»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.
RabidGravy I'n sitting here listening to a Perl 6 program sending random notes to a synthesizer 00:26
AlexDaniel RabidGravy: I am sitting here controlling a robotic vacuum cleaner with Perl 6 :) 00:33
RabidGravy this is a cool thing 00:34
RabidGravy so Perl 6 will rule the world 00:36
timotimo rule, token or regex. 00:40
AlexDaniel OK, so IO::Handle.split does not really work? 02:29
it looks like it returns a non-lazy list 02:30
SmokeMachine____ hi again! I have a question: 02:32
when I use [X] on a list with 3 tuples on it, it returns a list of tuples with 3 elements each
m: my @a = (1, 2, 3, 4, 5), (10, 20, 30, 40, 50), (100, 200, 300, 400, 500); ([X] @a).say
camelia rakudo-moar 4f7cb8: OUTPUT«((1 10 100) (1 10 200) (1 10 300) (1 10 400) (1 10 500) (1 20 100) (1 20 200) (1 20 300) (1 20 400) (1 20 500) (1 30 100) (1 30 200) (1 30 300) (1 30 400) (1 30 500) (1 40 100) (1 40 200) (1 40 300) (1 40 400) (1 40 500) (1 50 100) (1 50 200) (1 50 300) (1…»
SmokeMachine____ www.irccloud.com/pastebin/wIvpqV9X/ 02:33
AlexDaniel SmokeMachine____: and what do you want instead?
SmokeMachine____ when I do it with 2 tuples, it returns tuples with 2 elements: 02:36
m: my @a = (1, 2, 3, 4, 5), (10, 20, 30, 40, 50); ([X] @a).say
camelia rakudo-moar 4f7cb8: OUTPUT«((1 10) (1 20) (1 30) (1 40) (1 50) (2 10) (2 20) (2 30) (2 40) (2 50) (3 10) (3 20) (3 30) (3 40) (3 50) (4 10) (4 20) (4 30) (4 40) (4 50) (5 10) (5 20) (5 30) (5 40) (5 50))␤»
AlexDaniel SmokeMachine____: that's how X works :) 02:37
SmokeMachine____ so, why it returns only 1 tuple with all elements when I use a list with 1 tuple?
m: my @a = (1, 2, 3, 4, 5); ([X] @a).say
camelia rakudo-moar 4f7cb8: OUTPUT«((1 2 3 4 5))␤»
AlexDaniel m: my @a = (1, 2, 3, 4, 5); ([X] @a).perl.say
camelia rakudo-moar 4f7cb8: OUTPUT«((1, 2, 3, 4, 5),).Seq␤»
geekosaur seems to make sense to me, picks one element from each input tuple -- which, there being one, means one element. and there's no "1-tuple" 02:38
geekosaur m: ((1),(2),(3),(4),(5)).say 02:39
camelia rakudo-moar 4f7cb8: OUTPUT«(1 2 3 4 5)␤»
SmokeMachine____ but that doesn't "do what I mean"...
geekosaur that being? 02:40
SmokeMachine____ my problem is:
AlexDaniel m: .say for $*IN.lines 02:41
camelia rakudo-moar 4f7cb8: OUTPUT«Céad slán ag sléibhte maorga Chontae Dhún na nGall␤Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤I mo dhiaidh bhí gleanntáin ghlas’ G…»
AlexDaniel m: .say for $*IN.split(splitter => ‘ ’)
camelia rakudo-moar 4f7cb8: OUTPUT«␤C␤é␤a␤d␤ ␤s␤l␤á␤n␤ ␤a␤g␤ ␤s␤l␤é␤i␤b␤h␤t␤e␤ ␤m␤a␤o␤r␤g␤a␤ ␤C␤h␤o␤n␤t␤a␤e␤ ␤D␤h␤ú␤n␤ ␤n␤a␤ ␤n␤G␤a␤l␤l␤␤␤A␤g␤u␤s␤ ␤d␤h␤á␤ ␤c␤h…»
AlexDaniel ??? 02:41
AlexDaniel ah, I'm not reading the docs right 02:42
m: .say for $*IN.split(‘ ’)
camelia rakudo-moar 4f7cb8: OUTPUT«Céad␤slán␤ag␤sléibhte␤maorga␤Chontae␤Dhún␤na␤nGall␤Agus␤dhá␤chéad␤slán␤ag␤an␤Eireagal␤ard␤ina␤stua␤os␤cionn␤caor␤is␤coll;␤Nuair␤a␤ghluais␤mise␤thart␤le␤Loch␤Dhún␤Lúich’␤go␤ciúi…»
saaki psch: ah that was it, thanks! i was wanting to use "," rather than "|" there 02:44
trying*
SmokeMachine____ m: my @a = (1, 2, 3), (10, 20, 30), (100, 200, 300); say "call-func-with({ $_.join(q|, |) })" for [X] @a
m: my @a = (1, 2, 3), (10, 20, 30); say "call-func-with({ $_.join(q|, |) })" for [X] @a
m: my @a = (1, 2, 3); say "call-func-with({ $_.join(q|, |) })" for [X] @a
camelia rakudo-moar 4f7cb8: OUTPUT«call-func-with(1, 10, 100)␤call-func-with(1, 10, 200)␤call-func-with(1, 10, 300)␤call-func-with(1, 20, 100)␤call-func-with(1, 20, 200)␤call-func-with(1, 20, 300)␤call-func-with(1, 30, 100)␤call-func-with(1, 30, 200)␤call-func-with(1, 30, 30…»
rakudo-moar 4f7cb8: OUTPUT«call-func-with(1, 10)␤call-func-with(1, 20)␤call-func-with(1, 30)␤call-func-with(2, 10)␤call-func-with(2, 20)␤call-func-with(2, 30)␤call-func-with(3, 10)␤call-func-with(3, 20)␤call-func-with(3, 30)␤»
rakudo-moar 4f7cb8: OUTPUT«call-func-with(1, 2, 3)␤»
SmokeMachine____ AlexDaniel: you said you liked the idea of Test::Fuzz... just to let you know, the module is getting better... I did the fuzz method that you "asked for"... if you have a free time and want to critic: github.com/FCO/Test-Fuzz 03:14
zacts hi perl6 nerds 03:16
AlexDaniel SmokeMachine____: sounds great. Consider adding it to the ecosystem once it gets more or less ready :) 03:19
SmokeMachine____ AlexDaniel: I will! thanks! have a good night!! 03:20
perlawhirl hi perlers 06:08
who's awake
sortiz perlawhirl, Hi 06:09
perlawhirl how can i do a defined-or subscript? 06:10
m: my @a = <one two three>; say @a[5 || 1];
camelia rakudo-moar 4f7cb8: OUTPUT«(Any)␤»
perlawhirl or do i have to ne content with $var = @a[5] || @a[1]; 06:11
sortiz The later, but use // 06:12
perlawhirl ahh yeh.. the defined-or op :D
alright, just checking if there was a shorter way
sortiz m: my @a = <a b c d e>; say @a[10,4,3,5].first(*.defined); # Not shorter ;-) 06:17
camelia rakudo-moar 4f7cb8: OUTPUT«e␤»
perlawhirl hah 06:18
sortiz And I expect 'c' :-(, Need more coffee
perlawhirl maybe thats why you've got "c" on the brain... your bodies telling you C-offee 06:26
and i blame my poor grammar on lack of sleep and too much coffe 06:27
sortiz Coffee ready :-) 06:29
perlawhirl ok... more random defindedness related shenanegans... 06:30
m: my @a = <one two three>; 12 ~~ /(\d+) { ?@a[$0].defined } / 06:31
camelia ( no output )
perlawhirl m: my @a = <one two three>; say 12 ~~ /(\d+) { ?@a[$0].defined } /
camelia rakudo-moar 4f7cb8: OUTPUT«｢12｣␤ 0 => ｢12｣␤»
perlawhirl why does the regex succeed if @a[12] is not defined, and can/should it backtrack to 1?
read: is there a way to make it so
perlawhirl is off to read regexes doc again 06:32
PerlJam m: my @a = <one two three>; say 12 ~~ /(\d+) <?{ ?@a[$0].defined }> / 06:37
camelia rakudo-moar 4f7cb8: OUTPUT«｢1｣␤ 0 => ｢1｣␤»
PerlJam perlawhirl: like that?
perlawhirl oh right... the ? i totes forgot 06:40
i could have just referred to one of my other scripts
again... did i mention i'm sleeped deprived :D 06:41
PerlJam needs to get to bed actually 06:42
my brain hasn't worked quite right all day and no sleep doesn't help that for tomorrow
stmuk 08:55
azawawi hi 08:57
anyone with Mac OSX to test a Selenium::WebDriver? 08:58
re travis-ci.org/azawawi/perl6-seleni...1139#L3594 08:59
azawawi any news on 2016.03? 09:10
azawawi listens to the echo in the room 09:11
nine echo, echo, echo, echo, ... 09:23
azawawi :) 09:26
nine: do you have by any chance a macosx environment around you? :) 09:27
nine Nope...I'm a free software zealot 09:28
azawawi nine: cool. i respect that :) 09:29
Ven azawawi: cloning! 09:47
azawawi Ven: which MacOSX?
Ven 10.10.5 yosemite 09:48
(updating rakudo first)
azawawi cool thanks 09:49
azawawi basically phantomjs is refusing to start for some reason via Proc::Async 09:49
Ven sent a PR your way WRT phantomjs on mac os 09:52
well, `make install` is failing for me. "An exception occurred while evaluating a constant. Could not find symbol '&bool'" 09:53
azawawi Ven: link output please? 09:58
Ven cleaned up my install, rebuilding :) 09:59
updating panda now. 10:01
azawawi cool thx for your help. I really appreciate it :) 10:03
is it me or did they Apple remove the i7 CPU option from iMac? 10:04
s/did they/did/
Ven no idea, sorry :)
azawawi oh you get it in the custom builds section now :) 10:07
for only +200$ (hehe) 10:08
Ven well, HTTP::UserAgent's tests fail here 10:08
azawawi NETWORK_TESTING enabled? 10:10
Ven dunno, I just panda install'd it
(so: no)
azawawi link the output please? 10:11
Ven just "host unreachable" 10:11
azawawi panda update && panda install HTTP::UserAgent 10:12
Ven and now it worked on the second try :P. guessed it was just a spurious failure
azawawi most likely a HTTP::UserAgent test not adhering to "NETWORK_TESTING" being off 10:13
echo $NETWORK_TESTING # to be sure
Ven it's empty 10:14
okay, got it to "Cannot obtain a session after 30 attempts" on Selenium::WebDriver
azawawi good
phantomjs --webdriver=5555 --webdriver-loglevel=DEBUG 10:15
Ven same error 10:16
azawawi no did phantomjs give any output?
for the command above?
Ven no
well, it started, but nothing more. "GhostDriver - Main - running on port 5555"
azawawi telnet localhost 5555 10:17
azawawi can you clone the S:WebDriver repo please? 10:17
hankache morning * 10:18
azawawi Ven: Try to debug this please github.com/azawawi/perl6-selenium-...re.pm6#L44 10:18
azawawi Ven: maybe _empty-port is not getting us a port 10:19
hankache: morning
hankache hi azawawi 10:20
azawawi Ven: gist.github.com/azawawi/f9b8356ed8ed0a4c3c0f # Use this for testing it quickly 10:25
hankache: hey
dalek Iish/myGLR: 0055e68 | (Salvador Ortiz)++ | / (2 files):
GLR test target
 10:32
Iish/myGLR: 3849814 | (Salvador Ortiz)++ | lib/DBDish/StatementHandle.pm6:
The GLR of DBDIsh. $sth.allrows now returns Seq

Fetching rows is now lazy, plus:
  - named ':hash' of .row method is now implemented by the role.
  - legacy '.fetchrow' requirement removed from drivers, a simple one
   implement by the role.
  - Allows driver's ._row method return a simple list of typed values
Iish/myGLR: d20be30 | (Salvador Ortiz)++ | lib/DBDish/ (3 files):
Simplify drivers for post GLR role
Iish/myGLR: 4ae1c57 | (Salvador Ortiz)++ | / (4 files):
Bump to v0.5.0

DBDish's changes deserve the version incremented.
RabidGravy I'm happy, Perl 6 is fast enough to generate a passible acid house bassline at ~120bpm 10:54
azawawi RabidGravy: cool 10:55
RabidGravy: what's your machine specs btw?
timotimo neato
and good morning
RabidGravy azawawi, it's >5yo i7 10:56
with 8gig of memory
azawawi RabidGravy: Geoforce Titan X with 8 Xeon E7 CPUs :)
azawawi one can dream 10:57
RabidGravy well if I had one of those I could probably do the synthesis part as well
azawawi most probably your old i7 will beat my old AMD FX-8150 16GB on single thread performance :) 10:58
an i5 is already doing that
timotimo RabidGravy: oh, so that's only midi commands now? 11:01
RabidGravy yeah, well you know how hard it was to generate a sine in the time budget?
adding a LPF and a volume envelope I guess would be good for one note a second or something 11:02
instead of ~32 11:03
timotimo right 11:06
RabidGravy it may or may not be playing at rabidgravy.com:8000/stream
timotimo i'd think adding those things will only slow the thing down a little bit, compared to the overall overhead we already had
timotimo but it's still pretty bad to have all these gc pauses in there 11:07
(we have a way to make that faster, fortunately)
azawawi listens 11:09
RabidGravy is it actually playing, jack has the soundcard here so I can't actually listen to the stream 11:10
but fwiw, Perl6 generating the note data, perl6 reading the audio stream, perl6 encoding to mp3 and perl6 streaming to the icecast server 11:11
RabidGravy if it's doing all those things reasonably, then it's "fast enough" and I don't give a stuff about no ten million iteration loop 11:13
perlawhirl it is playing
psch and it definitely sounds somewhat like acid 11:14
perlawhirl i lasted about 3 mins before my eyes started bleeding
psch hard to tell if it actually drops out or if that's the synth
as in, there's minor glitchy bits occassionally
RabidGravy possibly, I've got the latency optimistically low on the jackd 11:15
psch the tempo also doesn't seem completely consistent...
RabidGravy it probably isn't streaming quite right, but I'm hand calculating the note length in millis so that could be faulty too :) 11:17
right, I'll leave that going while I have a shower 11:18
timotimo i wonder if we can find out some more good stuff by doing a C-level profile of your workload 11:21
ideally, when we have a big bunch of arithmetic operations, we'd be 100% in jit-land, but since things aren't always inlined nicely, we'll also be having some invocation overhead, increasing and decreasing frame reference counters etc 11:22
RabidGravy yeah 11:56
anyway I had a realisation when I was in the shower
timotimo keep it real! 11:57
perlawhirl the best code ideas happen in the shower, on the toilet, or while sleeping 11:58
timotimo i agree with that statement
perlawhirl who knows what would happen if you fell asleep in the shower and pee'd on yourself! :D
timotimo :\
RabidGravy I wasn't thinking hard enough about the realtime synthesis thing, and if the Meeblip can do a good virtual analgue synthesis on a 16Mhz AVR in ~1000 lines of AVR assembler then it *must* be possible in a Perl 6 program 11:59
going to kill the stream now 12:00
want to put it at a more sensible latency value and also have jack use the dummy driver so I can listen to the stream 12:01
also later I'm going to haul out a real synthesizer and a mixer as this could be quite awesome 12:05
or synthesizer*s* 12:08
right off out for a bit anyway 12:11
timotimo where was the code for the sine sound again?
timotimo twitter.com/HourlyCats/status/7110...8873508864 - this is approximately how awake i feel right now 12:17
andreoss m: my \i =$= 1; say ++i; 13:17
camelia rakudo-moar 319ec8: OUTPUT«2␤»
andreoss m: my \i =$(1); say ++i;
camelia rakudo-moar 319ec8: OUTPUT«Cannot call prefix:<++>(Int); none of these signatures match:␤ (Mu:D $a is rw)␤ (Mu:U $a is rw)␤ (Int:D $a is rw)␤ (int $a is rw)␤ (Bool $a is rw)␤ (Num:D $a is rw)␤ (Num:U $a is rw)␤ (num $a is rw)␤ in block <uni…»
andreoss m: my \i :=$= 1; say ++i; 13:18
camelia rakudo-moar 319ec8: OUTPUT«2␤»
psch m: use nqp; my \i =$= 1; say nqp::iscont(i); my \j =$(1); say nqp::iscont(j) 13:19
camelia rakudo-moar 319ec8: OUTPUT«1␤0␤»
andreoss what's $?
psch andreoss: 'i = $ = 1' assigns an anonymous scalar and assigns 1 to that
andreoss: 'i = $(1)' assigns an itemizes single-item list 13:20
jnthn An anonymous state variable. Which means that the =$= idiom is not threadsafe. :)
(Or at least, it's risky depending what you're doing with it)
jnthn Together with it looking darn weird that's a good technical reason to consider avoiding it. ;) 13:21
andreoss m: my \a =@= 1,2,3; a.push: 4; say a; 13:22
camelia rakudo-moar 319ec8: OUTPUT«[1 2 3 4]␤»
jnthn m: sub foo($i) { my \a =@= 1,2,3; foo(0) if $i; a.push: 4; say a; }; foo(1) 13:24
camelia rakudo-moar 319ec8: OUTPUT«[1 2 3 4]␤[1 2 3 4 4]␤»
andreoss when i should use this stuff? they do make sigiless variables more friendly for mutable use 13:25
psch m: my \a = [1,2,3]; a.push: 4; say a
camelia rakudo-moar 319ec8: OUTPUT«[1 2 3 4]␤»
andreoss m: my \a = 1, 2, 3; a.push: 4; say a 13:26
camelia rakudo-moar 319ec8: OUTPUT«Cannot call 'push' on an immutable 'List'␤ in block <unit> at /tmp/q_pAan7Buw line 1␤␤»
jnthn andreoss: Yes, but it goes against the cultural trend to use them for immutable things...
shinobicl r: my @rr1 = Nil xx 5; say @rr1.perl; my @rr2 = Any xx 5; say @rr2.perl 13:27
camelia rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
..rakudo-moar 319ec8: OUTPUT«[Any, Any, Any, Any, Any]␤[Any, Any, Any, Any, Any]␤»
shinobicl say Any ~~ Nil;
r: say Any ~~ Nil;
camelia rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
..rakudo-moar 319ec8: OUTPUT«False␤»
jnthn andreoss: It's unfriendly for a reason :) 13:28
shinobicl Hi all, i need to discriminate for an empty value and undefined value, Why i can not have an array of Nil?
jnthn shinobicl: Because Nil's assignment behavior is "set this thing back to its default"
shinobicl: Maybe use :exists?
shinobicl r: say Any ~~ Mu; say Nil ~~ Mu; 13:29
camelia rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
..rakudo-moar 319ec8: OUTPUT«True␤True␤»
jnthn m: my @a is default(42); @a = Nil xx 5; say @a;
camelia rakudo-moar 319ec8: OUTPUT«[42 42 42 42 42]␤»
jnthn m: my @a; @a[1] = 5; say @a[0]:exists; 13:30
camelia rakudo-moar 319ec8: OUTPUT«False␤»
shinobicl So, how can i be sure if something is undefined?
I mean 13:31
jnthn The .defined method, or the defined function, or use with/without
shinobicl Ok lets say that i have an array of 10 elements, but only with 5 defined, what should i put in the other 5 to say that is not yet assigned? 13:32
psch m: my @a[10] = ^5; say @a[5]:exists 13:33
camelia rakudo-moar 319ec8: OUTPUT«False␤»
psch m: my @a[10] = ^5; say @a[5]:exists; say @a[0..4]:exists
camelia rakudo-moar 319ec8: OUTPUT«False␤(True True True True True)␤»
jnthn shinobicl: Just don't assign to them at all 13:35
Or if you already did then :delete them 13:36
(I assume you need ordering and thus didn't go for the "just use a hash" approach) 13:37
vendethiel azawawi: sorry, had to run 13:44
shinobicl r: my @a[10] = ^5; say @a.perl 14:00
camelia rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
..rakudo-moar 319ec8: OUTPUT«Array.new(:shape(10,), [0, 1, 2, 3, 4, Any, Any, Any, Any, Any])␤»
shinobicl r: say Int ~~ Any; 14:01
camelia rakudo-moar 319ec8: OUTPUT«True␤»
..rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
shinobicl r: my @a[10] = ^5; say @a.perl; say @a[0] 14:02
camelia rakudo-moar 319ec8: OUTPUT«Array.new(:shape(10,), [0, 1, 2, 3, 4, Any, Any, Any, Any, Any])␤0␤»
..rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
shinobicl r: my @a[10] = ^5; say @a.perl; say @a[0]:exists; say @a[9]:exists; 14:03
camelia rakudo-moar 319ec8: OUTPUT«Array.new(:shape(10,), [0, 1, 2, 3, 4, Any, Any, Any, Any, Any])␤True␤False␤»
..rakudo-jvm 978754: OUTPUT«cannot connect to eval server: Connection refused␤»
shinobicl thanks! :) i think i got it now
FROGGS_ shinobicl: btw, you can use "m:" to query rakudo-moar only 14:05
shinobicl how can i define a method callable like :exists? 14:06
i'm not familiar with this $var:something syntax 14:07
geekosaur that's not a method, it's a named/keyword parameter to something else. in @x[n]:exists it is a named parameter to the postcircumfix:<<[ ]>> method 14:10
shinobicl Ahhhh nice! so when i define the postcincumfix op i just define a named argument called :exists and see if "$exists" it is defined or not 14:12
geekosaur yep
shinobicl :D
thanks! 14:13
psch m: sub postcircumfix:<' '>($it, $arg, :$double) { say $double ?? 2 * $it * $arg !! $it * $arg }; 2'3'; 2'3':double 14:14
camelia rakudo-moar 319ec8: OUTPUT«5===SORRY!5=== Error while compiling /tmp/AUrHjKD5d0␤Unable to parse expression in postcircumfix:sym<' '>; couldn't find final $stopper ␤at /tmp/AUrHjKD5d0:1␤------> 3 ?? 2 * $it * $arg !! $it * $arg }; 2'3'7⏏5; 2'3':double␤»
RabidGravy Rarr!
psch huh
m: sub postcircumfix:<' '>($it, $arg, :$double) { say $double ?? 2 * $it * $arg !! $it * $arg }; 2'3'
camelia rakudo-moar 319ec8: OUTPUT«6␤»
psch m: sub postcircumfix:<' '>($it, $arg, :$double) { say $double ?? 2 * $it * $arg !! $it * $arg }; 2'3':double
camelia rakudo-moar 319ec8: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lx3nwuMA77␤Unable to parse expression in postcircumfix:sym<' '>; couldn't find final $stopper ␤at /tmp/lx3nwuMA77:1␤------> 3 $it * $arg !! $it * $arg }; 2'3':double7⏏5<EOL>␤ expecting any of:…»
RabidGravy there, glitch free and everything - rabidgravy.com:8000/stream 15:04
masak RabidGravy: what's that? 15:11
RabidGravy a perl program sequencing nekobee live and another perl program encoding the output and streaming it 15:13
but it's lost it now
masak Perl 5? 15:15
RabidGravy er perl 6
masak woo
RabidGravy the problem happens when the GC kicks in and it doesn't have enough time headroom to catch back up again 15:16
masak today's autopun spotting: xkcd.com/1657/
lucs m: use Test; is(15 + 27, 42) # This is fine. 15:17
camelia rakudo-moar 319ec8: OUTPUT«ok 1 - ␤»
lucs m: require Test; is(15 + 27, 42) # But how does ｢require｣ work?
camelia rakudo-moar 319ec8: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Hk_O5lXaCZ␤Undeclared routine:␤ is used at line 1␤␤»
masak RabidGravy: I wonder what kind to GC setting might help in such cases where non-stuttering matters more than being conservative with memory.
ugexe require doesn't import anything
masak problem is, you have to GC at some point, I guess.
lucs m: require Test; Test::is(15 + 27, 42) # ? 15:18
camelia rakudo-moar 319ec8: OUTPUT«Could not find symbol '&is'␤ in block <unit> at /tmp/QgcI3c_ZTr line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/QgcI3c_ZTr line 1␤␤»
ugexe put your entire program in DESTROY and hope its not collected
`is` is not `our` scoped within Test
lucs ugexe: Not sure what to do with that info :/ 15:20
ugexe im not sure what you are trying to do 15:21
lucs I'm trying to ｢require Test｣ and use its functions. 15:21
RabidGravy lucs, you have to call import 15:21
RabidGravy otherwise the functions "don't exist" in your scope 15:22
ugexe m: require Test <&is>; is(1,1)
camelia rakudo-moar 319ec8: OUTPUT«ok 1 - ␤»
lucs Aha, thanks. 15:22
ugexe the import statement is implied by the <&is> 15:22
lucs How would I explicitly invoke the 'import'? 15:23
RabidGravy import Test 15:25
lucs Oh, okay.
ugexe design.perl6.org/S11.html#Exportation # most of this is valid 15:26
lucs Thanks
ugexe if `require Test <&is>;` doesn't work on your local perl6 you probably need to upgrade, as I think that type of symbol importing was added in the last month or two 15:28
lucs ugexe: Yeah, works fine (I have a rather recent build). 15:29
rindolf Hi all. Who here is Wga Van Dick?
s/Dick/Dijk/ - sorry. 15:30
lucs: hi , sup?
ugexe thats wendy isnt it?
lucs rindolf: Same old same old, but with Perl 6 :)
RabidGravy masak, I could foresee some sort of "I need 0.001 secs" directive for a block and the GC won't take time below that 15:31
lucs rindolf: Yes, Wendy (according to google).
rindolf lucs: ah.
RabidGravy and not here right now
rindolf lucs: what is her nickname?
RabidGravy: ah.
ugexe is IO::Spec extendable such that it could be made to accept file uri? (file://<...>)? 15:35
rudi_s Is there a way to get the native file descriptor of Async::Proc's stdin? 15:39
skids There wasn't as of a week ago IIRC. 15:40
jnthn rudi_s: No; we don't actually expose the handles at all
mst jnthn: I can't get the fdno out of a filehandle? 15:41
rudi_s jnthn: Ok. - So I guess there's no way to use a native C function which writes to a file descriptor and use that with Async::Proc? I tried to use it with plain run but I get deadlocks when the programs writes to stdout and stderr and reads from stdin. 15:42
Which is kind-of obvious when you think about. I then try to use Supply on $proc.out and $proc.err, but that hangs when I run $proc.out.Supply.tap(...). 15:43
jnthn mst: Out of a file handle yes, but Proc::Async doesn't expose the handles at all. In no small part 'cus it's hard to imagine anything going well when libuv is "owning" them 15:44
mst aha, ok, this is probably fair enough
jnthn rudi_s: Are you trying to work around something we're missing? 15:45
mst I suspect the use cases I'm thinking of will involve me nativecalling fork+exec myself anyway 15:46
rudi_s jnthn: Well, maybe.
rudi_s I want to run a program, write to its stdin and get its stdout/stderr. 15:46
In addition I need the native file descriptor of stdin because I want to call a C function which writes to an fd. 15:47
But as a I said, I can't get it to work with just run because of deadlocks and Proc::Async has no way to get the native file descriptor. 15:48
pmurias masak: there are different GC for situations where non-stuttering is important 15:49
RabidGravy: is your streaming program up for reading somewhere? 15:50
RabidGravy yeah, it's an example in Audio::PortAudio 15:51
github.com/jonathanstowe/Audio-Por...eam-source
jnthn rudi_s: Yeah, Proc::Async would handle that great except the native descriptor bit... 15:54
rudi_s jnthn: Yeah. Any idea how I could get it to work with run without the deadlock?
jnthn rudi_s: Well, do you know how much data you're going to get back? 15:55
rudi_s jnthn: No.
jnthn rudi_s: OK. Then not sure. 16:03
jnthn gotta go take care of home stuffs...
bbl
rudi_s jnthn: Ok, thanks anyway. For now I have a workound and write to a temp file and then read that in and use Proc::Async. But it's not nice. 16:04
azawawi hi 16:43
im getting some weird behavior on Proc::Async. Is it working on Mac OSX? 16:44
travis-ci.org/azawawi/perl6-seleni...9148#L3596
ugexe azawawi: all the problems i had with it on osx boiled down to: rt.perl.org/Public/Bug/Display.html?id=125758 16:49
azawawi reads 16:50
ugexe i.e. osx has some problem with `await Promise.allof(@promised)`, but `for @promised { await $_; }` works, with the difference being the first one doesnt seem to sink all the promises so they never get out of Planned state 16:51
and that would cascade into all sorts of other errors 16:53
azawawi do you have a MacOS X development environment?
RabidGravy perl6 channeling Junior Mance on speed rabidgravy.com:8000/stream
ugexe not now. i use macincloud.com when i need to test on osx
but $
ugexe a hack i found to also work was to call `?$p.result;` before using $p 16:55
so in your test, maybe `my $p = $process.start; ?$p.result; say("phantomjs...` 16:56
rudi_s jnthn: Btw. is it expected that tap on $proc.out.Supply will block or is this supposed to work? 17:00
azawawi ugexe: we should start doing rakudostar binary builds... because it is silly... we're using precious travis ci time (and developer time) waiting for test result
azawawi for rakudobrew 2016.xy, couldnt we update the travis ci container image to take ready-made binaries instead of slow source make/test 17:02
ugexe tony-o has shown you can archive the rakudo directory in a travis build and reuse it
azawawi ugexe: where is that? 17:03
ugexe github.com/tony-o/perl6-travis-bin
azawawi ugexe: thx
RabidGravy: awesome
RabidGravy that's what nearly a gig of piano sample sounds like ;-) 17:06
same program as before, using linux sampler instead of nekobee and streaming via darkice rather audio::portaudio 17:07
ugexe azawawi: there is also caching on travis, but i could never get it to work before my attention span moved on docs.travis-ci.com/user/caching/ 17:08
jnthn rudi_s: I suspect so, given handles currently can't be used across multiple threads
RabidGravy I say the same program, it's actually slightly weighted toward the third and fifth of the mixolydian 17:09
rudi_s jnthn: Multiple threads? I just have one thread which calls run. 17:10
jnthn rudi_s: I was explaining why $handle.Supply.tap blocks 17:13
rudi_s jnthn: I know. But why is this related to multiple threads? 17:14
TimToady prefers inverted hungarian, or whatever that's called that's used in Samson and Delilah, with a major third but a minor second
so two 1.5 step intervals, but between 2nd/3rd and 6th/7th 17:15
Phrygian or some such
RabidGravy I was just about to say phrygian, yep 17:16
TimToady likes raised 9th chords too :)
jnthn rudi_s: Well, *something* has to be reading from the handle and showing the read stuff into the Supply 17:17
*shoving
rudi_s jnthn: Oh, so this is handled by threads? I expected something like epoll or similar. 17:18
azawawi btw, Samson in Arabic is pronounced as Shamshon (شمشون)
RabidGravy sounds like Sean Connery saying it ;-) 17:19
jnthn rudi_s: Even then some thread needs to play event loop
rudi_s jnthn: The main "thread"?
jnthn rudi_s: Proc::Async, IO::Socket::Async, IO::Notification etc. all proxy operations to a single thread that serves as the event loop
rudi_s And that's not the normal main thread?
jnthn rudi_s: It then dispatches results into the queue of ThreadPoolScheduler 17:20
rudi_s: For non-async things, though, we don't handle them through an event loop-y thing at present
File handles are just normal buffered I/O
azawawi RabidGravy: it is like Shamshoon
jnthn (All this is subject to change.)
rudi_s jnthn: Ah, ok. Good to know. Thanks. 17:21
TimToady well, back to cleaning out the garage (and all the other infrastructural things I've been putting off for the last 15 years for some reason...)
jnthn ;-)
TimToady: 15 years would give a garrage time to accumulate a good amount of stuff, I'd guess :)
RabidGravy yeah we got a skip last year and dumped ten years worth of junk 17:22
TimToady oh, it was already accumulating for a number years before that :)
Ven azawawi: back
RabidGravy I spent three days shredding stuff 17:23
jnthn moved apartment after just *5* years in it last year, and was shocked how much more stuff he had than when he'd moved in there... :)
azawawi Ven: weclome back. So far I traced it to travis-ci.org/azawawi/perl6-seleni...9148#L3596
TimToady we'll probably wait till we get our Free Shredding Day in Mountain View
Ven I've seen you linked me a gist, but I've been unable to find it atm
RabidGravy :)
Ven azawawi: I don't even get that – I get "Starting phantomjs process" 17:24
jnthn I suspect I'll be moving again within the next 5 years...though for once I intend it *not* to be an international move. :-)
azawawi Ven: i will simplify it into a test scrit... give me a min plz 17:25
Ven no problem :). sorry for the delay.
(I reconnected a bit earlier but you were away)
azawawi Ven: gist.github.com/azawawi/c4c2f5b5b31906ea28ed 17:28
Ven: try this code and paste output please
RabidGravy If wanted to make something that was like an "enum Foo ..." but the values were lists is it just sub-class EnumMap? 17:29
azawawi Ven: sample output here (Linux) pasteboard.co/2o2oIRph.png 17:31
azawawi Ven: no need to worry about the delay. I was outside enjoying the beautiful spring sunny weather :) 17:33
RabidGravy Boo, it's freezing and yuck in London still 17:34
Ven azawawi: phantomjs path: /Users/ven/.opam/system/bin/phantomjs phantomjs port: 5555 phantomjs returned Proc::Async promise: Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned) 17:34
azawawi RabidGravy: hehe :)
RabidGravy we seem to be going through a period of one day of pleasant spring-like weather and three days of winter 17:35
azawawi Ven: Is there any DEBUG output like "Config - init - Configuration..."... see the above picture link 17:36
Ven azawawi: "Config - config.init - {"ip":"127.0.0.1","port":"5555","hub":null,"proxy":"org.openqa.grid.selenium.proxy.DefaultRemoteProxy","version":"","logFile":null,"logLevel":"DEBUG","logColor":false}"
azawawi RabidGravy: it was raining but today is sunny and green :)
Ven: strange 17:37
Ven: 'git pull' the selenium git repo please and test it 17:39
Ven trying t/02
azawawi: "phantomjs returned Proc::Async promise: Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned)" but it's using the wrong port 17:40
azawawi Ven: no it is using a random free empty port 17:40
Ven oke :) 17:41
azawawi Ven: github.com/azawawi/perl6-selenium-...e.pm6#L818
Ven hahaha 17:42
azawawi Ven: can you please pastebin the following command: prove -v -e 'perl6 -Ilib' t/02-phantomjs.t ? 17:48
Ven not sure why you go through `prove` in this case, but... 17:49
azawawi: gist.github.com/vendethiel/6df07e0ed75749643264 17:50
azawawi Ven: sleep is working correctly on Mac OS X? 18:01
Ven yes :) 18:02
azawawi Ven: does the test take time to finish or right away?
Ven azawawi: it takes 32.8s to run (real) 18:03
azawawi Ven: but no output from phantomjs process...
Ven indeed 18:04
azawawi Ven: try to curl it from another shell while the test is running gist.github.com/azawawi/e30986625e31cae4f2e9 18:06
Ven that worked 18:08
azawawi aha
same random port?
and killall phantomjs # just in case 18:09
Ven no, I mean – the curl worked. but that didn't make the test work.
azawawi make sure no phantomjs process is still there by killing them all: killall phantomjs 18:10
then try that curl command
dalek osystem: 3c5d499 | (Fernando Correa de Oliveira)++ | META.list:
Update META.list
 18:12
osystem: 33ae8ba | azawawi++ | META.list:
Merge pull request #174 from FCO/patch-1

Update META.list
TreyHarris is there shorthand for one grammar referring to another grammar's TOP in one of its rules? 18:13
azawawi Ven: Try add :port(5555) parameter to github.com/azawawi/perl6-selenium-...omjs.t#L45 18:14
Ven now there's only one test failing: "not ok 86 - Only One session should be there" line 57 18:15
azawawi Ven: maybe the free port range is somehow bad for osx
Ven: good 18:16
Ven: let it run them without curl-ing
Ven huh?
azawawi Ven: s/them/then
Ven: that curl http post is actually requesting a new session and the test checks for only one session which it has created earlier 18:17
Ven oh. I didn't curl it. It just worked 18:18
azawawi good then empty-port sucks on macosx :) 18:20
i wonder why
My guess is this assumption is bad on osx github.com/azawawi/perl6-selenium-...e.pm6#L829 18:22
Ven: thanks a lot :) 18:25
Ven++
Ven good!
azawawi Ven: also we're testing for an empty port against 127.0.0.1:XYZ ... maybe phantomjs is binding itself to something else 18:27
azawawi waits for travis ci results 18:28
dbohdan How production-ready is Perl 6 on the JVM for web development? I am looking for a fun, expressive JVM language in which to write a small (public) web application, and Perl 6 seems like fun. :-) 18:32
timotimo TreyHarris: no, you'll have to spell out the "TOP" part again i'm afraid
RabidGravy just for TimToady rabidgravy,com:8000/stream now playing in Phrygian mode 18:33
timotimo i have no idea what that means :|
RabidGravy the phrygian part? 18:34
it's one of the greek musical modes 18:35
TreyHarris timotimo: okay. is it an antipattern to be doing things like parsing a doc that can include subdocs in HERE-like format by referring to the subdoc's grammar's TOP? or just something which is too uncommon for syntactic sugar to be necessary?
timotimo well, we're using multiple grammars in rakudo, too 18:36
RabidGravy flattened second and fourth compared to a major scale
timotimo every quote is already a sub-language
so are regexes
lucs RabidGravy: How was that composed? (re. Phrygian stream) 18:37
RabidGravy it's a perl 6 program sequencing linuxsampler live 18:38
one of the examples from the forthcoming Audio::PortMIDI :)
TreyHarris timotimo: right, that's why I asked because I'd looked in nqp for examples of referring to another TOP and didn't find any 18:39
timotimo: but i was just doing 'ack --nqp TOP', so could've missed something 18:40
timotimo hm, look for something like nibble or something? 18:40
i haven't tried to understand the mechanism yet, because i never needed to modify anything about it 18:42
TreyHarris timotimo: yes, i see nibble. that seems more self-documenting anyway. thanks! 18:43
kid511 What is the best way to report a typo in doc.perl6.org/routine/spurt ? 18:47
RT? 18:48
timotimo nah, it's better to go to the github repo; there's a link at the far bottom
azawawi kid511: github.com/perl6/doc/issues 18:50
kid511 azawawi: Thanks. Created github.com/perl6/doc/issues/427 18:55
kid511 In the REPL, the following statement fails to compile: 19:13
kid511 given prompt "Enter a number: " { when * < 0 { say "Negative"; } when * > 0 { say "Positive"; } default { say "Zero"; } } 19:14
It will only compile when I put ';' after the '}' if the first 2 'when' blocks.
given prompt "Enter a number: " { when * < 0 { say "Negative"; }; when * > 0 { say "Positive"; }; default { say "Zero"; } }
However, those semicolons are *not* necessary if I enter this same code into a file and run it. Anyone know why? 19:15
timotimo because you've put newlines after the } probably 19:21
a } at the end of a line gives you a ; for free
kid511 thx 19:24
lucasb about the Bool/defined methods from yesterday, I think lizmat's commit only fixed half of the problem 19:27
lucasb only Bool(Mu:D:) was changed, but I think the same thing applies to Bool(Mu:U:) 19:27
timotimo please elaborate? 19:28
lucasb m: say ?class { method defined { True } }
camelia rakudo-moar 319ec8: OUTPUT«False␤»
lucasb star: say ?class { method defined { True } }
camelia star-m 2016.01: OUTPUT«True␤»
timotimo i don't remember 100%, but didn't we say you have to implement method Bool if you want ? to behave differently?
RabidGravy yeah I thought that was what was decided 19:29
lucasb if you append ".new" to my examples, you will see I don't need to define Bool 19:30
so, the behavior is different is you ask a type object or a instance
I can accept it, if someone says that's how things will be from now on :) 19:31
*the behavior is different if you ask
timotimo hmm 19:32
i'm not 100% convinced any more this is all right 19:33
lucasb yeah, it's a little confusing to be able to change the definedness of some object... 19:35
timotimo well, it's useful for things that act like proxies to other things 19:41
and are supposed to be transparent
what you can't change is the DEFINITEness
lucasb right
timotimo in general, it's bound to cause confusion one way or the other 19:42
masak accidentally writes !~ and gets told off by perl6 20:17
rudi_s Is it possible to automatically initialize a class member on .new call? I have something like has SetHash $.foo and I'd like to get a SetHash after I call new. Is this possible? 20:21
timotimo yeah, just write "has SetHash $.foo = SetHash.new"
masak m: class C { has $.foo = 42 }; say C.new.foo 20:23
camelia rakudo-moar 319ec8: OUTPUT«42␤»
masak rudi_s: ^^
masak m: class C { has $.foo = (1..6).roll }; say C.new.foo for ^3 20:24
camelia rakudo-moar 319ec8: OUTPUT«2␤3␤5␤»
rudi_s Oh, nice. Didn't know that this works. Awesome. Thank you timotimo, masak. 20:27
masak it's nice when you know about it, yes
one way to explain it is to say that the `= 42` and `= (1..6).roll` bits are *thunks* and they run in a separate *phase* (namely, at object construction time)
RabidGravy yeah and a nice implication of that is you can use other attributes in the initialiser 20:38
m: class F { has Int $.one is required; has Int $.two = do { $!one * 2 }; }; say F.new(one => 3).perl
camelia rakudo-moar 319ec8: OUTPUT«F.new(one => 3, two => 6)␤»
moritz m: class F { has Int $.one is required; has Int $.two = $!one * 2 }; say F.new(one => 3).perl 20:39
camelia rakudo-moar 319ec8: OUTPUT«F.new(one => 3, two => 6)␤»
moritz no need for that do block
RabidGravy yeah, superstitious on my part 20:40
BenGoldberg m: class Foo { my SetHash $.foo .= new; }; Foo.new.say;
camelia rakudo-moar 319ec8: OUTPUT«Foo.new␤»
BenGoldberg m: class Foo { my SetHash $.foo .= new; }; Foo.new.foo.say; 20:41
camelia rakudo-moar 319ec8: OUTPUT«SetHash.new()␤»
lucasb m: class C { has $.a = $!b; has $.b = 42 }; say C.new 20:42
camelia rakudo-moar 319ec8: OUTPUT«C.new(a => Any, b => Any)␤»
lucasb ^^ interesting, no? :D 20:43
RabidGravy the order of the attributes is important there
lucasb I could expect C.new(a => Any, b => 42) 20:43
masak the `do` block is unnecessary exactly *because* it's a thunk 20:44
RabidGravy lucasb, I missed that, I can however understand why it is happening 20:46
Begi Can we imagine, in the future, a translation of doc ? 21:00
*of the Perl 6 doc
TreyHarris "has Int(Cool) $.count is rw;" is "insufficiently type-like". Is there a way to make it dwim? 21:02
masak Begi: a translation in general, or to some particular language?
RabidGravy I can imagine it, but it won't happen unless someone else does it
masak Begi: either way, I'm pretty sure we can imagine it ;)
Begi It'd be super !
I'm French, so I can help with French
step by step 21:03
spider-mario so can I 21:04
masak merveilleux! 21:06
TreyHarris p6: class CoolInt { has Int(Cool) $.x is rw; }; say CoolInt.new(:x("3")).perl;
camelia rakudo-moar 319ec8: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tmpfile␤Coercion Int(Cool) is insufficiently type-like to qualify a variable␤at /tmp/tmpfile:1␤------> 3class CoolInt { has Int(Cool) $.x is rw7⏏5; }; say CoolInt.new(:x("3")).perl;␤ expecting …»
Begi masak : you're German, right ? 21:07
spider-mario I programmed with ints before they were cool
masak Begi: no, I'm from .se
Begi: you might be thinking of moritz; easy to confuse
Begi ah yes, wouupps :) 21:08
masak easy to confuse one for the other, I mean -- it's not moritz who is easily confused... :P
masak is easy to confuse, though
Begi so, do you think I should start the translation of some parts of the documentation ? (in French) 21:09
RabidGravy yes, just doing things is better than waiting for everyone to say yes :) 21:10
Begi sure :) 21:14
awwaiid m: for <100 10_000 100_000> -> $batch { my $start = now; ((10..99) X (10..99)).map({$_[0] * $_[1]}).grep(-> $n { $n.Str eq $n.Str.flip }).sort[* - 1].say ; my $end = now; say "Batch $batch took { $end - $start } seconds" } 21:16
camelia rakudo-moar 319ec8: OUTPUT«9009␤Batch 100 took 0.2074571 seconds␤This Seq has already been iterated, and its values consumed␤(you might solve this by adding .cache on usages of the Seq, or␤by assigning the Seq into an array)␤ in block <unit> at /tmp/w2TzV4qVr2 line 1␤␤»
awwaiid Which Seq is it talking about? the ((10..99) X (10..99)) ?
If so, why did it get re-used between iterations of the for-loop? 21:17
timotimo awwaiid: perhaps the sort made it problematic? 21:27
or the [* - 1]
TreyHarris Hm... I'm trying to parse a file format where the final line is special, and it can be anything, including something that would ordinarily be valid if the file were to continue. Is that a situation I can only deal with out-of-band (either by taking the last line off the string before parsing, or by removing it from the parse tree after parsing), or is there some way in a grammar to say that? 22:21
TreyHarris oh, duh, I just make TOP a regex instead of something racheting. I imagine that will kill my performance though. 22:23
mst 22:38

Powered by Cro and the Raku® Programming Language.

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

Thank you!