»ö« | perl6-projects.org/ | nopaste: sial.org/pbot/perl6 | evalbot: 'perl6: say 3;' | irclog: irc.pugscode.org/ | UTF-8 is your friend!
Set by Tene on 14 May 2009.
00:00 DemoFreak left 00:03 felipe joined 00:12 takadonet left 00:13 silug left 00:15 takadonet joined 00:17 cj left 00:19 cj joined 00:27 jferrero left 00:28 icwiener_ left 00:34 bacek joined 00:36 japhb left 00:39 Whiteknight joined 00:51 baest left 00:52 baest joined 00:54 cavelife^ left, cavelife^ joined 00:59 skids left 01:02 DanielC left
Sark hi 01:02
01:03 Chillance left 01:12 kimtaro left 01:13 kimtaro joined 01:22 viirya joined, takadonet left 01:25 dduncan joined
dduncan I'm trying to figure out the right terminology for something I once saw in Perl 6 ... 01:27
01:27 synth left
dduncan the kind of twigil you use in a closure to reference your caller's lexicals 01:27
unless that has gone away in favor of lift
specifically I refer to accessing the lexicals of the routine where the closure was *declared*
01:28 synth joined
dduncan or maybe I mean used ... 01:28
s1n dduncan: $*CALLER? 01:29
dduncan for example, when foo() calls "map", you are inline-declaring and passing a Callable to "map", which "map" then invokes on your behalf ... and that callable may want to reference lexicals in foo()
s1n viklund: does Web.pm effort (on github) not currently build? 01:30
01:32 kimtaro left
s1n rakudo: class A { method foo($bar) 01:32
say "oh hai"
end
}
doh
p6eval rakudo 77db80: OUTPUT«Malformed method definition at line 2, near "foo($bar)"␤in Main (src/gen_setting.pm:0)␤»
s1n viklund: it seems to not like the "method redirect(...) \n...\nend" format 01:33
is that a valid way of defining functions anymore?
01:34 kimtaro joined
StephenPollei rakudo: class A { method foo($bar) {...} } 01:35
p6eval rakudo 77db80: ( no output )
meppl good night
StephenPollei rakudo: class A { method foo($bar) { say 'cheese'} } ; my A $a= A.new(); a.foo() 01:39
p6eval rakudo 77db80: OUTPUT«Could not find non-existent sub a␤»
Tene s1n: has that *ever* been valid?
StephenPollei rakudo: class A { method foo($bar) { say 'cheese'} } ; my A $a= A.new(); $a.foo()
p6eval rakudo 77db80: OUTPUT«too few arguments passed (1) - 3 params expected␤in method A::foo (/tmp/YQYJOjgxaf:2)␤called from Main (/tmp/YQYJOjgxaf:2)␤»
Tene s1n: I don't remember ever seeing that in Perl 6 specs. 01:40
s1n Tene: dunno, it's in Web/Response.pm
pmichaud ...what's the question?
Tene in masak's Web repo? 01:41
s1n pmichaud: lemme linky...
StephenPollei rakudo: class A { method foo($bar) { say 'cheese'} } ; my A $a= A.new(); $a.foo('bar')
p6eval rakudo 77db80: OUTPUT«cheese␤»
01:41 meppl left
Tene method foo() \n...\n end 01:41
s1n github.com/masak/web/blob/56156ff72...nse.pm#L23
i've never seen that syntax before
rakudo currently barfs, as does the configure/make for web 01:42
swap it out with the usual {} suspects and it works fine
Tene s1n: I have an older version of that repo, and it doesn't have that.
s1n err it builds fine
Tene I suspect user error
s1n Tene: dunno, i assumed that was current
pmichaud that's a recent change, yes.
s1n Tene: you have commit access iirc, go ahead and change it :) 01:43
pmichaud it was added on 05-18
Tene pmichaud: ETA for protoregexes, btw?
pmichaud er, 05-19
Tene: I don't know at the moment. It may turn out that I need to redo the regex parser first.
Tene s1n: I'm on my phone right now... no free wifi at minneapolis airport
01:43 simcop2387 joined
Tene pmichaud: OK. :) 01:43
01:43 simcop2387-vnc joined, payload left
s1n Tene: okay, don't know how you and masak are handling issues, but i thought i'd point it out 01:44
01:44 simcop2387-vnc left
Tene s1n: thanks. :) 01:44
s1n Tene: any time
pmichaud My inner sense of timing is telling me not to push too hard on it yet. Usually that's because there's another change about to come up in the spec. :-)
At least, whenever I've listened to that inner sense of timing before, I've been glad for it shortly thereafter :-) 01:45
01:58 payload joined 01:59 kimtaro left, takadonet joined
takadonet hey everyone 01:59
TimToady dduncan: you're looking for OUTER::<$x> 02:00
or just skip the OUTER and use normal lexical scoping 02:01
takadonet what is an easy way to determine if iteration in a loop is the last one? 02:02
TimToady the easiest way to predict the future is to create it 02:03
takadonet Indeed 02:04
02:06 snarkyboojum left 02:12 silug joined 02:18 viirya left, viirya joined 02:24 kate21de left 02:35 Whiteknight left 02:38 agentzh left 02:54 payload left 02:55 viirya left, viirya joined, viirya left 02:59 takadonet left 03:00 viirya joined, viirya left 03:05 viirya joined 03:11 agentzh joined 03:15 pyrimidine joined 03:17 agentzh left, agentzh joined 03:18 agentzh left 03:19 agentzh joined 03:30 david___ joined
pyrimidine rakudo: for 1..3 -> $i { for $i..3 -> $j {}; say $i } 03:34
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤»
TimToady er, ow
pyrimidine yeah
but it loops three times 03:35
TimToady it's incrementing $i rather than a temporary
pyrimidine rakudo: for 1..3 -> $i { my $tmp = $i; for $i..3 -> $j {}; say $i }
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤» 03:36
pyrimidine rakudo: for 1..3 -> $i { my $tmp = $i; for $i..3 -> $j {}; say $tmp }
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤»
TimToady now that's bizarre
pyrimidine rakudo: for 1..3 -> $i { say $i }
p6eval rakudo 77db80: OUTPUT«1␤2␤3␤»
TimToady rakudo: for 1..3 -> $i { my $tmp = $i; for $tmp..3 -> $j {}; say $i } 03:39
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤»
dduncan TimToady, thanks for the response ...
TimToady rakudo: for 1..3 -> $i { my $tmp = $i; for $tmp..3 -> $j {}; say $i.WHAT }
p6eval rakudo 77db80: OUTPUT«Int()␤Int()␤Int()␤»
dduncan the situation I was thinking of was specifically how to deal with the general case where different scopes often have a variable with the same lexical name, such as $_
TimToady then OUTER::<$_> is fine 03:40
dduncan okay
TimToady OUTER::OUTER::<$_> etc
dduncan excellent, just the sort of thing I was hoping for
TimToady corresponding for the dynamic scope is CALLER::
pyrimidine rakudo: for 1..3 { for $_..3 -> $j {}; $_.say } 03:41
03:41 viirya left
p6eval rakudo 77db80: OUTPUT«1␤2␤3␤» 03:41
pyrimidine rakudo: for 1..3 { for $_..3 -> $j {$j.say}; }
p6eval rakudo 77db80: OUTPUT«1␤2␤3␤2␤3␤3␤»
03:42 davidad joined
TimToady rakudo: for 1..3 -> $i { my $tmp = $i; for $tmp..3 -> $j {}; $i.say } 03:42
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤»
pyrimidine rakudo: for 1..3 -> $i { for $i..3 -> $j {}; $i.say }
p6eval rakudo 77db80: OUTPUT«4␤4␤4␤»
pyrimidine ick 03:43
pyrimidine rakudobug
TimToady like $i.say is being way too lazy
03:43 FurnaceBoy left
TimToady rakudo: for 1..3 -> $i { my $tmp = $i; for $tmp..4 -> $j {}; $i.say } 03:43
p6eval rakudo 77db80: OUTPUT«5␤5␤5␤»
TimToady yeep, it's the inner loop
probably just clobbering a register 03:44
rakudo: for 1..3 -> $i { my $tmp = $i; for $tmp..4 -> $j { say "($j;$i)"}; $i.say } 03:45
p6eval rakudo 77db80: OUTPUT«(1;5)␤(2;5)␤(3;5)␤(4;5)␤5␤(2;5)␤(3;5)␤(4;5)␤5␤(3;5)␤(4;5)␤5␤»
TimToady yow
pyrimidine that's nasty
TimToady $i is being clobbered with the off-the-end range value, or something
pyrimidine yup 03:46
TimToady I think you win the Nasty Bug of the Day award
pyrimidine thanks, I think
03:50 xinming_ joined 03:51 xinming is now known as Guest55490 03:53 xinming_ is now known as xinming
pyrimidine rakudo: for 1..3 { for $_..4 -> $j { say "($j;$_)"}; $_.say} 03:54
p6eval rakudo 77db80: OUTPUT«(1;1)␤(2;1)␤(3;1)␤(4;1)␤1␤(2;2)␤(3;2)␤(4;2)␤2␤(3;3)␤(4;3)␤3␤» 03:55
TimToady rakudo: for 1..3 -> $i is copy { my $tmp = $i; for $tmp..4 -> $j { say "($j;$i)"}; $i.say } 03:56
p6eval rakudo 77db80: OUTPUT«(1;1)␤(2;1)␤(3;1)␤(4;1)␤1␤(2;2)␤(3;2)␤(4;2)␤2␤(3;3)␤(4;3)␤3␤»
TimToady heh
pyrimidine we have a winner
TimToady rakudo: for 1..3 -> $i is rw { my $tmp = $i; for $tmp..4 -> $j { say "($j;$i)"}; $i.say } 03:57
p6eval rakudo 77db80: OUTPUT«(1;1)␤(2;1)␤(3;1)␤(4;1)␤1␤(2;2)␤(3;2)␤(4;2)␤2␤(3;3)␤(4;3)␤3␤»
TimToady rakudo: for 1..3 -> $i is readonly { my $tmp = $i; for $tmp..4 -> $j { say "($j;$i)"}; $i.say }
p6eval rakudo 77db80: OUTPUT«(1;5)␤(2;5)␤(3;5)␤(4;5)␤5␤(2;5)␤(3;5)␤(4;5)␤5␤(3;5)␤(4;5)␤5␤» 03:58
TimToady rakudo: for 1..3 -> $_ is readonly { my $tmp = $_; for $tmp..4 -> $j { say "($j;$_)"}; $_.say }
p6eval rakudo 77db80: OUTPUT«(1;5)␤(2;5)␤(3;5)␤(4;5)␤5␤(2;5)␤(3;5)␤(4;5)␤5␤(3;5)␤(4;5)␤5␤»
TimToady something to do with the readonly proxy, maybe
pyrimidine appears so 03:59
TimToady rakudo: for 1..3 -> $_ { my $tmp = $_; for $tmp..4 -> $j { say "($j;$_)"}; $_.say } 04:00
p6eval rakudo 77db80: OUTPUT«(1;5)␤(2;5)»
TimToady o_O
rakudo: for 1..3 <-> $_ { my $tmp = $_; for $tmp..4 -> $j { say "($j;$_)"}; $_.say }
p6eval rakudo 77db80: OUTPUT«(1;1)␤(2;1)␤(3;1)␤(4;1)␤1␤(2;2)␤(3;2)␤(4;2)␤2␤(3;3)␤(4;3)␤3␤»
pyrimidine rakudo: for 1,3 -> $i { my $tmp = $i; for $tmp,4 -> $j { say "($j;$i)"}; $i.say } 04:04
p6eval rakudo 77db80: OUTPUT«(1;1)␤(4;1)␤1␤(3;3)␤(4;3)␤3␤»
pyrimidine rakudo: for 1,3 -> $i { my $tmp = $i; for $tmp..4 -> $j { say "($j;$i)"}; $i.say } 04:06
p6eval rakudo 77db80: OUTPUT«(1;5)␤(2;5)␤(3;5)␤(4;5)␤5␤(3;5)␤(4;5)␤5␤»
pyrimidine It's Range
04:07 xomas_ is now known as xomas, Guest55490 left 04:08 alester_ joined 04:13 hercynium left 04:14 japhb joined 04:17 pyrimidine_ joined 04:19 pyrimidine left, pyrimidine__ joined, pyrimidine__ is now known as pyrimidine 04:25 Limbic_Region left
pyrimidine RT #66280 on the pointy blocks issue above 04:29
TimToady pyrimidine++
pyrimidine sleep 04:30
TimToady night
04:32 pyrimidine left 04:36 pyrimidine_ left 04:37 nihiliad left 04:39 szabgab joined 04:43 sri_kraih left 04:50 ZuLuuuuuu left 04:53 lambdabot joined 05:09 cotto left, beggars joined 05:10 cotto joined 05:15 teun_ joined 05:16 teun_ left, justatheory left, Jaap__ joined, Jaap__ left, Jaaaap joined
Jaaaap Does anybody know how to fix an "Could not find non-existent sub if" error in rakudo? 05:17
05:19 kate21de joined 05:21 dduncan left
Tene Jaaaap: what sub are you trying to call? 05:36
Jaaaap no real sub just if (a == 3) { print 3; } 05:37
Tene that's trying to call a sub named 'a'. 05:38
05:45 davidad left
Jaaaap oh sorry i mean $a 05:46
05:46 finanalyst joined
finanalyst hi everyone 05:47
05:47 alester_ left
Tene rakudo: my $a 3; if $a == 3 { print 3 } 05:47
p6eval rakudo 77db80: OUTPUT«Statement not terminated properly at line 2, near "3; if $a ="␤in Main (src/gen_setting.pm:0)␤»
Tene finanalyst: hi
finanalyst i just send an email to p6l about slowness parsing a grammar. Was that a good place to send, or should I ask here
Tene rakudo: my $a = 3; if $a == 3 { print 3; }; 05:48
p6eval rakudo 77db80: OUTPUT«3»
Tene Jaaaap: like that?
finanalyst: parrot-dev is where I would have sent it, fwiw. 05:49
prolly fine where it it, though.
Jaaaap Tene, i'll post the code. hold on. 05:56
grammar Grmmr { .. }
class Actions { .. }
my $action = Actions.new(); 05:57
Grmmr.parse('alpha beta', :action($action));
if($action.in-a == 1){ print 'first action has been called'};
ERROR
strange eh? 06:17
TimToady what is the error?
Jaaaap Could not find non-existent sub if 06:18
TimToady replace the parens with whitespace
std: if($action.in-a == 1){ print 'first action has been called'};
Jaaaap ok hold on
p6eval std 26999: OUTPUT«Potential difficulties:␤ Variable $action is not predeclared at /tmp/ozzaGCqd2y line 1:␤------> if($action.in-a == 1){ print 'first action has bee␤Undeclared routine:␤ if used at 1 ␤ok 00:02 36m␤»
TimToady std: my $action; if($action.in-a == 1){ print 'first action has been called'};
p6eval std 26999: OUTPUT«Undeclared routine:␤ if used at 1 ␤ok 00:02 36m␤»
TimToady std: my $action; if($action.in-a == 1) { print 'first action has been called'}; 06:19
p6eval std 26999: 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/0RRaXKYKgW line 1:␤------> my $action; if($action.in-a == 1) { print 'first action has been
..calle…
TimToady std: my $action; if ($action.in-a == 1) { print 'first action has been called'};
p6eval std 26999: OUTPUT«ok 00:02 36m␤»
TimToady std: my $action; if (\$action.in-a == 1 { print 'first action has been called'};
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?) at /tmp/BZcifpOoF3 line 1:␤------> my $action; if (\$action.in-a == 1 { print 'first action has been called'};␤ expecting any of:␤
..infix or me…
TimToady std: my $action; if $action.in-a == 1 { print 'first action has been called'};
p6eval std 26999: OUTPUT«ok 00:02 36m␤»
Jaaaap std: grammar Grmmr { rule TOP { <a> <b> } token a { 'a' \w+ {*} } token b { 'b' \w+ {*} } } 06:20
std: sub c($c) { print $c, "\n"; }
std: class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b'; } }
std: my $action = Actions.new();
std: Grmmr.parse('alpha beta', :action($action));
std: if($action.in-a == 1){ print 'first action has been called'};
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Statements must be separated with semicolon at /tmp/ftQaxwhdxx line 1:␤------> grammar Grmmr { rule TOP { <a> <b> } token a { 'a' \w+ {*} } token b { 'b␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard
..stopper␤ statemen…
std 26999: OUTPUT«ok 00:04 39m␤»
std 26999: OUTPUT«##### PARSE FAILED #####␤Statements must be separated with semicolon at /tmp/4VkghgLZ7A line 1:␤------> hod a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b';␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard
..stopper␤ statem…
std 26999: OUTPUT«Undeclared name:␤ Actions used at 1 ␤ok 00:02 36m␤»
std 26999: OUTPUT«Potential difficulties:␤ Variable $action is not predeclared at /tmp/zOGSvoWZHO line 1:␤------> rmmr.parse('alpha beta', :action($action));␤Undeclared name:␤ Grmmr used at 1 ␤ok 00:02 37m␤» 06:21
std 26999: OUTPUT«Potential difficulties:␤ Variable $action is not predeclared at /tmp/6Ev8Uy6DWW line 1:␤------> if($action.in-a == 1){ print 'first action has bee␤Undeclared routine:␤ if used at 1 ␤ok 00:02 36m␤»
Jaaaap hmm guess i cant just remove the newlines ;-)
std: grammar Grmmr { rule TOP { <a> <b> } token a { 'a' \w+ {*} } token b { 'b' \w+ {*} } } 06:22
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Statements must be separated with semicolon at /tmp/fBetpKfNLt line 1:␤------> grammar Grmmr { rule TOP { <a> <b> } token a { 'a' \w+ {*} } token b { 'b' ␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard
..stopper␤ statemen…
TimToady not if you've assumed ; after }\n
06:22 iblechbot joined
Jaaaap std: grammar Grmmr { rule TOP { <a> <b> }; token a { 'a' \w+ {*} }; token b { 'b' \w+ {*} }; } 06:22
p6eval std 26999: OUTPUT«ok 00:03 36m␤»
Jaaaap class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b'; } }
std: class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b'; } } 06:23
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Statements must be separated with semicolon at /tmp/TiUr2GNQDJ line 1:␤------> hod a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b';␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard
..stopper␤ statem…
Jaaaap std: class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; }; method b($x) { $!in-b++; $!calls ~= 'b'; }; }
p6eval std 26999: OUTPUT«ok 00:03 38m␤»
Jaaaap td: grammar Grmmr { rule TOP { <a> <b> }; token a { 'a' \w+ {*} }; token b { 'b' \w+ {*} }; }; sub c($c) { print $c, "\n"; }; class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b'; } }; my $action = Actions.new(); Grmmr.parse('alpha beta', :action($action)); if($action.in-a == 1){ print 'first action has been called'}; 06:24
std: grammar Grmmr { rule TOP { <a> <b> }; token a { 'a' \w+ {*} }; token b { 'b' \w+ {*} }; }; sub c($c) { print $c, "\n"; }; class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b'; } }; my $action = Actions.new(); Grmmr.parse('alpha beta', :action($action)); if($action.in-a == 1){ print 'first action has been called'};
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Statements must be separated with semicolon at /tmp/TYW4A2p79h line 1:␤------> hod a($/) { $!in-a++; $!calls ~= 'a'; } method b($x) { $!in-b++; $!calls ~= 'b';␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard
..stopper␤ statem…
Jaaaap std: grammar Grmmr { rule TOP { <a> <b> }; token a { 'a' \w+ {*} }; token b { 'b' \w+ {*} }; }; sub c($c) { print $c, "\n"; }; class Actions { has $.in-a = 0; has $.in-b = 0; has $.calls = ''; method a($/) { $!in-a++; $!calls ~= 'a'; }; method b($x) { $!in-b++; $!calls ~= 'b'; }; }; my $action = Actions.new(); Grmmr.parse('alpha beta', :action($action)); if($action.in-a == 1){ print 'first action has been called'}; 06:25
p6eval std 26999: ( no output )
Jaaaap hmmm that's strange
TimToady timed out maybe
if() is still wrong
Jaaaap TimToady: what do you mean if() is still wrong? 06:26
TimToady any identifier followed by ( is a function call, not a keyword
std: if()
Jaaaap ah
p6eval std 26999: OUTPUT«Undeclared routine:␤ if used at 1 ␤ok 00:02 35m␤»
TimToady std: if() {...} 06:27
p6eval std 26999: 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/DRuBmob3xe line 1:␤------> if() {...}␤ expecting any of:␤ infix or meta-infix␤
..infix stopp…
06:27 davidad joined
Jaaaap thanks a lot. i didn't understand the "replace parens by spaces" but now i do ;) 06:28
06:41 snarkyboojum joined 06:46 DemoFreak joined 06:50 iblechbot left 07:04 dakkar joined 07:08 Jaaaap left 07:10 sparc joined 07:13 hanekomu joined
Matt-W Good morning 07:26
07:27 kate21de left 07:46 Muixirt joined 08:04 pochi left 08:09 beggars left 08:13 iblechbot joined 08:14 finanalyst left 08:15 unitxt left
cotto jnthn, something ate part of the uc example in your latest journal entry 08:19
Muixirt my @words = ("foo" "bar" "baz"); 08:23
cotto I just though he'd like to know. It's not much of stretch to fill in the blanks. 08:24
Muixirt :-)
cotto rakudo seems to complain about the comma-free version 08:27
08:27 amoc__ joined
cotto having .perl built in is *really* nice 08:28
Muixirt that was my first approach, don't rush me
cotto, why 08:29
?
08:31 amoc__ is now known as amoc_
cotto It's nice how easy .perl makes it to inspect a data structure. 08:31
It's very useful when learning a language, especially one as complex as perl6 08:32
08:32 DanielC joined, payload joined
Muixirt does the signatures of the candidates work? 08:36
08:37 finanalyst joined
Muixirt missing semicolons? 08:37
08:41 amoc left 08:43 amoc_ left, amoc joined 08:46 ZuLuuuuuu joined
Matt-W my @words = <foo bar baz? 08:49
my @words = <foo bar baz>
08:53 bacek left
cotto That's it. Thanks. 08:58
09:03 amoc__ joined 09:05 araujo left 09:06 amoc__ left 09:07 DemoFreak left 09:08 araujo joined 09:12 Jaaaa joined, Jaaaa left 09:13 Jaaaap joined
Jaaaap How do i work with newlines in a Grammar? If i add a \n to a working grammar (and the source), it stops working. 09:13
rule TOP { <a> \n <b> }; 09:14
never reaches b
jnthn Jaaaap: If you're writing a rule, there already is implicit whitespace matching anyway. 09:17
Including newline characters.
Jaaaap ah
jnthn Maybe you want to use regex instead of rule.
Which doesn't do that.
Jaaaap good idea. i'll try.
jnthn Alternatively, token gives the non-backtracking of rule but with the whitespace magic. 09:18
Jaaaap so token ha sautomatic whitespace stuff too?
jnthn cotto: Thanks for noticing - HTML makes a lousy blogging language... Fixed. :-) 09:20
09:20 luqui joined
luqui @seen diakopter 09:20
lambdabot diakopter is in #perl6. I don't know when diakopter last spoke. 09:21
jnthn Jaaaap: No, just rule has it.
Jaaaap: token and rule both disable backtracking.
regex is pretty much the backtracking semantics what you'd expect with regexes normally.
Jaaaap cool thx. 09:22
Muixirt jnthn, please explain what this means: proto 'term:' is precedence('z=') 09:27
it is at the end of src/parser/grammar.pg
is that valid perl6 code 09:31
09:34 amoc left
jnthn letenku mam :-D 09:34
oops, mischannel 09:35
Muixirt: Yes.
Muixirt: It's just defining term precedence level, with the operator precedence parser needs to know about. 09:36
Muixirt rakudo: proto test { } 09:38
p6eval rakudo 77db80: ( no output )
Muixirt rakudo: proto 'test' { }
p6eval rakudo 77db80: OUTPUT«Malformed routine definition at line 2, near "'test' { }"␤in Main (src/gen_setting.pm:0)␤»
jnthn Muixirt: Oh, hmm, the quotes are interesting. :-) 09:39
Muixirt: I don't know that this line will survive the protoregexes/LTM refactor though. 09:40
We'll end up with whatever STD.pm does for this.
09:41 jferrero joined
DanielC Btw, does Rakudo already support double precision numbers? I'm going to port another Debian shootout benchmark. This one is about double precision arithmetic. 09:42
Muixirt jnthn, i was curious and naive, and tried ./perl -c src/parser/grammar.pg 09:44
well ./perl6 of course, not perl :-) 09:45
btw. rakudo complains about the non matching =begin / =end pair 09:46
Matt-W jnthn: Good blog 09:51
DanielC rakudo: user constant PI => 3 09:52
p6eval rakudo 77db80: OUTPUT«No applicable methods.␤in Main (/tmp/37ajfGXymy:2)␤»
DanielC Help? :-(
09:52 masak joined
DanielC o/ masak 09:52
masak DanielC: \o 09:53
DanielC Does anyone know if "use constant" is implemented in Rakudo?
rakudo: user constant PI => 3
p6eval rakudo 77db80: OUTPUT«No applicable methods.␤in Main (/tmp/BAMlfuGuTH:2)␤»
Matt-W it'd help if you typed 'use' correctly
DanielC oh!
rakudo: use constant PI => 3
jnthn Rakudo: constant PIE = 3;
p6eval rakudo 77db80: OUTPUT«Can't find ./constant in @*INC␤in Main (src/gen_setting.pm:439)␤»
DanielC Matt-W: It might...
Matt-W but I think you might find Perl 6 has a different mechanism for constants
DanielC :-)
jnthn rakudo: constant PI = 3;
p6eval rakudo 77db80: ( no output )
jnthn rakudo: constant PI = 3; say PI;
p6eval rakudo 77db80: OUTPUT«3␤»
DanielC jnthn: thanks! 09:54
masak rakudo: constant PI = 3; PI = 4; say PI
p6eval rakudo 77db80: OUTPUT«Cannot assign to readonly variable.␤in Main (/tmp/cOkThmN31j:2)␤»
masak heh. :)
jnthn rakudo: constant π = 3; say π;
p6eval rakudo 77db80: OUTPUT«3␤»
Matt-W DanielC: first assumption is not 'does rakudo do 'use constant'', but 'how do I do constants in perl 6'. Saves time, I find.
jnthn Matt-W: nice try
erm, masak
Matt-W jnthn: tab miss?
hah
jnthn Matt-W: Yes. :-/ 09:55
Matt-W :)
Matt-W -> table tennis
masak jnthn: it used to work. :)
jnthn masak: Occasionally, we fix bugs. ;-)
masak jnthn: that, if I don't say it often enough, rocks. 09:56
09:56 alanhaggai joined
DanielC What does 'sub foo($) {...' mean in Perl 5? I'm converting a program from p5 to p6 and I had never seen that before. 09:59
jnthn It's some kinda prototype thingy but I never understood those... 10:00
masak DanielC: it takes a single argument, which must be 5.
oh, it's a $. sorry. :) 10:01
jnthn lol!
masak DanielC: see perldoc.perl.org/perlsub.html#Prototypes
jnthn ;-)
DanielC :-)
*click*
Ok, so foo($) means "takes a single argument". 10:03
dakkar DanielC: sub foo { $_[0] }; push @a, foo 1, 2, 3 => @a=(1);
DanielC: sub foo($) { $_[0] }; push @a, foo 1, 2, 3 => @a=(1,2,3);
masak I've heard various reliable sources say that prototypes do more harm than good.
dakkar prototypes change the way the call is parsed, so beware 10:04
DanielC Thanks.
dakkar masak: I'd put them in the same league as "local": use them when necessary, and only if you know what you are doing
DanielC Does Perl 6 have prototypes? Not that I plan to use them.
masak dakkar: right.
DanielC: Perl 6 has full signatures! 10:05
dakkar DanielC: Perl 6 has full-fledged signatures, and you should use them :)
DanielC That's why I was wondering.
masak DanielC: see S06.
dakkar sub foo(Num $a) { ... }
DanielC I was thinking that Perl 6 has no need for them...
10:05 skids joined
DanielC rakudo: my @a = (); sub foo($x) { $x }; push @a , foo 1, 2, 3; say @a.perl 10:06
p6eval rakudo 77db80: OUTPUT«too many arguments passed (3) - 1 params expected␤in sub foo (/tmp/jpOzhfgUqe:1)␤called from Main (/tmp/jpOzhfgUqe:2)␤»
masak DanielC: you need parens to disambiguate that. 10:07
DanielC You see why I was asking?
Yes
dakkar oh, p6 always parses sub call as listop?
DanielC What I mean is... this is why I asked if Perl 6 had prototypes. dakkar showed how prototypes change the way an expression is parsed. 10:08
dakkar and I didn't know signatures didn't, in p6
masak lunch &
dakkar but it makes much sense: it makes the parser simpler (and in the face of macros, this is very much important), and it makes reading the code easier 10:09
and you can add optional parameters to your signatures without messing up all your calls
hmmm. and multis wouldn't work at all
Matt-W exactly 10:13
DanielC rakudo: constant ARR => (1,2,3,4) 10:14
p6eval rakudo 77db80: OUTPUT«No applicable methods.␤in Main (/tmp/Lk3gkKKy1G:2)␤»
dakkar rakudo: constant @ARR => (1,2,3,4) 10:15
p6eval rakudo 77db80: OUTPUT«No applicable methods.␤in Main (/tmp/YcutHWi9aM:2)␤»
jnthn =
dakkar rakudo: constant @ARR = (1,2,3,4)
p6eval rakudo 77db80: ( no output )
dakkar I got confused with Readonly.pm :/
DanielC thanks
dakkar rakudo: constant @ARR = (1,2,3,4);@ARR[1]=4
p6eval rakudo 77db80: ( no output )
dakkar uh?
rakudo: constant @ARR = (1,2,3,4);@ARR[1]=4;say @ARR
p6eval rakudo 77db80: OUTPUT«1434␤» 10:16
DanielC Well.. it's read-only, but Rakudo should complain.
dakkar rakudo: constant $ARR = (1,2,3,4);$ARR[1]=4;say $ARR
DanielC oh, wait!
p6eval rakudo 77db80: OUTPUT«1 4 3 4␤»
DanielC my bad.
10:16 clintongormley joined 10:18 Jaaaap left
Matt-W umm 10:18
that seems... buggy
jnthn aye 10:25
10:26 synth left
DanielC Does Perl 6 have a Double type or do I have to use Num ? 10:29
I can't find a Double in S02... 10:30
10:38 zamolxes joined
jnthn DanielC: Use Num 10:39
DanielC k 10:40
jnthn I think it's double prec under the hood.
DanielC Ok, good.
10:46 agentzh left 10:47 agentzh joined 10:51 cj left
clintongormley DanielC - just checked out your blog - nicely written! 10:53
you should get involved with the p6 documentation effort 10:54
DanielC clintongormley: I would consider it, but I don't know Perl 6 very well yet.
afk & dinner
afk & lunch :-) 10:55
masak clintongormley: url?
DanielC: you have lunch after dinner? :P
clintongormley daniel.carrera.bz/ 10:56
masak clintongormley: thanks.
masak submits rakudobug
10:58 ZuLuuuuuu left
literal hm, where's DanielC's blog? 11:11
oh, I'm blind
DanielC back 11:14
clintongormley: Let me experiment more with Perl 6 first. Are you the documentation guy?
11:14 hercynium joined
clintongormley DanielC: No, I'm a watcher :) 11:15
DanielC :-)
clintongormley i think masak is the one you want to talk to
11:15 tulcod joined
DanielC I don't know if masak is writing much documentation, but he's working on a program called "grok" which will be kind of like perldoc but better. 11:16
masak I'm not, literal is. 11:18
DanielC Btw, I just updated my local copy of the Rakudo repository. Do I have to run "Configure.pl" again, or do I just do "make"?
masak and yes, I actually intend to write quite some documentation.
jnthn
.oO( who will literal in turn say it is? )
masak DanielC: if you need to ask... :) 11:19
DanielC: ...run Configure.pl.
DanielC just figured what jnthn's .oO means
skids \o/
DanielC masak: ok
DanielC runs perl Configure.pl --gen-parro 11:20
masak rakudo: class Object is also { method oO($s) { say $s } }; .oO( 'OH HAI' )
p6eval rakudo 77db80: OUTPUT«OH HAI␤»
skids masak++ That should be be used in the cookbooks :-) 11:21
jnthn "we'll deal with the future later"++ 11:22
masak jnthn: :P
hanging on #perl6 is really good for me sense of self-worth. :P 11:23
jnthn masak: Mine too!
masak :)
jnthn mmm...lunch. 11:24
skids masak: you mean .worth, no?
jnthn BTW, I now have a flight to YAPC::EU.
Don't have one to get home yet, but... :-)
jnthn almost looks forward to it...but he's got vacation to look forward to first :-) 11:25
patmat maybe has his first job in the programming world... but with PHP :/
masak skids: :)
patmat: better than nothing... I think. :/ 11:26
patmat yup and better than support...
masak :) 11:29
skids patmat: $support_job of Programming_job where { .release > 1.0 }; 11:31
patmat :d 11:32
skids Hrm method improvement () { .esteem++; .worth++; .motivation++; } 11:35
skids wonders how bad things explode if you define "class self" 11:36
DanielC @karma 11:40
lambdabot You have a karma of 7
DanielC pushes another benchmark to perl6-examples
@karma
lambdabot You have a karma of 7
DanielC hmm...
Committing to perl6-examples doesn't get you karma points :-( 11:41
skids There'd need to be a dalek message, I think.
DanielC Yeah... someone said that they were going to change it, but it doesn't seem to have happened.
skids @karma esteem 11:42
lambdabot esteem has a karma of 0
skids @karma .esteem
lambdabot .esteem has a karma of 0
skids I guess it's smartened. 11:43
rakudo: DanielC++ 11:44
p6eval rakudo 77db80: OUTPUT«Could not find non-existent sub DanielC␤»
skids @karma DanielC
lambdabot DanielC has a karma of 8
DanielC That's interesting.
skids Not entirely smartened. 11:45
DanielC rakudo: DanielC++
p6eval rakudo 77db80: OUTPUT«Could not find non-existent sub DanielC␤»
DanielC @karma
lambdabot You have a karma of 8
DanielC oh well...
skids That's probably the first if statement in the block :-) 11:46
DanielC heh
11:46 luqui left 11:52 mizioumt joined
DanielC Ok, on an arithmetic benchmark Rakudo is 720 times slower than Perl 5. 11:54
pmichaud only 720?
(good morning, #perl6)
masak that's great!
DanielC The benchmark is mostly double precision arithmetic.
masak as the saying goes, 720 times slower ought to be enough for anyone. 11:55
DanielC pmichaud I figure that your ++ improvements helped.
masak: :-)
If you are interested, this is the n-body benchmark in perl6-examples. 11:56
It is mostly double precision arithmetic, plus the evil ++
pmichaud (reading backscroll)
(also preparing @kid breakfasts) 11:57
viklund @karma masak
lambdabot masak has a karma of 212
viklund rakudo: say " masak++ "
p6eval rakudo 77db80: OUTPUT« masak++ ␤»
jnthn DanielC: Did you do an optimized build of Parrot also?
viklund @karma masak 11:58
lambdabot masak has a karma of 214
viklund ha
skids pmichaud: not %kid? You really should give them names you know, at least, by age 2 or so.
pmichaud They have names!
jnthn skids: It's an array of Kid objects. They know their names. ;-)
pmichaud @kid[0] and @kid[1]
lambdabot Unknown command, try @list
jnthn say @kid>>.name # see ;-)
masak everything is so much easier with Perl 6! :)
skids would not want to grow up as kid #0. 11:59
DanielC jnthn: I don't know Parrot.
jnthn DanielC: OK, in that case the answer is no. :-)
DanielC jnthn: oops, I mis-read your question.
jnthn: No, I didn't. How do you do that?
Can you just do something like 'make O3' ? 12:00
jnthn If you are building with --gen-parrot, then also give --gen-parrot-option=--optimize
DanielC Cool. 12:01
Is there any reason why I should /not/ optimize?
jnthn Parrot sometimes has some bugs exposed as a result of optimization.
Things that didn't get tracked down yet.
viklund masak: smoketesting all proto projects were trickier than I thought,
but here it is: 12:02
feather.perl6.nl/~viklund/
DanielC jntthn: Ok... So, should I stick with the non-optimized version?
jnthn DanielC: For normal use probably it's safest. I think how well it works out varies by platform. 12:03
DanielC OK
masak viklund: cool! 12:04
viklund: really, really cool!
viklund++
pmichaud I'd be fine with a --optimize option to Configure.pl (that is then passed along to parrot's build)
I'd even be fine if we did a little probing/detection to decide when it might be safe to automatically set --optimize by default
viklund masak: now I only need to cron it... 12:05
masak viklund: some of these projects target bleeding, some target release... 12:07
viklund masak: that may be 12:13
it's tested on bleeding...
Matt-W viklund: I've never got around to fixing that broken test in form yet :)
but I am aware of it 12:14
have to do that tomorrow
or maybe tonight
viklund nice ;)
now I really have to make a cron job, if it can motivate ppl ;)
Matt-W :) 12:15
12:15 ruoso joined
masak viklund: it can me, I'm sure. 12:15
Matt-W Well actually, I'm already motivated by having an appointment for a hackathon on Form with masak tomorrow
and it'd be nice to start on a tree that passes tests
masak yay, Friday evening hackathon appointment!
Matt-W before we add a load more
masak viklund: ONE failing test in November! 12:18
viklund: man, I hate that test! >:(
viklund: it could be argued that that test does nothing for us, since neither of us understand why it fails. and it sure annoys me. 12:19
Matt-W bah 12:20
just drink coffee until it makes sense
then declare it's irrelevant and delete it
viklund masak: yes, I know, need to look at that, but those test are important
oh well, time to do the dishes... 12:21
jnthn pmichaud: (optimize) For me I think the issue is that my compiler doesn't let me have an optimized build with debug symbols. 12:24
That may have changed though.
DanielC With --gen-parrot-option=--optimize Rakudo is only 550 times slower than Perl 5, instead of 720 times. 12:27
Of course, the benchmark has a significant margin of error.
jnthn That's a respectable improvement for setting a flag. :-) 12:28
DanielC yes 12:29
skids DanielC: you might get it down to 500 by setting MIN_BUCKETS to 4 in parrot/src/hash.c :-)
Matt-W DanielC: then you could go forward in time and bring back allison's calling conventions refactor so we don't have to wait 12:31
DanielC On the regex-dna benchmark (which tests regexes) Rakudo is 410 times slower than Perl 5 (using --optimize). 12:35
antiphase Time for premature optimization \o/ 12:36
DanielC waits for the computer to cool before running the last benchmark
12:37 beggars joined
jnthn -> slovak class, back later 12:40
DanielC WOW! On the last benchmark Rakudo was only 200 times slower than Perl 5.
This used to be the slowest benchmark. It used to be about 2,000 times slower. 12:41
So we have a factor of 10 improvement. Probably because of pmichaud's fixes to ++
pmichaud++
Matt-W no doubt there are some other big wins like that to be had 12:42
DanielC no doubt
12:45 ZuLuuuuuu joined
Muixirt pours water into the wine 12:48
speed of perl5 is really the goal? ;-) 12:49
masak no, but it's a reasonable comparison.
DanielC Muixirt: A line in the sand is a line in the sand. It's a good way to measure improvement. 12:50
Matt-W IT's something to compare to
I doubt we'll ever be that fast
But we'll be a lot faster than we are now! 12:51
DanielC Matt-W: That's depressing though. I would hope that Rakudo would be faster one day.
Why can't Rakudo ever be as fast as Perl 5?
Matt-W I don't know
maybe it can be
DanielC TimToady said yesterday that Perl 6 could one day be nearly as fast as C, if the programmer is willing to provide enough type info. 12:52
masak could be faster than C in some cases, if the optimizer is really clever. 12:53
Muixirt DanielC, i see that potential too, but ... optimizing compilers for languages as complex as perl6 are *really* hard to develop
Matt-W any kind of compiler's hard enough... 12:54
DanielC I know that Parrot is designed as a register-based machine because there is a lot of literature on optimization for register-based architectures.
Matt-W yup
DanielC Muixirt: Maybe, maybe not. You don't have to optimize everything. You have to optimize enough so that the programmer can refactor his code and make it fast. 12:55
12:56 zamolxes left 12:58 hanekomu left
DanielC Paul Graham opines that the most important thing for the speed of a language is not actually the compiler, but the profiling tools... because of the 80/20 rule. 12:58
pmichaud part of what makes optimization more difficult in Rakudo is the degree of dynamicism available 13:01
masak that also hinders static code analysis in many cases. 13:02
pmichaud there are a number of simplifying assumptions one could make that become no longer true when it's possible for a program to replace core functions / modify the grammar / override core functions / etc.
DanielC pmichaud: Maybe Rakudo shoudl have a flag where the programmer promises not to do any of the above. 13:03
That will be true 90% of the time.
pmichaud DanielC: Can't in the general case, because another programmer might undo it.
DanielC :-(
Sucks... how often do you override core functions? 13:04
masak it happens.
pmichaud anyway, we may be able to do some jit optimizations, and a variety of other things. 13:05
DanielC I think the flag would be a good idea. Let the programmer decide.
13:05 Whiteknight joined
pmichaud DanielC: but the problem is "which programmer"? 13:05
the one who wrote the CPAN module, or the one who is using it?
DanielC Can it be done per-module?
pmichaud yes, but it can always be undone.
it's kinda like the following from S12: 13:06
masak rakudo: say ?("say <OH HAI>" ~~ /<Perl6::Grammar::TOP>/); say $/.perl
p6eval rakudo 77db80: OUTPUT«1␤No result object␤in method Match::!_perl (src/gen_setting.pm:788)␤called from method Match::perl (src/gen_setting.pm:754)␤called from Main (/tmp/4P2ogeVaig:2)␤»
pmichaud Classes are open and non-final by default, but may easily be closed
or finalized not by themselves but by the entire application, provided
nobody issued an explicit compile-time request that the class stay open
masak pmichaud: I never reported this yesterday, but I think I perhaps should.
pmichaud: it is a bug in .perl after all, right?
pmichaud oh, yes.
masak submits rakudobug
pmichaud Although it could be a bug in Match somehow. 13:07
rakudo: say ?("say <OH HAI>" ~~ /<Perl6::Grammar::TOP>/); say ?$/
p6eval rakudo 77db80: OUTPUT«1␤1␤»
pmichaud rakudo: say ?("say <OH HAI>" ~~ /<Perl6::Grammar::TOP>/); say $/.ast
p6eval rakudo 77db80: OUTPUT«1␤say <OH HAI>␤»
pmichaud rakudo: say ?("say <OH HAI>" ~~ /<Perl6::Grammar::TOP>/); say $/.Str
p6eval rakudo 77db80: OUTPUT«1␤say <OH HAI>␤»
pmichaud yes, looks like a bug in .perl 13:08
DanielC I keep thinking that there must be a way that it's ok to say "yes, please optimize my code".
Some programs don't even use CPAN modules.
The benchmarks sure don't.
pmichaud oh, we'll probably have that. 13:09
I didn't say it wasn't possible, I just said it's hard.
DanielC I thought you said it was a bad idea :-)
13:09 pdcawley joined 13:14 beggars left 13:22 synth joined 13:25 hatsefla1s is now known as hatseflats 13:39 snarkyboojum left 13:41 jferrero left, FurnaceBoy joined
PerlJam good morning #perl6 13:42
DanielC good localtime()
13:42 decasm joined
masak gm, pj. 13:42
FurnaceBoy call DanielC for a good localtime()
DanielC :-) 13:43
Matt-W hi perljam
masak those were the localtime()s... 13:44
PerlJam It was the best of localtime()s, it was the worst of localtime()s 13:45
13:45 lichtkind joined
masak localtime() flies like an arrow... 13:46
somehow doesn't have the same ring to it...
PerlJam Is it just my perl6 build or does the interactive mode not stop when given a ^D anymore?
masak PerlJam: it was reported on RT a while ago. 13:47
by moritz_, I think. 13:48
pmichaud Good morning, PerlJam (and #perl6)
interactive mode doesn't stop? That's odd, I suspect a parrot change.
Matt-W masak: and weren't you upset that you didn't report it first :P 13:49
PerlJam I was guessing it's related to other IO weirdness (like .lines() returning an extra blank line at the end)
pmichaud that *really* should be fixable now. 13:50
masak Matt-W: I occasionally "let one slip" so that it won't look like a one-man show... :P
Matt-W masak: aah your secret is revealed!
masak dang.
DanielC rakudo: my $a = "hello"; $a = reverse $a; say $a; 13:51
p6eval rakudo 77db80: OUTPUT«hello␤»
DanielC :-( Shouldn't that be reversed?
PerlJam DanielC: perhaps you want .flip ?
pmichaud rakudo: my $a = 'hello'; $a = flip $a; say $a;
p6eval rakudo 77db80: OUTPUT«olleh␤»
pmichaud reverse is for lists.
DanielC ah
pmichaud rakudo: my @a = <hello world>; say reverse @a
DanielC thanks PerlJam & pmichaud
p6eval rakudo 77db80: OUTPUT«worldhello␤»
masak no, seriously. if the competition is about who first submits a bug, I think that can be very good for the RT converage. but I don't submit many bugs because of the statistics, I do it because I care for Rakudo.
pmichaud rakudo: my @a = <hello world>; say flip @a 13:52
p6eval rakudo 77db80: OUTPUT«dlrow olleh␤»
DanielC pmichaud: That's interesting.
PerlJam DanielC: you didn't expect flip to be distributive?
Matt-W rakudo: my @a = <hello world>; say @a>>.flip.reverse
p6eval rakudo 77db80: OUTPUT«dlrowolleh␤» 13:53
Matt-W rakudo: my @a = <hello world>; say @a>>.flip.perl
pmichaud rakudo: my @a = <hello world>; say @a.flip;
pdcawley <hello world>? Damn, but I'm behind the specs by _ages_
p6eval rakudo 77db80: OUTPUT«["olleh", "dlrow"]␤»
rakudo 77db80: OUTPUT«dlrow olleh␤»
Matt-W ah it did work, good
pdcawley: it's like qw//
pmichaud pdcawley: <hello world> is qw/hello world/;
pdcawley Nice
DanielC rakudo: my @a = (<one two>, <three four>); say flip @a;
p6eval rakudo 77db80: OUTPUT«ruof eerht owt eno␤» 13:54
DanielC kewl
PerlJam actually, that is interesting. I didn't notice before that it also reversed the order of the list 13:55
Matt-W pdcawley: do you know about %hash<key> as well?
masak PerlJam: it's stringified when sent to flip.
DanielC rakudo: my @a = (<one two>, <three four>); say (flip @a).perl;
p6eval rakudo 77db80: OUTPUT«"ruof eerht owt eno"␤»
PerlJam masak: oh, of course 13:56
pdcawley Matt-W Not sure.
How's that different from %hash{key} ?
PerlJam pdcawley: key isn't a bare word that just happens to be interpreted as a string. :) 13:57
Matt-W no barewords in Perl 6
masak pdcawley: only the <> form autoquotes.
Matt-W so %hash<key> is like %hash{'key'}
masak aye.
pdcawley and %hash{{:key<val>}} uses the anon hash as a key of the hash? 13:58
Or am I getting bits of my syntax muddled.
ZuLuuuuuu will we use also <elemet_1 element_2> instead of qw(element_1 element_2)
Matt-W ZuLuuuuuu: I certainly will! 13:59
masak ZuLuuuuuu: yes, why not?
Matt-W note that there's also the <<>> form, which is to <> as "" is to ''
pdcawley Schweet.
ZuLuuuuuu heh ok then, it looks better in my opinion, as well
Matt-W and its unicode equivalent, which I haven't learned how to type on a windows box yet
masak Matt-W: <<>>:<>::"":'' :)
pdcawley And <<>> can also use «» ? 14:00
Matt-W masak: I generally prefer to be more verbose than that. You might have noticed.
pdcawley: yes
pdcawley hasn't learned to type it yet, just copy/paste
masak Matt-W: yes, I just thought it was nice. :)
DanielC rakudo: my $s = "hi"; $s ~~ tr/hi/HI/;
p6eval rakudo 77db80: OUTPUT«Statement not terminated properly at line 2, near "/;"␤in Main (src/gen_setting.pm:0)␤»
masak Matt-W: not actually correcting you in any way.
Matt-W I'll worry about it when i start writing perl 6 on a windows system
DanielC Is 'tr' implemented in Rakudo?
pdcawley First time I came acros : :: : notation, I had the devil's own time working out what was meant by it.
Was in an online iq test and nowhere did the silly sods explain it. 14:01
Matt-W I don't think we have the syntax for tr/// like we don't have the syntax for s///
DanielC Matt-W: For s/// there is the alternate syntax .subst(...) Do you know if there is an equivalent for tr?
masak DanielC: .trans 14:02
Matt-W masak saves the day
masak DanielC: and yes, that one's implemented.
DanielC: see S05.
DanielC rakudo: my $s = "hi"; $s.trans(/hi/, 'HI');
p6eval rakudo 77db80: OUTPUT«Must pass a List of Pairs for transliteration␤in Main (/tmp/haaTVeYjxE:2)␤»
masak DanielC: no really. see S05.
don't guess. :)
DanielC goes to S05
14:02 dukeleto_ joined
finanalyst rakudo: my $x = 'hi'; my %z=<hithere a b c>Z 1,2,3,4; %z<<{$x}there>> 14:03
p6eval rakudo 77db80: ( no output )
masak it's under "Transliteration", I think.
DanielC I've been reading S05 for days and there's still so much I don't know...
pdcawley Can we write our own special forms in Rakudo yet?
masak pdcawley: special forms?
finanalyst rakudo: my $x = 'hi'; my %z=<hithere a b c>Z 1,2,3,4; %z<<{$x}there>>.say
pdcawley is thinking of porting his Test::Class::Sugar port to Rakudo for shits and giggles.
p6eval rakudo 77db80: OUTPUT«1␤»
masak pdcawley: um, probably not.
pdcawley Roll your own keywords.
macros.
masak pdcawley: nope, not yet.
pdcawley Dang.
Matt-W DanielC: It's okay, I think we told you we don't expect you to memorise it all this week 14:04
pdcawley: we do have custom operators though
DanielC rakudo: my $s = "hello"; $s.trans( 'a..z' => 'A..Z').say; 14:05
p6eval rakudo 77db80: OUTPUT«HELLO␤» 14:06
DanielC :-D
pdcawley Custom operators? Grab a unicode character and define an op?
DanielC rakudo: my $s = "hello"; $s.trans('aeiou' => 'AEIOU').say; 14:07
p6eval rakudo 77db80: OUTPUT«hEllO␤»
DanielC :-)
masak pdcawley: yes, you can have that. 14:08
pdcawley: with some not-too-large caveats.
pdcawley I always think it's a shame you can't do $s.trans('ae..u' => 'AE..U'), but Perl isn't turing complete yet.
masak pdcawley: you can't? 14:09
pdcawley: it isn't?
pdcawley Infix ops, prefix ops... can we do outfix ops? Grab a pair of brackets from unicode and make them into a custom object constructor?
In the sense of passing the turing test? No. Meant AI complete.
masak pdcawley: you're probably looking for 'postcircumfix ops'.
pdcawley: that's not the ordinary definition of "Turing complete". 14:10
pdcawley masak: Probably. Guessing names is always inferior to knowing them.
masak: I know. I'm an idiot.
and was proposing ae..u as a cleverdick 'shortcut' for aeiou.
masak pdcawley: ah. that wasn't too clear. :) 14:11
14:11 meppl joined
masak pdcawley: I don't know if I'd consider it a feature. :) 14:11
pdcawley grins. "It was in _my_ head"
14:11 dukeleto left
ruoso just sent a confusing message to p6-language 14:11
pmichaud note that qw(...) isn't the same as the qw quoting operator, though. 14:15
14:15 netsquire joined
pmichaud qw{foo bar} # okay 14:16
qw/foo bar/ # okay
qw(foo bar) # function call, not a quote.
ruoso pmichaud, btw... it would be nice if you could take a look in the message I just posted to p6-language... 14:17
14:17 iblechbot left
Matt-W ruoso: man, you really know how to make someone's head hurt 14:18
ruoso thanks?
14:19 hudnix joined
pmichaud ruoso: interesting post. I don't have any immediate comments on it, other than I'm curious to see how it plays out. 14:19
There's a part of me that still feels that 'for' and 'map' aren't completely isomorphic.
in particular, 'for' looks like an imperative statement to me, and I'd expect it to complete before moving on to the next statement. 14:20
i.e., I'd expect it to be a bit eager.
ruoso you know... I'm kinda getting used to this non-functional, non-imperative, non-declarative programming...
14:20 masak left
ruoso but it will certainly bite a lot of people 14:20
pmichaud but 'map' feels like it could be lazy.
ruoso I think we're in new territory here on the way you program "as if" you were programming in an imperative programming, but something completely different happens 14:21
I mean... erlang and io are languages that embrace this degree of lazyness.. but programming for them require you to have a different mindset 14:22
and maybe it is a good thing not having any imperative barrier -- I mean... maybe that's where Perl 6 is really going to make a difference... 14:25
jnthn lol I'm back 14:26
ZuLuuuuuu is the meaning of $/ variable changed in perl 6?
jnthn $/ in Perl 6 is match object.
$/ in Perl 5 was almost certainly not that since there was no such thing as a match object. 14:27
ZuLuuuuuu hmmm in which synopsis are these special variables explained?
jnthn It probably mean "moon phase" or something. :-)
ZuLuuuuuu yeah in perl 5 I guess it is the input line seperator
jnthn $/ will be in S05 surely.
ZuLuuuuuu ok thanks
jnthn Well, there is few specials now
$/, $_, $! - that's it.
literal there are some in S28 "Special Variables"
ruoso the term now is "pre-defined", not "special"
;)
14:28 mizioumt1 joined
ruoso it's not really magical anymore 14:28
14:28 mizioumt1 left
jnthn special and magical are different. ;-) 14:28
14:28 mizioumt1 joined
jnthn may well be special, but he sure ain't magical. 14:28
ZuLuuuuuu hmmm interesting
jnthn ;-)
ZuLuuuuuu only three, that's good :) 14:29
pmichaud @_ and %_ also. 14:34
lambdabot Maybe you meant: . ? @ v
pmichaud and maybe I meant "@go_jump_in_a_lake", lambdabot.
[particle]- burn. 14:35
14:36 viklund left, exodist joined 14:37 exodist left
rjh $/ is autoflush iirc :) 14:37
no that's $|
$/ is input record separator
14:37 exodist joined
[particle]- input recor... 14:37
rjh stupid perl 5 :/
[particle]- i use $/ all the time
but people yell at me to use "\n" instead
rjh do { $/ = undef; <> } :)
*local 14:38
TimToady std: do { $/ = undef; <> }
[particle]- print $foo, $/;
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Obsolete use of $/ variable as input record separator; in Perl 6 please use filehandle's :irs attribute instead at /tmp/zLBgluNXYU line 1:␤------> do { $/ = undef; <> }␤FAILED 00:02 36m␤»
rjh Is it ever legal to set $/ ?
[particle]- never liked do {...}, and used sub{ ...}->() instead
sure it's legal
rjh in perl6
DanielC rakudo: say "hello".flip.trans("aeiou" => "AEIOU")
p6eval rakudo 77db80: OUTPUT«too few arguments passed (2) - 3 params expected␤in Main (/tmp/QowrpTRHmg:2)␤»
DanielC ???
[particle]- ah, we're talking about perl 6 again :) 14:39
DanielC flip and trans don't get along.
[particle]- rakudo: $/ = 'foo'; $/.print
p6eval rakudo 77db80: OUTPUT«foo»
[particle]- std: $/ = 'foo'; $/.print
p6eval std 26999: OUTPUT«##### PARSE FAILED #####␤Obsolete use of $/ variable as input record separator; in Perl 6 please use filehandle's :irs attribute instead at /tmp/8Qr8lvp5u5 line 1:␤------> $/ = 'foo'; $/.print␤FAILED 00:02 36m␤»
DanielC rakudo: my $s="hello".flip; say $s.trans("aeiou" => "AEIOU")
TimToady rakudo: say "hello".flip.trans(("aeiou" => "AEIOU"))
p6eval rakudo 77db80: OUTPUT«OllEh␤»
rakudo 77db80: OUTPUT«too few arguments passed (2) - 3 params expected␤in Main (/tmp/vy6SakrVwW:2)␤»
rjh looks like a bug 14:40
[particle]- is $/ not an lvalue?
TimToady looks to me like "aeiou" => "AEIOU" is getting interpreted as a named argumetn
DanielC TimToady: Strangely, that only happens when you include the .flip
TimToady rakudo: say "hello".flip.WHAT 14:41
p6eval rakudo 77db80: OUTPUT«Str()␤»
TimToady rakudo: say "hello".trans("aeiou" => "AEIOU") 14:42
p6eval rakudo 77db80: OUTPUT«hEllO␤»
ruoso TimToady, I just sent a second message about implicit threading/event... if you happen to have some time, I'd appreciate your feedback
TimToady statement semi is the main barrier we have, because it creates void contexts, which are eager to lists 14:43
rakudo: say "hello".flip().trans("aeiou" => "AEIOU")
p6eval rakudo 77db80: OUTPUT«too few arguments passed (2) - 3 params expected␤in Main (/tmp/thZLp9GiYA:2)␤»
ruoso so a bare: "foo()" is the only way to make a trully-imperative call... 14:44
TimToady rakudo: say "hello".flip().trans(~"aeiou" => "AEIOU")
p6eval rakudo 77db80: OUTPUT«too few arguments passed (2) - 3 params expected␤in Main (/tmp/aqvohyj7kQ:2)␤»
pmichaud I suspect .trans is wrong/incomplete.
TimToady rakudo: say "hello".flip().trans(("aeiou" => "AEIOU"))
p6eval rakudo 77db80: OUTPUT«too few arguments passed (2) - 3 params expected␤in Main (/tmp/8lGuqYIGAy:2)␤»
pmichaud oh, in fact, I'm sure of it.
Parrot's ".trans" is interfering with Rakudo's .trans.
TimToady in that case, I will drink my coffee and backlog 14:45
DanielC found another bug o/
Soon I'll catch up with masak :)
Matt-W not a chance 14:48
unless we make him type his reports with his nose
14:48 alanhaggai left 14:49 alanhaggai joined, mizioumt left
DanielC Heh... on this benchmark Rakudo is 1600 times slower than Perl 5. But only because Perl does the benchmark in 0.09 seconds. 14:55
Rakudo does the benchmark fast, but when you divide by almost-zero you get a big number.
Matt-W DanielC: how do the factors change if you precompile your benchmark scripts to PIR and then time the runs from PIR? 14:57
DanielC How do you do that? 14:58
dalek kudo: b88c33d | tene++ | src/parser/actions.pm:
Register foreign types at compile time.
 14:59
kudo: 2653235 | tene++ | perl6.pir:
Fix an issue with foreign languages loading Perl6 libraries without exported symbols... probably the wrong fix
jnthn DanielC: What is the benchmark Rakudo is 1600 times slower testing? 15:01
15:01 nihiliad joined
DanielC jnthn: reads a large file and does .flip and .trans on every line. 15:01
StephenPollei ae..u as a 'shortcut' for aeiou reminds me of en.wikipedia.org/wiki/Fluid_Concept..._Analogies a book by Douglas Hofstadter .. I think that way too much dwim ;-) 15:02
DanielC jnthn: But Rakudo is fast enough, that the problem could just well be the overhead of ++ or the while loop and the like.
The program runs in 14.366s in Rakudo and 0.009s in Perl 5. 15:03
[particle]- danielc: i suggest you build an optimized parrot and re-run 15:04
DanielC [particle]- I am using an optimized Parrot. 15:05
15:06 viklund joined
DanielC After I spoke with jnthn I figured that the optimized Parrot would give more meaningful benchmark figures. 15:06
[particle]- good
15:08 pyrimidine joined
pmichaud I have a fix for #66280, spectesting now. 15:10
pyrimidine cool
I thought it might be something simple, just didn't get a chance to look into it 15:11
(re: #66280)
DanielC: re: .trans, it's implemented in PIR but I think it's suboptimally implemented 15:21
(I can say that b/c I implemented it)
15:21 hatseflats left
pyrimidine ;) 15:21
15:23 pmichaud_ joined 15:24 hatseflats joined, unitxt joined, dalek left 15:25 dalek joined, pmichaud_ left
dalek kudo: 5745c6b | pmichaud++ | src/ (7 files):
Convert 'deref_objectref' dynop to 'deobjectref' and 'descalarref'.
 15:29
kudo: 9e2b9ad | pmichaud++ | src/builtins/assign.pir:
Better handling of assignment of ObjectRef values. Fixes RT #66280.
kudo: 10a9b23 | pmichaud++ | :
Merge branch 'master' of [email@hidden.address]
15:31 justatheory joined
pyrimidine \o/ 15:32
pmichaud rakudo: for 1..3 -> $i { for $i..5 -> $j { }; say $i; } 15:34
p6eval rakudo 77db80: OUTPUT«6␤6␤6␤» 15:35
jnthn The bug of the beast
15:35 pmurias joined, ruoso left
pmichaud pmichaud@orange:~/rakudo$ ./perl6 15:35
> for 1..3 -> $i { for $i..5 -> $j {}; say $i; }
1
2
3
lambdabot <no location info>: parse error on input `..'
pmichaud better.
$lambdabot.set-location('lake') 15:36
$Pm.set-location('lunch') 15:39
or, more properly
$Pm.location = 'lunch'
TimToady the problem with writing our own bot in p6 is it will probably become sentient and secretly take over the world 15:40
pmichaud I suspect it will just take over the TimToady nick.
And then take over the world.
TimToady it certainly does a good job with speech synth on Wednesdays already...
pmichaud and like any good sentient being, it can't yet completely describe itself. :-) 15:41
pyrimidine pmichaud++ (for RT #66280). Now to look for something else to file a bug on...
pyrimidine turning in masak
TimToady try to find a pretty bug this time
pyrimidine ok 15:42
TimToady 'course, some people find rocket explosions beautiful...
15:44 sri_kraih joined
pyrimidine well, july 4th is not far off... 15:44
maybe christmas in july? Now, which july is another question... 15:46
15:47 ejs joined 15:48 araujo left
pugs_svn r27000 | pmichaud++ | [t/spec]: Add test for RT #66280. 15:48
15:52 veritos joined 15:53 iblechbot joined
diakopter tries to find my name highlighted in the backscroll 15:57
Tene 02:20 < luqui> @seen diakopter 15:58
02:21 < lambdabot> diakopter is in #perl6. I don't know when diakopter last spoke.
15:58 veritos left
diakopter Tene: thanks 15:58
15:58 araujo joined
pdcawley What're people using to edit perl6 code in? Is there an emacs mode, or should I bit the bullet and get padre installed? 15:59
ZuLuuuuuu I asked the same questiopn a few days ago, I guess only vim and padre has perl 6 syntax highlighting 16:00
pyrimidine using komodo myself 16:02
StephenPollei I wonder if the perl6 highlighting handles some of perl6 new features well 16:04
rakudo: say 377 #『 pare 』 +610
p6eval rakudo 10a9b2: OUTPUT«987␤»
ZuLuuuuuu pyrimidine: oh I didn't know komodo has it 16:05
will try now with some perl 6 files
pyrimidine no, komodo doesn't
but it could be done via UDL
www.perlmonks.org/?node_id=621135
16:06 dakkar left
DanielC I've had a lot of trouble getting Padre to run, so I think I'll stay with komodo for now. 16:09
16:12 finanalyst left 16:20 sparc left 16:23 agentzh1 joined 16:24 agentzh left 16:25 DanielC_ joined
Matt-W rakudo: "ABCD".comb.perl.say 16:30
p6eval rakudo 10a9b2: OUTPUT«["A", "B", "C", "D"]␤»
pdcawley comb? 16:31
16:33 DanielC is now known as Guest9589, DanielC_ is now known as DanielC 16:34 Psyche^ joined
Matt-W it's like split 16:36
but you specify what you want to keep, not what you want to throw away
rakudo: "ABCD".comb(/B/).perl.say; "ABCD".split(/B/).perl.say
p6eval rakudo 10a9b2: OUTPUT«["B"]␤["A", "CD"]␤»
[particle]- piers! long time no see. 16:37
StephenPollei rakudo: say "ABCDEIBKO".comb(/[BCDK][AEIO]/) 16:39
p6eval rakudo 10a9b2: OUTPUT«␤»
StephenPollei rakudo: say "ABCDEIBKO".comb(/[AEIO]/)
p6eval rakudo 10a9b2: OUTPUT«␤»
Tene StephenPollei: [] in Perl 6 Rules are noncapturing groups
rakudo: say "ABCDEIBKO".comb(/<[BCDK]><[AEIO]>/)
p6eval rakudo 10a9b2: OUTPUT«DEKO␤»
StephenPollei ki'e Tene thanks I need to reread the synopsis, especially before I start writing tests 16:40
hopefully by tuesday to friday of next week, given other things scheduled 16:42
pmurias pdcawley: you can always write a emacs mode for padre
16:43 Guest9589 left
szabgab pdcawley, pmurias there actually already is a partial emacs mode for padre 16:45
DanielC, can we help somehow on #padre
DanielC szabgab: I'll take you up on the offer, but later. I'm about to step out of the house. 16:47
szabgab ZuLuuuuuu, there is a /util/cperl-mode.el in the pugs directory
DanielC thanks
cheers...
16:47 DanielC left 16:50 Patterner left, Psyche^ is now known as Patterner 16:56 payload left, payload1 joined 17:07 ejs left 17:10 ejs joined 17:11 ejs left 17:12 payload1 left 17:13 payload joined
TimToady pmichaud: maybe Camelia's slogan should be "I'm a Second System. What's your problem?" 17:23
17:24 pdcawley left
viklund LOL!! 17:26
pmichaud TimToady: heh, I was just working on that.
I'll add that a bit. 17:27
TimToady on two lines, maybe
pmichaud Probably, yes.
I'm playing with converting the images now. I'll send you a preview link when I've got something semi-ready
TimToady what format do they use?
pmichaud png, for this. 17:28
so we send a *large* png :-)
if you want to make a couple I can set those up also.
TimToady yeah :)
a couple what?
pmichaud images
any particular font for the slogan? ;-)
TimToady let me ask Geneva 17:29
pmichaud oooh, good call.
TimToady mac fonts? 17:30
pmichaud I'm on linux, but I can probably get mac fonts.
or name a mac font typeface, and I can probably find a near-equivalent.
(or a selection of typefaces -- I'm usually pretty good about locating fonts once I know what I'm after :-) 17:31
17:32 DemoFreak joined 17:34 netsquire left 17:36 DemoFreak left 17:37 M_o_C joined
pmichaud (I also have Windows and Photoshop available to me via VMware) 17:37
[particle]- i have photoshop and illustrator, if you need. did the pvmw 2009 poster with those btw 17:38
pmichaud I have both of those. Education discount, and all that :-)
[particle]- yeah, i got the 'neighbor works for adobe' discount, which was *awesome*. 17:39
TimToady preliminary result is Forgotten Futurist 17:42
but her mac hung halfway through her font collection, which is...extensive... 17:43
pmichaud I suspect so. I remember when I had various font management utilities on my system :-)
desktoppub.about.com/library/fonts/...turist.htm # this one?
TimToady looks about right 17:44
the Second System looks really spectacular in that font
it looks cyber without looking dated 17:45
back to see if Geneva's mac is back up...
17:47 pyrimidine_ joined
ZuLuuuuuu what was the process in linux to add rakudo dir to system search path so that I can type perl6 and perl6 executable runs directly? 17:51
sorry if nooby I didn't do that in linux before :) 17:52
antiphase Some variant on adding the new path to the creation of the system PATH variable in /etc/profile
ZuLuuuuuu antiphase: erm, I'm looking into that file but it looks pretty complex? 17:54
szabgab I have it in .bashrc export PATH=$PATH:/path/to/rakudo/dir
17:55 mizioumt joined, DemoFreak joined
antiphase You can do that of course, if you'll only be using it yourself 17:55
szabgab don't forget to source .bashrc when you try it
ZuLuuuuuu szabgab: thank you, that solved the problem!
18:00 pyrimidine_ left 18:02 pmurias left 18:04 pyrimidine left 18:07 cj joined, M_o_C left 18:17 mizioumt1 left
jdv79 how would one do List::Pairwise::grep_pairwise() in p6? 18:24
TimToady with real pairs? 18:25
[particle]- that was my first thought, too 18:26
jdv79 sure, i like pears once in a while. 18:28
18:35 icwiener joined 18:39 mj41 left 18:47 kidd_ joined 18:51 Scorp1us left, Scorp1us joined 18:54 ruoso joined, jferrero joined, Trey left 18:55 payload left 18:56 decasm is now known as decasm_mtg 19:01 alexn_org joined 19:08 pjcj left 19:19 DanielC joined
DanielC After converting a few Debian benchmarks from Perl 5 o Perl 6 I have to say that Perl 6 really does feel a lot nicer. 19:20
19:20 masak joined
DanielC Invariably so far, the Perl 6 version is clearer, and easier to read. Usually shorter, never longer. 19:20
19:20 iblechbot left
viklund Nice 19:21
masak what features does the Associative role deliver?
(the one connected with Hash)
19:21 clintongormley left
PerlJam DanielC: yep. Every time I write some perl 5 code after having written some perl 6 code, I'm always making brain-o type mistakes in perl 5 because I *wish* it was like perl 6 :) 19:22
DanielC PerlJam: :-)
jnthn masak: Requirement of postcircumfix:<{ }> 19:23
masak jnthn: what about .exists?
jnthn No.
Not afaik.
masak ok. then this isn't necessarily a bug:
rakudo: say Match ~~ Associative; say ?("42" ~~ /<Perl6::Grammar::integer>/); say $/.can("exists")
p6eval rakudo 10a9b2: OUTPUT«1␤1␤0␤»
jnthn It just means "can this be indexed into associatively"
masak (oh, and I'm on neighbournet again. caveat IRC-or.) 19:24
19:24 pochi joined
skids DanielC: mainly it's because it is no longer vogue to use regular expressions to make code that looks like line noise. In perl6, you're supposed to use hyperopeators to do that. :-). 19:25
jnthn Not in the sense that Associative should promise you an exists method, no.
masak jnthn: right. oh well.
jnthn masak: I do suspect exists should have one though.
Though in the future postcircumfix:<{ }> probably should accept the :exists adverb too.
masak oh, right. .exists is deprecated. :) 19:26
I forget that sometimes...
jnthn It's deprecated?
masak it isn't?
jnthn I hadn't realized that. OK
I didn't know either way.
I'm not surprised to hear it though.
masak I thought it was to be replaced entirely by :exists.
jnthn Given the adverbial forms exist too.
DanielC skids/PerlJam: A lot of what I like about Perl 6 are the "little things". Loops look better. "say" is better than "print". Rewriting some functions as method calls ($str.substr) often makes the code more clear.
jnthn That'd make sense.
masak but I might be wrong.
skids It's unknown whether the method form is deprecated. We've tried repeatedly to pry that detail out of TimToady, but his lips are sealed. :-) 19:27
masak ok.
19:27 exodist left
lichtkind do backtics also work in perl 6? 19:28
19:28 explorer__ joined, exodist joined
masak lichtkind: yes, but they're spelled 'qx()' now. 19:28
backticks are reserved for user-defined sublanguages IIRC.
skids But it is pretty well established that %h{@a}.kv won't do what you expect, for that you need the adverb form. 19:29
19:29 Trey joined
DanielC skids: I hope that the method call is not deprecated. Often (not always) it makes the code less ambiguous. 19:29
pmichaud qx{}, not qx() 19:30
qx() is a function.
masak oh, right.
of course.
PerlJam DanielC: It's almost just a matter of syntax. there may not be a method, but there will be an adverb.
DanielC PerlJam: Remind me what an adverb is for the purpose of Perl 6? Do you have an example? 19:31
PerlJam DanielC: either way, there will be a "nice" way to say what you mean :)
DanielC: %h<foo>:exists
pmichaud DanielC: the canonical example is 1..200 :by(2)
DanielC Looks great.
pmichaud the ":by(2)" is an adverb.
lichtkind masak: so no more direct `` ? 19:32
19:32 jferrero left
PerlJam lichtkind: nope 19:32
pmichaud the backticks are reserved for user-defined thingies, I think.
although you can probably do qx `cmd` if you really want them :-)
rakudo: say qx`ls1
p6eval rakudo 10a9b2: OUTPUT«Statement not terminated properly at line 2, near "`ls1"␤in Main (src/gen_setting.pm:0)␤»
pmichaud rakudo: say qx`ls`
p6eval rakudo 10a9b2: OUTPUT«operation not permitted in safe mode␤in Main (lib/Safe.pm:25)␤»
lichtkind thanks @all
PerlJam p6eval++ 19:33
19:33 alanhaggai left
pmichaud std: `ls` 19:33
p6eval std 27000: OUTPUT«##### PARSE FAILED #####␤Can't understand next input--giving up at /tmp/64bWCtPDnj line 1:␤------> `ls`␤ expecting any of:␤ prefix or noun␤ statement end␤ statement list␤ whitespace␤FAILED 00:02 35m␤»
DanielC PerlJam: Anyways. I find that with Perl 6 I can say what I mean better.
pmichaud I wonder if there should be a default `` operator that says "did you mean qx{} ?"
masak s1n: thanks for discovering the lingering Rubyism in Web.pm. I must have gotten sloppy in porting the corresponding function from Ruby. :/
PerlJam probably
masak pmichaud: definitely, unless it interferes too much with the user-defined thing. :) 19:35
pmichaud it might
I'm not sure we want to assume that `` is always circumfix.
maybe defining term:sym<`> would be sufficient to still be easily overridden 19:36
PerlJam pmichaud: if there's <.ws> before the first `, then I think you can safely ask if they meant qx{}
pmichaud PerlJam: say 3 ` 5
or
say 3``` # postfix
although that doesn't have the <ws> 19:37
but one might want infix:<`>
or meta...:<`>
rakudo: sub prefix:<`> { '\o/' }; say `3; 19:38
p6eval rakudo 10a9b2: OUTPUT«\o/␤»
pmichaud rakudo: sub circumfix:<` '>($x) { "`$x'" }; say `4'; # just curious 19:39
p6eval rakudo 10a9b2: OUTPUT«`4'␤»
masak DanielC: did you submit the .flip.trans bug?
StephenPollei if term:sym<`> is in STD can it later be undefined if someone whats to extend the grammar their own way? 19:40
pmichaud I'm thinking it might end up being shortest (and thus last) token. 19:41
DanielC masak: This would be a good time for someone to show me how to submit a bug :-) I know that there is an email address, but I'd rather go to the issue tracker.
pmichaud the only way to submit rakudobugs is via email.
PerlJam DanielC: the email address is it
masak DanielC: what pmichaud and PerlJam said.
DanielC But how do I check the status of the bug?
masak DanielC: rt.perl.org/rt3/ 19:42
DanielC ok
PerlJam DanielC: rt will give you the number and you go where masak just said
DanielC thanks
PerlJam DanielC: if you're submitting a bug, prefix your subject with [BUG]
masak also, you'll get all the relevant conversation as emails.
DanielC [email@hidden.address] ...
masak DanielC: [email@hidden.address] 19:43
DanielC thanks
masak PerlJam: oh really? there's such a rule? :P
DanielC masak: I haven't received any conversation emails for the bug I submitted earlier.
PerlJam no, I may just be conflating lists.
masak PerlJam: I hope so, or I need to go back and change my 400 bugs... 19:44
DanielC PerlJam: [BUG] seems a bit redundant if it the email address is rakudobug@
PerlJam heh
pmichaud README:130 says
=head2 Reporting bugs
Bug reports should be sent to C<rakudobug@perl.org> with the moniker
[BUG] (including the brackets) at the start of the subject so that
it gets appropriately tagged in the RT system (rt.perl.org/rt3/).
DanielC It would be like adding [Rakudo]
PerlJam ah, there you go!
pmichaud DanielC: patches, todos, and other similar things also go to rakudobug
PerlJam (reading the docs)++
DanielC ok
masak :)
oh well. 19:45
DanielC "When all else fails, read the instructions"
rakudo: my @pairs = ((1,2), (2,3)); for @pairs -> @pair { } 19:46
19:46 ruoso left
p6eval rakudo 10a9b2: OUTPUT«Parameter type check failed; expected something matching Positional() but got something of type Int() for @pair in call to _block62␤in Main (/tmp/JpXB9pIXoI:2)␤» 19:46
DanielC rakudo: my @pairs = ((1,2), (2,3)); for @pairs -> $pair { }
p6eval rakudo 10a9b2: ( no output )
DanielC But but but...
@pairs[$x] is an array...
lambdabot Unknown command, try @list
19:48 decasm_mtg is now known as decasm
pmichaud rakudo: my @pairs = ((1,2), (2,3)); say @pairs.perl; 19:50
p6eval rakudo 10a9b2: OUTPUT«[1, 2, 2, 3]␤»
pmichaud note: not pairs
DanielC rakudo: my @pairs = ((1,2), (2,3)); say @pairs[0].perl 19:51
p6eval rakudo 10a9b2: OUTPUT«1␤»
DanielC wtf
pmichaud rakudo: my @pairs = (1 => 2, 2 => 3); say @pairs.perl;
p6eval rakudo 10a9b2: OUTPUT«[1 => 2, 2 => 3]␤»
Tene rakudo: my @pairs = ([1,2], [2,3]); say @paris.perl;
p6eval rakudo 10a9b2: OUTPUT«Symbol '@paris' not predeclared in <anonymous> (/tmp/Cx9xLMxNES:2)␤in Main (src/gen_setting.pm:3207)␤»
pmichaud rakudo: my @pairs = ((1,2), (2,3)); say @pairs.elems
p6eval rakudo 10a9b2: OUTPUT«4␤»
Tene rakudo: my @pairs = ([1,2], [2,3]); say @pairs.perl;
p6eval rakudo 10a9b2: OUTPUT«[[1, 2], [2, 3]]␤»
Tene rakudo: my @pairs = ([1,2], [2,3]); say @pairs.elems;
p6eval rakudo 10a9b2: OUTPUT«2␤»
Tene that's what you want to do.
pmichaud not if you really want Pairs 19:52
@pairs is an Array with two Array elements
Tene well, sure.
lambdabot Not enough privileges
Tene @shutuplambdabot
lambdabot Unknown command, try @list
DanielC Tene: Thanks, but what is [ ] ? Is it a scalar that points to an array? Or is it an array proper? Does it take the $ sigil or the @ sigil?
rakudo: my @pairs = ([1,2], [2,3]); for @pairs -> $pair { } 19:53
p6eval rakudo 10a9b2: ( no output )
DanielC rakudo: my @pairs = ([1,2], [2,3]); for @pairs -> @pair { }
p6eval rakudo 10a9b2: ( no output )
PerlJam DanielC: you're asking questions that make no sense in perl 6 :)
(mainly because perl 6 blurs the distinctions you're asking about)
pmichaud rakudo: my @pairs = ([1,2], [2,3]); for @pairs -> @pair { $_.say; }
p6eval rakudo 10a9b2: OUTPUT«Use of uninitialized value␤␤Use of uninitialized value␤␤» 19:54
pmichaud rakudo: my @pairs = ([1,2], [2,3]); for @pairs -> @pair { @pair.say; }
p6eval rakudo 10a9b2: OUTPUT«12␤23␤»
DanielC PerlJam: Surely, something can be an array or a scalar but not both... no?
pmichaud a scalar can be an array
DanielC :-(
pmichaud my $array = [1,2]; say $array.WHAT
rakudo: my $array = [1,2]; say $array.WHAT
p6eval rakudo 10a9b2: OUTPUT«Array()␤»
DanielC my $array = (1,2); say $array.WHAT
19:54 sri_kraih_ joined
DanielC rakudo: my $array = (1,2); say $array.WHAT 19:55
pmichaud the (1,2) case is now a bug in Rakudo (didn't used to be)
p6eval rakudo 10a9b2: OUTPUT«Array()␤»
pmichaud (spec changed, rakudo hasn't caught up)
19:55 sri_kraih left
PerlJam wait, what is it now? 19:55
pmichaud a Capture
except that we aren't sure we'll have captures anymore
at any rate, it doesn't autopromote to Array
PerlJam oh, because of the parens. 19:56
pmichaud well, we'll have captures, but they may be called "List"
PerlJam Sure, but maybe they should be called MagicalListlikeThingy
pmichaud the more we've talked about it, the more they start to look like "List", except List is also allowed to have named information attached to it 19:57
19:57 explorer__ left, jferrero joined
pmichaud in fact, List is something of a shorthand for "argument list" 19:57
(keeping in mind that this is just my interpretation of discussions and might not be accurate with respect to current spec, the spec that is about to be, or whatever TimToady++ may actually be thinking as of today's date/hour/minute :-) 19:58
PerlJam calling them captures has more value as far as indexing text goes. "List" is too generic. 19:59
(and *way* too overloaded already :) 20:00
pmichaud but it's much more natural to say that (1,2,3) produces a list 20:02
masak pmichaud: speaking of 'what is it now?', you mentioned there's now a difference between $x = <a b c> and $a = [<a b c>]. could you point me to the new semantics? I must have missed when the change happened.
pmichaud masak: $x = <a b c> is the same as $x = ('a', 'b', 'c')
for the rest, see above.
masak ok.
20:02 Whiteknight left
pmichaud $a = [<a b c>] is the same as $a = [('a', 'b', 'c')] 20:03
and since [...] imposes list context, that makes it act like $a = ['a', 'b', 'c']
masak so the first is a Capture, and the second an Array?
pmichaud yes.
masak ok, worksforme.
pmichaud the change occurred when $x = (1,2,3) no longer promoted to an Array. 20:04
20:06 alexn_org left 20:08 skids left 20:14 Scorp1us left, Scorp1us joined
DanielC Which synopsis explains captures? 20:16
pmichaud none of them at the moment explain it very well
DanielC Ok. In a way that makes me feel better about not knowing what they are. 20:17
masak DanielC: IIUC, they are the container type that return values come in.
sort of like a list, but with more structure to it. 20:18
DanielC And (1,2,3) is a capture/list but [1,2,3] is an array
masak right. 20:19
DanielC Is there any practical difference? They seem to behave the same.
pmichaud a capture can have named parts
PerlJam DanielC: captures can hold named values too
masak List is immutable.
Array is not.
pmichaud (1, 2, 3, :foo, :bar) # has 3 positional elements, two named elements 20:20
DanielC masak: That's semi-acadameic because when you assign a list to a variable it becomes an array.
pmichaud: That's new!
pmichaud when you assign a list to an array variable it becomes an array
if you assign it to a scalar, it's still immutable
DanielC ah..
pmichaud my $x = (1,2,3); $x.push # fails
masak DanielC: I don't think it's semi-academic. it's still important to know the difference. 20:21
DanielC rakuo: $a = (1,3,4); $a[1] = 8;
pmichaud fails.
(although that might fail for a reason other than $a being immutable)
PerlJam DanielC: you're thinking of the perl5 list/array distinction. Perl 6 is different
DanielC masak: Yeah. I had misunderstood something. I thought that you would never actually see a list in real code because you couldn't assign them to variables. But pmichaud explained that I was wrong. 20:22
masak nod.
DanielC rakudo: $a = (1,3,4); $a[1] = 8;
p6eval rakudo 10a9b2: OUTPUT«Symbol '$a' not predeclared in <anonymous> (/tmp/C8Xm4L5d8w:2)␤in Main (src/gen_setting.pm:3207)␤»
pmichaud note that rakudo still gets them wrong, though.
DanielC rakudo: my $a = (1,3,4); $a[1] = 8; # BAH!
p6eval rakudo 10a9b2: ( no output )
pmichaud what I'm not entirely certain of at this point 20:23
masak hm.
DanielC pmichaud: So, Rakudo should yell at me but doesn't.
pmichaud my $b = 5; my $a = (1, $b, 3); $a[1] = 2; say $b; # 2 ?
masak might that be something worth reporting?
pmichaud masak: you can report it, yes, but it's a very well known bug at the moment, at least to me.
masak submits rakudobug 20:24
better to have duplicates in RT than bad coverage.
DanielC rakudo: my $b = 5; my $a = (1, $b, 3); $a[1] = 2; say $b; # 2 ?
p6eval rakudo 10a9b2: OUTPUT«5␤»
pmichaud I think rakudo may be wrong there also.
masak I'll include that one.
pmichaud (for the same reason -- it mistakenly promotes the List to an Array, per the earlier spec)
DanielC pmichaud: It would be cool if one got karma points for submitting bugs... 20:25
pmichaud I don't control the karmas :-)
DanielC ok :)
pmichaud you'll have to ask the various @botmasters for those.
DanielC Maybe I'll just make my own bot that gives a ++ every few minutes :-D 20:27
masak rakudo: say <jnthn pmichaud DanielC>.pick() ~ '++'
p6eval rakudo 10a9b2: OUTPUT«jnthn++␤»
pmichaud ... 20:28
... 20:29
that took an awfully long time.
masak Rakudo likes jnthn++. and suspense. :)
pmichaud I suspect rakudo was a little frightened of returning the wrong answer :-)
PerlJam rakudo doesn't like you :)
ZuLuuuuuu it loves you all couldn't pick one up, it is like "who do you love most, dad or mom?"
pmichaud rakudo: say <jnthn pmichaud DanielC>.pick() ~ '++' # faster this time?
p6eval rakudo 10a9b2: OUTPUT«DanielC++␤»
ZuLuuuuuu put me in and it will pick me in just a few miliseconds 20:30
DanielC @karma 20:31
pmichaud rakudo: say <jnthn pmichaud DanielC Zuluuuuuu>.pick() ~ '++' # faster this time?
TimToady you could just change your name to Daniel C++
masak :P
DanielC he he
pmichaud I have no idea why rakudo/p6eval is taking forever on that.
lambdabot You have a karma of 8 20:32
ZuLuuuuuu hmmmm
masak all bots are slow today.
DanielC Hmm... even lambdabot is taking its sweet time.
p6eval rakudo 10a9b2: OUTPUT«DanielC++␤»
Tene rakudo: say (<jnthn pmichaud Danielc ZuLuuuuuu> >>~>> '++').perl
pmichaud speedy on my local system, so I suspect p6eval slowness.
japhb Some set of servers lagging, perhaps?
(irc servers)
masak yes, probably.
but only for bots.
pmichaud I couldn't connect to feather.perl6.nl earlier -- might be related.
p6eval rakudo 10a9b2: OUTPUT«["jnthn++", "pmichaud++", "Danielc++", "ZuLuuuuuu++"]␤»
jnthn hey, stop picking on me!
;-)
TimToady would you rather be picked on or picked off?
pmichaud TimToady: my wife's comment upon seeing Camelia: "It's a bug."
TimToady indeed
pmichaud followed by "Why is it a bug?" 20:33
japhb pmichaud: some claim the universe itself is just one giant bug that got loose ...
TimToady there's always a bug
pmichaud ...but does it have to be this big?! ;-)
TimToady we want everyone to get bit by the P6 bug
pmichaud aha!
jnthn I've got like huge lag here now and then...
pmichaud it could be fatal.
japhb Camelia: Vorpal Butterfly
mberends boobot: spack Whatever 20:34
masak buubot: spack Whatever
buubot masak: S02-bits.pod:10 S03-operators.pod:2 S12-objects.pod:1 S29-functions.pod:2
mberends :)
boohoo 20:35
TimToady buubot: spack *
buubot TimToady: S01-overview.pod:14 S02-bits.pod:215 S03-operators.pod:280 S04-control.pod:15 S05-regex.pod:357 S06-routines.pod:91 S07-iterators.pod:1 S09-data.pod:72 S10-packages.pod:16 S11-modules.pod:18 S12-objects.pod:19 S13-overloading.pod:5 S16-io.pod:18 S17-concurrency.pod:7 S19-commandline.pod:40 S22-package-format.pod:9 S26-documentation.pod:5 S29-functi
masak if we were the Rails community, we'd probably have a boob bot...
Tene buubot: spack boobs 20:36
buubot Tene: Sorry, I couldn't find any matches for: boobs
TimToady n
jnthn
.oO( must resist temptation to @seen ... )
pmichaud Camelia wingspan ~==~ 3m, yes? 20:37
TimToady yes
japhb The little guy on my left shoulder is trying to convince me to put easter eggs into spec comment blocks ....
TimToady while the guy on the right is recommending christmas eggs?
Camelia is a butterfly that was caused by a storm flapping its wings in China :) 20:38
DanielC @seen DanielC++
lambdabot I haven't seen DanielC++.
DanielC @karma
lambdabot You have a karma of 8
DanielC :(
it was worth a shot. 20:39
japhb I'm somehow reminded of an Eric Carle book about a little cloud ....
masak DanielC: I can assure you, winning karma points honestly is more fun than trying to game the system. :)
japhb ... except when the karma bots reset.
masak bah. karma is just a fiction anyway.
jnthn Yeah! karma-- 20:40
masak the real karma is within us.
jnthn oh, wait...
PerlJam karma PerlJam
20:40 mncharity joined
PerlJam @karma PerlJam 20:41
lambdabot You have a karma of 10
PerlJam @karma duff
lambdabot duff has a karma of 7
japhb I think Raph Koster would say that karma (as we use it) is a way to game our own pleasure centers to get us to have fun doing useful stuff
TimToady @karmel apples
lambdabot Maybe you meant: karma karma+ karma-
DanielC It would be funny if the karma points were stored as an unsigned byte, so that after 255 they reset back to 0. :)
TimToady @karmal apples
lambdabot Maybe you meant: karma karma+ karma-
TimToady @karma apples
lambdabot apples has a karma of 0
PerlJam @karma TimToady 20:42
lambdabot TimToady has a karma of 45
pmichaud @karma ran over my dogma
lambdabot ran has a karma of 0
20:42 Scorp1us left
TimToady @karma sutra 20:42
lambdabot sutra has a karma of 0
TimToady @karma+ 20:43
lambdabot usage @karma(+|-) nick
masak ZuLuuuuuu: thanks for your question use.perl.org/comments.pl?sid=43098&cid=68951 -- I'll have to think about that one for a bit.
diakopter karma++
PerlJam so I finally got my team of hot-shot coders together for a hackathon and what happens? Everything that could possibly happen to make us not hack. 20:44
ZuLuuuuuu masak: oki doki
diakopter karma karma
@karma karma
lambdabot karma has a karma of 1
masak ZuLuuuuuu: I haven't been thinking very much about performance and target groups, I just want to give Perl 6 web applications a head start before Christmas.
diakopter karma--
@karma karma
lambdabot karma has a karma of -1
PerlJam What was supposed to be a day long hack session has turned into a day long series of distraction
20:44 Scorp1us joined
diakopter @karma ++ 20:44
lambdabot ++ has a karma of 0
diakopter ++++
@karma ++ 20:45
lambdabot ++ has a karma of 1
jnthn ++++++
@karma ++++
lambdabot ++++ has a karma of 1
jnthn hah
ZuLuuuuuu masak: I see
jnthn std: ++++
p6eval std 27000: OUTPUT«##### PARSE FAILED #####␤Can't understand next input--giving up at /tmp/KVGPipzYNI line 1:␤------> ␤ expecting noun␤FAILED 00:02 37m␤»
jnthn phew
diakopter no! 20:46
jnthn rakudo: my $x = 1; $x++ ++; say $x;
p6eval rakudo 10a9b2: OUTPUT«2␤»
jnthn (yes, masak, you already ticketed that one)
;-)
masak jnthn: I thought postfix ops couldn't have space after the term. 20:47
jnthn masak: Right, and you probably can't stack 'em either.
masak: Thus it's a Rakudo bug.
masak 'probably'?
jnthn masak: Just an already ticketed one. :-)
std: $x++++
std: $x++ ++
masak I thought the bug was that it only increments once.
p6eval std 27000: OUTPUT«Potential difficulties:␤ Variable $x is not predeclared at /tmp/AqAB5Jl3BI line 1:␤------> $x++++␤ok 00:02 36m␤»
std 27000: OUTPUT«##### PARSE FAILED #####␤Can't understand next input--giving up at /tmp/v0WFDCUwvK line 1:␤------> ␤ expecting noun␤Other potential difficulties:␤ Variable $x is not predeclared at /tmp/v0WFDCUwvK line 1:␤------> $x++ ++␤ Variable $x is not
..predecla…
jnthn No, parse bug.
masak right.
mncharity Hi folks. Random observation: rosettacode.org/wiki/Main_Page seems to have no Perl 6 presence. I've added it to rosettacode.org/wiki/Help:Request_a...g_language . So if someone wants to churn out
err, 20:48
masak "churn out..."?
jnthn churns out
masak *lol*
mncharity write a bunch of illustrative Perl 6 code, here's a nice place for it. The setup is flexible enough to permit including multiple versions, eg, the strongly typed version, the terse version, etc. 20:49
lots and lots of p5 rosettacode.org/wiki/Category:Perl 20:50
masak oh, and a reminder to everybody: write Perl 6 code! preferably as modules/projects, uploaded to github. it's a great way to learn Perl 6, and maybe even find your very own rakudobug! 20:51
TimToady $x++++ is legal syntax, but a semantic error, since you can't increment the temp value returned by the ++ 20:52
++++$x should be fine though
jnthn std: $x++++ # is a fail though
p6eval std 27000: OUTPUT«Potential difficulties:␤ Variable $x is not predeclared at /tmp/FQgdE38OSO line 1:␤------> $x++++ # is a fail though␤ok 00:02 36m␤»
jnthn oh no
TimToady no, it's only complaining about $x
jnthn right
mis-parsed
the red scared me
TimToady rakudo: my $x = 0; ++++$x; say $x; 20:53
p6eval rakudo 10a9b2: OUTPUT«2␤»
TimToady rakudo: my $x = 0; ++$x++; say $x;
p6eval rakudo 10a9b2: OUTPUT«2␤»
TimToady that's arguable, but probably correct
masak only golfers would do that anyway. 20:54
TimToady rakudo: my $x = 0; say ++$x++; say $x;
p6eval rakudo 10a9b2: OUTPUT«1␤2␤»
TimToady std: ++$x++
p6eval std 27000: OUTPUT«##### PARSE FAILED #####␤"++" and "++" are not associative at /tmp/k5B9OrRF19 line 1:␤------> ++$x++␤ expecting any of:␤ POST␤ postfix␤ postfix_prefix_meta_operator␤ standard stopper␤ terminator␤Other potential difficulties:␤ Variable $x is not
..predeclared at /tmp…
TimToady right 20:55
nm
masak 'are not associative'?
so it is a bug in Rakudo?
TimToady well, it's not up to STD's std
diakopter std: ++++() 20:56
p6eval std 27000: OUTPUT«ok 00:02 37m␤»
TimToady semantic error, but not syntactic
those are nested, not associated :)
masak I usually equate STD's std with 'objectively correct', unless there's a known STD bug.
21:01 decasm left, masak left 21:02 pmurias joined
mncharity hey pmurias. 21:10
how goes?
StephenPollei rakudo: say $PROCESS::PID 21:11
p6eval rakudo 10a9b2: OUTPUT«Use of uninitialized value␤␤»
21:15 skids joined
mncharity perl6: say $*PID; 21:18
p6eval rakudo 10a9b2: OUTPUT«Use of uninitialized value␤␤»
..pugs: OUTPUT«␤»
..elf 27000: OUTPUT«18045␤»
21:18 ejs joined
StephenPollei perl6: say $PROCESS::PID 21:20
p6eval rakudo 10a9b2: OUTPUT«Use of uninitialized value␤␤»
..elf 27000, pugs: OUTPUT«␤»
21:20 viklund_ joined 21:21 gegloe joined, mncharity left
StephenPollei I also don't know how to test if a package/module exists at all 21:21
21:22 viklund left, viklund_ is now known as viklund
StephenPollei rakudo: my $foo;say $foo.WHAT 21:23
p6eval rakudo 10a9b2: OUTPUT«Failure()␤»
21:26 lucs left 21:28 agentzh1 left
StephenPollei rakudo: package FOO { my $bar=1; } ; say $FOO::bar; 21:29
p6eval rakudo 10a9b2: OUTPUT«Use of uninitialized value␤␤»
21:30 Chillance joined 21:37 Whiteknight joined
skids seems to recall something about Rat maybe being made into a hiesentype, or is it still goin to be as specced? 21:38
21:38 meppl left 21:39 agentzh joined 21:42 davidad left 21:53 tulcod left 21:54 eternaleye joined, alester left 21:57 Limbic_Region joined 21:59 mberends left, hercynium left 22:03 ejs left 22:04 eternaleye_ joined 22:09 IRSeekBot joined, mj41 joined 22:14 eternaleye left 22:16 mizioumt left
viklund rakudo: say "—" 22:20
p6eval rakudo 10a9b2: OUTPUT«—␤»
viklund *sigh*
my perl6 binary produces squiggles when I do that 22:21
anyone knows how to fix that?
it works in perl5, the exact same thing
DanielC works here
are you using perl -e ? 22:22
viklund perl -E 'say "—"'
works
but, ./perl6 -e 'say "—"'
doesn't
DanielC hm
viklund yes, hmm
DanielC Just for the sake of testing, could you put the "say" in a file? Just to check that it's not the terminal. 22:23
It shouldn't be. If it was the terminal, I'd expect to see the same problem with Perl 5.
viklund sure thing
me too 22:24
22:24 pmurias left
viklund look at that, it worked 22:24
now I'm confused
DanielC Interesting indeed. 22:25
viklund why did the perl5 version work?
DanielC dunno
My guess is that the terminal is sending a slightly funny character to Perl.... Perl 5 knows what to do with it and Rakudo does not.
But you probably want to speak with pmichaud or jnthn. 22:26
viklund we have the same problem in november
DanielC You know, someone who actually knows something about Rakudo.
viklund ;)
yes
DanielC In any case, it looks like you found a bug. \o/
jnthn It's interesting thtat 5.10 gets it right. 22:27
viklund but you could do it from the terminal?
jnthn (if that's what you're saying)
viklund jnthn: yes
DanielC viklund: yes... perl6 -e 'say "-"' => works for me
22:28 pmurias joined
viklund but that's not the Unicode dash I used up there? (just checking) 22:28
DanielC it is!
viklund ok
DanielC Your dash is different from mine!
I hadn't noticed that.
viklund ha! 22:29
DanielC Good thing you thought of that.
viklund if you copy paste it to the terminal, does it work then?
DanielC On my terminal, using your hash, I get the letter "a" with a little hat on top.
s/hash/dash/
viklund I'm not alone in this world!
DanielC â
That's what I get.
viklund me, too, and when I used "gnome-terminal" I got "â" 22:30
DanielC And like you, if I put it in a file, then it works correctly.
viklund and if you cat the file? 22:31
DanielC And like you, Perl 5 does the right thing.
viklund it looks ok here
DanielC cat => ok here
viklund probably a bug with -e then?
lets try some other unicode character
DanielC yeah, all broken. 22:32
I just tried ñ and ü
viklund ☺
ascent_ ń
viklund and from a file it works 22:33
DanielC So we have a bug that happens with Unicode + terminal (aka "-e" )
yup, ñ and ü work from a file.
jnthn: Are you following this thread? 22:34
jnthn: Just making sure that at least one Rakudo hacker is seeing this.
viklund doesn't work in interactive mode either...
DanielC Terminal.
viklund DanielC: can you submit a bug? 22:35
DanielC And not just gnome-terminal. I get the same result with xterm.
viklund my gmail is blacklisted at perl.org for some reason, otherwise I would do it
DanielC Ok, I'll submit the bug.
viklund my normal terminal is urxvt, that's were I tried it first
I've already checked for it in rt, it's not filed 22:36
I need to go to bed now 22:37
c u
DanielC good night.
jnthn DanielC: no, sorry, I fixing up the build of some POS .Net project... 22:38
DanielC jnthn: Ok. I'm about to send a bug report. Somebody will see it.
jnthn (where POS obviously means Point Of Sale...oh, wait...) 22:41
DanielC: Yes, please do ticket it.
DanielC: I'm not sure what the fix will be, but note that Perl 5.10 gets it right in the report.
DanielC ok 22:42
jnthn Then we've got a lead of hunting soembody down who knows The Answer. :-)
DanielC sent 22:44
22:47 nihiliad left 22:52 nsh_ joined
jnthn -> sleep, rakudo day tomorrow 22:55
Whiteknight (rakudo day)++ 22:56
nsh_ what's rakudo day?
Whiteknight that's where jonathan kicks ass 22:57
he gets funded to work on Rakudo one day per week 22:58
DanielC where does jnthn work?
DanielC just added a new benchmark for Rakudod o/ 23:00
skids from home I think
DanielC *Rakudo
skids: Ok... who does he work for?
skids I think he has an assortment of clients.
DanielC As of now I have 5 benchmarks ported from the Debian shootout. 23:01
skids: Ok... who pays him to work on Rakudo?
23:01 ZuLuuuuuu left
skids There are a couple of funding sources. He has a Hague grant and something else -- I'm not the one to ask, but the details are probably scattered around blogs and wikis. 23:02
23:03 nsh_ left
skids Basically one applies for a grant, proposing something they want to do, the applications are reviewed, and some get funded, others don't. 23:04
DanielC ok
lichtkind DanielC: are you still awaiken ? 23:05
DanielC yup
lichtkind DanielC: yould you take a look my english isnt very native 23:08
DanielC: www.perlfoundation.org/perl6/index....ial_part_1
DanielC lichtkind: Not right now, and honestly I have other Perl 6 projects competing for my time. I'll be happy to help, but I probably won't be able to give it a lot of time. 23:09
s1n DanielC: i'll have to keep an eye on that benchmark now to make sure it's doing what it's supposed to :)
(ironman blog inside joke) 23:10
DanielC lichtkind: but I have bookmarked the page. I'll try to find time this weekend.
lichtkind DanielC: right its just some lines
DanielC: i try to find some one tomorrow i want to be some pages further :) 23:11
DanielC lichtkind: Ok. Just some lines, I can do tomorrow when I'm awake.
23:11 Muixirt left
DanielC I'm half asleep right now, but I see it's just one paragraph. 23:11
I'll be happy to review that tomorrow.
lichtkind DanielC: thanks
23:13 gegloe left 23:18 icwiener left 23:24 pmurias left 23:29 DemoFreak left 23:31 amoc joined 23:33 FurnaceBoy_ joined 23:40 payload joined 23:50 FurnaceBoy left
lichtkind whats wrong with run.pugscode.org? 23:58
23:58 exodist left, exodist joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!