, aindilis joined
skids kinda wonders what's up with Hash::reverse, both in the Synopsis and as it has been copied there, the setting. 23:14
I could understand gather for %hash.kv -> $k, $v { take $v, $k; } for use to reverse hash explosions in List::reverse but... 23:16
with the value to key aliasing if someone wants that it shouldn't be called reverse 23:17
s/that/the thing that is now implemented/
dalek kudo: 36314e0 | pmichaud++ | src/parser/grammar.pg:
Refactor check for curly brace statement terminator.
 23:18
kudo: 1e22a68 | pmichaud++ | src/parser/grammar.pg:
Fix RT #57876 -- if/while/unless/for/etc on line following closing brace
pmichaud skids: I think that .reverse needs a substantial re-think. 23:34
I commented on this last year -- just a sec.
groups.google.com/group/perl.perl6....8c5d9509fd 23:36
skids Well, whatever is done in that respect, surely what we have now is dead wrong. 23:41
rakudo: my %f = <a 2 b 2 c 3>; %f.reverse.say 23:42
p6eval rakudo 1e22a6: OUTPUT«2 b3 c␤»
23:49 ZuLuuuuuu left
skids Yeah looking at this, hashes explode to pairs, in no particular order, unless fully flattened. Since they have no guaranteed order between calls, reversing them makes no sense. And you have sort.reverse using List's reverse if you really need it. So Hash should not export reverse. 23:53
Though it might be nice to have a less-computationally-intense-than-sort method that keeps order intact between modifications of the underlying data. 23:54
That is, so long as it isn;t modified.
pmichaud there might be some discussion about that in S09. 23:55
I could potentially see method reverse() { self.kv.reverse; } 23:56
so that %hash.reverse doesn't end up being the same as @(%hash).reverse 23:57

Powered by Cro and the Raku® Programming Language.

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

Thank you!