»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.
japhb pTk is a *p*ortable version of the Tk part of TCL/TK. It was made so that people could do perl5 + Tk 00:00
timotimo oh
so the tcl portion ripped out and bound via XS?
japhb yeah, with a bunch of magic because Tk *really* didn't expect that particular ripping apart. 00:01
timotimo hah
i can imagine that
i'm not sure i'd be excited for any Tk to be ported to perl6
it's just ... it's tk! :P
japhb Uh, no.
This was far more interesting in the 90's than today. :-)
timotimo have you followed opengl in the recent years? 00:04
and looked at vulkan at all yet?
japhb I stopped watching OpenGL closely around the mid-3.x period 00:05
Basically, I don't know any 3D artists, and don't have time to become one myself, so an engine without cool content becomes kinda boring after your third or fourth one. :-) 00:06
timotimo 'k
yeah, i can really imagine ;(
just today i looked at godot; there's a silly benchmark that just spawns a bunch of bunnies that bounce around the screen like bouncy balls 00:07
japhb Heh 00:08
Today I recommended a very good graphics theory book to someone, and they said "Yeah, the person who wrote that was my thesis advisor"
Talk about a deflating retort ... 00:09
timotimo and i apitraced it and it turns out it generates vector arrays for every single one of them and pushes them to the driver
on the way redundantly setting state over and over again
japhb Eww.
timotimo t.h8.lv/godot_apitrace.png 00:10
that's the better part, there's worse parts :)
japhb Oh dear
There are extensions (and heck, base APIs) for that ... 00:11
timotimo yes, absolutely
this targets OpenGL ES 2, iiuc
japhb Oh ah.
Yeah, ES 2 is barely in the shader era 00:12
Lots of stuff people expect just isn't there.
timotimo oh, is that so?
that's rough :\
japhb I remember when ES 2 came out, wondering if I should get into handheld engine development, and then realizing if I did it on Android I'd have to do it in Java, and going "Nope, not gonna happen." 00:13
timotimo you don't have to; you can do it in C or C++ 00:14
japhb timotimo: Yes, now.
timotimo ok that may be new
japhb IIRC at the time that was not so well supported.
That was years ago.
geekosaur even back then people were playing with android programming with other jvm languages (I recall helping someone figure out clojure for it) 00:17
and, hm, something else that ended up being a non-starter because the runtime glue was too big and they hadn't worked out a way to minimize to the stuff absolutely needed yet 00:18
timotimo scala?
geekosaur no
00:19 cdg joined
timotimo well, mobiles are an "interesting" target for game development anyway, because the fill rates are just so abysmal on typical hardware 00:21
so overdraw is ridiculously costly
skink You ever write something, and not really know why you wrote it?
$ dd if=/dev/urandom bs=1 count=8 of=/tmp/bin 00:22
$ p6-pgp-words /tmp/bin
wallet concurrent python vacancy ...
00:23 vendethiel joined 00:24 BenGoldberg joined
geekosaur can think of reasons for what that appears to be doing... 00:25
although count=8 seems rather low entropy :p
skink The PGP wordlist is a set of 512 words, representing each byte in odd and even positions 00:28
Phonologically distinct, so you can transmit a key or fingerprint spoken
timotimo neat 00:32
geekosaur international radio phonetic alphabet for pgp keys, cute >.>
skink I decided to make a Crypt::Misc to just throw random crap in 00:34
00:34 BenGoldberg left 00:35 BenGoldberg joined
timotimo Cryptic Mist 00:39
00:41 nbrown joined, wamba left
timotimo ... helps to compile moarvm with an --optimize higher than 0 00:44
00:46 cdg left 00:47 cdg joined 00:50 nbrown left 00:51 cdg left 00:52 vendethiel left 00:53 wamba joined
skink You know it's quality when you've got a 520-line module that's literally just a lookup table 00:55
00:57 arlenik joined
skink (Fun fact: the fastest library for interpreting HTML involves over 20k lines of hand-written lookup tables) 00:58
arlenik I have a function that takes an argument whose type should only be of TypeA or TypeB. What would the signature of that function look like?
psch m: sub f($a where Int|Str) { $a }; f 5; f "foo"; f 5.4 00:59
camelia rakudo-moar a45224: OUTPUT«Constraint type check failed for parameter '$a'␤ in sub f at /tmp/E77i0CBgCC line 1␤ in block <unit> at /tmp/E77i0CBgCC line 1␤␤»
psch hm, is that compile time foiling me? 01:00
i'd think so...
m: sub f($a where Int|Str) { $a }; f 5; f "foo";
camelia ( no output )
01:00 matiaslina joined
psch arlenik: anyway, generally that's the pattern 01:00
arlenik: the alternative is declaring a subset that fits exactly your two types
m: subset StrInt of Any where Int|Str; sub f(StrInt $) { }; f "foo"; f 5 01:01
camelia ( no output )
arlenik psch: thanks. What if the functio had two arguments with the same exact constraints. Is there anyway to apply the constraint to both arguments without repeating 'where TypeA|TypeB' for both?
psch arlenik: although then you'd have to cheat (as i did) a bit
as in, a subset needs a base type first, and that base type can be constraint further 01:02
01:02 wamba left
arlenik psch: ok, i see. thanks very much. 01:02
psch in my example, i'm using Any, which is our "root type", as in what everything inherits from if there's no specific parent type
alternatively you could probably just create a new class that inherits from both of your allowed types 01:03
unless they conflict in some manner
m: class C { }; class D { }; class E is C is D { }; sub f(E $) { }; f C.new; f D.new
camelia rakudo-moar a45224: OUTPUT«Type check failed in binding <anon>; expected E but got C (C.new)␤ in sub f at /tmp/nU0VgIJk2D line 1␤ in block <unit> at /tmp/nU0VgIJk2D line 1␤␤»
psch oh, no 01:04
psch is actually not great with OOP vOv
maybe it'd work with subsets or some role shennanigans...
ZoffixWin Your E would fit into a C or D restriction, not the other way around. 01:07
psch oh, right
ZoffixWin m: class C { }; class D { }; class E is C is D { }; sub f(C $) { }; sub x(D $) { }; f E.new; x D.new
camelia ( no output )
psch m: class C { }; sub f(C $x, $y where *.isa($x.WHAT)) { }; f C.new, C.new # another somewhat weird way 01:10
camelia ( no output )
psch hm, i wonder if that can be parameterized...
eh, no need, actually
but the semantics are somewhat different anyway i suppose
because that one means $y has to be of $x type or a subtype of the type of $x 01:11
m: say Int.isa(Any)
camelia rakudo-moar a45224: OUTPUT«True␤»
01:18 cpage_ left, molaf_ joined 01:22 molaf left
timotimo m: say 9.9e0.Int 01:24
camelia rakudo-moar a45224: OUTPUT«9␤»
01:31 MasterDuke left
skink m: subset MyTypeA of Str where *.chars > 1; subset MyTypeB of Int where 1..10; sub thingy ($thing where MyTypeA|MyTypeB) { say $thing }; thingy 4 01:32
camelia rakudo-moar a45224: OUTPUT«4␤»
01:33 kid51 joined
skink Ah, yeah, psch has the shorter, cheatier way 01:34
arlenik, Does the function do different things depending on the input type? You may be able to use a multi sub
BenGoldberg m: my ([$a]); $a.WHAT.say 01:37
camelia rakudo-moar a45224: OUTPUT«Cannot call method 'say' on a null object␤ in block <unit> at /tmp/FMj7GpxRQj line 1␤␤»
BenGoldberg m: my ([$a]); say $a.WHAT
camelia rakudo-moar a45224: OUTPUT«(Mu)␤»
BenGoldberg wonders why it's Mu instead of Any. 01:38
m: Mu.say
camelia rakudo-moar a45224: OUTPUT«(Mu)␤»
BenGoldberg m: my Mu $a; $a.say;
camelia rakudo-moar a45224: OUTPUT«(Mu)␤»
01:47 adhoc joined 01:59 MasterDuke joined 02:00 kid51 left 02:07 MasterDuke left 02:09 MasterDuke joined
BenGoldberg m: my @a; say @a.splice: 0, 1; 02:09
camelia rakudo-moar a45224: OUTPUT«[]␤»
BenGoldberg m: my @a; say @a.shift;
camelia rakudo-moar a45224: OUTPUT«Cannot shift from an empty Array␤ in block <unit> at /tmp/_S4v8hLWS1 line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/_S4v8hLWS1 line 1␤␤» 02:10
02:10 BenGoldberg left, pecastro joined 02:12 BenGoldberg joined
ZoffixWin m: my @a = [],; say @a.shift; 02:17
camelia rakudo-moar a45224: OUTPUT«[]␤»
BenGoldberg m: my @a; say @a.shift; 02:18
camelia rakudo-moar a45224: OUTPUT«Cannot shift from an empty Array␤ in block <unit> at /tmp/AMtIHyCI_d line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/AMtIHyCI_d line 1␤␤»
BenGoldberg m: [].shift;
camelia rakudo-moar a45224: OUTPUT«Cannot shift from an empty Array␤ in block <unit> at /tmp/ZufodnO_Bg line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/ZufodnO_Bg line 1␤␤»
BenGoldberg What's wrong with shifting from an empty array?
What if I'm happy getting an (Any) or whatever? 02:19
ZoffixWin What would you return?
m: my @a = Any,; say @a.shift;
camelia rakudo-moar a45224: OUTPUT«(Any)␤»
ZoffixWin it can be a legit element inside of an array and now you have an ambiguity
You can do @a and @a.shift
BenGoldberg Seems silly. 02:20
ZoffixWin Why?
BenGoldberg I'd also be happy with getting a failure object.
Err, unless that's what it's doing? 02:21
m: [].shift.WHAT.say;
camelia rakudo-moar a45224: OUTPUT«(Failure)␤»
BenGoldberg ooh
ZoffixWin :D
BenGoldberg m: so([].shift).say;
camelia rakudo-moar a45224: OUTPUT«False␤»
BenGoldberg m: gist.github.com/BenGoldberg1/467df...d16fb4af23 02:24
camelia rakudo-moar a45224: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229 233 239 241 251 257 263 269 271 277 281 283 293 307 311 313 317 331 337 347 349 353 3…»
02:24 vendethiel joined
BenGoldberg I know a falure in sink context creates an exception... and prefixing my '@!factors.shift' with 'so' defuses this... is there a better/more idiomatic way of doing this? 02:26
02:27 pecastro left 02:29 pecastro joined 02:30 pecastro left 02:31 pecastro joined, pecastro left
ZoffixWin It's kinda weird there's .uniname*s* but no .uniprop*s* 02:36
02:38 pecastro joined, pecastro left 02:39 noganex_ joined
BenGoldberg m: gist.github.com/BenGoldberg1/467df...d16fb4af23 02:41
camelia rakudo-moar a45224: OUTPUT«In 3 seconds, produced 9779 primes, from 2 to 102103␤» 02:42
02:42 pecastro joined
BenGoldberg m: my $howlong = 3; my $end = $howlong + now; my $cnt = 0; for ( ^Inf ) { ++$cnt if .is-prime; last if now > $end }; say $cnt; 02:42
camelia rakudo-moar a45224: OUTPUT«1099␤»
BenGoldberg grins.
02:42 noganex left
BenGoldberg m: 9779 / 1099 02:42
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/peG9nMYig5:␤Useless use of "/" in expression "9779 / 1099" in sink context (line 1)␤»
BenGoldberg m: say (9779 / 1099)
camelia rakudo-moar a45224: OUTPUT«8.898089␤»
BenGoldberg 's prime generator is almost 9x faster than repeated primality generation. 02:43
Err, repeated primality testing. 02:44
02:44 pecastro left 02:45 pecastro joined
ZoffixWin cool 02:45
m: say (^Inf .grep: *.is-prime)[0..9779]; say now - INIT now; 02:47
camelia rakudo-moar a45224: OUTPUT«(2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229 233 239 241 251 257 263 269 271 277 281 283 293 307 311 313 317 331 337 347 349 353 …» 02:48
ZoffixWin m: my $x = + eager (^Inf .grep: *.is-prime)[0..9779]; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«20.1472873␤» 02:49
sortiz_ m: my @a; with @a.shift { say "data" } else { say .DEFINITE ?? "failure" !! "undef" } ; # with is good to handle failures 02:51
camelia rakudo-moar a45224: OUTPUT«failure␤»
03:05 _notbenh left 03:09 _notbenh joined
BenGoldberg m: (^Inf .grep: *.is-prime)[9779].say; say now - INIT now; 03:18
camelia rakudo-moar a45224: OUTPUT«(timeout)»
BenGoldberg m: (^Inf .grep: *.is-prime)[0].say; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«2␤0.00637310␤»
BenGoldberg m: (^Inf .grep: *.is-prime)[1099].say; say now - INIT now; 03:19
camelia rakudo-moar a45224: OUTPUT«8831␤1.35728061␤»
BenGoldberg m: (^Inf .grep: *.is-prime)[9000].say; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«93187␤20.1244585␤»
BenGoldberg m: say now - INIT now; 03:20
camelia rakudo-moar a45224: OUTPUT«0.00159358␤»
BenGoldberg m: (^Inf .grep: *.is-prime && now < INIT { now + 3 } )[*-1].say; say now - INIT now; 03:21
camelia rakudo-moar a45224: OUTPUT«Cannot use Bool as Matcher with '.grep'. Did you mean to use $_ inside a block?␤ in block <unit> at /tmp/7Cwhu2b0yB line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/7Cwhu2b0yB line 1␤␤»
BenGoldberg m: (^Inf .grep: { *.is-prime && now < INIT { now + 3 } } )[*-1].say; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«(timeout)» 03:22
BenGoldberg m: (^Inf .grep(*.is-prime).first: now > INIT { now + 3 } ).say; say now - INIT now; 03:23
camelia rakudo-moar a45224: OUTPUT«Cannot use Bool as Matcher with '.first'. Did you mean to use $_ inside a block?␤ in block <unit> at /tmp/din7593oon line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/din7593oon line 1␤␤»
BenGoldberg m: (^Inf .grep(*.is-prime).first({ now > INIT { now + 3 } )).say; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BNUa1cpbfa␤Missing block␤at /tmp/BNUa1cpbfa:1␤------> 3s-prime).first({ now > INIT { now + 3 } 7⏏5)).say; say now - INIT now;␤ expecting any of:␤ statement end␤ statement modi…»
ZoffixWin I wonder why this times out: m: (^Inf .grep: { *.is-prime && now < INIT { now + 3 } } )[*-1].say; say now - INIT now;
Ah
BenGoldberg Cause it's ^Inf.grep
ZoffixWin Precedence
Oh 03:24
OK
BenGoldberg It's never going to create an end-of-list condition
ZoffixWin Right
m: (^Inf .grep: { *.is-prime and now < INIT now + 3 } )[*-1].say; say now - INIT now;
camelia rakudo-moar a45224: OUTPUT«(timeout)» 03:25
BenGoldberg m: my @naive := ^Inf .grep: *.is-prime; my $f = @naive.first: now > INIT now + 3;
camelia rakudo-moar a45224: OUTPUT«(timeout)» 03:26
BenGoldberg m: my @naive := ^Inf .grep: *.is-prime; my $f = @naive.first: now > INIT (now + 3);
BenGoldberg shrugs
camelia rakudo-moar a45224: OUTPUT«(timeout)WARNINGS for /tmp/JFqjN12hv3:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤»
03:27 cpage_ joined
BenGoldberg m: my @naive := ^Inf .grep: *.is-prime; my $f = @naive.first: (now > INIT (now + 3)); 03:27
camelia rakudo-moar a45224: OUTPUT«(timeout)WARNINGS for /tmp/YeFnSptnqR:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤» 03:28
BenGoldberg m: my @naive := ^13 .grep: *.is-prime; my $f = @naive.first: (now > INIT (now + 3));
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/iGf5g3txYU:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤Type check failed in binding; expected Positional but got Seq ((2, 3, 5, 7, 11).Seq)␤ in block <unit> at /tmp/iGf5g3txYU line 1␤␤»
BenGoldberg m: my @naive := ^13 .grep: *.is-prime; my $f = @naive.first: { now > INIT (now + 3) };
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/K1MDjvy0_r:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤Type check failed in binding; expected Positional but got Seq ((2, 3, 5, 7, 11).Seq)␤ in block <unit> at /tmp/K1MDjvy0_r line 1␤␤»
BenGoldberg m: my @naive := ^13; my $f = @naive.first: { now > INIT (now + 3) };
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/cs7jtG2_tl:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤»
BenGoldberg m: my @naive := ^13 .grep: True; my $f = @naive.first: { now > INIT (now + 3) }; 03:29
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/6GO2CTxrQm:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤Earlier failure:␤ Cannot use Bool as Matcher with '.grep'. Did you mean to use $_ inside a block?␤ in block <unit> at /tmp/6GO2CTxrQm line 1␤␤Fi…»
03:29 vendethiel left
BenGoldberg m: my @naive := ^13 .grep: .so; my $f = @naive.first: { now > INIT (now + 3) }; 03:29
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/eJSB0BwKMj:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤Earlier failure:␤ Cannot use Bool as Matcher with '.grep'. Did you mean to use $_ inside a block?␤ in block <unit> at /tmp/eJSB0BwKMj line 1␤␤Fi…»
BenGoldberg m: my @naive := ^13 .grep: *.so; my $f = @naive.first: { now > INIT (now + 3) };
camelia rakudo-moar a45224: OUTPUT«WARNINGS for /tmp/pxZRbAvokI:␤Useless use of "+" in expression "now + 3" in sink context (line 1)␤Type check failed in binding; expected Positional but got Seq ((1, 2, 3, 4, 5, 6, 7,...)␤ in block <unit> at /tmp/pxZRbAvokI line 1␤␤»
BenGoldberg m: 1.so.say
camelia rakudo-moar a45224: OUTPUT«True␤»
BenGoldberg m: my @naive := ^13 .grep: *.so; 03:30
camelia rakudo-moar a45224: OUTPUT«Type check failed in binding; expected Positional but got Seq ((1, 2, 3, 4, 5, 6, 7,...)␤ in block <unit> at /tmp/PF1i9JVKnA line 1␤␤»
BenGoldberg huh
m: my @naive := flat ^13 .grep: *.so;
camelia rakudo-moar a45224: OUTPUT«Type check failed in binding; expected Positional but got Seq ((1, 2, 3, 4, 5, 6, 7,...)␤ in block <unit> at /tmp/j9d3mDmDCQ line 1␤␤»
BenGoldberg m: my @naive := eager ^13 .grep: *.so;
camelia ( no output )
[Coke] aigh.
BenGoldberg m: my \naive = ^13 .grep: *.so; 03:31
camelia ( no output )
BenGoldberg m: my \naive := ^13 .grep: *.is-prime; my $f = naive.first: { now > INIT {now + 3} }; 03:32
camelia ( no output )
BenGoldberg m: my \naive := ^13 .grep: *.is-prime; say naive.first: { now > INIT {now + 3} };
camelia rakudo-moar a45224: OUTPUT«Nil␤»
BenGoldberg m: my \naive := ^Inf .grep: *.is-prime; say naive.first: { now > INIT {now + 3} };
camelia rakudo-moar a45224: OUTPUT«15581␤»
[Coke] m: say " Gentle reminder, dear reader, that you can use camelia with private messages."
camelia rakudo-moar a45224: OUTPUT« Gentle reminder, dear reader, that you can use camelia with private messages.␤»
skink What's the Perl6 syntax for a value I don't care about? e.g. 'let _ = whatever' in OCaml 03:42
Timbus $ = whatever() 03:43
doc.perl6.org/language/variables#T...4_Variable 03:44
[Coke] $ isn't quite nothing, it's an anonymous state variable. 03:48
skink: if you don't care about it, don't save it in a variable.
skink I had a given $x, $y where True, False and True, True did the same thing 03:49
So I wanted True, _ for that case basically
03:51 pecastro left
Timbus so I think you want Any ? 03:52
03:52 matiaslina left 03:54 pecastro joined, pecastro left 03:55 pecastro joined
skink Oh, hm 04:02
If I include a script in bin/ with a module, does it not see the same ?%RESOURCES as the modules?
%? *
04:03 SalamiTactics joined 04:07 aries_liuxueyang joined
aries_liuxueyang Hello, everyone. 04:07
I have a question: `my $a; my $b; $b := $a;` this can be run in script. but it can not be run on repl. 04:08
whey.
why
04:10 skink left
[Coke] works fine in my repl. 04:12
what does your "perl6 --version" say?
aries_liuxueyang `This is Rakudo version 2016.03-119-ga452244 built on MoarVM version 2016.03-108-gca1a21a 04:13
implementing Perl 6.c.`
this error: `===SORRY!=== Error while compiling <unknown file>
Cannot use bind operator with this left-hand side
at <unknown file>:1`
04:14 Cabanossi left
[Coke] Are you doing anything other than typing "perl6" and then typing in that line between the `` on a single line on the REPL? 04:14
aries_liuxueyang I type one line a time by hand. 04:15
04:16 Cabanossi joined
[Coke] I can duplicate that error message if I hit enter after each of the ; 04:16
aries_liuxueyang Yes, I hit enter after each of the ; 04:17
why? does it because of the repl? 04:18
04:18 skids joined
[Coke] I don't know exactly what the root cause there is, but in general, the REPL isn't quite the same as running a script. the context as you go from line to line isn't quite the same. 04:18
aries_liuxueyang okay, thanks. 04:20
besides, anyone use vim? my vim can not show syntax highlight 04:22
my vim version is 7.4
[Coke] I do, but I don't use syntax highlighting, usually. You probably want: github.com/vim-perl/vim-perl/blob/.../perl6.vim 04:29
04:29 nige1 joined
[Coke] Note that it's so old it references a VM we no longer use, so it might be out of date. 04:29
aries_liuxueyang [Coke]: thank you, I have seen that and will have a try now ;-) 04:32
04:34 sortiz_ left 04:37 nige1 left 04:45 molaf_ left 04:48 mr-fooba_ left 04:51 mr-foobar joined 05:09 skids left 05:11 BenGoldberg left 05:13 BenGoldberg joined 05:18 maybekoo2 left 05:27 pecastro left, pecastro joined 05:33 pecastro left 05:35 rurban joined
[Coke] m: say nativecast.WHAT 05:42
camelia rakudo-moar a45224: OUTPUT«5===SORRY!5=== Error while compiling /tmp/anO1BgKLLX␤Undeclared routine:␤ nativecast used at line 1␤␤»
[Coke] nativecast is referenced in the NativeCall doc
05:56 Vlavv_ left 06:05 wamba joined 06:08 cognominal_ joined 06:09 Vlavv_ joined 06:10 CIAvash joined
aries_liuxueyang [Coke]: did you use Pathogen to install vim-perl? 06:17
06:17 BenGoldberg left
[Coke] It was so long ago, I have no idea, sorry 06:19
I presume I did whatever the README told me.
aries_liuxueyang it's weird and it does not start to work by Pathogen way. 06:21
Maybe I can try another way. thank you. ;-) 06:22
06:29 cognominal_ left, Cabanossi left 06:31 cognominal joined 06:32 Cabanossi joined 06:35 wamba left 06:43 jjido joined 06:48 rindolf joined
ZoffixWin I'm guessing there's something special with exceptions when the caller is a sub in .nqp file? Looking at #127883 and see a call from load_module() in src/Perl6/World.nqp is involved. 06:49
m: CompUnit::RepositoryRegistry.head.need( CompUnit::DependencySpecification.new( :short-name<Foo> ) ); 06:50
camelia rakudo-moar a45224: OUTPUT«Could not find Foo in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-2/share/perl6␤ CompUnit::Repository::AbsolutePath<140328410…»
ZoffixWin ^ and that does have a line number, but with `use Foo`, there's no line number
06:52 rurban left, rurban joined
ZoffixWin Screw it. Over my head. 06:55
06:59 azawawi joined
azawawi hi 06:59
ZoffixWin: ping
ZoffixWin pong
azawawi ZoffixWin: my mind is attached to my email. i just woke up :) 07:00
ZoffixWin lol
azawawi ZoffixWin: re no line number info, it is evident a lot when using atom perl6 editor tools plugin 07:01
ZoffixWin: generally the error is at line number 0 which is the default for undefined line numbers
ZoffixWin: or line 1
ZoffixWin: the problem occur on scripts or not one liners also 07:02
ZoffixWin It occurs with one liners too 07:03
There's no line info at all.
azawawi i see
07:04 SalamiTactics left, KotH joined
azawawi ZoffixWin: github.com/rakudo/rakudo/blob/nom/....nqp#L2378 ... we're not passing line number information? 07:09
ZoffixWin: something like github.com/rakudo/rakudo/blob/nom/...d.nqp#L609 07:10
ZoffixWin ¯\_(ツ)_/¯ it's over my head. I gave up :)
azawawi i bet we need to handle the exception inside the grammar instead of instead of the world 07:11
s/instead of//
and that's what happens when you dont drink your morning coffee :) 07:12
moritz or pass $/ to world
azawawi moritz: good morning
moritz so that it knows the line number
azawawi: \o
azawawi $*LINE_NO is not set though. that's is passed using $/ ? 07:13
my nqp foo is weak, master :)
ZoffixWin load_module() actually has my $line := self.current_line($/); but $line isn't used anywhere in it 07:15
moritz azawawi: what do you mean by "it's not set though"? Is it set to an undefined value? Or is there a place where it's NQPMu or something? 07:16
azawawi moritz: ignore my comment. im just woke up :) 07:19
moritz: s/im/i/
azawawi starts a rakudo fork to try to fix it :) 07:22
ZoffixWin m: ."\b\b\b\b\b\b"() 07:25
camelia rakudo-moar a45224: OUTPUT«Method '' not found for invocant of class 'Any'␤ in block <unit> at /tmp/WXndyRcnbX line 1␤␤»
ZoffixWin
.oO( camelia ain't a terminal... )
 07:26
07:26 vendethiel joined
vendethiel masak: are you sure you know me? :P 07:27
07:32 Ven joined 07:34 jjido left 07:36 AlexDaniel left
ZoffixWin wow, Proxy is cool 07:42
07:43 CIAvash left 07:44 CIAvash joined
ZoffixWin m: my $s = 0; my $v := Proxy.new( FETCH => method () { $s }, STORE => method ($new) { $s = ^25 .pick } ); .say for ++$v, $v; say $v; .say for ++$v, $v; 07:44
camelia rakudo-moar a45224: OUTPUT«6␤6␤6␤19␤19␤»
ZoffixWin contemplates a text web browser using a Proxy... :P 07:45
To go to a URL, you assign it to a variable and to view the results you simply read it
Wait. That's the simplest LWP::Simple! :o
07:46 jack_rabbit joined 07:49 Ven left 07:57 xfix joined 08:02 firstdayonthejob joined 08:03 ocbtec joined 08:19 darutoko joined
ZoffixWin m: (sub (FatRat) {})(class FatCat{}); 08:24
camelia rakudo-moar a45224: OUTPUT«Type check failed in binding <anon>; expected FatRat but got FatCat (FatCat)␤ in sub at /tmp/kYUx5sVasF line 1␤ in block <unit> at /tmp/kYUx5sVasF line 1␤␤»
ZoffixWin 😂
masak vendethiel: enough to make the "hug a stranger" punchline ;) 08:27
08:33 _mg_ joined 08:35 RabidGravy joined 08:37 _mg_ left
RabidGravy harr 08:38
08:40 hankache joined
masak ponders how compatible server-side events are with Bailador and PSGI 08:43
RabidGravy the WebSocket module is a PSGI app thingy 08:45
masak ooh metacpan.org/pod/distribution/PSGI...aming-Body
RabidGravy you can alternatively return a Channel instead of content from a PSGI app
masak yes, this is what I want 08:46
finally a chance to use Channels! :D
RabidGravy :)
08:46 vendethiel left
RabidGravy be a love amd make a an EventSource *client* too, I want to rip this rubbish implementation out of Sofa soon 08:47
masak m: sub (FatRat) {}(class FatCat{})
camelia rakudo-moar a45224: OUTPUT«Type check failed in binding <anon>; expected FatRat but got FatCat (FatCat)␤ in sub at /tmp/9NBeKEEHAT line 1␤ in block <unit> at /tmp/9NBeKEEHAT line 1␤␤»
masak ZoffixWin: no need for that first set of parentheses
RabidGravy: my client is in JS, though.
RabidGravy :-\ BOO 08:48
masak heh 08:55
hm, only Bailador::Test mentions p6sgi.streaming... and it sets it to Bool::False 08:58
ufobat: ...halp :)
09:00 jjido_ joined
RabidGravy I'd make a quick example but I haven't had a coffee yet 09:14
azawawi RabidGravy: Greetings from non-coffee land 09:15
RabidGravy :-O
09:15 Relsak joined 09:27 Relsak left 09:32 hankache left
dogbert17 hello #perl6 09:39
continuing my documentation spree, it seems as if only about 75 % of the methods in Cool have been documented at this point ;( 09:42
09:42 CIAvash left
dogbert17 on the plus side there is some low hanging fruit to be found here I think 09:42
moritz dogbert17: many of the methods have been documented in the classes where the more specific variant lives 09:44
for example, .uc coerces to Str, so .uc is documented in Str
09:44 lizmat joined
dogbert17 does that mean that we shouldn't document it in cool, I guess the difference is coercion of the invocant? 09:45
Stuff like that would be low hanging fruit but perhaps it doesn't have to go into Cool 09:46
moritz dogbert17: it can be documented in Cool, we'll just duplicate lots of stuff that way 09:47
dogbert17 True, should I try to look at other methods instead, i.e. stuff that is not documented elsewhere? 09:48
RabidGravy I guess one could knock up a quick script to find all the methods in cool that aren't over-ridden by sub-classes
also there are probably some that are over-ridden by some sub-classes and not others 09:49
moritz dogbert17: and if ou haven't seen it yet, the Cool docs have a table of which method coerces to what type
extendind that would be the real LHF
09:49 TEttinger left
moritz dogbert17: yes, documenting stuff that hasn't been documented at all would be more beneficial 09:50
RabidGravy I guess documenting *all* the methods in Cool would be a boon for anyone who had a desire to sub-class it themself for some reason 09:51
dogbert17 moritz: I actually added a few methods to that table yesterday
and removed one as well ... capitalize
moritz dogbert17: ok, great 09:53
just shows how much I'm out of the loop :/
dogbert17 found a changelog claiming that capitalize was removed in 2013 09:54
09:54 CIAvash joined
moritz yes, wordcase is the new capitalize 09:55
dogbert17 I'll snoop around a bit in the docs, find something that is not documented, and then pester you with a suggestion :-)
09:56 vendethiel joined 09:58 spider-mario joined 09:59 vendethiel left 10:00 Ven joined
moritz dogbert17: have you looked into the issue tracker? 10:01
dogbert17 no (blushes) 10:03
rt.perl.org ?
moritz github.com/perl6/doc/issues?utf8=%...writing%22 10:04
dogbert17 thanks, will take a look
moritz dogbert17: some of them might be outdated (so the doc might already exists), do a git grep before actualling documenting stuff 10:05
dogbert17 actually found one that should be closed, i.e. already implemented, github.com/perl6/doc/issues/113 10:07
moritz dogbert17: will you close it, or should I? 10:08
dogbert17 can I close it? 10:09
I'll try :-)
moritz dogbert17: if you can push to the repo, you can also close issues 10:10
dogbert17: if it doesn't work, tell me
dogbert17 done :-) 10:11
moritz dogbert17++
masak RabidGravy: I'd really appreciate a quick example of SSE. 10:18
10:22 Ven left 10:25 dpk left
azawawi github.com/azawawi/rakudo/commit/c...749dc2ced2 # First attempt at fixing RT #127883 10:30
feedback is welcome 10:32
10:32 ssm_ is now known as ssm
moritz azawawi: I find $.line as the attribute a bit short 10:38
dogbert17 do these suggested doc additions seem correct or am I in over my head? gist.github.com/dogbert17/6deed8b9...aaae08ebf1 10:39
moritz what kind of line is it?
azawawi $.line-number ?
moritz source-line-number?
dunno
dogbert17: the first three are good; the others not so much (more) 10:40
azawawi moritz: i used the same conventions as X::Comp
10:40 Ven joined
moritz azawawi: yes, but in an exception, it's a bit clearer IMHO what it refers to 10:40
dogbert17 argh
moritz dogbert17: coercion to Int is not soething that Int actually implements 10:41
dogbert17: it's more like, several types that implement Cool do it in a different way
azawawi moritz: ok... fixing ... thanks for the feedback
moritz for example Str.Int parses the str, Array.Int looks at the number of elements
lizmat waits for azawawi++ to make changes and then will merge the PR 10:42
dogbert17 ah, I'd better remove these for the time being :-) got three right at least
10:43 firstdayonthejob left
dogbert17 my proficiency in perl6 is still a bit sketchy, that's why I'm trying to find LHF 10:44
10:46 Ven left
dogbert17 one more question, Cool.pod documents a routine called EVALFILE but that does not show up when typing 'say Cool.^methods'. What am I missing here? 10:50
10:51 chris2_ is now known as chris2, vendethiel joined
lizmat dogbert17: EVALFILE is a sub 10:53
dogbert17 lizmat: thx 10:54
azawawi lizmat: first green travis-ci.org/rakudo/rakudo/builds/123540811 :) 11:08
dalek c: d8a76bd | (Jan-Olof Hendig)++ | doc/Type/Cool.pod:
Added chomp, chop and codes to the method coercion table
 11:09
11:09 kid51 joined 11:10 pecastro joined
lizmat azawawi: I had already tested it myself... :-) so it's merged now 11:10
azawawi lizmat: supercomputer machine? :)
lizmat naaah... combination of looking at code and seeing it compiles :-) 11:11
azawawi dreams of ark.intel.com/products/85766/Intel-...e-2_10-GHz 11:13
11:16 travis-ci joined
travis-ci Doc build errored. Jan-Olof Hendig 'Added chomp, chop and codes to the method coercion table' 11:16
travis-ci.org/perl6/doc/builds/123541890 github.com/perl6/doc/compare/8e218...a76bddbfe4
11:16 travis-ci left
dogbert17 nooooo 11:16
lizmat azawawi: hmmm... some places don't specify the line number :-( 11:17
azawawi spectest?
dogbert17 so how do I figure out what went wrong here, I could build the docs locally 11:18
azawawi lizmat: EVAL?
lizmat not sure yet
dogbert17 strange, it complains that it couldn't build panda ?!? 11:21
11:25 jeek joined
lizmat azawawi: there are about 6 test files that fail with source-line-number being required: I made it optional for now 11:25
azawawi ok :)
dogbert17 as a newbie it's not obvious to me why a change in doc/Type/Cool.pod could lead to a build failing with the msg 'The command "rakudobrew build-panda" failed and exited with 2 during .'
11:26 jeek left
lizmat dogbert17: wasn't that caused by me mergening azawawi's PR ? 11:29
*merging
dogbert17 perhaps :-)
it said something about # Looks like you planned 16 tests, but ran 14, t/02-shell-command.t .. 11:31
azawawi pasteboard.co/ec2dlgV.png # Finally...
lizmat if it was azawawi's PR, then it should be fixed now 11:32
11:32 rurban left
dalek rl6-most-wanted/add-bindings-for-nlp: b3fcfa3 | titsuki++ | most-wanted/bindings.md:
Add "NLP and Machine Learning" to the list
 11:32
dogbert17 it is fixed, lizmat++ 11:33
azawawi lizmat: so im a core contributor? :)
lizmat azawawi: yes :-) 11:37
psch azawawi++ 11:38
azawawi :)
lizmat azawawi: would appreciate it if you could check all occurrences of DependencySpecification in roast that lack a :source-line-number parameter 11:40
RabidGravy wonders if titsuki would also like the moon on a stick and a pony 11:42
azawawi lizmat: sure 11:44
dogbert17 The round method/sub is defined in Real and Cool but I can't find any description of which tie-breaking rule is use, i.e how to round 2.5 or -0.5, is it documented somewhere?
s/use/used/ 11:45
psch m: say Cool.^can('round').candidates>>.signature 11:46
camelia rakudo-moar 674186: OUTPUT«Method 'candidates' not found for invocant of class 'List'␤ in block <unit> at /tmp/CwGOMQtgmG line 1␤␤»
psch m: say Cool.^can('round')>>.candidates>>.signature
camelia rakudo-moar 674186: OUTPUT«(((Cool $: *%_) (Cool $: $base, *%_)))␤»
psch m: say Real.^can('round')>>.candidates>>.signature
camelia rakudo-moar 674186: OUTPUT«Too many positionals passed; expected 2 arguments but got 3␤ in block <unit> at /tmp/844TT2hxEd line 1␤␤»
psch ah, doesn't pun there
m: say Real.new.^can('round')>>.candidates>>.signature
camelia rakudo-moar 674186: OUTPUT«(((Real:D $: *%_) (Real:D $: $scale, *%_)))␤»
psch m: say Cool ~~ Real; say Real ~~ Cool
camelia rakudo-moar 674186: OUTPUT«False␤True␤»
psch dogbert17: Real is Cool, so it's narrower
11:47 jjido_ left
psch i also think :D would be narrower than no type smiley, but that doesn't matter here 11:47
11:47 jeek joined 11:48 MasterDuke left, kid51 left 11:52 _mg_ joined
lizmat goes enjoying the great outdoors again 11:58
11:59 lizmat left
dogbert17 psch: interesting but I'm not sure how it helps me here, do the different types use different tie breaking rules en.wikipedia.org/wiki/Rounding#Tie-breaking 11:59
psch ohh 12:01
yeah, i thought tie-breaking in a MMD sense, my bad
dogbert17 nah, I should have been clearer when i posted my question 12:02
s/i/I/
psch dogbert17: in any case, Cool.round calls self.Numeric.round, and Real does Numeric 12:03
so they both essentially do the same thing, except for the fact that Cool.round first coerces to Numeric
dogbert17 psch: thanks, that's good to know
looks as if the rule used is 'Round half away from zero' 12:05
12:10 mane joined
azawawi .tell lizmat nothing so far but github.com/perl6/roast/blob/master...fication.t 12:13
yoleaux azawawi: I'll pass your message to lizmat.
12:14 azawawi left 12:17 molaf joined 12:19 mane left
dogbert17 Hmm, I must suffer from a serious lack of coffee, the rounding tie-breaker rule p6 uses is 'Round half up' not 'Round half away from zero' as I wrote earlier 12:21
12:23 rurban joined
dogbert17 m: say -1.5.round # this fooled me :( 12:23
camelia rakudo-moar 4b0e5b: OUTPUT«-2␤»
psch m: say (-1.5).round 12:24
camelia rakudo-moar 4b0e5b: OUTPUT«-1␤»
psch m: say -1.5 .round
camelia rakudo-moar 4b0e5b: OUTPUT«-1␤»
dogbert17 parenthesis FTW 12:25
psch: why did your last example return -1 ?
psch m: say -1.5\ .round 12:26
camelia rakudo-moar 4b0e5b: OUTPUT«-2␤»
psch dogbert17: there's slightly different precedence for postfix methodop and infix methodop i think
dogbert17 aha, I have a lot to learn it seems 12:27
12:29 Actualeyes joined 12:31 Ven joined
psch finally got the chord variant stuff working 12:32
RabidGravy psch++ 12:34
psch example.pl6 doesn't care about playing mode-unfitting chords, though, so sometimes there's stuff like Dminmaj7 in e.g. C major... vOv 12:35
12:37 CIAvash left
dalek rl6-most-wanted: b3fcfa3 | titsuki++ | most-wanted/bindings.md:
Add "NLP and Machine Learning" to the list
 12:38
rl6-most-wanted: 1a7abda | titsuki++ | most-wanted/bindings.md:
Merge pull request #27 from perl6/add-bindings-for-nlp

Add "NLP and Machine Learning" section to the list
RabidGravy masak, server side events gist.github.com/jonathanstowe/cf91...37f7137427 12:57
Now, that *appears* to send all that right things, but for some reason the node EventSource thingy doesn't recognise it, despite it being effectively the same response to one it *does* recognise 12:59
oh wait, just edited 13:02
that *does* work 13:03
now, how one might integrate that with Bailador is left as an exercise for the reader 13:05
[Coke] waves from Baltimore. 13:24
(documenting things duplicatively - be nice if we documented it once and then built the docs so you could see it both ways) 13:25
"I have no idea what I'm doing" - me in about 3.5 hours 13:27
RabidGravy that's me all the time 13:28
[Coke] wonders if baltimore has an IRC channel. 13:29
[Coke] 's stomach is not happy about the coffee/food ratio atm. 13:30
rjbs: let me be the first to publicly hope that your free time will now be spent hacking on Rakudo! ;) 13:33
rjbs 🙂 13:35
13:36 Ven left 13:39 jameslenz joined
[Coke] .u 🙂 13:40
yoleaux No characters found
13:48 Ven joined, Ven left 13:54 sue joined 13:57 kid51 joined 14:14 colomon joined
colomon timotimo: octree, actually. 14:14
colomon thinks he can hack his way around the issue without an octree, based on looking more closely at an actual case. 14:15
14:15 tharkun left 14:17 Ikke_ joined 14:19 Ikke_ left 14:23 tharkun joined 14:24 CIAvash joined 14:25 skids joined 14:32 skids left
colomon O(N^2) operations are bad when N=766, m’kay? 14:35
14:37 arlenik left
[Coke] Nice. the Perl 6 installfest had a few people who were interested in getting perl 6 installed, but that were basically able to get it done on their own. one panda snag from a week ago that magically worked today. 14:43
14:44 BenGoldberg joined
RabidGravy wahay! 14:45
14:51 _mg_ left 14:52 pecastro left 14:53 Skarsnik joined
RabidGravy Is something like "role Sofa::Document::All[::Doc = (Hash but Sofa::Document::Wrapper)] does JSON::Class {" a bit too weird? 14:55
Juerd Yes. 14:57
14:58 colomon left 14:59 colomon joined
arnsholt It's definitely weird enough that it deserves an explanatory comment, but whether it's worth it or not depends on the interface you get to implement 15:03
15:05 kid51 left 15:07 jjido joined
masak RabidGravy: `::Doc =` -- I don't think I understand that bit. is it really two colons? 15:07
RabidGravy Really two colons, a "type capture" 15:08
it works, the actual API is a bit nicer :) 15:09
[Coke] Hi to the people in baltimore who are seeing #perl6 on the board. 15:10
15:10 kid51 joined
mst _ _ ___ ____ _ _ _____ ___ __ __ ___ ____ _____ 15:11
| | | |_ _| | __ ) / \ | | |_ _|_ _| \/ |/ _ \| _ \| ____|
| |_| || | | _ \ / _ \ | | | | | || |\/| | | | | |_) | _|
| _ || | | |_) / ___ \| |___| | | || | | | |_| | _ <| |___
|_| |_|___| |____/_/ \_\_____|_| |___|_| |_|\___/|_| \_\_____|
teatime lol 15:12
teatime has a sudden urge to watch The Wire.
15:16 kurahaupo joined
TimToady dogbert17: you apparently got warnocked yesterday..."capitalize" turned into "wordcase" quite some time ago, in part because some people can't remember how to spell it :) but mainly because tclc is just the default policy you might apply word by word 15:18
mst TimToady: like, say, you. given it's spelled 'capitalise' :D 15:19
mst goes to fish his tea back out the harbour
15:21 sue left
ZoffixWin #perl6 on the board? What board? 15:28
mst I would presume the board in the room hosting the baltimore installathon 15:29
ZoffixWin I see
Also, you misspelt 'capitalize'
15:31 sue joined
mst in soviet zoffix, ALL words spelled with Z 15:33
15:33 rurban left
TimToady dogbert17: oh, I see moritz unwarnocked you, moritz++ 15:33
ZoffixWin :)
15:34 rurban joined
dogbert17 TimToady: thanks for the capitali[s|z]e explanation, much appreciated 15:34
and I also learned the meaning of 'warnocked' :) 15:35
RabidGravy masak, did you see the gist of the really dumb SSE psgi app?
ZoffixWin m: ^0x1FFFF .grep({ .uniname ~~ m:i/web/ })».say 15:39
camelia rakudo-moar 4b0e5b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/EWdvojzrDj␤Missing « or »␤at /tmp/EWdvojzrDj:1␤------> 031FFFF .grep({ .uniname ~~ m:i/web/ })».7⏏5say␤»
ZoffixWin bug?
15:42 _mg_ joined
TimToady arguable both ways :) 15:45
ZoffixWin TimToady, what's the non-bug argument?
TimToady . is an infix that is looser than ». so how is it supposed to figure out the head of that is really a method? 15:46
ZoffixWin m: (^0x1FFF).grep({ .uniname ~~ m:i/web/ })».say 15:47
camelia ( no output )
TimToady when you write ^42 .grep it's really ^42 . grep
ZoffixWin Ah
TimToady so switching back from infix processing to postfix processing is a bit fraught
otoh, people think of infix . as a way to write a funny postfix, hence the argument to fix it too :) 15:48
ZoffixWin :)
BenGoldberg m: (^0x1FFF).grep({ .uniname ~~ m:i/web/ }).map: .say 15:50
camelia rakudo-moar 4b0e5b: OUTPUT«(Any)␤Cannot call map(Seq: Bool); none of these signatures match:␤ ($: Hash \h, *%_)␤ (\SELF: &block;; :$label, :$item, *%_)␤ (HyperIterable:D $: &block;; :$label, *%_)␤ in block <unit> at /tmp/ukvrre4GEy line 1␤␤»
BenGoldberg m: (^0x1FFF).grep({ .uniname ~~ m:i/web/ }).map: { .say } 15:51
camelia ( no output )
ZoffixWin is trying to find a good unicode char for "browser in a variable" thingie :) 15:52
timotimo 💻? 15:54
ZoffixWin m: '💻'.uninames.say
camelia rakudo-moar 4b0e5b: OUTPUT«(PERSONAL COMPUTER)␤»
ZoffixWin That doesn't display neither in my IRC client nor in my code editor :)
timotimo hehe.
TimToady you need a better 💻 15:55
ZoffixWin heh
timotimo well, moar knows it, so it's not that new :)
15:56 firstdayonthejob joined 15:57 _mg_ left
ZoffixWin m: sub term:<%💻> { 42 }; say "%💻[]" 15:57
camelia rakudo-moar 4b0e5b: OUTPUT«42␤»
ZoffixWin freaky :)
timotimo what's that character?
is that the personal computer? 15:58
ZoffixWin Yes
TimToady finds it fascinating that we spent millenia trying to reduce the number of ideographs in our writing system, but now that we can look things up on a computer, we're reinventing many new ideographs
ZoffixWin: that's...freaky...yeah... 15:59
overloading sigils via LTM seems like a bad idea for some reason
m: say ${foo} 16:01
camelia rakudo-moar 4b0e5b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/YQulP65Blp␤Unsupported use of ${foo}; in Perl 6 please use $foo␤at /tmp/YQulP65Blp:1␤------> 3say ${foo}7⏏5<EOL>␤»
colomon Is there a nice idiomatic way to initalize a Hash value to the empty array / list / whatever if the value doesn’t currently exist in the Hash?
TimToady though that's exactly what we're doing to get that error, I guess
16:01 rurban left
grondilu m: say [1, 1] ~~ Array[Real] 16:02
camelia rakudo-moar 4b0e5b: OUTPUT«False␤»
ZoffixWin m: my %h = bar => 42; %h<foo> //= Empty; %h<bar> //= Empty; say %h; 16:03
camelia rakudo-moar 4b0e5b: OUTPUT«{bar => 42, foo => ()}␤»
ZoffixWin colomon, something like that, ^ perhaps?
Hm, I guess that's "false" not "not exists"
or not defined
TimToady note that .push will do something like this automagically 16:04
colomon I think it will work either way?
grondilu m: subset Metric of Array where .all ~~ Real; say [1, 1i] ~~ Metric; say [1, 2] ~~ Metric
camelia rakudo-moar 4b0e5b: OUTPUT«False␤True␤»
colomon TimToady: right, but I’m dealing with the results of a classify call, and don’t actually want to add anything else to the list, I just want the Empty list to be there.
TimToady seems a bit X/Y to me 16:05
grondilu m: subset Metric of Array where { .all ~~ Real and .map(&abs).all == 1 }; say [1, 1i] ~~ Metric; say [1, 2] ~~ Metric; say [1, -1] ~~ Metric
camelia rakudo-moar 4b0e5b: OUTPUT«False␤False␤True␤»
colomon X/Y?
TimToady S99:XY 16:06
or so
ZoffixWin colomon, x/y problem. You're trying to solve the wrong problem.
timotimo er. i need to find tuits to fix synopsebot6
16:07 synopsebot6 joined
timotimo S99:XY 16:07
synopsebot6 Link: design.perl6.org/S99.html#XY
RabidGravy there
timotimo S99:XY Problem
synopsebot6 Link: design.perl6.org/S99.html#XY_Problem
TimToady if you only want to know the existence of a key, seems more like a Set 16:08
or are you overloading two different behaviors on one structure?
colomon TimToady: okay, here’s the problem. I have two lists of 3D coordinate systems, and a test that can classify whether or not a system is the standard global origin. I need to know how many of the systems are the standard global origin in each list, and get a list of all the ones which are not for each list. 16:09
TimToady S99:MEGO
synopsebot6 Link: design.perl6.org/S99.html#MEGO
TimToady hah, found one that isn't there :) 16:10
ZoffixWin colomon, you should probably using a Set and one of those fancy set operators 16:11
Check out these: docs.perl6.org/language/setbagmix
TimToady I suspect colomon++ already knows about those, which is why I said "seems" :)
ZoffixWin Alright :) 16:12
16:12 kid51 left
colomon TimToady: two different behaviors on one structure, as my longer message says. :) 16:12
TimToady okay, though my eye gloze over 16:13
*eyes
colomon Quick version: Have list. Need to know how many things in the list pass a test, and get a list of all the ones that fail it. 16:15
… and when I put it like that, the answer is obvious.
TimToady++
ZoffixWin grep for failures and subtract the number you got from original number to get successes 16:17
colomon ZoffixWin: right, like I said, obvious.
16:17 Herby_ joined
TimToady 😹 16:17
Herby_ Morning, everyone!
o/
colomon ZoffixWin++ 16:18
stmuk does anyone know any SF.pm names? 16:19
16:21 cdg joined
TimToady I believe so :) 16:22
stmuk is Fred Moyer the best contact? 16:25
sue sigh, typical perl geek response. let's get breakfast 16:28
TimToady no clue; SF is on the other side of Palo Alto from me, which is too massive to transmit much
sue this is the way it works when people ask questions on the list. you get 20 different people saying they have no freaking idea 16:29
16:29 skink joined
sue CHEERS FOR THAT :-) 16:29
TimToady well, I thought one such answer was okayish
colomon There’s more than one way not to know it.
TimToady just to be social and all
16:30 CIAvash left
colomon got the running time for his program from 20 minutes to 45 seconds with this round of changes. Hup! 16:30
sue i'm mostly joking
TimToady 😎 16:31
skink colomon, Reminds me of this i.imgur.com/U3kDo3W.png 16:33
dogbert17 This usage description of sort is backwards, isn't it? sort LIST, SORT_BY 16:34
its from List.pod 16:36
Usage:
sort LIST
sort LIST, SORT_BY
LIST.sort
LIST.sort(SORT_BY)
ZoffixWin Yeah 16:37
TimToady if it's a named parameter, it doesn't matter where it goes really
timotimo the usage sections were not terribly well received and not too many of them were actually added 16:38
or reviewed for that matter
dogbert17 m: say sort &abs, (3, -4, 7, -1, 2, 0)
camelia rakudo-moar 4b0e5b: OUTPUT«(0 -1 2 3 -4 7)␤»
16:38 sue left
TimToady if it's position, then it starts mattering more :) 16:38
*tional
16:38 rurban joined
dogbert17 was thinking of changed that part of the docs 16:39
s/changed/changing/
I'll change it then :) 16:41
timotimo: are you a p6 profiler virtuoso?
timotimo just a little bit 16:42
16:42 colomon left
timotimo what do you need? 16:42
dogbert17 do you know why this fails perl6 --profile -e 'my @m[5]'
timotimo no, haven't looked into it yet
could be because it doesn't run any code and the profiler doesn't expect the result set to be empty 16:43
so maybe it's running into a null pointer and asplodes
16:43 colomon joined
dogbert17 the above is a golfed version [Coke]++ a 'real' program 16:44
timotimo what kind of failure are you seeing? "knowhow methods must be called on an object instance with repr KnowHOWREPR"? 16:45
and "don't know how to dump a BOOTCode"?
16:45 araujo_ joined
dogbert17 I see the first one 'knowhow ...' 16:46
timotimo i can try to look into it later today; for now i'll be AFK
16:47 araujo_ left 16:48 araujo_ joined, Herby_ left 16:49 araujo left, araujo_ left 16:50 araujo_ joined
dogbert17 timotimo: thx, let me now if you need the code and I'll paste it 16:50
16:51 araujo_ left 16:52 araujo_ joined 16:55 Actualeyes left 17:03 BenGoldberg left, riatre left, riatre joined, zakharyas joined 17:04 wamba joined, BenGoldberg joined
dalek c: db65e03 | (Jan-Olof Hendig)++ | doc/Type/List.pod:
Removed two typos and changed an inconsistency in the usage section for sort
 17:08
17:08 colomon left 17:11 rurban1 joined 17:12 rurban1 left
RabidGravy renames a bunch of tests so they are numbered 100 apart 17:12
mst the skills we learned from BBC BASIC will always be with us 17:13
17:14 rurban left
Skarsnik Hello 17:14
vendethiel o/
timotimo dogbert17: you can paste it right now, i'll look at it when i get to it 17:15
17:16 Emeric joined
dogbert17 timotimo: gist.github.com/dogbert17/c510cc42...8624acc66d 17:17
17:19 zakharyas left
timotimo dogbert17: if you're after performance, you can get a decent improvement by not using multi-dimensional arrays with fixed dimensions; those are not yet optimized at all 17:19
just drop the shape and it ought to be faster already
dogbert17 cool, will try 17:20
17:21 Ven joined 17:23 travis-ci joined
travis-ci Doc build passed. Jan-Olof Hendig 'Removed two typos and changed an inconsistency in the usage section for sort' 17:23
travis-ci.org/perl6/doc/builds/123584211 github.com/perl6/doc/compare/d8a76...65e03d930c
17:23 travis-ci left
skink Hey ZoffixWin, if have bin/script in a module, can it see %?RESOURCES? 17:23
you* have 17:24
ZoffixWin No idea
RabidGravy I don't think it can, but it's easy to test
timotimo just make a tiny module that just returns %?RESOURCES :P 17:25
RabidGravy yeah, that's what I've actually done
17:27 kid51 joined
RabidGravy No, just tested and it doesn't appear that they can 17:29
skink mhm 17:30
RabidGravy but as timotimo says you can just have "module MyResources { sub resources() is export { %?RESOURCES }}" in your lib and jobs a good-un 17:32
skink Yeah I can try that after fixing panda again 17:35
panda --force install . does not actually overwrite previous installations, even if you bump versions, so testing this is a tad annoying 17:36
RabidGravy A little gimmick I've used is github.com/jonathanstowe/Lumberjac...rovider.pm
just specify the new version in the use
skink The script included with the module isn't using the module itself :D 17:37
17:37 kid51 left
ugexe it should have access to resources, afterall if you use a local module via -I its %?RESOURCES are still available to the local directory loaded copy of the module 17:48
i.e. if you can access a (possibly not-loaded) CompUnit you should be good 17:49
17:50 kid51 joined 17:51 BenGoldberg left 17:52 Sgeo_ left, mane joined
grondilu jeez the EVAL warning message is very dissuasive indeed. 17:52
17:52 BenGoldberg joined
ugexe perl6 -e 'EVAL()' \ "FUCK YOU!!! only kidding" 17:54
17:58 domidumont joined 18:05 djbkd joined 18:07 kid51 left 18:09 Ven left 18:11 Ven joined 18:12 Ven left 18:16 colomon joined 18:18 domidumont1 joined, domidumont left 18:28 djbkd left, domidumont1 left 18:31 kid51 joined
[Coke] "Brian Duggan"++ 18:33
colomon sort of knows a Bryan Duggan… 18:36
18:43 mane left
[Coke] hopes he seppled this Duggan's name right - he's giving a presentaitn on writing a web app in P6. 18:44
colomon Bryan is a tech guy, but I haven’t heard anything about him working with p6, and he lives in Ireland. 18:45
[Coke] this guy is in Baltimore right now. :) 18:46
colomon :)
[Coke] likes This presentation so much better than the one Coke did!
colomon itunes.apple.com/us/app/tunepal/id...35033?mt=8
[Coke] github.com/bduggan/hamna :) 18:58
19:02 colomon left 19:05 kid51 left 19:07 Ven joined, Ven left 19:12 BenGoldberg left 19:13 BenGoldberg joined 19:15 jjido left 19:18 obfusk joined 19:29 colomon joined, tharkun left, tharkun joined
colomon [Coke]++ 19:31
19:31 dwarring joined 19:33 colomon left 19:34 regreg joined
regreg hello 19:34
i have a clean rakudo star installation
I'm trying to run "panda install Readline"
and I get a failure at the "==> Testing LibraryMake" step
no such file or directory 19:35
in sub run-and-gather-output at C:\rakudo\share\perl6\site\sources\5AEF9DA5AE15E5AB5CB2ADB58A455E007FA783
no such file or directory in sub run-and-gather-output at C:\rakudo\share\perl6\site\sources\5AEF9DA5AE15E5AB5CB2ADB58A455E007FA7839E line 85 19:36
19:36 djbkd joined
regreg any idea on how to proceed? 19:36
19:37 kaare_ left 19:38 kaare_ joined 19:39 Ven joined
[Coke] which version of rakudo star? 19:47
what OS, btw?
19:56 jjido joined
regreg windows 10 64bit 19:57
rakudo-star-2016.01-x86_64 (JIT)
19:58 jjido left 19:59 TimToady joined
regreg [Coke]: 20:00
20:01 jjido joined 20:02 kaare_ left 20:03 Ven left
[Coke] sorry, don't have a win box to test on at the moment. Anyone else? 20:04
20:04 colomon joined, colomon left
regreg [Coke]: don't I need some sort of C compiler to install the Readline library? I have MSVC but probably it needs mingw- gcc 20:05
20:09 cpage joined
regreg i think i'll try perl6 in a linux vm then 20:11
20:24 mane joined
timotimo i need a beefier computer so i can play those DOS games 20:31
20:33 kurahaupo left 20:34 djbkd left
geekosaur regreg, if you can find a built readline.dll then NativeCall can use it directly 20:35
regreg the point of Readline library is to offer a better console wrapper i think, not to be called from code
geekosaur it needs to be called from code to provide things like intelligent completion 20:36
rlwrap could be used to wrap an existing program (sort of; not sure that even flies on Windows) but it can't distinguish filename completion from command completion from variable completion from etc. 20:37
20:37 kid51 joined 20:45 djbkd joined
[Coke] release process will start tomorrow about 9 am eastern 20:47
20:57 darutoko left 21:04 mane left
[Coke] .seen dha 21:05
yoleaux I saw dha 11 Apr 2016 21:26Z in #perl6: <dha> That doesn't really look like Larry.
[Coke] dha was registered to be here at baltimore, but he's not. :(
ZoffixWin How many people are there? And is it all about P6? 21:11
21:13 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg 21:15 AlexDaniel joined 21:19 sortiz joined, kid51 left
sortiz \o #perl6 21:19
ZoffixWin \o
skink Anyone know any cool/neat/weird ideas for the crypto trashbag? 21:24
Got a passphrase generator and the PGP wordlist convert in so far
sortiz Digging into regreg problem with in R* on Windows found that a) LibraryMake needs an environment with cl and nmake, b) as Readline don't include readline's sources, the dependency on LibraryMake is needless. 21:28
21:29 Skarsnik left 21:32 wamba left 21:42 jjido left 21:45 TEttinger joined, leont joined 21:50 BenGoldberg left 21:51 ZoffixLappy joined
leont jnthn: Test::Harness is so bad a being parallel on Windows you might as well not use the parallelism 21:58
Also, I think we're one bugfix away from parallel TAP::Harness in pure-perl6
22:05 ocbtec left 22:06 BenGoldberg joined
timotimo ooooh 22:08
leont: you got some details? :)
well, for today i promised i'd look at a profiler invocation going boom
ZoffixLappy m: role Base { has $.name = 'Base; }; role Curiosity does Base { has $.name = 'Curiosity' }; my $curi = Curiosity.new; say $curi 22:09
camelia rakudo-moar 4b0e5b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/bG0GyDWpIA␤Two terms in a row␤at /tmp/bG0GyDWpIA:1␤------> 3ole Curiosity does Base { has $.name = '7⏏5Curiosity' }; my $curi = Curiosity.new; ␤ expecting any of:␤ infix␤ infi…»
ZoffixLappy TTIAR??
22:09 khw joined
leont timetime: still working on figuring out if it's rakudo or moar, I suspect the latter 22:09
timotimo, 22:10
ZoffixLappy Ah, damn. missing quote.
m: role Base { has $.name = 'Base'; }; role Curiosity does Base { has $.name = 'curiosity'; }; my $curi = Curiosity.new; say $curi
camelia rakudo-moar 4b0e5b: OUTPUT«Attribute '$!name' conflicts in role composition␤ in any apply at gen/moar/m-Metamodel.nqp line 1974␤ in any compose at gen/moar/m-Metamodel.nqp line 2057␤ in any specialize_with at gen/moar/m-Metamodel.nqp line 2497␤ in any at gen/moar/m-Meta…»
22:11 rindolf left
ZoffixLappy How can I make it so I have a $.name I can use in the Base, but other roles (like Curiosity here) can change the value of it? 22:11
leont Proc::Async file handles don't give an EOF signal, so the Supplies never have their :done handlers called 22:12
22:12 cdg left
ZoffixLappy m: role Base { method foo { say self!name } }; role Curiosity does Base { has $!name = 'Curi'; }; say Curiosity.new.foo 22:13
camelia rakudo-moar 4b0e5b: OUTPUT«No such private method '!name' for invocant of type 'Curiosity'␤ in method foo at /tmp/LPoj1_AvsP line 1␤ in block <unit> at /tmp/LPoj1_AvsP line 1␤␤»
ZoffixLappy Ah. I get it.
Never mind me :)
leont Hmmm, it seems the problem is Rakudo after all. If my suspicion is correct I have a fix :-) 22:15
sortiz m: role Base { has $.name = 'Base'; }; role Curiosity does Base { method new { self.bless(:name<curiosity>) }}; my $curi = Curiosity.new; say $curi 22:17
camelia rakudo-moar 4b0e5b: OUTPUT«Curiosity.new(name => "curiosity")␤»
ZoffixLappy Cool. Thanks. 22:18
timotimo leont: sounds great! 22:22
22:24 zakharyas joined
sortiz ZoffixLappy, When B does A, all methods and attributes of A, even private ones, are now an integral part of your class/role B. 22:25
ZoffixLappy 👍 I need to read up more on that stuff later on. 22:29
Curious, is there a away to specify either-or in the signature? e.g. sub query (Str :$sol, Str :$earth-day) { ... } and I want either $sol or $earth-day to be a required parameter. 22:33
Right now I'm just checking in the body of the sub
timotimo no, but you can build two multi candidates 22:34
ZoffixLappy ! timotimo++ right
22:44 dolmen joined
timotimo dogbert17: did you get around to writing the diagonals-sum thing without shaped arrays yet? 22:46
22:52 firstdayonthejob left
sortiz Some of the ideas I've been testing in DBDish for search and load of libraries at runtime, now on an independent module, not ready for ecosystem yet but for comments and discussion: github.com/salortiz/NativeLibs 22:53
AlexDaniel sortiz: that looks very promising 22:54
22:54 RabidGravy left 22:58 Emeric left
sortiz AlexDaniel, Thanks! 22:59
AlexDaniel sortiz: any reason why this cannot be built into nativecall itself? 23:01
ZoffixLappy m: sub foo (:$foo where { $_ > 0 } ) { say $foo ?? 'foo' !! 'no foo' }; foo 23:11
camelia rakudo-moar 4b0e5b: OUTPUT«Use of uninitialized value of type Any in numeric context in block at /tmp/nGdR66G6h_ line 1␤Use of uninitialized value of type Any in numeric context in block at /tmp/nGdR66G6h_ line 1␤Constraint type check failed for parameter '$foo'␤ in sub f…»
ZoffixLappy This is kinda LTA, TBH. No `foo` parameter was passed to the sub, so it shouldn't be checked for validity 23:12
leont The cycle of "change one bit, rebuild the entire prelude", repeat, repeat is a bit slow :-/ 23:13
ZoffixLappy Especially since you can't `return` from within a where clause, so if you're writing neat subsets, they turn out awkward
dogbert17 timotimo: I changed my @mat[$size; $size]; to my @mat; instead but perf went down. Should I change the way I access the matrix as well? 23:14
23:15 dolmen left
ZoffixLappy Oh, it's actually failing the constraint ;/ 23:15
m: subset Foo of Any where { $_ ~~ Any ?? True !! ($_ > 5 or warn 'Need more than five') }; sub foo (Foo :$foo) { say $foo ?? 'foo' !! 'no foo' }; foo :2foo 23:18
camelia rakudo-moar 4b0e5b: OUTPUT«foo␤»
ZoffixLappy I can't master this. How do I define a subset for an optional named arg that also able to `warn`?
m: subset Foo of Any where { .defined ?? ($_ > 5 or warn 'Must be more than 5') !! True }; sub foo (Foo :$foo) { say $foo ?? 'foo' !! 'no foo' }; foo :2foo 23:19
camelia rakudo-moar 4b0e5b: OUTPUT«Must be more than 5 in block at /tmp/v9LYrPAHDk line 1␤Must be more than 5 in block at /tmp/v9LYrPAHDk line 1␤Constraint type check failed for parameter '$foo'␤ in sub foo at /tmp/v9LYrPAHDk line 1␤ in block <unit> at /tmp/v9LYrPAHDk line 1…»
geekosaur smartmatch against Any seems wrong 23:20
m: (5 ~~ Any).so.say
camelia rakudo-moar 4b0e5b: OUTPUT«True␤»
ZoffixLappy That's the answer ^. My original compaint still stand though. Unspecified named args should not be checked
23:21 sue joined 23:22 zakharyas left
ZoffixLappy This can probably be addressed via module-space. 23:23
ZoffixLappy gets cracking 23:25
23:28 molaf left 23:33 colomon joined
sortiz AlexDaniel, That is the plan, after intensive testing. 23:33
AlexDaniel sortiz: cool
timotimo dogbert17: hm, dunno, should be all right 23:34
23:34 djbkd left
timotimo dogbert17: i put a piece of safety code in the right spot to make that code run under the profiler again 23:34
23:35 BenGoldberg left
timotimo dogbert17: you can pull the very latest nqp and compile a rakudo with that, it should work then 23:35
23:36 BenGoldberg joined, spider-mario left
timotimo dogbert17: what $size did you use to get it to actually take enough time? :) 23:41
dogbert17 timotimo: the problem calls for the size to be 1001, see projecteuler.net/problem=28 23:42
timotimo ouch :)
ZoffixLappy m: sub subset-is (&check, Str $message = '') { return sub ($v){ $v.defined ?? ( &check($v) or warn $message ~ " Got $v" and False ) !! True }; }; subset Positive of Int where subset-is { $_ > 0 }, "Must be a positive Int."; my Positive $x = -2; 23:43
camelia rakudo-moar 4b0e5b: OUTPUT«Must be a positive Int. Got -2 in sub at /tmp/uRSn9hmKzP line 1␤Type check failed in assignment to $x; expected Positive but got Int (-2)␤ in block <unit> at /tmp/uRSn9hmKzP line 1␤␤»
ZoffixLappy m: sub subset-is (&check, Str $message = '') { return sub ($v){ $v.defined ?? ( &check($v) or warn $message ~ " Got $v" and False ) !! True }; }; subset Positive of Int where subset-is { $_ > 0 }, "Must be a positive Int."; sub foo ( Positive :$foos ) { }; foo foos => -2;
camelia rakudo-moar 4b0e5b: OUTPUT«Must be a positive Int. Got -2 in sub at /tmp/KCxnFKVH0X line 1␤Must be a positive Int. Got -2 in sub at /tmp/KCxnFKVH0X line 1␤Constraint type check failed for parameter '$foos'␤ in sub foo at /tmp/KCxnFKVH0X line 1␤ in block <unit> at /tmp/…»
ZoffixLappy Any idea what the 'must be a positive..' message is printed TWICE when using the subset as a sig in the sub?
23:44 djbkd joined
ZoffixLappy s/what/why/; 23:44
leont Managed to make Proc::Async call :done (and presumably :quit), but await on stdout still hands :-/
dogbert17 yeah, it takes quite some time to run with the size set to 1001, hopefully things will improve in the coming months 23:45
there is also a high chance that I'm using a lousy method in order to solve the problem 23:46
ZoffixLappy m: sub foo ($foos where { $_ > 2 or warn "fail" }) {}; 23:47
camelia ( no output )
ZoffixLappy m: sub foo ($foos where { $_ > 2 or warn "fail" }) {}; foo -2
camelia rakudo-moar 4b0e5b: OUTPUT«fail in block at /tmp/z3WXOD2Mgv line 1␤fail in block at /tmp/z3WXOD2Mgv line 1␤Constraint type check failed for parameter '$foos'␤ in sub foo at /tmp/z3WXOD2Mgv line 1␤ in block <unit> at /tmp/z3WXOD2Mgv line 1␤␤»
ZoffixLappy Seems like sub signature subsets are evaluated TWICE :S
dogbert17 timotimo: thanks for looking into this problem, timotimo++ 23:50
23:51 colomon left 23:56 sue left
leont Ow wait, that was supposed to hang, because that cast doesn't make sense to me. 23:58

Powered by Cro and the Raku® Programming Language.

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

Thank you!