»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | tinyurl.com/p6contest
Set by moritz_ on 28 December 2010.
sjohnson (っ´ω｀)っ 00:13
snarkyboojum coldhead: aye, they could cover for "Inversion of Control" 00:32
sorear good * #perl6 01:47
snarkyboojum o/ 01:48
colomon \o 01:51
sorear colomon: for a while it was fashionable to put that constant in x86 release binaries, because IDA has the same bug 01:59
colomon: a related bug, at least. 02:00
sorear shortcircuit: "mono" on Debians is a stripped down version that can't run any real programs. You need "mono-complete". 02:05
shortcircuit: that said, niecza has been tested on Mono 2.4.4, 2.6.4, 2.6.7, 2.8.2, and HEAD 02:06
Tene: Niecza has a fork of the IO class. I'm interested in discussing philosophy with you. 02:08
sorear whew. backlog finished. 02:12
colomon sorear: I'm pleased to say I had only the smallest difficulty getting niecza v2 running on my Windows box, and that problem was completely the result of my own impatience. Whee! 02:19
sorear colomon: Is that a +1 to "I need to find a way to avoid the initial setting compile on all new installs"? 02:51
colomon errr... only the gentlest +1. It's not exactly arduous, if you know to expect it. 02:52
I just didn't read quite far enough down the README. 02:53
Actually, if you could just have it print a message saying what it is doing before it starts, that would be all you need, IMO.
I didn't know if I was invoking it the right way, so when it just sat there, I assumed I had done something wrong. 02:54
sorear how long did it take on what hardware, ooc? 02:59
perl6: sub infix:<L+>($x, $y) { $x ~ $y }; my $z = "f"; $z L+= "g"; say $z 03:07
p6eval niecza v2-1-gc6596ea: OUTPUT«fg␤»
..rakudo d3e9a3: OUTPUT«Null PMC access in invoke()␤ in main program body at line 1␤»
..pugs: OUTPUT«*** ␤ Unexpected "L"␤ expecting operator␤ at /tmp/PKSv33MRrm line 1, column 53␤»
colomon It took a minute or two, perhaps? It's a quad core machine with 8 gigs RAM, I'm not sure the exact processor type or speed. 03:11
It really wasn't unreasonably long, it was just surprisingly long because I wasn't expecting any significant pause. 03:12
sorear huh. 03:13
it takes about 30 seconds on my one-core (P4 Northwood 2003), 0.375 gigs RAM machine. 03:15
I wonder what's so pessimal about your setup.
diakopter says it takes 5 seconds for him
colomon Huh. It was definitely more than 5 seconds. Might have been 30 seconds, though, I certainly wasn't timing it closely. Is there any way I can redo it so I can time it properly? 03:22
colomon afk # bed 03:30
snarkyboojum niceza takes about 10-15 seconds for a first run, on my lappy 03:43
fwiw :) 03:44
Util rakudo: say "line ", $. ; 03:52
p6eval rakudo d3e9a3: OUTPUT«===SORRY!===␤Unsupported use of $. variable; in Perl 6 please use the filehandle's .line method at line 22, near " ;"␤»
Util rakudo: say "line ", $*IN.line ;
p6eval rakudo d3e9a3: OUTPUT«Method 'line' not found for invocant of class 'IO'␤ in main program body at line 22:/tmp/TTTFk9sbXm␤»
Util rakudo: say $*IN.^methods».name.grep({ m:i/ line | input | record | number / }).Str;
p6eval rakudo d3e9a3: OUTPUT«lines lines␤»
Util Error message says ".line", S28 says ".input_record_number", Rakudo lists neither as methods on a file handle.
snarkyboojum rakudo: say 'niecza'.comb(/./).pick(*).join 03:52
p6eval rakudo d3e9a3: OUTPUT«ziecan␤»
Tene sorear: I recommend you look through the IRC logs for my last conversation with timtoady about IO 05:17
sorear phenny: tell jnthn Is 6model/dotnet or nqp-rx/nom the current state of the art in nomminess? 07:07
phenny sorear: I'll pass that on when jnthn is around.
coldhead nomminess <3 07:12
sorear colomon: so, you say you're not good enough? 07:13
moritz_ good morning 07:46
moritz_ funny, the top 4 stories on hackernews are all about bing vs. google and alleged search result stealing 09:26
masak g'day, zebras. 10:17
moritz_ o/ 10:18
frettled whinnies in two tones like a good zebra.
mathw thinks about recorder technique 10:22
masak thinks about $x % 2 10:26
frettled That's very zebraish of you. 10:28
masak phenny: tell daxim ok, I think I solved your problem, bonus points and all: gist.github.com/807554 11:08
phenny masak: I'll pass that on when daxim is around.
takadonet morning all 13:00
tadzik o/ 13:03
colomon \o 13:04
abraxxa i'm having a problem with nested hash access 13:07
my $matching_rank = %rank_for{$matching_number_cnt}.{$matching_star_cnt};
when printing $matching_rank the output is Any()
what happens in perl 6 if a key doesn't exist?
abraxxa when i look at github.com/perl6/roast/blob/master...fication.t it seems the dot isn't needed 13:10
arnsholt IIRC if you're reading you'll get an undefined value, if you assign it'll get auto-vivified
abraxxa is my syntax correct?
or do i need <> around vars?
that's how i've defined my hash: my %rank_for = { 5 => { 2 => 1 }}; 13:11
this works my $matching_rank = %rank_for{$matching_number_cnt}{$matching_star_cnt}; 13:16
but it returns Any() and not undef if the key's don't exist 13:17
daxim perl6: my %rank_for = { 5 => { 2 => 1 }}; %rank_for.perl.say;
phenny daxim: 11:08Z <masak> tell daxim ok, I think I solved your problem, bonus points and all: gist.github.com/807554
p6eval rakudo d3e9a3: OUTPUT«{"5" => {"2" => 1}}␤» 13:18
..pugs: OUTPUT«{("5" => {("2" => 1),}),}␤»
..niecza v2-1-gc6596ea: OUTPUT«Unhandled exception: Attempted to access slot key of type object for Hash␤ at /home/p6eval/niecza/lib/SAFE.setting line 0 (SAFE Hash.LISTSTORE @ 0)␤ at /tmp/HtlCApg8Lv line 1 (MAIN mainline @ 8)␤ at /home/p6eval/niecza/lib/SAFE.setting line 1037 (SAFE C432_ANON @
..1)␤ at /home/p…
tadzik there is no undef in Perl 6
std: undef
p6eval std 625303c: OUTPUT«===SORRY!===␤Unsupported use of undef as a value; in Perl 6 please use something more specific:␤ Mu (the "most undefined" type object),␤ an undefined type object such as Int,␤ Nil as an empty list,␤ :!defined as a matcher,␤ Any:U as a type constraint␤
.. or fail() as…
abraxxa daxim: i did that and the hash looks ok to me 13:20
the encoding of p6eval isn't ok for me 13:21
tadzik it's not the encoding, these are the ANSI escape sequences for colors
abraxxa tadzik: ah, thanks for the info
colomon abraxxa: Any() is (one of) Perl 6's way(s) of saying undef. 13:23
abraxxa what's the perl 6 equivalent of cond ? true : false; ? 13:24
masak cond ?? true !! false; 13:25
abraxxa ah, ?? !!
i've read it somewhere yesterday and thought it was ?? ||
masak std: my ($a, $b, $c); $a ? $b : $c; 13:26
p6eval std 625303c: OUTPUT«===SORRY!===␤Unsupported use of ?: for the conditional operator; in Perl 6 please use ??!! at /tmp/KlsAZcIyvq line 1:␤------> my ($a, $b, $c); $a ?⏏ $b : $c;␤Parse failed␤FAILED 00:01 122m␤»
abraxxa gist.github.com/806301 13:26
masak: i tried that with my say line but got a different output 13:27
std: my $foo; say $foo ? 'yes' : 'no';
p6eval std 625303c: OUTPUT«===SORRY!===␤Unsupported use of ?: for the conditional operator; in Perl 6 please use ??!! at /tmp/t6fEDYHAgL line 1:␤------> my $foo; say $foo ?⏏ 'yes' : 'no';␤Parse failed␤FAILED 00:01 121m␤»
abraxxa i got Unable to parse blockoid, couldn't find final '}' at line 77 13:28
colomon abraxxa: that would be a Rakudo error, not a std: error, I'm betting.
rakudo: my $foo; say $foo ? 'yes' : 'no';
p6eval rakudo d3e9a3: OUTPUT«===SORRY!===␤Confused at line 22, near "say $foo ?"␤»
abraxxa colomon: i see
masak abraxxa: it's ??!!, not ?: 13:29
abraxxa: I just did ?: to show you the informative error message STD gives you.
abraxxa masak: i tried to trigger the error message i got
masak Rakudo sometimes says "confused" where it should give more informative error messages.
daxim masak, what's the expression mex in solution.txt? 13:32
masak daxim: mex means "minimum excluded". it's the smallest natural number that's not among the arguments. 13:34
daxim: for an (impartial) game position with a number of "options", positions to which you can move, the value of that game position is the mex of its options. 13:35
daxim I fail to see a flaw in your written reasoning. I still need an explanation how you arrive at picking up a certain number of sticks when the opponent has already forced you into an unwinnable situation, e.g. when 6 are remaining 13:37
daxim if I see this correctly, the formula would pick 1 13:38
masak looks 13:43
right, 6-sticks-remaining is an unwinnable situation. 13:44
hence, there is no winning move.
daxim but you must play until the bitter end
masak right, so pick an arbitrary strategy.
it doesn't matter at that point.
not if the opponent's good.
daxim so, what do you pick? I usually say rand(1..4)
masak it doesn't matter. 13:45
there's nothing to recommend any particular strategy in a losing position.
also, that wasn't part of your original problem.
daxim I read that nim games are always played out because one could always make a mistake at the last moment (humans being involved and all…) and giving up would squander that chance 13:46
masak I'm not arguing against that. I'm saying that at the point where you find yourself in a losing position, there's no strategy that will save you. only "luck". 13:47
...which doesn't exist when you play against a good player.
Util rakudo: our regex serial_number { <[A..Z]> \d**8 }; $_ = "ABCDEFGH"; my $z = /<serial_number>/ ?? "Y" !! "N"; say $z; 14:04
p6eval rakudo d3e9a3: OUTPUT«Method 'serial_number' not found for invocant of class 'Cursor'␤ in <anon> at line 22:/tmp/l6FgU8fGKV␤ in 'Cool::match' at line 2631:CORE.setting␤ in 'Regex::Bool' at line 6211:CORE.setting␤ in main program body at line 1:/tmp/l6FgU8fGKV␤»
Util ??? 14:05
snarkyboojum rakudo: our regex serial_number { <[A..Z]> \d**8 }; $_ = "ABCDEFGH"; my $z = /serial_number/ ?? "Y" !! "N"; say $z; 14:15
p6eval rakudo d3e9a3: OUTPUT«N␤»
Util snarkyboojum: without the angle brackets, "serial_number" is just a series of characters to search for. 14:18
snarkyboojum Util: Sure. Nfi here - just playing :)
Util thx 14:19
masak Util: I'm momentarily mystified as well. 14:39
Util: that, or something very much like it, used to work.
rakudo: our regex serial_number { <[A..Z]> \d**8 }; $_ = "ABCDEFGH"; my $z = /<&serial_number>/ ?? "Y" !! "N"; say $z;
Util It works within a grammar.
p6eval rakudo d3e9a3: OUTPUT«N␤»
masak there you go.
Util My reading of S05 is that the & should be legal (to mean a lexical regex), but not *necessary* in this code. 14:44
thanks, though; it is just the work-around that I needed.
OHNO! the &-form supresses capture. 14:45
PerlJam Util: <foo=&foo>
Util PerlJam: thx
and Doh! 14:46
PerlJam Util: what part of S05 gives you the impression that the & isn't necessary?
Util PerlJam: "This form always gives preference to a lexically scoped regex declaration" 14:51
(I was testing with "my" instead of "our"). 14:52
rakudo: my regex serial_number { <[A..Z]> \d**8 }; $_ = "ABCDEFGH"; my $z = /<serial_number>/ ?? "Y" !! "N"; say $z;
masak Util: I think you're right, and that the &-less shortcut is simply unimplemented in nqp-rx.
p6eval rakudo d3e9a3: OUTPUT«Method 'serial_number' not found for invocant of class 'Cursor'␤ in <anon> at line 22:/tmp/MsLFjfYuxL␤ in 'Cool::match' at line 2631:CORE.setting␤ in 'Regex::Bool' at line 6211:CORE.setting␤ in main program body at line 1:/tmp/MsLFjfYuxL␤»
Util rakudo: my regex serial_number { <[A..Z]> \d**8 }; $_ = "ABCDEFGH"; my $z = /<&serial_number>/ ?? "Y" !! "N"; say $z;
p6eval rakudo d3e9a3: OUTPUT«N␤»
PerlJam Hmm. 14:53
In my reading that sentence applies to either <foo()> or <foo:>, but not <foo>
masak I haven't read the context in S05, but it feels "right" to me that serial_number should defer to &serial_number. 14:54
PerlJam ah, but the last paragraph in that section says "A call to C<< <foo> >> will fail if there is neither any lexically scoped routine of that name it can call, nor ..." 14:55
masak: yes, it does. 14:56
now I've got to lookup what the problems with implementing it were last time we talked about this. 14:59
masak :)
Util PerlJam: I just arrived at the same thought; that the last paragraph implies my interpretation. Also, all the intervening text makes sense (as in "is coherent") if you read it as generic <foo>, and take all the extra stuff in each <foo...> example as being extra detail given to match the discussion in previous section. 15:00
Util is not awake enough to make that last missive more intelligible :( 15:02
PerlJam Util: looked fine to me :)
Util thx, and big thanks to you both for your help. 15:03
I will have more bugs (of a different nature) later this morning.
masak yay 15:04
PerlJam Looks like we talked about this subject (named regex outside of a grammar) on 2010-07-20: irclog.perlgeek.de/perl6/2010-07-20 15:10
I don't know if that was the last time though :)
PerlJam The answer to Pm-21/22 in github.com/perl6/mu/blob/master/misc/pm.txt seem to be the last state of everyone's thinking. 15:16
and NYI of course 15:17
Util PerlJam: thanks! 15:21
masak PerlJam++ # research 15:22
pmichaud good morning, #perl6 15:32
takadonet pmichaud: morning 15:33
masak morning, pm
jnthn morning, pmichaud 15:35
phenny jnthn: 07:07Z <sorear> tell jnthn Is 6model/dotnet or nqp-rx/nom the current state of the art in nomminess?
colomon \o 15:36
jnthn phenny: tell sorear the CLR one is a bit cleaner for understanding it. Only nqp-rx/nom knows the native types bit though.
phenny jnthn: I'll pass that on when sorear is around.
pmichaud note that the answer to Pm-21/22 are not only NYI, but also "not yet specced" 15:38
*answers
sorear good * #perl6 16:18
phenny sorear: 15:36Z <jnthn> tell sorear the CLR one is a bit cleaner for understanding it. Only nqp-rx/nom knows the native types bit though.
sorear jnthn: thanks. 16:20
Util pmichaud: thanks 16:24
(unrelated) 16:25
Strange example in S05: given $source_code { $parsetree = STD.parse($source_code) }
Why use a given{...} block?
colomon thinko 16:26
PerlJam also, that code might have undergone modifications that left it in that state but was originally more complex
colomon I... yeah, what PerlJam said. 16:27
or the standard switched and it was corrected without looking at it closely or something.
Util Ah, I did not think to look at the Spec's history.
`git show a8aa081` is the commit that did it. 16:31
colomon what was it before? 16:32
Util $parsetree = m:keepall/<Perl::prog>/;
I will update the spec 16:33
pmichaud Util++
dalek ecs: 4c4b50f | util++ | S05-regex.pod:
Remove block left-over from a8aa081e5acf4116026f20152fd50a8d78e7fd03
 16:37
sorear jnthn: In a STable, how many of the fields are REPR-dependant? Which are they? 16:38
hello pmichaud
pmichaud sorear: o/ 16:39
sorear rakudo: subset Even of Int where * %% 2; my Even $x; say $x.WHAT; say $x ~~ Int; say $x.^methods; 16:46
p6eval rakudo d3e9a3:
..OUTPUT«Int()␤Bool::True␤perlpredsuccWHICHStrBridgeIntRatNumsignACCEPTSRealBoolComplexrealsisNaNabsexplnsqrtrootsfloorceilingtruncateroundcisunpolarrandsinasincosacostanatansecaseccosecacoseccotanacotansinhasinhcoshacoshtanhatanhsechasechcosechacosechcotanhacotanhatan2Numericloglog10to-rad…
sorear rakudo: subset Even of Int where * %% 2; my Even $x; say $x; $x = Int; say $x 16:47
p6eval rakudo d3e9a3: OUTPUT«Int()␤Int()␤»
sorear rakudo: say Int %% 2;
p6eval rakudo d3e9a3: OUTPUT«Bool::True␤»
sorear rakudo: say Int !%% 2;
p6eval rakudo d3e9a3: OUTPUT«===SORRY!===␤Infix !% is deprecated in favor of infix %% at line 22, near "% 2;"␤»
sorear rakudo: say Int ![%%] 2;
masak submits rakudobug
p6eval rakudo d3e9a3: OUTPUT«Bool::False␤»
masak std: say Int !%% 2; 16:48
p6eval std 625303c: OUTPUT«ok 00:01 120m␤»
sorear rakudo: subset Odd of Int where * ![%%] 2; my Odd $x; say $x; $x = $x; say $x 16:48
p6eval rakudo d3e9a3: OUTPUT«Int()␤Type check failed for assignment␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/fNWcjWwAcl␤»
sorear Is this how it's supposed to work?
perl6: sub infix:<L+>($x,$y) { $x ~ $y }; my $z = 'f'; $z L+= 'g'; say $z; # masak, is this one known? 16:49
p6eval niecza v2-1-gc6596ea: OUTPUT«fg␤»
..rakudo d3e9a3: OUTPUT«Null PMC access in invoke()␤ in main program body at line 1␤»
..pugs: OUTPUT«*** ␤ Unexpected "L"␤ expecting operator␤ at /tmp/8nysncGppZ line 1, column 52␤»
masak sorear: yes. 16:50
sorear: it's an oldie-but-goodie.
IIRC, use 'our' on the sub and it goes away.
pmurias perl6: say 3 ![%%] 2 16:53
p6eval rakudo d3e9a3: OUTPUT«Bool::True␤»
..niecza v2-1-gc6596ea: OUTPUT«Unhandled exception: Modulus calculation NYI␤ at /home/p6eval/niecza/lib/SAFE.setting line 377 (SAFE die @ 2)␤ at /home/p6eval/niecza/lib/SAFE.setting line 1003 (SAFE infix:<%> @ 2)␤ at /home/p6eval/niecza/lib/SAFE.setting line 983 (SAFE infix:<%%> @ 2)␤ at
../tmp/pewVv5Ji_r line…
..pugs: OUTPUT«*** ␤ Unexpected "![%%]"␤ expecting operator, ":" or ","␤ at /tmp/K8eVkslczG line 1, column 7␤»
pmurias rakduo: say 3 %% 2 16:54
[Coke] why is the output frmo niecza in 2 sends there?
(why isn't it just a single send with another ␤ ?
pmurias ? 16:55
[Coke]: what do you mean?
pmichaud yay, snow days!
pmurias [Coke]: ah, i get it
masak [Coke]: because it's so long. 16:56
masak it's broken off somewhere along the way. 16:56
pmurias eval_bot should just nopaste errors
* p6eval
[Coke] pmichaud: I'm just glad it looks like we're not going to lose power this time. :P 16:57
pmurias niecza, pugs: see nopaste-url 16:57
[Coke] ... wait, YOU have a snow day? 16:58
pmichaud our second in a row 16:58
and there's a good possibility for a third tomorrow
we got hit by ice storms, and the temperature has yet to go above 20 degF, so the ice on the roads is just.... sitting there
the whole state of Texas is going through "rolling blackouts" because the utilities can't keep up with demand 16:59
takadonet bah we never have snow days :( 17:01
sorear hasn't had any either :( 17:02
[Coke] you're welcome to come shovel us out, you can have our snow day. :) 17:03
takadonet [Coke]: see we do have lots of snow but we just have enough machines to clear it out 17:06
pmichaud No machines here. No road salt. Just ice pellets all frozen together on the road surface, and a bunch of Priuses trying to maintain traction on it. :-P 17:10
(no "snow tires" either, of course :) 17:11
arnsholt My girlfriend is headed for Texas ATM. She'll be really unhappy if there's snow =) 17:13
pmichaud That's further south -- it doesn't snow quite as often there.
arnsholt Ah, right. Good for her then =)
sorear is still trying to figure out subsets- see above 17:15
pmichaud oops, but it looks like it's forecasting snow there on Friday :-)
arnsholt Fun, fun. Austin looks like it should miss it, but the temperature forecasts are a bit chilly 17:17
pmurias sorear: what's the part you are figuring out? 17:19
tadzik \o/ 17:21
sorear rakudo: subset Odd of Int where * ![%%] 2; my Odd $x; $x = $x
p6eval rakudo d3e9a3: OUTPUT«Type check failed for assignment␤ in '&infix:<=>' at line 1␤ in main program body at line 22:/tmp/2g5di8RELL␤»
tadzik is now on the 3rd semester, disregarding the two remaining exams
sorear pmurias: stuff like that
sorear phenny: tell jnthn thanks. 17:27
phenny sorear: I'll pass that on when jnthn is around.
sorear phenny: tell jnthn In a STable, how many and which of the fields are REPR-dependant? 17:28
phenny sorear: I'll pass that on when jnthn is around.
sorear phenny: tell jnthn How do subsets fit into the 6model type_check cache paradigm?
phenny sorear: I'll pass that on when jnthn is around.
sorear eeek! nqp-rx/nom hardcodes "8" instead of "CHAR_BIT" 17:41
sorear phenny: tell jnthn At P6str.c line 128, why don't you mark the STRING? 17:43
phenny sorear: I'll pass that on when jnthn is around.
dukeleto magic constants!
sorear I'm sure 9-bit chars will become fashionable again someday 17:44
Or perhaps 32
dukeleto I am holding out for 36bit words 17:46
ash_ does IO let you access the file descriptor? 17:55
ash_ or am i going to have to dig into PIR to get to that? 17:55
masak likely. 17:57
perigrin colomon++ # submitted a talk! 17:58
colomon perigrin: I was also unable to pay when I tried. Clicking on the "credit card" link took me to an error page. 18:01
perigrin *sigh*
I'll talk with the ACT peopel
er people
colomon: your talk will probably be accepted so don't pay yet :) 18:02
colomon perigrin: yeah, I figured out after the fact that I shouldn't have been trying, but thought you'd want to know. 18:03
perigrin colomon: thanks, mail sent to the ACT people 18:05
ash_ is there an unbuffered io object? 18:16
and or a bare file descriptor object?
sorear Int, use POSIX:from<perl5> 18:17
ash_ eh, well not object but i hope you understand what i am asking
i am hoping to be able to do this with just parrot/rakudo (i'd like to just do it with rakudo to be honest though) 18:18
i was going to see about writing a serial library, for opening serial ports (like the USB kind)
ah, parrot's Handle object has get_fd, which will let me access the file descriptor directly 18:20
IO should let you get the fd as well, for ffi purposes
sorear phenny: tell jnthn How does the NQP 'native' type_declarator work? 18:30
phenny sorear: I'll pass that on when jnthn is around. I'll have to use a pastebin, though, so your message may get lost.
sorear agggh.
aloha: msg jnthn How does the NQP 'native' type_declarator work?
aloha: tell jnthn How does the NQP 'native' type_declarator work?
phenny is configured to use a pastebin that no longer exists, so "may" means "will" 18:32
ash_ can we fix phenny? 18:33
pmurias sorear: what's the hard part with that subset? 19:51
![%%] seems strange 19:52
[Coke] masak: is that %% ticket different from the existing %% ticket? 19:55
[particle] potty mouth 19:56
masak [Coke]: url?
dalek tpfwiki: (Razan Abbass)++ | www.perlfoundation.org/perl6/index....ssd_weblog 20:05
tpfwiki: (Razan Abbass)++ | www.perlfoundation.org/perl6/index....ssd_weblog 20:11
tpfwiki: (Razan Abbass)++ | www.perlfoundation.org/perl6/index....ssd_weblog
colomon ![%%] seems insane, I'd say. 20:13
or at least, [%%] does.
rakudo: say [%%] 5, 4, 6, 3 20:14
p6eval rakudo d3e9a3: OUTPUT«Bool::False␤»
colomon rakudo: say [%%] 6, 3
p6eval rakudo d3e9a3: OUTPUT«Bool::True␤»
colomon rakudo: say [%%] 6, 3, 1 20:15
p6eval rakudo d3e9a3: OUTPUT«Bool::True␤»
colomon rakudo: say [%%] 6, 3, 2
p6eval rakudo d3e9a3: OUTPUT«Bool::False␤»
masak ![%%] just means !%%
it's just a way of disambiguation.
don't panic :)
colomon ah
rakudo: say 5 ![%%] 2 20:16
p6eval rakudo d3e9a3: OUTPUT«Bool::True␤»
colomon rakudo: say 5 !%% 2
p6eval rakudo d3e9a3: OUTPUT«===SORRY!===␤Infix !% is deprecated in favor of infix %% at line 22, near "% 2"␤»
[particle] ELTMFAIL
masak exactly. 20:17
pmichaud std: say 5!%2; 20:20
p6eval std 625303c: OUTPUT«===SORRY!===␤Can't negate % because multiplicative operators are not iffy enough at /tmp/XMjHuV6Ugh line 1:␤------> say 5!%⏏2;␤Parse failed␤FAILED 00:01 120m␤»
[Coke] masak: i haven't found the old one yet. :)
masak searching RT is a fine art.
[Coke] what's the syntax for looping over an array while still having access to the index if you need it?
pmichaud for @array.kv -> $index, $value { ... } 20:21
[Coke] danke.
pmichaud std: say 5 !+ 2; 20:24
p6eval std 625303c: OUTPUT«===SORRY!===␤Can't negate + because additive operators are not iffy enough at /tmp/8gz0BTMnxe line 1:␤------> say 5 !+⏏ 2;␤Parse failed␤FAILED 00:01 120m␤»
masak "iffy" == "returning something that's either a boolean, or would be if it wasn't a more interesting value" :) 20:26
pmichaud right 20:27
I'm just trying to add it to Rakudo to see what happens :)
colomon pmichaud: are you fixing the !%% in the process?
pmichaud yes
since %% is iffy 20:28
colomon then I will undo my local (rather brute force) changes. ;)
pmichaud let's see what happens :-) 20:29
gist.github.com/808367 # patch I'm currently testing, if anyone is interested :-) 20:30
colomon so you're switching the specific !% error message to a generic "not iffy enough" one. 20:31
pmichaud I'm eliminating the specific !% token altogether. STD.pm6 doesn't have one. 20:31
I'm generally following the way that STD handles it 20:32
colomon very reasonable policy.
pmichaud pmichaud@orange:~/rakudo$ ./perl6 20:39
> say 5 !%% 2;
Bool::True
> say 5 !+ 2;
Can't negate + because it is not iffy enough at line 1, near " 2;\n"
tadzik pmichaud++
pmichaud spectesting now
tadzik pmichaud: ooc, what are the chances for Rakudo using STD one day, like niecza does? 20:40
pmichaud tadzik: I still expect STD and Rakudo's grammar to converge, more than it being "one or the other"
colomon do we have a spectest for !%% ?
pmichaud colomon: ack doesn't reveal any 20:41
there are still some things that I think Rakudo's grammar handles slightly better than STD, although that's becoming less and less so thanks to niecza
but yes, someday I expect us to be using a very common grammar at some point in the not-to-distant future 20:42
colomon huh. there's a block of %% tests in S03-metaops.not.t. Leftover from !%, perhaps? 20:43
pmichaud it's even more likely to occur if/when we get roles into nqp
+1 for colomon++ to fix the tests :-) 20:44
colomon crazily enough, there appear to be no other direct tests of %% -- all other incidences are using %% to test some other feature.
We should have done tests that day at YAPC::NA. :)
pmichaud I wonder if I should start using audreyt's approach "You want this fixed? Patch in a test first!" :-P
colomon I'll get to work on it.
pmichaud or, as I think she said: "I produce code in exchange for tests" 20:45
something like that
colomon what do we call %%, anyway? divisible-by.t ?
pmichaud C<< infix:<%%> >>, is divisible by 20:46
(from S03)
colomon is-divisible-by.t it is, then.
dalek ast: 799b9e9 | (Solomon Foster)++ | S03- (2 files):
Move infix:<%%> tests from S03-metaops/not.t to a new S03-operators/is-divisible-by.t.
 20:56
ast: 30833cd | (Solomon Foster)++ | S03-operators/is-divisible-by.t:
Temporarily fudge file until pmichaud gets his infix::<!%%> patch merged in.
colomon done. 20:58
colomon is suspicious he's at least partially responsible for the previous sloppy test.
dalek kudo: 88fa70d | colomon++ | t/spectest.data:
Add S03-operators/is-divisible-by.t.
 21:00
pmichaud almost finished with spectests 21:07
one failure in charset.t 21:08
....which appears to be due to a nqp-rx string mismatch 21:09
so, committing.
pmichaud@orange:~/rakudo$ make t/spec/S03-operators/is-divisible-by.t 21:10
All tests successful.
masak \o/ 21:11
colomon did you unfudge it? I wasn't able to run the !%% tests here (obviously).
pmichaud colomon: did you update the plan in not.t? plan says 36 tests but I think there's only 30.
colomon pmichaud: good catch.
I'll fix it.
dalek ast: 5a1dd86 | (Solomon Foster)++ | S03-metaops/not.t:
Fix plan to reflect the moved tests.
 21:12
pmichaud Resolved ticket :-) 21:16
colomon pmichaud++
tadzik 527 open tickets, a bit like Parrot :)
dalek kudo: 3851302 | pmichaud++ | / (2 files):
Fix RT #83250, enabling !%% to work properly, removing the bogus !% panic, and eliminating the invalid cases of meta-negation on non-iffy infix operators.
 21:17
kudo: 2666b6f | pmichaud++ | /:
Merge branch 'master' of github.com:rakudo/rakudo
colomon probably resolved at least one other with the iffy fix, too, no?
pmichaud if there was another, then yes.
[Coke] tadzik: you have to count new+open
tadzik ouch
masak sorry... :P
pmichaud I blame masakbug. 21:18
tadzik rakudo: $_ = "test"; s/t/x/; .say;
p6eval rakudo d3e9a3: OUTPUT«xest␤»
tadzik rt.perl.org/rt3/Public/Bug/Display.html?id=62064 closable then, no?
pmichaud tadzik: looks like yes to me, if we have spectests. 21:19
832 tickets, it looks like
masak we have tr/// ?
[Coke] down from about 850 earlier in the week.
colomon [Coke]++
pmichaud [Coke]++ and tadzik++ # ticket wranglers
tadzik dozens of them are opened, but fixed
[Coke] tadzik: might be less than dozens, now, but I'm hopeful. 21:20
tadzik I remember finding like 20 of them one afternoon, but had no one to close them, and didn't think about responding to them, shame on me
[Coke] tadzik: you have privs now, yes? 21:21
tadzik [Coke]: dunno, don't remember I must say
I remember asking pmichaud about them
[Coke] huh. new "mark spam" link on the toolbar. 21:22
(used to be you had to change the queue to spam manually)
tadzik will investigate after my exams 21:22
everything on my schedule is for "after my exams"
pmichaud std: my %v = q=>10; 21:25
p6eval std 625303c: OUTPUT«ok 00:01 121m␤»
dalek ast: fbe953c | (Solomon Foster)++ | S03-operators/is-divisible-by.t:
Unfudge now that pmichaud++ has it working.
 21:26
moritz_ rakudo: say 1/6 * [+] 0..5 21:28
p6eval rakudo d3e9a3: OUTPUT«2.5␤»
moritz_ rakudo: say 1/6 [+] (0..5 X- 2.5) X** 2 21:31
p6eval rakudo d3e9a3: OUTPUT«38.0277777777778␤» 21:32
colomon huh 21:35
?
rakudo: say (1/6 [+] (0..5 X- 2.5) X** 2).perl
p6eval rakudo d3e9a3: OUTPUT«(1369/36)␤»
moritz_ context: mean and variance for www.perlmonks.org/?node_id=885823 21:36
colomon but how is 1/6 [+] (list) not two terms in a row?
std: say 1/6 [+] (0..5 X- 2.5) X** 2
moritz_ a growth model where each member of the population spawns between 0 and 5 children (uniform distribution)
p6eval std 625303c: OUTPUT«ok 00:01 120m␤»
moritz_ rakudo: say 1/6 * [+] (0..5 X- 2.5) X** 2 21:37
p6eval rakudo d3e9a3: OUTPUT«2.91666666666667␤»
moritz_ much saner :-)
rakudo: say (1/6 * [+] (0..5 X- 2.5) X** 2).perl
p6eval rakudo d3e9a3: OUTPUT«35/12␤»
pmichaud std: say (1/6) [+] (0..5 X- 2.5) X** 2
moritz_ doesn't know how it's parsed
p6eval std 625303c: OUTPUT«ok 00:01 121m␤»
moritz_ oh
OH
pmichaud that's indeeded a bit weird. I'm guessing it's the operator [] instead of the reduction [] 21:38
moritz_ it's jsut infix<+>
pmichaud right
colomon ah.
moritz_ rakudo: say 1 [+] 3
p6eval rakudo d3e9a3: OUTPUT«4␤»
pmichaud [+] in infix position is same as +
moritz_ rakudo: say 1 [[[+]]] 5
colomon but isn't it a list on the RHS?
p6eval rakudo d3e9a3: OUTPUT«6␤»
colomon rakudo: say 1 + (2, 3)
moritz_ colomon: yep. In numeric context
p6eval rakudo d3e9a3: OUTPUT«3␤»
colomon ah. right.
moritz_ colomon++ # spotting bug in my code
pmichaud do we have tests for <time> and <now>? I'm thinking RT #77904 can be closed 21:41
rakudo: say now
p6eval rakudo d3e9a3: OUTPUT«Instant:2011-02-02T21:41:41.700150Z␤»
pmichaud I thought there was a 'needtests' status or something? 21:42
pmichaud looks
pmichaud found it 21:43
time for a break 21:44
[Coke] I called it testneeded to match the existing test* 21:46
pmurias sorear: is niecza powerfull enough for writing a moderatly complex gui app (a behaviour trees editor)? 21:49
moritz_ would be surprised if it weren't 21:50
masak pmurias: how will you do the GUI bindings? 21:51
pmurias that's the hard part 21:52
that is the one thing turing completness isn't enough for 21:53
masak: the best option would be to have the work magically 21:54
as hand binding stuff is annoying and doesn't give a better binding if it's not a C library 21:55
masak I'm all for things working magically. 21:57
pmurias but it's possible to call C# stuff by some thing to embed niecza IR (NAM) 21:59
moritz_ I've now created rakudo.de/testneeded and rakudo.de/testcommitted as redirects the corresponding RT queries 22:00
and 'needstest' and 'hastest' as aliases 22:01
masak moritz_++ 22:01
Tene For any vim users, I just discovered and rather like code.google.com/p/conque/wiki/Usage 22:06
Tene seems like a nice way to embed a terminal in a vim window, and I've had a few nice uses of sending Perl 6 code to a repl from a buffer. 22:07
jnthn phenny: tell sorear answers in gist.github.com/808572 22:19
phenny jnthn: I'll pass that on when sorear is around.
jnthn: 17:27Z <sorear> tell jnthn thanks.
jnthn: 17:28Z <sorear> tell jnthn In a STable, how many and which of the fields are REPR-dependant?
jnthn: 17:28Z <sorear> tell jnthn How do subsets fit into the 6model type_check cache paradigm?
jnthn: 17:43Z <sorear> tell jnthn At P6str.c line 128, why don't you mark the STRING?
[Sorry, some messages were elided and lost...]
arnsholt Tene: That -does- seem useful
jnthn phenny: :P
pmichaud++ # fixing bugs \o/ 22:21
jnthn is muchly distracted by $teaching this week :)
2 more days, then normal life/hacking resumes. ;)
Tene arnsholt: I'm also trying to get in the habit of using surround.vim 22:22
jnthn: what are you teaching?
jnthn Tene: Some .Net crap 22:23
Tene: Well, some of it's good too :)
jnthn It's a web dev course. It's nice that at least one of the people in the class actually thinks about usability issues along the way. :) 22:24
Tene Nice.
jnthn Yeah, the students are a very nice bunch. :)
Makes it much more fun that way.
Tene I'm starting to miss teaching a bit.
jnthn You don't do it so much these days? 22:25
Tene No, I've been full-time sysadmin for about a year.
jnthn I couldn't cope with doing it every week, but it's nice to do.
Aww.
Tene Yeah, previous job was a bit too much. I was teaching linux sysadmin classes almost every week.
Lots of travel.
Lots of fun, but was getting tiring after 2.5 years. 22:26
jnthn *nod8
I tend to teach about one class a month, roughly.
Normally locally, but they sent me up to Stockholm for this one. :) 22:27
masak oh dear, I think I've gotten used to Mu.
jnthn oh no, you accidentally Mu in Java instead of Object? ;) 22:28
masak not that used to it.
jnthn :)
masak I'm reading github.com/perl6/mu/blob/master/misc/pm.txt , a very nice read. 22:29
in Ml-1, the reply had a few "Object"s in it. I found myself thinking "heh, that's old... that should be Mu..."
masak I complained yesterday about the lack of good error reporting for multis. 23:02
there's a cute workaround that's worth knowing about. 23:03
define a most-general multi and have it give a good error :)
jnthn masak: Huh? Multi errors are quite informative these days, no?
masak jnthn: yes, they are. 23:04
they're just missing one essential part: what you passed in. :/ 23:05
jnthn: but don't worry, the awesome vaporware state machine binder will solve that ;)
masak haz plans
jnthn uh-oh :) 23:06
sleep & 23:07
masak oh, and I did my first C<(for ... { ... }) // ...> yesterday. pretty neat.
masak reads the reply to Pm-18 with widening eyes 23:41
masak rakudo: (say "Found $_!" if $_ %% 10 for 7..12) or say "Aww" 23:45
p6eval rakudo 2666b6: OUTPUT«Found 10!␤»
masak rakudo: (say "Found $_!" if $_ %% 10 for 11..12) or say "Aww" 23:45
p6eval rakudo 2666b6: OUTPUT«Aww␤»
masak rakudo: (say "Found $_!" if $_ %% 10 for 11..12) // say "Aww"
p6eval rakudo 2666b6: ( no output )
masak hm. so it really is 'or' there, and not '//' ?
masak looks in S04 23:46
masak I can't find the place I'm looking for, but I guess it should be 'or', since an empty Seq isn't undefined. 23:52
colomon n 23:57
masak nn
colomon whoever thought it was a good idea to shift focus from the "enter password" window to the normal irc window in the middle of initialization needs a lesson in UI design. 23:58
masak most UIs suck. film at eleven. 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!