»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:09 captain-adequate left 00:14 cpage_ joined 00:20 troys is now known as troys_ 00:26 Xliff left 00:30 lookatme joined
lookatme morning \o. 00:30
Herby_ o/ 00:31
00:35 skids joined 00:37 [2300mAh] left 00:40 lookatme left, lookatme joined 00:44 lookatme left, lookatme joined
Juerd ♥ given $input { when any(@stack) { note "Already selected!" } ... } 00:44
00:48 lookatme left, lookatme joined 00:51 kurahaupo joined 00:53 Cabanossi left 00:54 astj joined 00:55 Cabanossi joined 01:00 lookatme left 01:02 astj left, astj joined 01:03 astj left 01:04 lookatme joined 01:08 lookatme left, baest left, lookatme joined, baest joined 01:09 astj joined 01:13 Zoffix joined
Zoffix Any idea why when I run this program it always prints "Killing" three times, but frequently it doesn't end, suggesting the murder did not succeed? gist.github.com/zoffixznet/79543ee...p6-L37-L41 01:14
As you can see, I even tried sending three different kill signals, to no avail
And indeed `ps ax` shows the procs unkilled, still running 01:16
01:17 cyphase left
kurahaupo Zoffix: do you own the processes in question? 01:18
Zoffix kurahaupo: they're started by that very script
lookatme Use SIGKILL ?
kurahaupo check for error returns?
lookatme check $! ? 01:19
Zoffix Error returns from where?
lookatme: at which point?
And yeah, I'm already sending SIGKILL
MasterDuke if you add signal handlers to the code you're running in the Proc, do they fire?
kurahaupo error returns from the kills 01:20
Zoffix kurahaupo: it just returns the signal number
MasterDuke run under strace and see what that shows happening?
01:21 cyphase joined
Zoffix adding `signal(SIGINT).tap: {say "got a signal"};` to the spawned procs and then sending $proc.kill: SIGINT don't print that message :/ 01:23
I don't know how to use strace.
kurahaupo strace -o newtracefile program args ... 01:24
MasterDuke you'll want -f also
Juerd github.com/Juerd/shalog/blob/master/tracking.p6 # Beginning of something usable
kurahaupo but for this you'll want -f or -ff
MasterDuke: snap
01:25 astj_ joined
Zoffix Hm. The code's `say "Killing!"; say $proc.kill: SIGSEGV;` and it always prints 3 killing but only 1-3 "11"s 01:25
Juerd At some point I want to spread my code over several files. Are there any best practices regarding putting multiple classes in a single file? 01:26
Zoffix Juerd: no
Juerd Great, at least I can stop looking for them then :)
kurahaupo Zoffix: if you just want something to stop normally, SIGTERM should be your first choice. SIGHUP is typically caught by daemons and used to trigger reloading 01:27
not that you've made the relevant arrangements, but that's the convention
01:27 astj left
MasterDuke you could add an `fprintf(stderr, "Kilroy was here")` to src/io/procops.c:1073 and see if the moarvm op is getting called 3 times 01:28
kurahaupo for interactive processes SIGHUP indicates that its terminal has disconnected -- the "hup" or "hangup" refers to a phone line with a dial-up modem 01:30
Zoffix Here's the strace. What am I looking for gist.githubusercontent.com/zoffixz...tfile1.txt ? 01:31
here's the program that I used to generate it: gist.github.com/zoffixznet/c737b79...75d9b5339f
01:37 BenGoldberg joined
Zoffix Golfed down a bit: await ^2 .map: {start { with Proc::Async.new: $*EXECUTABLE, "-e", "sleep" -> $p { Promise.in(2).then: {say "Killing"; $p.kill: SIGTERM}; await $p.start } }} 01:38
What's weird is it doesn't appear to happen if that `^2` is a `^1`
Zoffix files in hopes jnthn++ would know a fix 01:39
01:40 lookatme left
Zoffix .ask jnthn any idea what may be causing a Proc::Async.kill to fail killing when there's more than one Proc::Async involved? This is sort of blocking my ecosystem toaster script :( would be sweet if it could be fixed: rt.perl.org/Ticket/Display.html?id=131479 01:43
yoleaux Zoffix: I'll pass your message to jnthn.
Zoffix Man, need to learn C so I could fix stuff like this on my own :*
pilne c is dark and full of terrors... but at least is isn't c++ (: 01:45
01:46 ilbot3 left
pilne one bad thing about perl6 is that after dipping my toes in it... trying to learn anything else for a change of career (i don't do anything programming related at the moment), feels pooptastic 01:46
MasterDuke i added a print to that moar function. i see two "Killing"s, but only one print from moar
01:48 ilbot3 joined, ChanServ sets mode: +v ilbot3 01:50 cpage_ left 01:52 lookatme joined 01:53 Cabanossi left 01:55 Cabanossi joined, kurahaupo_ joined 01:56 aborazmeh joined, aborazmeh left, aborazmeh joined 01:57 kurahaupo left 02:00 cpage_ joined
Zoffix pilne: yeah, that's why we say Perl 6 is the last language you'll ever learn :) 02:00
Zoffix goes into hybernation mode
I see I first suspected .kill was busted last September: github.com/perl6/roast/issues/158 02:01
lookatme I send SIGTERM can killed one of **Promise** which tag with "bar" 02:02
Zoffix & 02:03
lookatme @_@
02:04 kurahaupo_ left
pilne i just have to keep hoping i'll be in the right place at the right time to jump on something looking to try "anything" as a coding backend >.< but there is a lot of buzzword hype and shenanigans locally. 02:05
but just barely getting to know perl6, it feels fantastic for "big data" and even machine learning/AI stuff. 02:06
02:11 jcallen_ is now known as jcallen 02:13 Herby_ left
curt_ performance isn't quite there yet, but NativeCall is my best friend. I love how trivial it is to drop into C. 02:16
Juerd "The following packages were stubbed but not defined:", but if I remove the "..." it says about the definition later in the file: "Redeclaration of symbol 'Nowhere'" 02:18
Any ideas?
(Nowhere is the name of the class)
lookatme Is machine learning/AI need performance ?
Juerd, can you paste code to somewhere 02:19
Juerd lookatme: gist.github.com/Juerd/27947ac0e756...abf69011fd 02:20
02:20 zhongjin616 joined
Juerd It worked when I didn't have it in a .pm file 02:20
curt_ I'm not doing machine learning/AI, my 'big data' is just volume.
lookatme Juerd, I comment on the gist 02:27
pilne for research purposes, technically, machine learning and AI don't need speed, they need "flexibility", but speed is always nice in that area (: 02:28
02:29 raschipi joined
TEttinger pilne: there's also bearlibterminal if you don't need some of the stuff libtcod provides. it's probably much easier to call from a not-C language 02:29
lookatme The document about that forward declared not very clear
TEttinger (libtcod defines their own kinds of list and such) 02:30
foo.wyrd.name/en:bearlibterminal
pilne ahhhh, TEttinger, i was mainly interested in libtcod for it's terminal stuffs, tyvm
TEttinger BLT's terminal is definitely better in a lot of ways
pilne and it reminds me of a good meal
so it's like... doubleplusgood 02:31
TEttinger it allows movement of each char outside its position, plus layers
and works with TTF fonts
pilne project tcodperl: killed before germination
TEttinger BLT only does terminal output, since most games want their own map gen anyway 02:32
pilne this will allow me to just write all the "roguelike" stuff (ai, etc.) in perl6, and use that for terminal.
TEttinger yep
roguebasin has good resources
pilne tyvm, no need to lug around the horse trailer full of horses i ain't using
yeah, it's where i get most of my stuff, not sure how i missed that
TEttinger I don't know if BLT has much presence there. libtcod had a monopoly on full-color terminal-likes for many years 02:33
but it stopped mac support and various other oddities before finally picking up some pieces
www.roguebasin.com/index.php?title=Dijkstra
pilne yeah, and it was kinda "more" than i wanted from a library, even when using it in a tutorial for python a longtime ago.
TEttinger (useful but easy to miss AI technique) 02:34
pilne i actually read a comparison of fill, dijkstra, and a* recently
02:34 xiaoyafeng joined
TEttinger I work on a very "more" library for roguelikes but it's in Java 02:34
02:34 aborazmeh left
TEttinger yeah, it's crazy the differences 02:34
xiaoyafeng hello, I have a question. 02:35
pilne i just have a hard time feeling comfy as a one-man-dev in javaland, everything just feels so "big"
TEttinger currently in my benchmarks, I have A* beating everything hand-down for short simple paths, but surprisingly a breadth-first search with no heuristic can beat A* on long and moderately convoluted paths 02:36
xiaoyafeng what's different between ".".io.dir.grep: {d} and ".".io.dir.grep({d});
please help
pilne TEttinger: a* "thinks too much" on the long and convoluted paths i've noticed.
TEttinger hi xiaoyafeng. there's some experts around recently, but I am not one of them 02:37
pilne likewise, my perl-fu is weak like cooked spaghetti
xiaoyafeng :(....
02:37 Cabanossi left
TEttinger (I'm rather sure Zoffix is an expert, since I've seen a lot of commits tied to the name) 02:38
pilne m: ".".io.dir.grep: {d}
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
d used at line 1
lookatme xiaoyafeng, no different. 02:39
curt_ xiaoyafeng: subroutine(...) and subroutine: ... are just two ways of calling the sub
pilne it might just be alternate syntax, have you examined them?
TEttinger also you might not have access to current dir in the bot
xiaoyafeng ".".io.dir.grep: {.d}
sorry
raschipi m: ".".io.dir.grep: {.d}
camelia No such method 'io' for invocant of type 'Str'
in block <unit> at <tmp> line 1
lookatme m: say ".".IO.dir.grep:{.d};
camelia (".cpanm".IO ".local".IO ".npm".IO ".perl6".IO ".perlbrew".IO ".rcc".IO ".ssh".IO "Perlito".IO "evalbot".IO "log".IO "nqp-js".IO "p1".IO "p2".IO "perl5".IO "std".IO ".cpan".IO "rakudo-m-1".IO "rakudo-m-2".IO "rakudo-j-inst-1".IO "rakudo-j-1".IO "rakudo-j…
xiaoyafeng IO
lookatme m: say ".".IO.dir.grep({.d}); 02:40
camelia (".cpanm".IO ".local".IO ".npm".IO ".perl6".IO ".perlbrew".IO ".rcc".IO ".ssh".IO "Perlito".IO "evalbot".IO "log".IO "nqp-js".IO "p1".IO "p2".IO "perl5".IO "std".IO ".cpan".IO "rakudo-m-1".IO "rakudo-m-2".IO "rakudo-j-inst-1".IO "rakudo-j-1".IO "rakudo-j…
raschipi It's just two ways of saying the same thing.
02:40 Cabanossi joined
TEttinger oh, one was {.d} and one was {d} 02:40
pilne god i love this bot
TEttinger it's a pretty great bot
lookatme It just two ways of function/sub call, you can chose your prefer style
xiaoyafeng so : and () is identical
raschipi What is different yould be ".".IO.dir.grep ({.d}); The space makes a difference. 02:41
m: ".".IO.dir.grep ({.d}); 02:42
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3".".IO.dir.grep7⏏5 ({.d});
expecting any of:
infix
infix stopper
statement end
statement modifier
st…
lookatme m: say ".".IO.dir.grep : {.d};
camelia Cannot resolve caller grep(Seq: ); none of these signatures match:
($: Bool:D $t, *%_)
($: Mu $t, *%_)
in block <unit> at <tmp> line 1
raschipi m: say ".".IO.dir.grep\ ({.d});
camelia (".cpanm".IO ".local".IO ".npm".IO ".perl6".IO ".perlbrew".IO ".rcc".IO ".ssh".IO "Perlito".IO "evalbot".IO "log".IO "nqp-js".IO "p1".IO "p2".IO "perl5".IO "std".IO ".cpan".IO "rakudo-m-1".IO "rakudo-m-2".IO "rakudo-j-inst-1".IO "rakudo-j-1".IO "rakudo-j…
xiaoyafeng well it's not allowed space. between sub and args 02:43
raschipi m: say ".".IO.dir.grep\ ␤ ({.d});
camelia (".cpanm".IO ".local".IO ".npm".IO ".perl6".IO ".perlbrew".IO ".rcc".IO ".ssh".IO "Perlito".IO "evalbot".IO "log".IO "nqp-js".IO "p1".IO "p2".IO "perl5".IO "std".IO ".cpan".IO "rakudo-m-1".IO "rakudo-m-2".IO "rakudo-j-inst-1".IO "rakudo-j-1".IO "rakudo-j…
raschipi You can even split into multiple lines, as long as you put a \ there. Otherwise it means something else. 02:44
xiaoyafeng but why ".".IO.dir.grep: ({.d}) is allowed too?
02:44 curt_ left
raschipi That is an alternative format. 02:45
xiaoyafeng ok, now it's 3 ways to call a sub
raschipi Gotta have at least three ways of doing something in Perl6.
xiaoyafeng TIMTTWTDI 02:47
;)
lookatme m: say ".".IO.dir.grep {.d};
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected block in infix position (missing statement control word before the expression?)
at <tmp>:1
------> 3say ".".IO.dir.grep7⏏5 {.d};
expecting any of:
infix
infix…		 02:48
lookatme m: say grep {.d}, ".".IO.dir;
camelia (".cpanm".IO ".local".IO ".npm".IO ".perl6".IO ".perlbrew".IO ".rcc".IO ".ssh".IO "Perlito".IO "evalbot".IO "log".IO "nqp-js".IO "p1".IO "p2".IO "perl5".IO "std".IO ".cpan".IO "rakudo-m-1".IO "rakudo-m-2".IO "rakudo-j-inst-1".IO "rakudo-j-1".IO "rakudo-j…
02:49 noganex joined
lookatme m: say ({.d}).WHAT 02:51
camelia (Block)
lookatme xiaoyafeng, ({.d}) also a callable Block
02:51 noganex_ left
lookatme like {.d} does 02:52
xiaoyafeng yeah, it makes sense
02:56 aborazmeh joined, aborazmeh left, aborazmeh joined
raschipi m: lc('Y','N'); lc 'Y','N' 03:02
camelia 5===SORRY!5===
Calling lc(Str, Str) will never work with declared signature (Cool $s)
at <tmp>:1
------> 3<BOL>7⏏5lc('Y','N'); lc 'Y','N'
Calling lc(Str, Str) will never work with declared signature (Cool $s)
at <tmp>:1
------> 3lc…
raschipi m: lc ('Y','N'); lc 'Y','N'
camelia 5===SORRY!5=== Error while compiling <tmp>
Calling lc(Str, Str) will never work with declared signature (Cool $s)
at <tmp>:1
------> 3lc ('Y','N'); 7⏏5lc 'Y','N'
lookatme m: say lc('Y', 'N'); 03:03
camelia 5===SORRY!5=== Error while compiling <tmp>
Calling lc(Str, Str) will never work with declared signature (Cool $s)
at <tmp>:1
------> 3say 7⏏5lc('Y', 'N');
lookatme m: say lc('Y');
camelia y
raschipi bisectable6: say lc ('Y','N'); 03:04
bisectable6 raschipi, On both starting points (old=2015.12 new=b667e81) the exit code is 0 and the output is identical as well
raschipi, Output on both points: «y n»
lookatme m: say lc ('Y', 'N');
camelia y n
lookatme m: say &lc.WHAT;
camelia (Sub)
lookatme m: say &lc('Y', 'N'); 03:05
camelia Too many positionals passed; expected 1 argument but got 2
in block <unit> at <tmp> line 1
raschipi I see, in the second, it was the version without parens that it didn't like.
03:06 Cabanossi left
lookatme m: say &lc.signature.perl; 03:06
camelia :(Cool $s)
lookatme m: say &lc(('Y', 'N')); 03:07
camelia y n
lookatme m: say ('Y', 'N') ~~ Cool; 03:08
camelia True
lookatme Oh.. @_@
03:08 Cabanossi joined 03:11 simonm left
raschipi m: say lc( ('Y','N') ); 03:12
camelia y n
lookatme m: say &lc(('Y', 'N')).WHAT; 03:13
camelia (Str)
03:14 khw left 03:28 aborazmeh left 03:41 simonm joined 03:49 Xliff joined
Xliff \o 03:49
Is there a way to make a class that is only available to a specific scope?
m: class A { my class B { method bb { say "bb"; } }; method new( B.bb; }; }; A.new; 03:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed parameter
at <tmp>:1
------> 3method bb { say "bb"; } }; method new( B7⏏5.bb; }; }; A.new;
expecting any of:
constraint
formal parameter
Xliff m: class A { my class B { method bb { say "bb"; } }; method new( B.new.bb; }; }; A.new;
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed parameter
at <tmp>:1
------> 3method bb { say "bb"; } }; method new( B7⏏5.new.bb; }; }; A.new;
expecting any of:
constraint
formal parameter
Xliff m: class A { my class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new;
camelia bb
Xliff m: class A { my class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; A::B.new.bb;
camelia bb
Could not find symbol '&B'
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
Xliff Hmm... that should work.
03:52 Cabanossi left 03:53 Cabanossi joined
raschipi m: class A { our class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; A::B.new.bb; 03:54
camelia bb
bb
raschipi Xliff: ^^ 03:55
m: class A { our class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; B.new.bb; 04:02
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
B used at line 1
raschipi m: class A { our class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; my B.new.bb; 04:03
camelia 5===SORRY!5===
Type 'B' is not declared
at <tmp>:1
------> 3} }; method new{ B.bb; }; }; A.new; my B7⏏5.new.bb;
Malformed my
at <tmp>:1
------> 3; } }; method new{ B.bb; }; }; A.new; my7⏏5 B.new.bb;

»
Xliff raschipi: Why "our" instead of "my"? 04:07
raschipi They say where the symbol will be visible. 'my' means it's visible only inside the block, 'our' means it will be visible outside. 04:08
Xliff But if I only want B visible to methods in A, isn't that the same? 04:09
raschipi If you want B visible only inside, use my. If you want it visible outside, use our. 04:10
Xliff m: class A { my class B { method bb { say "bb"; } }; method new{ B.bb; }; method aa { say "aa" B.bb; }; A.new; A::B.new.bb; 04:11
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3ethod new{ B.bb; }; method aa { say "aa"7⏏5 B.bb; }; A.new; A::B.new.bb;
expecting any of:
infix
infix stopper
postfix…
Xliff m: class A { my class B { method bb { say "bb"; } }; method new{ B.bb; }; method aa { say "aa"; B.bb; }; A.new; A::B.new.bb;
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3{ say "aa"; B.bb; }; A.new; A::B.new.bb;7⏏5<EOL>
Xliff m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; method aa { say "aa"; B.bb; }; }; A.new; A::B.new.bb;
camelia bb
Could not find symbol '&B'
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
raschipi m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; method aa { say "aa"; B.bb; }; }; A.new; A.new.aa; 04:12
camelia bb
bb
No such method 'aa' for invocant of type 'Bool'
in block <unit> at <tmp> line 1
raschipi m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; method aa { say "aa"; B.bb; }; }; A.new; A.aa;
camelia bb
aa
bb
Xliff m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; method aa { say "aa"; B.bb; }; }; A.new.aa;
camelia bb
No such method 'aa' for invocant of type 'Bool'
in block <unit> at <tmp> line 1
Xliff m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; }; method aa { say "aa"; B.bb; }; A.new.aa;
camelia 5===SORRY!5===
Undeclared name:
B used at line 1

Other potential difficulties:
Useless declaration of a has-scoped method in mainline (did you mean 'my method aa'?)
at <tmp>:1
------> 3b"; } }; method new { B.bb; }; };…
Xliff m: class A { my class B { method bb { say "bb"; } }; method new { B.bb; }; }; method aa { say "aa"; B.bb; }; }; A.new.aa; 04:13
camelia 5===SORRY!5===
Unexpected closing bracket
at <tmp>:1
------> 3b; }; }; method aa { say "aa"; B.bb; }; 7⏏5}; A.new.aa;
Other potential difficulties:
Useless declaration of a has-scoped method in mainline (did you mean 'my method aa…
Xliff OK, this needs an editor.
04:14 geekosaur left 04:15 geekosaur joined
Xliff m: class A { 04:16
my class B {
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3class A {7⏏5<EOL>
Xliff method new { B.bb; };
method bb { say "bb"; };
};
method aa { say "aa"; B.bb; };
};
A.new.aa;
Feck.
class A { my class B { method new { B.bb; }; method bb { say "bb"; }; }; method aa { say "aa"; B.bb; }; }; A.new.aa; 04:17
raschipi Use a pastie
Xliff m: class A { my class B { method new { B.bb; }; method bb { say "bb"; }; }; method aa { say "aa"; B.bb; }; }; A.new.aa;
camelia aa
bb
Xliff That would have been my next step if that hadn't worked.
m: class A { my class B { method new { B.bb; }; method bb { say "bb"; }; }; method aa { say "aa"; B.bb; }; }; A.new.aa; A.B.new;
camelia aa
bb
No such method 'B' for invocant of type 'A'
in block <unit> at <tmp> line 1
Xliff m: class A { my class B { method new { B.bb; }; method bb { say "bb"; }; }; method aa { say "aa"; B.bb; }; }; A.new.aa; A::B.new;
camelia aa
bb
Could not find symbol '&B'
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
raschipi do a /msg camelia 04:18
Xliff OK, so I don't really see a noticible difference between using "my" and "our"
04:19 xtreak_ joined
raschipi m: class A { our class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; A::B.new.bb; 04:19
camelia bb
bb
raschipi m: class A { my class B { method bb { say "bb"; } }; method new{ B.bb; }; }; A.new; A::B.new.bb;
camelia bb
Could not find symbol '&B'
in block <unit> at <tmp> line 1

Actually thrown at:
in block <unit> at <tmp> line 1
raschipi The second one doesn't work because B doesn't exist outside of A 04:20
Xliff Heh.
Which is exactly what I want.
I don't want B to work outside of A
raschipi So use it. What's the problem?
Xliff No problem. I just wanted to know if I was doing it the right way. 04:21
You suggested "our" and I was confused if I had missed something.
No worries.
Thanks for the input, though. 04:22
raschipi I see. Yeah, using 'my' is the way to go. If you want something else, there are other options. 04:23
04:23 Actualeyes left
Xliff 'k 04:26
04:28 pilne left 04:36 Cabanossi left 04:39 Cabanossi joined 04:40 xtreak_ left, xtreak joined 04:41 fatguy joined 04:43 raschipi left 04:45 aborazmeh joined, aborazmeh left, aborazmeh joined 04:47 aindilis joined
Xliff m: class A { my class B { method new { B.bb; }; method bb { say "bb"; A.aa;}; }; method aa { say "aa"; B.bb; }; }; 04:56
camelia ( no output )
04:56 xtreak left
Xliff m: class A { my class B { method new { B.bb; }; method bb { say "bb"; A.aa;}; }; method aa { say "aa"; B.bb; }; }; A.new.aa 04:56
camelia (timeout)aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
bb
aa
…
Xliff m: class A { my class B { method new { B.bb; }; method bb { say "bb"; A.aa}; }; method aa { B.bb }; }; A.new.aa 04:57
camelia (timeout)bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
bb
…
Xliff m: class A { my class B { method bb { say "bb"; A.aa }; }; method a { say B.bb }; method aa { say "aa" }; }; A.new.a 04:59
camelia bb
aa
True
Xliff m: class A { my class B { method bb { say "bb"; A.aa }; }; method a { B.bb }; method aa { say "aa" }; }; A.new.a
camelia bb
aa
Xliff Rawk. 05:00
05:00 finanalyst joined 05:04 astj_ left 05:07 astj joined, Cabanossi left 05:08 Cabanossi joined 05:10 pegop left
finanalyst can someone tell me what had happened to perl6-debug? I have just rerun 'rakudobrew build moar', then 'perl6-debug' and got 'perl6-debug: command not found' 05:10
05:11 astj left 05:14 troys_ is now known as troys 05:19 BenGoldberg left 05:25 aborazmeh left 05:30 ufobat joined 05:33 domidumont joined 05:36 domidumont left
zengargoyle finanalyst: i think it's broken or going away or being changed or something. you probably have a `perl6-debug-m` for the moar build but that gives some other error. 05:52
finanalyst zengargoyle: if it's going away, it would be nice to know. 05:53
05:53 MasterDuke left
zengargoyle quotable6: perl6-debug 05:53
quotable6 zengargoyle, gist.github.com/3ef7cdc97a78f145bf...78c134949d
finanalyst If being changed, when will it be back?
05:53 ufobat left
finanalyst perl6-debug-m does not work at all - gives an error for all input 05:54
zengargoyle gues that doesn't do what i thought. :) too many bots to keep track of. i remember seeing a bit of discussion about perl6-debug in the past few days here in IRC.
05:55 Mithaldu left, Mithaldu joined
finanalyst how to find discussion? 05:55
zengargoyle if you search the IRC logs you'll probably find the answer. it was just a few days or so ago.
irclog.perlgeek.de/perl6/2017-06-02 05:57
zengargoyle thinks theres a bot that can search the logs but i don't know which one or how it works. :/
05:58 cpage_ left
zengargoyle irclog.perlgeek.de/perl6/search/?n...erl6-debug 05:58
05:59 zhongjin616 left
Xliff Does FALLBACK take positional args? 06:00
Oh, nevermind. 06:02
It can.
zengargoyle finanalyst: i think perl6-debug vs perl6-debug-m is a little implementation detail of rakudobrew vs rakudo-star or whatever. not sure if the brokeness is terminal or just a bit of "it hasn't caught up yet". :) but you're not alone, it doesn't work. 06:03
06:04 troys left
finanalyst zengargoyle: just looking through the last few logs 06:04
basically debug & Grammar:: Debug stopped working at the last update 06:05
06:05 domidumont joined 06:06 astj joined
zengargoyle finanalyst: cool. BTW i'm probably going to steal your Task::Popular as a template to make my own little set of favorite modules because i'm too lazy to track down the old Task::Star. :P 06:08
finanalyst zengargoyle: :)
06:09 stmuk_ joined
finanalyst zengargoyle: By the way, I created a local directory on my computer, put a META6.json file in it with some modules I like to have, and then used zef on the directory URL 06:09
It nicely installed all the modules 06:10
list the required modules in the depends list
zengargoyle but i wonder if having an actual Task::Popular module that has nothing but a bare actual module with a version so it can somehow be tested against is more cromulent than just a Null install with dependencies....
like the p5 'perl -MTask::Whatever\ 99999' quick check of is it installed or not??? 06:11
06:12 stmuk left
finanalyst zengargoyle: not sure what you mean about "testing against" a distribution 06:13
zengargoyle package manager satisfying dependencies is a bit different than being able to 'use Task::Whatever' and then check Task::Whatever's version inside the code. 06:14
it's sortof a meta thing that's probably not too important anyways... :) 06:15
06:16 nadim_ joined
lookatme perl6 -MTask::Whatever ? 06:16
zengargoyle `perl -MGetopt::Long\ 9999 -e1` -> Getopt::Long version 9999 required--this is only version 2.48 at /usr/share/perl/5.24/Getopt/Long.pm line 1530. 06:18
if Task::Popular (or Task::Whatever) doesn't actuall install a module named Task::Popular with a version, then as far as perl6 is concerned it doesn't exist. 06:19
only the package manager can tell by checking wheter all dependencies are installes. 06:20
if there's an actual Task::Poplular module that does nothing but is installed then the check is quick and easy. require Task::Popular. or query the version installed. 06:21
06:22 wamba joined
zengargoyle lookatme: doing the -MSome::Module\ 9999999 is just a really quick and easy way to see if a module is installed and which version is actually installed. you ask perl to load a reallly high version and it fails and tells you which version it actually has or that there's no such module installed. 06:25
lookatme zengargoyle, I see. I dont' know much about module version control. 06:27
zengargoyle IMHO, it's best practice for every module/package/bundle/whatever to have an actual module that gets installed (even if it doesn't do anything) just so you can tell that it's there and what version it is.
but i may be alone in this thinking :) 06:29
lookatme Is there a way use specify version module in command `perl6 -M..` or in Perl 6 script..
zengargoyle i'm not really sure about p6. but that's pretty much exactly what's going on in p5. in a script 'use Foo 4;' is the command line 'perl "-MFoo 4" (because spaces). 06:32
06:33 darutoko joined 06:35 kyan left 06:36 lowbro joined
zengargoyle but the point is can you tell that Task::Whatever is installed or not? if Task::Whatever actually installs a Task::Whatever module then you can tell. if Task::Whatever is just a list of dependencies and doesn't install an actual Task::Whatever module, then you can't check that it has been installed except by checking all of the dependencies. 06:36
06:36 lowbro left, lowbro joined
zengargoyle one way is easy, one is a PITA. :P 06:38
lookatme zengargoyle, There are something about module version in document. Such as `module mymodule:ver<0.0.1> { }`, but I dont' think it's exactly way manager module version. 06:39
I gotta do a test about that. 06:41
docs.perl6.org/language/modules#Ex..._Importing
zengargoyle i don't think i've seen many command line (or scripts) that actually use the version information.... github.com/dnmfarrell/Perl6-One-Liners this page has a few uses of -M but not that many. 06:42
lookatme Use that module would be like that `perl6 -MSOme::Module::ver<0.0.1>` 06:43
I think they add the `-M` option recently 06:44
The command **perl6** not have that option before
zengargoyle i don't think i've ever used a version in a -M other than checking whether module was installed or not.... 06:46
one-liner's are usually using what you know work. :) 06:47
fatguy anyone using IO-Notification-Recursive ? i need to catch file/dir event (create/modify/delete), how can i do that ? 06:50
06:50 parv joined
zengargoyle `perl6 '-MTask::Popular' -e ''` --> No output 06:58
`perl6 '-MTask::Popular' -e ''` --> ===SORRY!=== Could not find Task::Popular at line 1 in: ...... 06:59
gah, `perl6 '-MPSGI' -e ''` --> No output. 07:00
07:02 cpage_ joined 07:04 astj left
parv zengargoyle, what is the difference in commands with Task::Popular so that you get the output only on the second time? 07:08
lookatme Zef can install specify version module like this `zef install Foo:ver<1.0.0>` 07:09
And I found it's work when I use `use Foo:ver<1.0.0>` in script file, but not in `-M` option. 07:12
zengargoyle lookatme: the first time was a cut-n-paste error.... -MPSGI shows nothing -MTask::Popular returns error (even though Task::Popular is installed) because Task::Popular doen't actually install a modules named Task::Popular, it just has a list of dependencies that must be met (which `zef install Task::Popular` will happily do. 07:13
Voldenet fatguy: start with IO::Notification class, it has .watch-path command and a good example on consuming the supply :)
zengargoyle lookatme: me too. i haven't been able to make -MFoo:ver<x.y>' work after trying many variants.... 07:14
Voldenet watch-recursive pretty much does the same, except it sets up watches for the tree and merges the supplies 07:15
fatguy Voldenet: where can i find the example and types of event doc ? 07:17
zengargoyle parv: oops, that cut-n-paste error reply was to your question... 07:18
Voldenet hm, docs.perl6.org/type/IO::Notification 07:19
parv zengargoyle, don't worry; i read that.
zengargoyle :)
fatguy Voldenet: i don't understand what is FileRenamed event classification
i did look at p6doc 07:20
07:20 wamba left
Voldenet Hmm, It's not very well-documented 07:20
github.com/rakudo/rakudo/blob/c194...ication.pm
but at least the underlying enums and classes are short ;)
07:21 nadim_ left
Voldenet apparently "event" is FileRenamed or FileChanged, no idea why inotify events like IN_CREATE are not described 07:22
07:26 skids left
fatguy Voldenet: thanks ! i will use p5 inotify2 for my case 07:27
07:33 lookatme left, Sgeo_ joined, lookatme joined 07:35 Sgeo left 07:36 Cabanossi left 07:38 rindolf joined, Cabanossi joined 07:47 dct joined 07:48 astj joined 07:49 astj_ joined, astj left 07:50 astj_ left, finanalyst left, astj joined 07:51 geekosaur left 07:52 xtreak joined, geekosaur joined
fatguy how can i print current line number and namespaces ? 07:52
07:53 astj_ joined 07:56 astj left, xtreak left
lookatme Do you mean $?LINE and $?PACKAGE 07:57
07:57 salv0 joined
lookatme fatguy, it's package, there no namespace. 07:58
08:02 Sgeo__ joined
fatguy lookatme: yes, how about printing the method ? the $?PACKAGE print the class name, i need to print module name then class , then method for debugging msg purpose 08:03
08:03 xtreak joined 08:04 Sgeo_ left, dwarring left 08:05 zapwai left
lookatme fatguy, there are many compile-time variable about that: docs.perl6.org/language/variables#..._variables 08:07
The method is &?BLOCK 08:08
m: class A { method f-print-name( ){ say &?BLOCK; } }; A.new.f-print-name;
camelia f-print-name
fatguy lookatme: thanks! 08:14
08:14 wamba joined
lookatme :) 08:15
08:15 xtreak left, AlexDaniel joined 08:16 xtreak joined 08:18 nadim_ joined
AlexDaniel quotable6: ‘perl6-debug’ 08:24
quotable6 AlexDaniel, gist.github.com/f23c1cb3afddef7ac0...d8142244ce 08:25
08:29 xiaoyafeng left 08:32 zakharyas joined 08:35 Cabanossi left 08:38 Cabanossi joined 08:40 Ven joined, Ven is now known as Guest35264 08:44 zakharyas left 08:45 zakharyas joined 08:47 parv left 08:55 AlexDaniel left 09:01 xtreak_ joined 09:02 xtreak left
rightfold m: (my $foo) = (); 09:04
camelia ( no output )
09:04 Guest35264 left, sivoais left 09:05 vimal2012 joined, sivoais joined 09:06 xtreak_ left 09:07 xtreak joined 09:09 xtreak left 09:10 xtreak joined
vimal2012 <:Lt> does not match anything, but <:Lu> matches 'T', 'V', 'M'. Aren't the letters 'T', 'V', 'M' titlecase letters in the following string? 09:10
p6: "Trump Visits Mexico" ~~ m:g/<:Lt>/
camelia ( no output )
vimal2012 p6: "Trump Visits Mexico" ~~ m:g/<:Lu>/
camelia ( no output )
vimal2012 p6: say "Trump Visits Mexico" ~~ m:g/<:Lu>/
camelia (｢T｣ ｢V｣ ｢M｣)
09:10 Ven joined
vimal2012 p6: say "Trump Visits Mexico" ~~ m:g/<:Lt>/ 09:11
camelia ()
09:11 Ven is now known as Guest39608
vimal2012 None of the 26 Uppercase letters in English match <:Lt>. It looks like English doesn't have titlecase letters? Then what other languages have titlecase letters? 09:15
moritz vimal2012: there's a greek title case latter, iirc 09:16
lookatme m: say "Trump Visits Mexico" ~~ m:g/<:Lt>/; 09:18
camelia ()
moritz m: say (1..0x10FFFF).grep({ .chr ~~ /<:Lt>}).map(&uniname)
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse regex; couldn't find final '/'
at <tmp>:1
------> 3say (1..0x10FFFF).grep({ .chr ~~ /<:Lt>7⏏5}).map(&uniname)
expecting any of:
infix stopper
term
moritz m: say (1..0x10FFFF).grep({ .chr ~~ /<:Lt>/}).map(&uniname)
camelia (LATIN CAPITAL LETTER D WITH SMALL LETTER Z WITH CARON LATIN CAPITAL LETTER L WITH SMALL LETTER J LATIN CAPITAL LETTER N WITH SMALL LETTER J LATIN CAPITAL LETTER D WITH SMALL LETTER Z GREEK CAPITAL LETTER ALPHA WITH PSILI AND PROSGEGRAMMENI GREEK CAPITAL… 09:19
lookatme m: say "Trump Visits Mexico" ~~ /<:Lt>/;
camelia Nil
09:19 Guest39608 left
rightfold m: say "Dž" ~~ /<:Lt>/; 09:19
camelia ｢Dž｣
rightfold Seems to be those ligatures where the first letter is uppercase and the second one is lowercase 09:20
moritz it's 4 latin and 27 greek capital letters that match <:Lt>
lookatme m: say "a Trump Visits Mexico" ~~ m:g/<:Lt>/;
camelia ()
09:22 Cabanossi left 09:23 TEttinger left, Cabanossi joined
rightfold There's no ASCII in Lt 09:23
09:25 Ven_ joined 09:41 finanalyst joined 09:44 robertle joined 09:45 mr-foobar left 09:50 mr-foobar joined 09:52 fatguy left 09:55 Ven_ left 10:03 ufobat joined 10:06 vimal2012 left, vimal2012 joined
Zoffix finanalyst: perl6-debug just bittrotted. It's not going away. Gonna try to fix it before next release (though I recall saying the same like 3 releases ago :P) 10:06
10:08 lookatme left
Zoffix xiaoyafeng, no difference, as both are compile time errors unless you augmented Strs to have .io and implemented `d` routine or constant :) There's, however, no difference between '.'.IO.dir.grep: {.d} and '.'IO.dir.grep({.d}); the `:` is just an alternative way to write `()` in this case, except you can't follow it with another method call in the same chain. 10:09
TEttinger, nah, I'm a total n00b. Started with Perl 6 like a week ago. 10:11
And it's hrd
10:11 Zoffix left 10:21 ChoHag left, Cabanossi left 10:23 Cabanossi joined 10:36 astj_ left 11:00 sammers left, aborazmeh joined, aborazmeh left, aborazmeh joined 11:06 Cabanossi left 11:08 Cabanossi joined
tbrowder .tell dwarring i installed latest version and both courier and times-roman show True for IsFixedPitch. i'm not positive about your methods but i think the default value of IsFixedPitch should be False, and the Adobe spec says the absence of it is False. 11:16
yoleaux tbrowder: I'll pass your message to dwarring.
11:23 grondilu left
tbrowder .tell dwarring the is-deeply test on the metric may be giving an unexpected result 11:24
11:27 labster left 11:34 wamba left, xtreak left, cpage joined 11:43 curt_ joined 11:46 vimal2012 left 11:59 pmurias joined 12:01 aborazmeh left 12:04 salv0 left 12:06 rindolf left 12:09 ufobat left 12:11 pmurias left
tbrowder .tell dwarring you need to actually test the module with a Font::AFM object, then the IsFixedPitch test fails for Courier using ok (expected True) 12:12
yoleaux tbrowder: I'll pass your message to dwarring.
12:14 finanalyst left
Voldenet m: say "Zażółć gęślą jaźń λ" ~~ m:g/<:Lt>/; 12:14
camelia ()
Voldenet m: say "Λ" ~~ m:g/<:Lt>/; 12:15
camelia ()
12:17 pmurias joined 12:18 ufobat joined 12:19 mcmillhj joined 12:28 pmurias left, pmurias joined 12:32 simonm left 12:34 pmurias left, pmurias joined, simonm joined 12:39 pmurias left 12:41 pmurias joined 12:42 pilne joined
nebuchadnezzar hello 12:43
tbrowder .tell dwarring i have submitted a PR for new tests that show the problem 12:54
yoleaux tbrowder: I'll pass your message to dwarring.
nebuchadnezzar I have a NFG character and I want the list of code points of the NFD form 12:55
timotimo m: say "ö".NFD.perl 12:56
camelia Uni.new(0x006f, 0x0308).NFD
timotimo m: say "ö".NFD.list.perl
camelia (111, 776).Seq
nebuchadnezzar thanks 12:57
13:08 aborazmeh joined, aborazmeh left, aborazmeh joined 13:09 pmurias left 13:23 nadim_ left 13:24 Zoffix joined
Zoffix When I Lock.protect something, does it matter where the lock is instantiated? I see the docs list it as `my $l = Lock.new; $l.protect({ #`( some unsafe operations here ) 1 + 1; });`, but it wonder why `$l` is involved there 13:25
Can it be written as just Lock.new.protect({ #`( some unsafe operations here ) 1 + 1; }); or does the Lock need to be created outside the threaded area?
Aha! My hunch was right. Lock must be created outside the threaded area. 13:29
This spams stuff in wrong order: perl6 -e 'await ^100 .map: { start { Lock.new.protect: { print "Foo"; sleep rand; print "Bar"; sleep rand; say "Ber"; } } }'
While this works right (note how lock is outside the promises): perl6 -e 'my $l = Lock.new; await ^100 .map: { start { $l.protect: { print "Foo"; sleep rand; print "Bar"; sleep rand; say "Ber"; } } }'
[Coke] foo bar ber? heretic! 13:30
as a foo bar baz-ian, I must protest.
Zoffix :) 13:31
jnthn
.oO( foo bar ber-ians at the gate... )
yoleaux 01:43Z <Zoffix> jnthn: any idea what may be causing a Proc::Async.kill to fail killing when there's more than one Proc::Async involved? This is sort of blocking my ecosystem toaster script :( would be sweet if it could be fixed: rt.perl.org/Ticket/Display.html?id=131479
Zoffix jnthn: never mind about that ^. I just found where the bug's at. Race in $*KERNEL.signal
jnthn d'oh 13:32
Zoffix++ # bug hunting
Zoffix \o/
13:38 liweigan joined
Geth doc: f98e28f667 | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Type/Lock.pod6
Expand Lock.protect docs

To explain the lock must be instantiated outside the threaded area		 13:40
13:40 Zoffix left 13:45 fatguy joined 13:47 araraloren joined
Voldenet Could locks work with 6.0d's continuations? Would it be possible to use locks in async context? (schedule execution when lock available) 13:51
13:51 Cabanossi left
Voldenet something like: my $l = AsyncLock.new; await $l.lock(); ... $l.unlock(); 13:52
13:53 mcmillhj left 13:54 Cabanossi joined
Voldenet or even: my $l = AsyncLock.new; await $l.protect({ ...piece of async code }); 13:54
13:54 mcmillhj joined
timotimo ZofBot: if you $l = Lock.new inside of a multi-threaded area, you'll just be creating one lock per thread, which won't do very much :) 13:59
13:59 astj joined
jnthn Voldenet: I've pondered a Lock::Async at some point 14:00
It can be a kinda useful mechanism
14:04 astj left, ChoHag joined 14:13 dct left 14:16 nadim joined 14:21 rindolf joined 14:24 dct joined 14:38 aborazmeh left 14:41 holyghost joined
araraloren evening .o/ 14:45
dudz \o. 14:46
14:52 mls left 14:53 zakame joined 14:54 mls joined 14:55 AnX joined, Herby_ joined
Herby_ o/ 14:56
AnX o/ 14:57
how to start with perl 6 14:58
14:58 AnX left 14:59 AnX joined, amerlyq joined, AlexDaniel joined
Herby_ Anx: here's a new book that was just released on getting started with Perl 6: greenteapress.com/wp/think-perl-6/ 14:59
14:59 khw joined
moritz AnX: do you have previous programming experience with other languages? 15:02
AnX Hi 15:03
yes
i worked sometime in perl 5, little tcl, shell, C/C++ 15:04
thanks for the link Hereby_
moritz then in a few weeks, www.apress.com/us/book/9781484228982 will be out
that targets folks that are already familiar with the basic programming constructs
AnX ok i see .. that would be great 15:05
15:05 ovechkin joined
robertle moritz: is yours really 140 pages as the apress page says? 15:06
moritz robertle: it's 125 pages content in A4 format
robertle: so a bit more in a smaller format
(A4 is pretty big, as far as books go)
15:06 Cabanossi left
robertle ok, thanks 15:07
15:09 Cabanossi joined
Herby_ moritz: can't wait to read it 15:11
15:13 darutoko- joined, darutoko- left
Herby_ question: I'm using zoffix's Twitter module. I used Zef to install it. How do I find the local path of the module on my pc? 15:15
15:15 Zoffix joined
Herby_ and he appears 15:16
Zoffix I have a bot ping me 'cause you mentioned me
Herby_ ahhh
15:16 darutoko left
Zoffix Herby_: why do you need a path? Theoretically there don't have to be one 15:16
Like, what are you trying tod o?
Installed modules get precompiled, so I doubt the path will help you much 15:17
Herby_ I don't really know how to use git. I like your module and would like to try and add some more functionality to it
figured if i could find the local file i could start adding additional GET methods
Zoffix right, module's path will be useless for that
Herby_ :(
Zoffix Just do git clone github.com/zoffixznet/perl6-Twitter Twitter 15:18
And then you'll have the Twitter dir. Go into it and you'll be able to load your local changes when you use `perl6 -Ilib`
or do `perl6 examples/bot.pl6` that script already adds `lib` topaths 15:19
Herby_ ok. does it matter what folder I'm in when I do the: git clone github.com/zoffixznet/perl6-Twitter Twitter command?
Zoffix nope 15:20
Herby_ awesome. I'll give it a shot, thanks
15:21 cdg joined
Zoffix And if you end up with something that you'd want to add to the module, you need to "fork" my repo and then do a PR ("Pull Request"). Appears to be a guide for the process: gist.github.com/Chaser324/ce0505fbed06b947d962 15:22
Zoffix vanishes into the shadows
15:22 Zoffix left, Actualeyes joined 15:25 cdg left, cdg joined 15:30 ovechkin left, liweigan left 15:31 zakharyas left 15:32 hythm left, hythm joined, lowbro left, hythm left 15:33 hythm joined 15:34 hythm left, hythm joined 15:36 hythm left 15:41 setty1 joined
zakame hi, is Task::Star gone from modules.perl6.org list? I can see its github still though 15:47
tadzik yeah, it's been removed from the ecosystem 15:48
zakame ah
tadzik some people actually noticed its disappearance though, so perhaps a fork would be desirable 15:49
zakame there's still Task::Galaxy etc
15:51 Cabanossi left 15:54 Cabanossi joined 15:57 cdg left 15:59 domidumont left 16:01 robertle left 16:06 hythm joined
zengargoyle resists urge to summon the Z 16:12
Herby_ I think I'm missing an essential step on something. I cloned the perl 6 twitter module and am trying to add a new GET method 16:14
16:15 astj joined
Herby_ none of the changes I'm making appear to be taking affect. its like my script is still pointing to the Twitter module I installed with zef 16:15
instead of the local lib that I'm changing
ahh nevermind
ugexe are you using -Ilib
Herby_ disregard
ugexe or -I. 16:16
Herby_ looks like i needed to run it from the parent directory
16:19 astj left
ugexe or you could -Ipath/to/parent-directory 16:28
or `PERL6LIB="/path/to/parent-directory" perl6 ...` 16:29
16:29 cibs left
Herby_ ugexe: thanks 16:30
16:31 cibs joined 16:32 zakame left
zengargoyle don't export PERL6LIB=lib or similar, it makes things strange and slow whenever you change directories. :P 16:32
it took me way to long to figure out that setting PERL6LIB in my .profile was why p6 was so slow all the time but only sometimes. :) 16:34
16:35 devmikey joined 16:37 robertle joined
tbrowder i've not noticed any ill effects using PERL6LIB defined with multiple ditrctories 16:37
16:37 holyghost left 16:38 devmikey left
zengargoyle if they're relative then it's like a new repo chain each time. (i think). i noticed by rsyncing rakudobrew installs (and ~/.perl6) across machines and noticing the startup lag after every sync. 16:40
16:42 Actualeyes left 16:43 troys joined
eater our can't have type constraints? 16:43
AlexDaniel eater: seems so 16:44
eater m: module Aliens { our Str $type-me = "ayyyy"; } say Alliens::typ-me;
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot put a type constraint on an 'our'-scoped variable
at <tmp>:1
------> 3module Aliens { our Str $type-me7⏏5 = "ayyyy"; } say Alliens::typ-me;
expecting any of:
constraint
eater is this an active design choice?
geekosaur I think the problem is that you 'our' it to use it in multiple modules; which one gets the type annotation, or do you need to check for matching annotations every time you compile *or load* a module that 'our's it? 16:47
jnthn Yes; `our` is just a lexical alias to something in a package. And packages are really just a hash table of keys to values, which you can poke whatever you like into without even needing declaration.
eater ah 16:48
thanks
sounds like a nice feature tho
but I guess it's a bitch to implement in what a module is now?
16:48 cdg joined
jnthn Even without a module... 16:48
module Foo { sub a { our Str $x }; sub b { our Int $x }; sub c { Foo::<$x> = 4.2 } } # uhh :) 16:49
All of those are talking about the same slot.
Uh, I meant even *within* a module.
It's not so much a "hard to implement" as a "what would we even implement" ;) 16:50
eater you can define in a sub an our D:
makes sense if you look at a module as hash
jnthn That's precisely what a module's symbol table is.
16:53 ChoHag left
eater hmm, but from a non-"i know all the insides" standpoint, it seems rather confusing. you could just forbid our definition's outside the module block 😇 16:53
keeps most code cleaner, and it looks like it would make more sense
but idk, maybe Im missing something 16:54
16:55 st_elmo joined
eater hmm maybe my idea of module's is just wrong 16:57
16:58 cdg_ joined, cdg left 17:06 dct left
TreyHarris anybody here ever tried to learn duodecimal (dozenal) or sexagesimal? I just read something suggesting that some people with dyscalculia have had success with them. (I'm rather good with math but abysmal with arithmetic.) 17:08
17:19 mr-foobar left, troys is now known as troys_
Voldenet well, I use base-60 for hours, it's quite neat 17:19
if that's what you mean 17:20
17:20 AlexDaniel left
TreyHarris lol 17:21
Voldenet but if someone asked me to added 34 to 48 i wouldn't be too sure 17:22
s/added/add/ 17:23
17:23 mr-foobar joined 17:24 troys_ is now known as troys 17:33 AnX left 17:42 skids joined 17:49 st_elmo left 17:55 dct joined 18:05 ufobat left
geekosaur I remember playing with duodecimal a bit as a kid, but that's about it 18:07
dylanwh base 6 is more fun 18:08
geekosaur and the link to time _is kinda obvious for both of those bases :)
(in fact, I kinda wonder the vintage of the claim that people have had success with those bases; they're much easier when you can just glance at the nearest clock... which doesn't work in the age of digital clocks :) 18:09
18:09 mcafee joined
geekosaur then again, the flip side of _that_ one is that the Sumerians used those bases for a reason... 18:10
18:12 domidumont joined
BooK wasn't the reason for 12 or 60 is that it's easily divisible by 2, 3, 4 (and 5) 18:17
mcafee Today I evaluated the expression [Z+] ((1,2,3),), expecting it to return (1,2,3) since I'm zipping a single element, but instead I got (6). What's going on here? 18:18
moritz something flattens :( 18:19
m: say [Z+] ([1, 2, 3], )
camelia (6)
pilne m: say [+] ([1, 2, 3],) 18:20
camelia [1 2 3]
pilne m: say [Z+] ([1, 2, 3]) 18:21
camelia (6)
pilne m: say [Z+] ([1, 2, 3], )
camelia (6)
pilne hrm
moritz m: say [Z+] $(1, 2, 3) 18:22
camelia (6)
moritz m: say [Z+] $(1, 2, 3),
camelia (6)
mcafee So it's not just me that's surprised? Whew!
moritz wow, it seems *really* eager the flatten the shit out of any single-element list it gets
mcafee: your sanity isn't in question this time! :-) 18:23
mcafee It's really puzzling because I thought Z+ is supposed to combine elements from different lists, but here it seems to degenerate to [+] on a single list...
I'm golfing and have written some code like: [Z+] @_.map: { ... } and it's falling over if @_ has just one element. 18:25
18:33 ufobat joined
b2gills A work around [Z+] 0 xx *, |@_.map: { ... } 18:34
18:35 espadrine_ joined 18:36 TeamBlast left
b2gills I've come across this a few times, but I was always in the process of creating a golf, so I just quickly moved on to the next algorithm. 18:36
18:39 TeamBlast joined
perlpilot Though you'd have to change the identity element depending on the operation. 18:39
i.e. [Z*] 1 xx *, |@_.map: { ... } 18:40
mcafee Yeah, I came up with that workaround too, but I hate to lose all those strokes!
18:43 TeamBlast left 18:45 Zoffix joined
Zoffix How specify a constraint for "all elements of array are either Cool:D or Nil or Any:U" ? 18:45
m: my @stuff = (1, Any, 3); -> @a where {@a.all.ACCEPTS: Cool:D|Nil|Any:U} { say "got it"}( @stuff )
camelia Constraint type check failed in binding to parameter '@a'; expected anonymous constraint to be met but got Array ($[1, Any, 3])
in block <unit> at <tmp> line 1
Zoffix Reading the Junction.ACCEPTS code suggests the above should work, but it doesn't 18:46
18:47 TeamBlast joined
Zoffix m: my @stuff = (1, Any, 3); say @stuff.map(*.ACCEPTS: Cool:D|Nil|Any:U) 18:47
camelia (any(False, False, False) any(True, True, True) any(False, False, False))
Zoffix Ah, ok
well, I see why that way don't work, but what's the way?
18:48 mr-foobar left 18:49 mr-foobar joined
Zoffix Actually, I don't want just any Any:U but THE Any:U 18:51
moritz then you need { $_ === Any:U } 18:52
Zoffix Well, that's the easy part. How to combine all that into something that doesn't look insane :D 18:53
18:53 zacts left
dogbert17 is this statement from the docs still correct: 18:54
Zoffix Also, you need === Any, not === Any:U :)
dogbert17 "Coercion types are supposed to work wherever types work, but Rakudo currently (2015.02) only implements them for subroutine parameters."
Zoffix dogbert17: yes 18:55
m: my Int() $x
camelia 5===SORRY!5=== Error while compiling <tmp>
Coercion Int(Any) is insufficiently type-like to qualify a variable
at <tmp>:1
------> 3my Int() $x7⏏5<EOL>
expecting any of:
constraint
dogbert17 Wasn't it implemented for return types or something to that effect
Zoffix Well, yes as in it don't work; I guess they're supposed to work
m: -> --> Int() { '42' }().say
camelia 42
Zoffix m: -> --> Int() { '42' }().^name.say
camelia Int
Zoffix dogbert17: looks like it
dogbert17: oh, well, "subroutine" should read just "routine", since they work in methods too 18:56
Zoffix wildly guesses they work in unpacking too
m: my (Int() $x) := ('42',); dd $x
camelia Int $x = 42
dogbert17 thx Zoffix, will update the text
Zoffix Haha yes :)
dogbert17 is unpacking the right term to use in the text 18:57
Zoffix No idea
moritz isn't that just binding an anonymous signature?
m: -> Int() $_ { say .^name }('42') 18:58
camelia Int
Zoffix m: my @stuff = (1, 42, 3); -> @ where {so all .map: {$_ ~~ Cool:D|Nil or $_ === Any}} { say "got it" }( @stuff ) 18:59
camelia got it
Zoffix Well, that's the shortest I got it to. Any saner way?
(`so ` part can be removed, but what else) 19:00
I guess it ain't that bad. /me goes with that
dogbert17 how about: "Coercion types are supposed to work wherever types work, but Rakudo currently (2017.05) only implements them for routine parameters and return types."
moritz ... "only implements in signatures, for both parameters and return types" 19:01
*them
19:01 Zoffix left
dogbert17 moritz++, will fix and commit 19:01
moritz on a completely different topic... 19:02
I have the domain perl6book.com, which I intended as a landing page for my book
but, apress has one, so I can reuse it for another purpose
I thought I might give an overview of all Perl 6 books there 19:03
and maybe even a small "which Perl 6 book to read" wizard
any thoughts on that? Does anybody want to contribute? 19:04
perlpilot moritz++ sounds like an excellent idea. 19:05
Geth doc: 2ebb405467 | (Jan-Olof Hendig)++ | doc/Language/functions.pod6
Fixed a few errors. Zoffix++, moritz++
 19:06
19:12 zakharyas joined 19:14 itaipu joined 19:15 cpage_ left 19:16 cpage_ joined 19:19 setty1 left
moritz I'm starting to collect data and thoughts at github.com/moritz/perl6book-web/bl...edesign.md 19:21
if anybody else wants to contribute, send a PR. Or if that's too much hassle, I can transfer the repo the perl6 org 19:22
(though I feel bad doing that as long as there's so much out-of-date stuff on the front page)
19:24 wamba joined
moritz ok, I'm filling out github.com/moritz/perl6book-web/bl...edesign.md now 19:29
if you think some of the characterizations are not fair or balanced, please speak up 19:30
19:32 dct left
tbrowder moritz: how about the book Metagenomics? 19:33
moritz tbrowder: that's a bit of a weird one. It doesn't advertise any Perl 6, but contains lots of it
tbrowder: I'm having trouble coming up with a description that doesn't sound totally weird 19:34
tbrowder ok, Perl 6 in the trenches?
moritz link?
19:34 grumble left 19:35 zakharyas left
tbrowder sorry, my attempt at humor, description of metagenomics... 19:35
19:35 troys is now known as troys_
moritz oh :-) 19:35
19:36 grumble joined
moritz anyone want to create a generic Perl 6 book logo, with camelia reading a book or something? :-) 19:37
zostay i just found a weird bug in DateTime::TimeZone's tools for generating classes from the Olsen TZ database that looks like a grammar doing something a grammar should not be able to do 19:40
moritz zostay: it became self-aware? 19:41
19:41 troys_ is now known as troys
zostay maybe 19:41
when you run the script, it outputs this error: Substring length (-3194) cannot be negative
in any Str at /Users/sterling/.rakudobrew/moar-nom/install/share/nqp/lib/QRegex.moarvm line 1
19:42 zacts joined
zostay the line causing it is: my $yfrom = +$rule<from>; 19:42
where $rule<from> is a hash lookup in a Match
moritz try changing it to +~$rule<from> 19:43
just to see if it makes a difference
zostay same error
moritz right, it tries to call Str already 19:44
zostay the more i work with it, the weird it behaves... like dd hangs trying to show the returned parse tree 19:45
19:46 kyan joined
zostay if i have Grammar::Tracer loaded, it starts tracing infinitely somewhere in the middle on the dd line 19:46
i'm wondering if one of the very common names used for the tokens might be clashing with a built-in
but that's just a guess at this point 19:47
19:48 cdg_ left
moritz zostay: try inspecting $rule<from> for its .from, .to and .orig properties 19:49
19:51 zacts left
zostay i think i just confirmed my hypothesis doing a search/replace and insert "tz-" on to the front of all the rule names fixed it... i wonder if having a rule named to or from is a no-no 19:52
moritz erm, yes 19:53
and i can't believe I didn't notice
tokens/rules are really methods on the cursor 19:54
at least "orig", "target" and "pos" would be verboten 19:55
but it seems of the Match methods/attributes are too
19:55 mcmillhj left
zostay to and from are definitely the problem, i changed just those to tz-to and tz-from and the problem goes away 19:55
19:56 zacts joined
zostay this seems like a place where having final methods could be helpful 19:56
19:59 mcmillhj joined
moritz agreed 20:00
or having to be explicit about masking parent classes method 20:01
zostay yeah, true final would be anti-perl right? there'd have to be something like augment to get around it, right? ;)
moritz right :-)
20:02 zapwai joined 20:03 mcafee left 20:06 zapwai is now known as Zapwai\lettuce
moritz my Perl 6 book decision tree: github.com/moritz/perl6book-web/bl...ision-tree :-) 20:11
20:18 cpage_ left
ingy is there a way to access all args like @_? 20:20
20:20 domidumont left
moritz |c in the signature 20:20
then c becomes a capture 20:21
20:22 bdmatatu joined
ingy gist.github.com/anonymous/5a7c24b1...c8e2a5eed3 something like that 20:25
20:26 cpage_ joined
moritz m: sub foo(*@_ ($x, $y)) { die @_ }; foo('a', 'b') 20:27
camelia a b
in sub foo at <tmp> line 1
in block <unit> at <tmp> line 1
timotimo m: sub foo(|c, $x, $y) { die c.list }; foo("a", "b") 20:29
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot put required parameter $x after variadic parameters
at <tmp>:1
------> 3sub foo(|c, $x7⏏5, $y) { die c.list }; foo("a", "b")
expecting any of:
constraint
timotimo ah, right 20:30
ingy so no way without changing the sig?
moritz no official way, at least 20:31
I guess if you dive deep enough into nqp:: ops, you'll find something
Xliff \o
ingy :)
20:31 mcmillhj left 20:33 mcmillhj joined
ingy p5 has to be better at something ;) 20:33
Xliff What's the best way to check type against all elements of an array?
moritz fsvo "better"
Xliff: all(@array) ~~ TheType 20:34
Xliff LOL
I was doing TheType ~~ all(@array)
And it wasn't working.
Why is ~~ not commutative? 20:35
Or is it the junction that is not commutative? 20:36
timotimo because that went very, very, very wrong in perl5 :)
Xliff LOL!!!
timotimo you remember how they got smartmatch, but then decided to toss it again?
Xliff No.
I smell a story
Is there a blog post, or is it timotimo story time? :)
timotimo not only was it a bad fit because of the lack of types in perl5, but it was also bad in general because they stole it before perl6ers realized commutative smartmatch is not a good idea at all 20:37
20:37 troys is now known as troys_
Xliff Ooooooo 20:37
So it's always better to do: <Thing you are checking> ~~ <Type>?
m: say 4 = any(2, 3, 4, 5) 20:38
camelia Cannot modify an immutable Int (4)
in block <unit> at <tmp> line 1
moritz that's the only correct way to do it :-)
Xliff m: say 4 == any(2, 3, 4, 5)
camelia any(False, False, True, False)
Xliff m: say any(2, 3, 4, 5) == 4
camelia any(False, False, True, False)
20:38 mcmillhj left
Xliff OK, so it is smartmatch. 20:38
moritz it's ~~ that's not commutative
right
Xliff Up to this point, I've only used == with junctions.
timotimo++ && moritz++ 20:39
20:39 itaipu left 20:41 ufobat left 20:42 mcmillhj joined
awwaiid ingy: you can get some interesting things with MY::, but it is a superset of what you requested 20:42
m: sub my_params($x, $y) { say MY::.perl } ; my_params(2, 4) 20:43
camelia PseudoStash.new(("\$!" => Nil,"\$*DISPATCHER" => Mu,"\$/" => Nil,"\$_" => Any,"\$x" => 2,"\$y" => 4,"\$¢" => Nil))
ingy awwaiid: awesome
awwaiid m: sub my_params($x, $y) { say MY::.perl; my $z = 7 } ; my_params(2, 4) 20:44
camelia PseudoStash.new(("\$!" => Nil,"\$*DISPATCHER" => Mu,"\$/" => Nil,"\$_" => Any,"\$x" => 2,"\$y" => 4,"\$z" => Any,"\$¢" => Nil))
awwaiid $z in scope but not defined, interesting.
moritz declaration is a compile-time operation 20:45
awwaiid ah yes
moritz but the initialization is run time
awwaiid ingy: I hope you are doing horrible horrible things... whatchya up to? 20:46
ingy I'm writing new languages in a dialect or perl6 for $work. 20:47
awwaiid I cannot parse that sentence without more parenthesis, but sounds cool
geekosaur change "or" to "of" 20:49
awwaiid ahh
ingy openresty.org has a (soon to be open-sourced) internal programming language called FanLang that uses (mostly) Perl6 syntax to compile to LuaJIT. And with that lanugage they are writing several other languages/DSLs internally.
awwaiid madness 20:50
20:50 troys_ is now known as troys
moritz that's what agentzh has been working on, right? 20:51
ingy at the moment I'm implementing Template-Toolkit (Jemplate/Lemplate) in FanLang
which itself is a language (DSL)
yeah it's agentzh's startup company 20:52
I bet I could port TT2 from Fan to Perl6 quite easily. 20:53
20:53 TEttinger joined
awwaiid feel free 20:53
20:53 lizmat left
ingy :) 20:53
mostly I'm working on getting the p6 YAML stuff done ASAP
zostay \o/
ingy Make YAML Great Again!!!
[Coke] "too soon" 20:54
ingy YAML == SAD
moritz s/Again// # SCNR :)
ingy Hillary told me that YAML is already great.
awwaiid needs more exclamation marks. sad! 20:55
zostay YAML is a polarizing force... like ingy... maybe ingy should run for President and have YAML as his running mate
20:55 grondilu joined
zostay i'd vote for that duo 20:55
20:56 grondilu left, grondilu joined
ingy tinita has YAML.pm6 working (over the libyaml stuff) 20:56
we should get that out this weekend
awwaiid I tell all my friends that ingy taught me in-person that YAML is a superset of JSON. Mostly they just stare at me because they haven't met Ingy so haven't felt the mad-science inspiration that drips from his soul
ingy heh
awwaiid also many of my friends are not programmers, and thus even further removed from being able to relate to my cool story 20:57
ingy I feel totally "Ingy" trying to write a new language in a new language in a ...
awwaiid and getting paid to do it, so that's good 20:58
20:58 bdmatatu left
ingy +10! 20:58
Herby_ whats the simplest way to pretty print a hash?
or print a hash in a pretty state :)
awwaiid m: say {a => 2, b => 4}.perl 20:59
camelia {:a(2), :b(4)}
awwaiid m: say {a => 2, b => 4}
camelia {a => 2, b => 4}
timotimo well, Data::Dump::Tree isn't simple, but it is really pretty
moritz awwaiid: that kinda reminds me of a story a colleague told me. At the dinner party of some ruby conf, he talked for over an our with some friendly Japanese guy
timotimo you can use json::fast's to-json with :pretty
zostay m: dd {a => 2, b => 4}
camelia Hash % = {:a(2), :b(4)}
moritz awwaiid: ... and only later learned that it was Matz :-)
timotimo that's rather simple and looks good
awwaiid hahaha
How pretty is pretty, Herby_? 21:00
Herby_ thanks i'll take a look at
sorry, missend
pilne i'd love to find an opportunity to jump ship from where i'm working now (non-coding) to something coding, but every time i sit down and try and work on something coding related that isn't perl6 now, i get frustrated at how much I feel like i'm appeasing the language instead of utilizing it lol.
timotimo i can only hope that when someone meets me at a party they aren't intimidated by me being the author of the #1 most cited perl6 module :)
ingy I can see that. I've had beers with Matz. He's totally chill.
Herby_ a json response from twitter, trying to print it out so i can see whats going on
right now its just a blob
pilne and a good... 100% of the stuff around here is java/python/php/ruby (not in that order necesarily).
zostay pilne: where you at -ish? 21:01
moritz Herby_: say .key, "\t", .value for %hash;
pilne ruby is "ok" but i feel like everything in that ecosystem is webdev
[Coke] pilne: makes you feel any better, I work somewhere where that is about 60/39/1 java/cold fusion/php
pilne zostay=> closest major metro (and it is close) is saint louis, missouri.
your poor sanity coke
Herby_ moritz: thanks 21:02
pilne there's some perl5 stuff around, but tbh, i barely know perl5.
zostay i'm near KC, assuming you won't relo, if you are open to working remote, we are hiring perl hackers like crazy
ingy I remember the first time I had lunch with TimToady. I was sitting next to him and I asked him a question and he just stared off into space. After a minute I just went back to eating. Then 5 minutes later he turns to me and gives me a 5 minute answer.
pilne p5 or p6 zostay? 21:03
zostay p5, i don't know of much p6 work out there
ingy He never even nodded that he heard me or anything. It was awesome awkward. :)
zostay or python
pilne relocation might not be impossible, missouri end-to-end isn't that bad, i used to live near kingdom city in missouri. 21:04
21:05 mcmillhj left, AlexDaniel joined
pilne i'm comfortable with python, i'm sure i could learn perl5 since well, perl6 is coming easy, but i'm not great in either, i took a long hiatus from programming when i got ptsd from c++ and java in 2000 lol. 21:05
ingy ah nostalgia. ok back to the almost-p6-salt-mines...
perlpilot learning programming languages is easy ... learning how to /think/ is hard :) 21:06
moritz perlpilot: you'd think that :-)
Xliff zostay: I am in DC and cannot relocate. Can I pass you a resume? ;) 21:07
perlpilot++
pilne perlpilot=> my love of perl6 stems from the fact that it generally allows me to think how i want and it (often) easily translates to code. 21:08
Xliff pline++ # Once you learn the syntax, this is (mostly) true.
moritz I'll be teaching a programing course at $work, starting next week
(python though)
perlpilot indeed. That's one of the reasons I've stuck with Perl 5 since forever--it fits my brain well. Perl 6 more so.
Xliff python-- 21:09
moritz looking forward to it
Xliff I have tried my hand at python. Was able to grok it enough to get it to do what I wanted in a large project.
moritz the initial response was pretty good: 25 out of about 200 employees signed up
Xliff Did not enjoy the experience.
perlpilot Xliff: python is okay until you try to do something perlish and it fights you ;) 21:10
moritz it assumes a different mindset than Perl
Xliff Yes. I got Lisp flashbacks.
moritz it took me about 2 years to kinda like it
timotimo i built a pretty cool cellular automata simulator framework in python
Xliff I'm not supposed to have Lisp flashbacks!!!
python: Lisp lite, without all of the bloody parens. 21:11
perlpilot
.oO( <Xliff> I haven't even ever used Lisp! )
Xliff O_O
21:11 espadrine_ left
Xliff I never said that! 21:11
I disbelieve?
(should have not ended that in a ?)
perlpilot It's funny that you equate Python and Lisp (when I used to hang out on #python, they did too), but I never thought they were that close 21:12
Python seems more like a high-level C to me.
Lisp at least groks closures; python still struggles 21:13
moritz I still hate python's scoping (or lack thereof)
Xliff I have made my feelings on Lisp known here, before.
irclog.perlgeek.de/perl6/search/?n...amp;q=Lisp
tadzik it's awful :|
moritz and implicit declaration of variables
21:13 mcmillhj joined
pilne python's "do it this way or else" just kinda irks me... 21:13
Xliff pilne++
perlpilot: The problem there is that C is not as pedantic as python. At least C knows its limitations. 21:15
Python implies less restrictions than C, but in reality there are more
:/
pilne i mean, there's usually a lib or something to be able to to just about anything in python, but sometimes i look at it and go "but.... why?" (note: i've always loved scheme) 21:16
Xliff You could say the same about C/C++
pilne tbh i'd rather use python/ruby/perl on top of c, instead of c++ if i want more abstractions. 21:17
Xliff NativeCall FTW!
That was one of the things that drew me to Perl6.
And maybe I am just more used to c++, but I would prefer that than python. 21:18
pilne i read through the docs on nativecall and still have a lot to learn about it, but damn it looks fun.
Xliff It is.
Until your code blows up MoarVM
21:18 mcmillhj left
pilne to be fair, the last time i was doing c++ it was c++98 21:19
Xliff Oooooooo
perlpilot Xliff: what are you doing with NativeCall? Anything interesting?
Xliff was doing C++ up until 2003 or 4ish?
Then it was all Java and C#
(and Perl <5 and 6> whenever I got some spare time)
perlpilot: I am still trying to work on a P6 wrapper for libxml2 21:20
perlpilot cool.
Xliff But Perl6's design has changed since then and everything that FROGGS and I have written will need to be rewritten.
FROGGS is the current maintainer. I just contribute.
I haven't touched it in months cos.... life 21:21
I need to get back to it.
21:21 Cabanossi left
ingy how many people here get paid to write perl6 (not to hack on the language itself)? 21:22
Xliff ingy: Not much of a Perl6 market, yet.
pilne i write code snippets for 5/6 on my phone at work when there's downtime? lol.
perlpilot The 2 modules that I think allow Perl 6 the most flexibility and power are NativeCall and Inline::Perl5. If I can't do something directly in P6, I know I can delegate to P5 for most things and then NativeCall for everything else. 21:23
moritz ingy: does getting paid for a Perl 6 book count? :-)
ingy no :)
21:24 Cabanossi joined
perlpilot ingy: I've been trying come up with a good excuse to use P6 at work, but haven't found the right project yet. So ... not getting paid for P6 now, but I hope to in the not-too-distant future 21:24
AlexDaniel ingy: I've never heard of anybody using just perl 6 on their job, but some people seem to use it for little scripts and stuff
timotimo i was being paid for perl6 work for a few months
perlpilot ingy: Although a few years ago, I got paid for P6 in that I used it to help with a fortran -> java translation. P6 did some parsing and provided a test framework. 21:25
ingy perlpilot: awesome 21:26
pilne my problem with trying to use anything at my job code-related is that the company will own it all with no obligation to compensate/promote me lol.
ingy yeah I think using P6 to make internal DSLs is a future
agentzh/openresty are shooting for top performance so having full control of their runtime is critical. 21:27
when writing new languages is trivial, everyone will do it 21:29
unless it's in Haskell ;)
writing new languages in Haskell is trivial (once you know Haskell...)
21:30 mcmillhj joined
ingy or so I've heard 21:30
pilne haskell does have incredible parsing/lexing abilities 21:31
perlpilot Do you mean in the sense of using Parsec, or warping Haskell's syntax to suit your needs?
pilne but lord is it a mindbender of a language
ingy was actually started the perl6 channel (back when au was trivially implementing p6 in Haskell)
that's why it's on freenode
s/was//
timotimo you haven't seen "mindbender of a language" until you've worked with prolog, IMO 21:32
ingy (at the time I didn't consider p6 to be a Perl language :D 21:33
pilne now *that* would be a fun inline module :p
ingy 's mind is too bent already :(
pilne i love all the older smaller more-focused languages, and feel that perl, and perl6 have done the most when it comes to learning from those languages (well, moreso perl6, but for it's roots, perl5 can do some incredible things). 21:34
timotimo why don't we get some Inline::Forth? :) :)
21:35 mcmillhj left
ingy I think I wrote a CPAN Forth module 21:35
cen't remember
metacpan.org/pod/Pegex::Forth
21:35 perlpilot left
Xliff Someone mentioned Haskell? 21:35
YIKES! I'm outta here! ;) 21:36
It's Friday anyways.
ingy Xliff: Perl6 came from Haskell!
21:36 lizmat joined
Xliff ingy: Perl6 is less Haskelly than... Haskell 21:36
ingy TimToady was said "People say that Perl 6 is Haskell's scripting language. I say that Haskell is Perl 6's functional language." 21:37
*once said
japhb timotimo: As a (years ago) Forthwright, I have to ask: Why would you want an Inline::Forth? Its use cases are pretty much diametrically opposed to Perl 6's. 21:38
ingy at least that's what I remember (which isn't that much these days)
timotimo you're right, it was a little facetious
ingy japhb: Perl never asks Why... :) 21:39
Perl asks Why not...!
21:39 mcmillhj joined
pilne acme anyone? (: 21:39
ingy for better or worse.
(or best)
timotimo a perl6-based forth should be called sixth 21:40
ingy Forx
japhb Now, doing a teaching version of Forth with a debugger and visual editor and all that, written in Perl 6, that could be fun.
lizmat perltricks.com/article/on-sigils/ # also about Perl 6 21:41
news.ycombinator.com/item?id=14462863 ^^^ HN comments 21:45
Xliff ??? 21:47
$ is "no type constraint"?
According to that sigils article.
moritz what type constraint do you think that $ implies? 21:48
Xliff Well. This is probably more Perl5 constraint. But I always thought of $ as a singular object or reference. 21:49
s/Perl5 constraint/Perl5 thinking/
@ isn't singular.
% isn't either 21:50
timotimo stacks and queues are totally provided natively by perl5 though?
Xliff You can implement them with @ 21:51
pilne singular/plural isn't really a "type"
timotimo i wouldn't say "can implement them" 21:52
by using .push and .shift you have a queue
by using .push and .pop you have a stack
Xliff Exactly.
timotimo more like "you can ignore parts of @ to have a stack/queue"
Xliff LOL
Fair enough.
pilne $, @, and % are more about the shape of the data being held, not what that data is made of.
b2gills a $ containing a @, contains a single @ value 21:53
Xliff ^^ THIS
pilne makes sense to my brain 21:54
timotimo i find the paragraph about variables with "the same name but different sigils" really confusing or misleading
pilne hell a $ containing a @ that contains %s still makes sense
i try to avoid same name with different sigils, just cuz i can do it don't mean i should.
timotimo it's not about "the way the variable is stored in the namespace" 21:55
$foo and @foo are stored the exact same way in the namespace
Xliff pilne: But in that case, the $ still contains a singular @ value
21:56 ChoHag joined
pilne yeup, i've never had a problem with that notion 21:56
21:57 mcmillhj left
pilne although using $x = @array to get the length of an arrray (context?) in perl5 is something i still see as one of those "wow cool" things but still trips me up sometimes 21:57
and lets not mention how i put $ instead of $_ twice the other day when working on some rosetta code for perl6....
ugexe there is always $x += @array 21:58
pilne lol 21:59
timotimo and also, the paragraph above the OrderedHash.new code claims "if the sigil truly were an interface only in perl6 we could apply any of the sigils to the same variable" 22:00
but you *can*
you just have to keep the name and put a sigil in front
i.e. not my $ordered, then %ordered and @ordered
but my $ordered, then %$ordered and @$ordered
this really grinds my gears a little
moritz oops, midnight coming around suddenly. Time for sleep here :-) 22:04
22:05 MasterDuke joined
timotimo gnite moritz :) 22:06
22:11 mcmillhj joined
timotimo can someone reach david farrell? 22:13
22:15 astj joined, mcmillhj left 22:20 astj left 22:21 skids left
pilne i wonder if i'd bugger up perlbrew trying to install two versions of perl simultaneously? 22:21
22:23 mcmillhj joined
[Coke] I've used perl 6 for a $dayjob script that is still in production 22:28
22:28 cgfbee left
Xliff pilne: Have two .rakudobrew directories. 22:28
[Coke] (and some one offs to extract data from .docx)
Xliff .rakudobrew and .rakudobrew2 ... repeat for as many versions as you need. 22:29
[Coke]: What did you use to extract data from .docx?
22:29 geekosaur left, cgfbee joined
pilne nahhhhh i'll just try and focus my nearly non-existent patience (: 22:30
22:31 geekosaur joined
Xliff pilne: You're (patience's) funeral. :> 22:31
22:31 DrForr_ joined, pochi joined
pilne LOL 22:31
22:31 zengargo1le joined 22:32 grondilu_ joined, sivoais_ joined 22:34 mcmillhj left 22:36 Cabanossi left 22:37 jantore_ joined
[Coke] Xliff: don't have access to the repo atm, but basically: unzip the .docx and dig through the xml 22:38
22:39 Cabanossi joined 22:40 jantore left, pochi_ left, DrForr left, sivoais left, kent\n left, bonsaikitten left, araraloren left, xiaomiao joined
timotimo one of perlbrew's features is having multiple perl5 versions installed and easily switching between them, isn't it? 22:40
22:40 Some-body_ joined, araraloren joined
[Coke] aye. 22:40
22:41 kent\n joined, Some-body_ is now known as DarthGandalf 22:42 simonm left 22:44 grondilu left, zengargoyle left
pilne yes, i was just wondering if it would barf if i used two terminals to install two different versions at the same time 22:45
22:46 simonm joined
timotimo Xliff: what. rakudobrew's feature is to have multiple rakudo versions installed at the same time and switch with a single command 22:49
22:50 nadim left
pilne i think there was a bit of a misunderstanding as to what I was trying to do earlier. 22:53
22:56 lizmat left, troys is now known as troys_
Xliff timotimo: Oh. I've never used that feature so I didn't know. 23:02
Change of topic. Is there a ^remove_method method in the MOP?
timotimo don't think so, no 23:03
Xliff Crud. 23:04
So if I have a dynamically created method x() on an object.
And I want to change the definition of x()... do I still use ^add_method and ^compose?
timotimo change the definition of it, eh ... 23:05
Xliff yeah, eh
timotimo really, i'd probably use wrap on the routine object
Xliff "use wrap"
???
Is there a helpful linky that explains that? 23:06
timotimo you use find_method and call .wrap(-> some-signature-or-something { new code })
normally you'd be calling to the original function inside the wrapper
but you can abuse this to replace methods 23:07
Xliff And is this abusable behavior going to stick around, or is a later version of Rakudo going to close that hole? 23:08
Like... say.... 6.e?
timotimo i think you're supposed to have to "use soft" for this to work 23:09
Xliff perl6advent.wordpress.com/2011/12/...rprofiler/ 23:12
^^ Something like that... but it doesn't use .wrap
timotimo yeah that changes how methods are returned from lookup
Xliff OK. So it looks like my FALLBACK method, which adds methods based on certain conditions, will need to handle all use cases in the method being added. 23:14
*Whee*
I was hoping to just define the useful method when data existed and just have a dummy method when it didn't.
Then I added the ability to load data at run time. 23:15
timotimo ah, hmm
Xliff Which means that the methods that were set to the dummy routine might then become available, later.
timotimo oh, hm.
Xliff I was hoping I could check the instance for all of the dummy methods and replace them.
23:15 wamba left
timotimo watch out, though, instances don't have their own method tables and such 23:16
Xliff Erp!
timotimo you'll want to mix in roles for that
Xliff so: class A { my class B {}; has B $!b; }; 23:17
Class B will have same method tables across each instance of A?
timotimo yup
the method table is a property of the type
Xliff EEP!
OK, but that doesn't hold true for Roles? 23:18
(but if method table is property of the type... then that doesn't make sense.) 23:19
timotimo++ # Gives me something to think about. 23:21
Now... back to my regularly scheduled drining.
s/drining/drinking/
pilne ya'll are definitely further down the rabbit hole than I am
Xliff And if I was cogent enough to make that correction... I have not drunk enough!
pilne a nice craft beer or 3 sounds good for tonight
23:21 Cabanossi left 23:24 Cabanossi joined
timotimo you can create roles dynamically through the MOP 23:24
23:26 setty1 joined
Xliff timotimo: OK. That is something. 23:32
timotimo and by parameterizing a non-dynamic role you can also do some stuff
Xliff However I think I may just re-implement things so that the dynamic methods Do The Right Thing, regardless of whether the data exists or not. 23:33
Maybe when I understand more about Roles and their dynamic-ness I will write something that takes advantage of it.
Any links on that particular subject would be good. :) 23:34
Thanks.
timotimo oof
23:57 labster joined 23:59 setty1 left

Powered by Cro and the Raku® Programming Language.

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

Thank you!