»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo:, alpha:, pugs:, std:, or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by lichtkind on 5 March 2010.
ash_ oh, yah, i can understand that 00:00
1e60 is big
colomon I've got it working in my project by simply skipping all the initial 0s, and expressing 1e60 as 10 ** 60, so that it gracefully switches to floating point. 00:01
00:10 [Coke] left
sorear There must be a cleverer algorithm out there somewhere 00:12
00:13 sorear sets mode: +o colomon
colomon sorear: probably in general, but I dunno about with the Perl 6 specific characteristics. (like 0.000001 should be a Rat, but 0.000000000000000000000000000001 a Num) 00:14
rakudo: say 3.14159265358979323846264338327950288
p6eval rakudo 11cbd4: OUTPUT«===SORRY!===␤PAST::Compiler can't compile node of type BigInt␤»
00:15 sorear sets mode: +vvvv dalek p6eval _ilbot2 IRSeekBot
sorear even just str2num conversion is a bit interesting 00:16
at one point the GHC parser converted strings to Num-equivalent through a FatRat-equivalent
leading to much hilarity a few years later when some smartalec tried to use 1e-999999 in a program 00:17
00:18 lestrrat is now known as lest_away, [Coke] joined
sorear rakudo: foo(1,2,3; 4,5,6) 00:19
p6eval rakudo 11cbd4: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 11␤»
sorear std: foo(1,2,3; 4,5,6) 00:20
p6eval std 31419: OUTPUT«===SORRY!===␤Undeclared routine:␤ 'foo' used at line 1␤Check failed␤FAILED 00:01 112m␤»
sorear alpha: foo(1,2,3; 4,5,6)
p6eval alpha 30e0ed: OUTPUT«Confused at line 10, near "(1,2,3; 4,"␤in Main (file <unknown>, line <unknown>)␤»
sorear pugs: foo(1,2,3; 4,5,6)
p6eval pugs: OUTPUT«*** No such subroutine: "&foo"␤ at /tmp/qbSAxXLWNk line 1, column 1 - line 2, column 1␤»
colomon sorear... huh. Actually, it occurs to me that a smart version of a naive conversion routine might automatically Do The Right Thing wrt Rat versus Num. 00:22
00:25 azert0x left 00:26 kensanata left
colomon okay, I'm not sure whether to be proud or ashamed of this code: 00:32
$int-part.Int + [+] $frac-part.comb.kv.map(-> $i, $d { $d eq '0' ?? 0 !! $d.Int / (10 ** ($i + 1) });
well, certainly mildly ashamed for the missing paren. 00:33
00:35 micahjohnston joined
colomon rakudo: say ~("14159265358979323846264338327950288".comb(/. ** 1..4/)) 00:36
p6eval rakudo 11cbd4: OUTPUT«1415 9265 3589 7932 3846 2643 3832 7950 288␤»
00:36 micahjohnston left, [Coke] left 00:40 masonkramer_ joined, masonkramer left, masonkramer_ is now known as masonkramer 00:51 \shade\ left 00:52 shade_ joined
colomon if a few people could pound on github.com/colomon/str2num, I'd appreciate it. I think I've got it basically working now. Not the most efficient algorithm for sure, but as far as I can tell (tired as I am) this version works pretty well. 00:59
writing more tests would be good. 01:00
:)
off to my in-laws' to get some sleep now....
01:03 cjk101010 left 01:05 colomon left 01:07 uniejo joined 01:13 japhb left 01:14 whiteknight left
uniejo perl6: role A { method a { say "a" } }; class B does A {}; B.new().a() 01:17
p6eval rakudo 11cbd4: OUTPUT«Could not find sub &say␤ in 'a' at line 11:/tmp/w7XZWIStA8␤ in main program body at line 11:/tmp/w7XZWIStA8␤»
..elf 31419, pugs: OUTPUT«a␤»
sorear pugs: \(1,2,3; 4,5,6).WHAT.say 01:26
p6eval pugs: OUTPUT«*** ␤ Unexpected "\\("␤ expecting program␤ at /tmp/I4zvlSCKzd line 1, column 1␤»
sorear pugs: say WHAT \(1,2,3; 4,5,6)
p6eval pugs: OUTPUT«*** ␤ Unexpected "\\("␤ at /tmp/4EbqdDPmI5 line 1, column 10␤»
diakopter pugs: (\.(1,2,3; 4,5,6)).WHAT.say 01:27
p6eval pugs: OUTPUT«*** Cannot cast from VUndef to Pugs.AST.Types.VCode (VCode)␤ at /tmp/XHf5g0iSDN line 1, column 2-18␤»
diakopter O.o
01:28 redicaps joined 01:33 felliott joined 01:37 plobsing joined 01:40 shade_ left, shade_ joined 01:41 drbean_ is now known as drbean 01:50 japhb joined 01:52 snarkyboojum joined 01:55 ashleydev left 01:57 jaldhar joined 02:00 zorgnax joined 02:08 lest_away is now known as lestrrat 02:09 meppl left, christine left, nadim left, tomaw left, felipe left, Gothmog_ left, ingy left, opx^away left 02:10 Grrrr left, Grrrr joined, meppl joined, christine joined, nadim joined, tomaw joined, Gothmog_ joined, ingy joined, opx^away joined, card.freenode.net sets mode: +o ingy 02:11 uniejo left 02:14 shade_ left, shade_ joined 02:16 x19290 joined
x19290 Question about Rakudo/Parrot: 02:17
When I
$ perl6 --target=pir -o 123.pir -e 123.say
and
$ parrot 123.pir
I get syntax errors. What's wrong?
sorear 1. using --target with -e is not working properly last I heard 02:18
x19290 echo 123.say > 123.p6 02:19
sorear 2. you just tried to load 123.pir in a clean Parrot enviroment - Rakudo installs custom opcodes, which changes the assembly syntax
x19290 How to install the 'custom' opcode? 02:20
sorear by running 'perl6' instead of 'parrot' 02:21
123.pir can only be loaded from inside a rakudo process
x19290 is there no easy way to run 123.pir? 02:22
sorear I don't think there's /any/ way to run 123.pir 02:23
precompilation is not suitable for general use at this time
it's only useful for bootstrapping
x19290 I think perl6 will have a compiler mode like javac 02:24
Is it wrong?
sorear our build system uses perl6 --target=pir to compile the standard library, then pastes the output into the compiler
oh yes, it *will*
plobsing sorear: have you tried using PBC instead? dynops work in that medium (modulo load order issues)
x19290 you mean --target=pbc? 02:25
sorear plobsing: 1. it's not possible to generate PBC from PCT-based compilers 2. even if it did work, it wouldn't, because rakudo needs to control its own initialization 02:26
02:26 ciphertext left
sorear getting AOT compilation to work in the general case is a GSoC-sized project 02:26
you're welcome to take it on now, but I'd wait till 2011
sorear is having trouble with slicels 02:27
02:27 colomon joined, sorear sets mode: +o colomon
pmichaud good evening, #perl6 02:27
sorear good evening, pmichaud 02:28
pmichaud: can you briefly explain how foo(1,2,3; 4,5,6) is expected to work?
pmichaud sorear: that's still a bit of speculation (more) 02:31
sorear: my plan at this point is that multiple statements in an arglist results in a slicel instead of a parcel
frew pmichaud: where do you keep your slides again? 02:32
pmichaud and converting a slice to a Capture causes things to not collapse
s/collapse/flatten
frew is making links etc in blog post
pmichaud frew: pmichaud.com/2010/pres/
colomon o/
frew perfect
thanks
02:33 ciphertext joined
frew pmichaud: what are the keys for this? I see that space and enter go forward, what about backwards? 02:34
pmichaud N and P work
so should pgup and pgdn
and there are links at the bottom, if you can make the window tall enough
frew ahhh
I was wondering why I couldn't see those
my laptop must be too short 02:35
x19290 sorear: I've learned that in somedayperl6 will have
pmichaud yeah, they're done with floating blocks in the HTML
x19290 a compiler mode like javac. Thank you.
pmichaud Spork has been really good for generating slides for my presentations; I have lots of things I'd like to do differently with it, but ENOTUITS
(fixing the layout and navigation is one of them) 02:36
frew pmichaud: eh, that's how it goes.
mine are using javascript etc and that has other issues
pmichaud one of my first significant Perl 6 projects (besides the compiler itself) is likely to be a WikiCreole to HTML translator, followed shortly by a slideshow presenter based on that :) 02:37
I sure wish those Perl 6 folks would hurry up and release something so that I can do that. :) :)
frew hahahaha
shouldn't you be able to do that now?
that's just a matter of tuits right? 02:38
02:38 x19290 left
pmichaud yes, it's still ENOTUITS on that :) 02:38
frew of course
sorear What day was -perl6? 02:39
pmichaud although maybe I'll work on it during my plane trip in July.
sorear wonders what good a hyperlink in a slide does
pmichaud It would be *really* cool if I could give a presentation that was labeled "powered by Rakudo Perl" :)
frew sorear: so you can follow doc at home?
pmichaud sorear: while giving a presentation, I can click on the link to jump to the site directly 02:40
and yes, what frew++ said
frew pmichaud: I'm actually pretty amazed at how easy it is to work on a plane, maybe because of the lack of distractions?
I got 3 things that I'd been planning on doing for a couple months done on the flight up and down
pmichaud Sometimes working on a plane works for me, sometimes not. Often I'm just glad to have some time to rest :)
frew sure 02:41
pmichaud but I've also been known to do some pretty significant branch work on a plane :)
(much easier now that we have git)
frew indeed
colomon I've never had a seat / laptop combination I could comfortable open on a plane. :( 02:42
but I don't fly much.
otoh, I get good work done when we're driving somewhere on vacation.
frew colomon: I find it interesting that just shaving a couple inches off my 13 inch laptop made it comfortable
pmichaud plane work is one of the (many) reasons I went for an ultra-small notebook. 02:43
sorear Frighteningly, I don't think I've ever been on an airplane
pugssvn r31420 | lwall++ | [STD] differentiate dup prefix message from bad infix for bbkr++
pmichaud afk for a moment, running up to my room to grab a dr pepper. They seem to be hard to find around here.
TimToady never flew till he was 19 02:44
colomon TimToady: don't know if you saw, I came up with a beautiful, terrible way to handle the decimal number issue: 02:46
sorear has a few months yet, then
colomon [+] $int-part.comb.reverse.kv.map(-> $i, $d { $d.Int * (10 ** $i) }) + [+] $frac-part.comb.kv.map(-> $i, $d { $d.Int / (10 ** ($i + 1)) }) 02:47
02:47 ciphertext left
TimToady wonders if colomon has Tourette's... 02:47
colomon It works properly because **, *, /, and + do the Right Thing wrt to types.
TimToady yay (I think) 02:49
colomon it's good news, but for sure we'll have to find a more efficient approach in the long run.
possibly even the middle run. :)
pmichaud ...and maybe even in the short run. :)
I'm not a fan of those **'s, though. 02:50
colomon ;)
pmichaud Seems like we should be able to do things incrementally.
colomon yeah, I'm sure we can. just lose the beautiful functional programming style of it if we do it the sensible way.
*we just
sorear just as long as 1e-99999999 doesn't attempt to allocate a 12 megabyte integer 02:51
TimToady well, the compiler should spot it and do strength reduction for us :)
colomon actually, there is a functional way to do it in p6:
$int-part.comb.reverse Z 1, 10, 100 ... * -> $d, $pot { $d.Int *$pot } 02:52
TimToady \o/ 02:53
colomon (that's untested, but it ought to work)
pmichaud I do like that form much better :) 02:54
TimToady .comb(/\d/) unless you've removed _
colomon TimToady: nice. that is a good way to remove it all right.
does \d only match '0' .. '9'?
pmichaud matches any unicode digit
colomon (ie do we have to worry about funky unicode digits)
pmichaud we eventually have to worry about funky unicode digits 02:55
(iirc)
sorear Is there a simple, exact algorithm to parse C numbers?
TimToady but what do .Int do with it, is the Q
sorear Emphasis on the exact
0.1 cannot be represented exactly as a Num; have fun parsing 0.123456789
colomon well, I'm thinking once this works it should probably be used for .Int, too.
sorear: 0.1 is a Rat in p6
02:56 Limbic_Region left
colomon TimToady: which is why I was wondering how to write a "single-digit to Int" algorithm. Easy enough for ascii digits... 02:56
TimToady should be in the unicode tables 02:57
sorear it's the NumericValue property
colomon do we have an easy way of getting at that?
TimToady blizkost to Perl 5? :) 02:58
pmichaud colomon: no, not presently.
colomon: I suspect ICU provides that property somewhere, but Parrot doesn't expose it yet.
sorear ICU?
TimToady you could borrow the tables that sorear++ came up with for STD
colomon oh, interesting. sorear++ 02:59
pmichaud colomon: this is a place where I'd not have perfect become the enemy of good enough (for now) :-)
colomon :)
TimToady and if you grab those tables you can probably dump ICU, for the most part... 03:00
pmichaud ...at the startup cost of initializing the tables
unless they're in C...?
TimToady or implementing mmap
pmichaud or mmap might work 03:01
if they're in C we could make something work.
colomon rakudo: (1, 10, 100 ... *).munch(20).map({ .WHAT.say }).eager
sorear the STD tables are designed to be processed using a minimum number of PerlIO traps
p6eval rakudo 11cbd4: OUTPUT«Int()␤Int()␤Int()␤Rat()␤Rat()␤Rat()␤Rat()␤Rat()␤Rat()␤Rat()␤Num()␤Num()␤Num()␤Num()␤Num()␤Num()␤Num()␤Num()␤Num()␤Num()␤»
03:01 pnate left
sorear they aren't quite properly aligned for mmap() 03:01
colomon hmmm, not quite right yet.
TimToady yet
03:01 pnate joined
pmichaud well, they're static, so we could certainly make a static C data structure for them, and some opcodes to do what we want. 03:01
sorear also, they currently only contain Blank and GeneralCategory 03:02
pmichaud That might in fact be a Big Win, rather than waiting for Parrot to figure out how it's going to handle things.
03:02 snarkyboojum left 03:03 meppel joined
sorear ok, I'll do that soon 03:04
sorear loves designing data structures and has already worked out a nice format for this
colomon sorear++ 03:06
03:08 meppl left 03:10 Guest39268 left 03:17 snarkyboojum joined 03:18 agentzh joined
colomon 'night, all. 03:22
pmichaud 'night, colomon++
colomon pmichaud, TimToady: hope to see you in the morning for a bit.
pmichaud colomon: same here. I think we'll be in the union, either downstairs in the lobby area or in the BOF room 03:23
(but perhaps check on irc first, or if you're unable to find us)
(cell # sent in /msg)
03:35 molaf joined 03:46 skangas joined 03:49 snarkyboojum left 03:51 ashleydev joined 03:56 ciphertext joined 04:04 snarkyboojum joined
sorear Maybe we should just use ICU... I don't feel very good about stealing my STD data source hack for Rakudo 04:11
It involves die unless $] == 5.012001; and lots of scraping of files in @INC/unicore/lib 04:12
04:13 JimmyZ joined
dalek meta: r342 | stefa...@cox.net++ | trunk/vicil/ (2 files):
[vicil] Adopt Parrotish arguments model - native array + native hash. It's

anyway. Also, it cuts the circularity issues involved in Parcel having methods which take arguments.		 04:21
04:24 skids left 04:25 mjk joined, shade_ is now known as shade\ 04:26 ashleydev left 04:30 patspam left 04:36 skids joined 04:38 mjk left 05:00 molaf left 05:09 ash_ left 05:16 kaare joined 05:17 kaare is now known as Guest10194
dalek p-rx: 9157c78 | pmichaud++ | (2 files):
Allow hyphens and single-quotes in subrule calls.
 05:32
p-rx: f0bed9b | pmichaud++ | (2 files):
Allow hyphen and single-quotes in NQP identifiers.
p-rx: a3d5c5b | pmichaud++ | build/PARROT_REVISION:
Bump PARROT_REVISION to get debugging fix.
p-rx: ebaf06c | pmichaud++ | (2 files):
Add ResizablePMCArray .delete and .exists .
p-rx: 4ee17c4 | pmichaud++ | src/stage0/ (4 files):
Update bootstrap.
05:40 orafu left 05:42 orafu joined 05:57 tylercurtis left 05:59 meppel left, Su-Shee joined 06:07 cjk101010 joined 06:09 jakk joined, jakk left 06:11 nadim left 06:12 nadim joined 06:17 Su-Shee_ joined 06:21 Su-Shee left 06:25 nadim left 06:26 justatheory left 06:29 nadim joined 06:32 zorgnax left
quietfanatic rakudo: class A {our $.x = 4}; say A.new.x 06:33
p6eval rakudo 11cbd4: OUTPUT«Method 'x' not found for invocant of class 'A'␤ in main program body at line 11:/tmp/HfotnSJZki␤»
quietfanatic I'm gonna submit that as a bug.
06:35 agentzh left 06:43 gfx joined 06:48 snarkyboojum left 06:49 Su-Shee_ is now known as Su-Shee 07:02 mberends joined 07:16 thepler left, TiMBuS joined 07:28 Trashlord joined, mberends left 07:33 quietfanatic left 07:42 mberends joined 07:55 kensanata joined 07:58 fridim joined 07:59 thepler joined 08:01 [particle] left 08:02 ejs joined 08:03 spinclad left 08:05 [particle] joined 08:06 tadzik joined 08:07 ejs left 08:08 ejs joined 08:13 thebird joined 08:20 skids left, skids joined 08:24 JimmyZ left 08:29 stepnem joined 08:30 Su-Shee left, Su-Shee joined 08:43 plobsing left 08:57 kensanata left 09:23 lestrrat is now known as lest_away 09:27 mjk joined 09:31 redicaps left 09:40 agentzh joined 09:46 masak joined 09:47 stepnem left, fridim left
mberends new directions to the Belgian Perl Workshop: brussels.the-hub.net/public/contact.html 09:48
masak o/ 09:49
mberends o/
masak I made moderate progress with the pls proof-of-concept yesterday. 09:50
found two bugs which I promptly saved away and ignored. 09:51
09:51 pmurias joined
mberends at least the TDD finds the bugs, instead of the end user :) 09:53
masak :)
09:54 ciphertext_ joined, ciphertext left, ciphertext_ is now known as ciphertext 09:55 nsh left 09:56 nsh_ joined
pmurias sorear: re vicil is intended to be used together with perlesque or a standalone thing? 09:57
sorear I don't know yet 10:01
it's more an exercise in clarifying the future than anything else
10:01 ciphertext left 10:03 ciphertext joined 10:04 azert0x joined
jnthn morning o/ 10:12
moritz_ \o 10:13
"morning"
masak grins
jnthn It is still in my current timezone. :P
11:23 or so here in England
moritz_ wow, England is UTC + 1:10 now! :-) 10:14
masak moritz_: don't tease jnthn, he's just woken up... :P
jnthn :P 10:15
True. :P 10:16
masak jnthn: it's OK, I didn't wake up long ago either. :) 10:17
jnthn :-)
Yes, I seem to recall our preferred waking up times worked out very well last hackathon. :-) 10:18
masak not to mention our preferred going-to-bed times! 10:19
jnthn :-) 10:20
masak if I do `run("something that works") or otherwise()`, when is &otherwise called?
jnthn I forget what run returns... 10:21
rakudo: say &run.perl.signature
p6eval rakudo 11cbd4: OUTPUT«Method 'signature' not found for invocant of class 'String'␤ in main program body at line 11:/tmp/gcljS5uMUl␤»
jnthn oh ffs
I forget what run returns...!
masak coffee for jnthn!
jnthn rakudo: say &run.signature.perl
p6eval rakudo 11cbd4: OUTPUT«:(Any $commandline)␤»
jnthn That's not going to answer what I wanted anyway.
masak jnthn: it returns an Int but a Bool.
jnthn oh 10:22
Well, I guess or cares about truth?
masak I hope so.
otherwise, what would be the point?
jnthn I'd imagine it does, in which case the answer is "when it's but True"
10:22 mjk left
jnthn However, but True only started working very recenlty in Rakudo. 10:22
So run may be out of date with respect to that. 10:23
masak submits rakudobug
jnthn oh huh, the Belgian Perl Workshop has moved?!
mberends o/ jnthn. yes, brussels.the-hub.net/public/contact.html 10:24
jnthn moritz_: I plan to be at YAPC::EU before the conf
moritz_: Some of the Monday, all of the Tuesday. 10:25
mberends: Cool 10:27
mberends: Don't suppose you have heard anything about any pre-conference meetup?
10:28 agentzh left
mberends jnthn: no, we probably all received the same email 10:28
jnthn: but I'd be happy to meet up near there quite early
jnthn mberends: I more meant, night before the conference, for pivo ;-) 10:29
mberends heh, jnthn is not and early-morning person :) 10:30
jnthn omg cake van?!
brb
10:34 [Coke] joined
jnthn wow 10:34
The village my family lives in now has a van that drives around the village selling cake. :-)
When I was a kid we only got ice-cream that way. :-) 10:35
colomon jnthn: when I was a kid, we lived far enough away from town that the ice-cream truck only came by a couple of times a summer. :) 10:37
jnthn wow :-)
colomon so, I went to a quick talk on functional programming in Perl 5 yesterday.
most everything they talked about is built-in / easier in p6. 10:38
except, we don't have memoize yet!
10:38 JimmyZ joined
arnsholt colomon: I'm sure that can be arranged =) 10:38
colomon arnsholt: me too, I'm just trying to make sure that the Right People are thinking about it. 10:39
jnthn colomon: I'm sur ethere's an is remembered or similar trait
That you can put on a sub to get that effect
I think it's in spec somewhere.
colomon seriously?!
TimToady++
jnthn Unless it was a bit of the spec I read in a dream. ;-)
jnthn tries to find it
colomon hmmm, it's not "remembered"
arnsholt If it was in a dream we can always update reality to reflect that =D
colomon it seems like memoize should be easy to do in p6, whether it's a module or a trait we need to implement 10:40
jnthn Found 10:41
is cached
colomon and given Rakudo's current speed, it seems like it would be very useful. ;)
jnthn In S06
If only we had .wrap working again, we could probably implement it already. :-)_ 10:43
colomon .wrap? 10:44
jnthn colomon: aye 10:45
colomon: The thing that lets you wrap things around a routine
Then it's just 10:46
tadzik rakudo: my @arr = (1, 2, 3, 4, 5); @arr ~~ 5
p6eval rakudo 11cbd4: ( no output )
tadzik rakudo: my @arr = (1, 2, 3, 4, 5); say 'ok' if 5 ~~ @arr
p6eval rakudo 11cbd4: ( no output )
tadzik hrm, repl is more verbose
rakudo: my @arr = (1, 2, 3, 4, 5); say 'ok' if @arr ~~ 5 10:47
p6eval rakudo 11cbd4: OUTPUT«ok␤»
jnthn multi trait_mod:<is>(Routine $r, :$cached!) { $r.wrap(sub (|$c) { if is_cached($c) { ...retrun cached... } else { run and cache } }); }
tadzik what's the difference between this two?
10:47 JimmyZ left
colomon @arr ~~ 5 converts @a to a scalar, the number of elements, ie 5 10:47
jnthn If you want to grep through an array for values, you want 10:48
@arry ~~ *,$value,*
tadzik looks weird, but I think I get it 10:49
jnthn * menas "match anything"
tadzik how about 5 ~~ @arr, how does it work?
jnthn Puts 5 in list context, rendering a one-item list
So it's true if @arr only contains one element and it's 5
Smart-match is not symmetric.
tadzik mhm 10:50
jnthn The type of the thing on the right hand side always decides.
10:54 cognominal left 10:56 cognominal joined 11:00 redicaps joined 11:01 cognominal left, cognominal joined
pugssvn r31421 | pmurias++ | [smop] much faster handling of .ri files 11:06
11:09 gfx left
colomon loliblogged: justrakudoit.wordpress.com/2010/06/24/euler-2/ 11:09
11:11 stepnem joined 11:13 Intensity left
jnthn afk for a bit 11:17
tadzik ~(1, 1, *+* ... 4000000) 11:23
I don't get it, the *+* part
looks a bit like Camelia, but what does it do?
colomon tadzik: if * is used as a term in for an operator, it generates a closure. 11:24
so *+* is the same as { $^a + $^b } 11:25
tadzik hmm
colomon the same trick is used there with * !% 2
the only difference is the * + * closure takes two arguments, and the * !% 2 takes only one 11:26
that's based on the number of *s in the expression.
tadzik alright, I get that one. But this *+*. In the closure you wrote, it gets the first one from the left and first from the right?
redicaps rakudo: grammar word {token TOP{\w+}}; word.parse("foo");
p6eval rakudo 11cbd4: ( no output )
tadzik no, that makes no sense
redicaps rakudo: grammar word {token TOP{\w+}}; word.parse("foo");
tadzik how does it know to take the last two?
p6eval rakudo 11cbd4: ( no output ) 11:27
colomon tadzik: the closure just takes two
tadzik hmm
colomon the series operator, if it has a closure with N arguments, takes the last N arguments that were generated
tadzik ah
what is the series operator here?
colomon ...
tadzik mhm 11:28
at the first glance it looks more magical than Haskell to me
masak rakudo: grammar word {token TOP{\w+}}; say word.parse("foo"); 11:29
p6eval rakudo 11cbd4: OUTPUT«foo␤»
colomon it's actually built out of very simple ideas, but it does take a while to get used to them.
tadzik it's magic anyway ;) 11:30
like bogosort. It's easy to understand, but it's magic 11:31
masak bogosort is magic? :) 11:32
tadzik well, the wikipedia implementation is, at the first glance
masak rakudo: my @a = 3, 2, 1, 5, 4; @a.=pick until [<=] @a; say @a.perl
p6eval rakudo 11cbd4: OUTPUT«[3]␤»
masak uhm :)
tadzik especially when you see the Java solution below
masak rakudo: my @a = 3, 2, 1, 5, 4; @a.=pick(*) until [<=] @a; say @a.perl
p6eval rakudo 11cbd4: OUTPUT«[1, 2, 3, 4, 5]␤» 11:33
masak \o/
tadzik hmm
arnsholt masak: Bogowin! =D
tadzik rakudo: my @a = 3, 2, 1, 5, 4; @a.=pick * until [<=] @a; say @a.perl
p6eval rakudo 11cbd4: OUTPUT«Could not find sub &until␤ in main program body at line 11:/tmp/FdJZ19PMQl␤»
tadzik there has to be () around *, hm?
masak tadzik: that's two-terms-in-a-row.
tadzik mhm
masak tadzik: and that's not allowed :)
tadzik: so yes, method calls must have either the colon or parens.
moritz_ pick: *
tadzik: either .pick(*) or .pick: * 11:34
tadzik hmm
what kind of syntax is 'pick: *'?
rakudo: my @a = 3, 2, 1, 5, 4; @a.=pick: * until [<=] @a; say @a.perl
moritz_ method call syntax
p6eval rakudo 11cbd4: OUTPUT«[1, 2, 3, 4, 5]␤»
tadzik yeah, but how it works?
arnsholt You put a colon after the method name instead of parens around the arglist 11:35
moritz_ it takes the rest of the statement as argument list
arnsholt $obj.method: $a, $b, $c
tadzik nice
moritz_ with the exception of statement modifiers, which are not part of the argument list
11:35 [Coke] left
tadzik even more TIMTOWDY 11:35
*towtdy 11:36
masak "Yt"?
:)
moritz_ y.
wow, just read about a wimbledon match 11:37
that was aborted after ten hours because it got too dark
score: 59:59 in the 5th set, not yet decided :-)
masak crazy.
moritz_ that's something I really admire, that none of the players gave up 11:38
after all even a standard 3 hours game of tennis is hard work
masak how can I do a MAIN sub which optionally accepts a --foo flag with no value, the presence of which translates into a Bool? 11:40
tadzik (1..*).list.reverse.shift
is reverse what breaks everything?
moritz_ masak: (Bool :$foo) 11:41
in the signature
colomon tadzil: yes
*tadzik
masak tadzik: what do you expect that to return? :)
moritz_ reversing an infinite list is bound to cause problems
colomon you can't revers... what moritz_ said.
11:41 ejs left
tadzik masak: and what do you think? :) pop didn't work either, so I tried shift :) 11:41
masak tadzik: I'm asking because I don't know. :) 11:42
moritz_ what's the last item of an infinite list anyway?
masak moritz_: that didn't work for me. Nominal type check failed for parameter '$foo'; expected Bool but got Any instead
moritz_: that's when it's not supplied.
I think there's an RT ticket for that.
moritz_ oh.
I think there was a patch that was supposed to fix the case of --/value 11:43
maybe that broke it
rakudo: sub MAIN(Bool :$foo) { }
p6eval rakudo 11cbd4: ( no output )
moritz_ works here when not present
masak hm.
moritz_ are you sure you didn't forget the : ? 11:44
masak seems I get that error, but not there.
rakudo: sub MAIN(Bool :$foo) { say $foo.WHAT }
p6eval rakudo 11cbd4: OUTPUT«Any()␤»
masak right. I tried to use it as a Bool() later, and it blew up.
and that's what the RT ticket was about, too. 11:45
so, nevermind. this is all workaroundable.
moritz_ urks
is this specific to MAIN?
masak nope.
moritz_ ah
masak all non-supplied tyoed nameds.
moritz_ that explains why I wasn't aware of it
11:47 [Coke] joined 11:51 envi^home joined 11:52 tadzik left
takadonet morning all 11:58
masak takadonet: mroing! 11:59
takadonet masak: how are u? 12:00
moritz_ who is that mr. oing?
masak :)
takadonet: I'm good. thankful for the relative efficiency of the climate control inside. :) 12:01
takadonet: how are u?
takadonet masak: doing good, extremely busy lately 12:02
masak takadonet: aye, same here. 12:03
12:05 felipe joined
dalek p-rx: a915b78 | bacek++ | (2 files):
Add bitwise operators
 12:07
masak rakudo: say ~(1, 1 ... &[+])[^10] 12:08
p6eval rakudo 11cbd4: OUTPUT«1 1 1 1 1 1 1 1 1 1␤» 12:09
moritz_ huh.
masak closures bug?
moritz_ wait
masak or am I doing something wrong?
moritz_ that's old series syntax
masak rakudo: say ~(1, 1, &[+] ... *)[^10]
p6eval rakudo 11cbd4: OUTPUT«get_attr_str() not implemented in class 'Perl6MultiSub'␤ in 'infix:<...>' at line 1␤ in main program body at line 11:/tmp/x5Trgy3XKZ␤»
masak submits rakuodbug
moritz_: if that's the old syntax, then I like the new syntax better. 12:10
colomon rakudo: say ~(1, 1, &[+] ... 50)
p6eval rakudo 11cbd4: OUTPUT«get_attr_str() not implemented in class 'Perl6MultiSub'␤ in 'infix:<...>' at line 1␤ in main program body at line 11:/tmp/9OU1aHEom6␤»
colomon rakudo: say ~(1, 1, *+* ... 50)
p6eval rakudo 11cbd4: OUTPUT«1 1 2 3 5 8 13 21 34␤» 12:11
moritz_ I guess it's related to &[+] referring to a multi
colomon moritz_: yes
I'm not sure the &[+] can actually work here.
masak why not?
colomon ... needs to know how many arguments the closure takes.
masak ah.
colomon how many does &[+] take?
masak two.
colomon or zero
masak zero!? 12:12
colomon rakudo: say infix:<+>()
moritz_ for [+] to return 0
p6eval rakudo 11cbd4: OUTPUT«0␤»
moritz_ and for [*] to return 1 etc.
the default values
colomon rakudo: say infix:<*>()
masak huh.
p6eval rakudo 11cbd4: OUTPUT«1␤»
masak well, the ... operator clearly chooses the two-args one here, no? 12:13
moritz_ where?
or for [~] to return an empty string, for that matter
masak moritz_: in the case of the ... series above.
moritz_ in which case?
12:14 [Coke] left
masak 1, 1, &[+] ... * 12:14
colomon looks to me like it doesn't know what one to choose
masak I mean, ideally.
moritz_ why?
colomon what's your rule for that?
masak that this isn't a zero-args case.
moritz_ what if there's another infix:<+> with three arguments?
colomon rakudo: say (1, 1, { rand } ... *).munch(10) 12:15
moritz_ the general solution is probably known, and neither specced nor implemented
p6eval rakudo 11cbd4: OUTPUT«110.03712539919385450.5533102571881830.2906026487207280.5188719889966540.7733171349245680.4927070330798780.3392575986627760.39404857995979␤»
masak then choose that for all but the third element.
this sounds a lot like ruoso++'s thinking around the reduce sub.
moritz_ and that is to do a partial binding, taking off as many elements of a list as can possibly used by the dispatch
colomon (ie, zero-arg closures are legal in series op)
TimToady &[+] is specced to choose the 2-arg case
colomon TimToady++ 12:16
masak that settles it. thanks.
colomon there's the bug, then. :)
TimToady S03:4686 12:17
shower & 12:18
moritz_ rakudo: multi sub a($a, $b) { $a + $b }; multi sub a(Int $a, Int $b) { $a + $b }; say ~(1, 1, &a ... *)[^10] 12:19
p6eval rakudo 11cbd4: OUTPUT«get_attr_str() not implemented in class 'Perl6MultiSub'␤ in 'infix:<...>' at line 1␤ in main program body at line 11:/tmp/UHpiQT8yte␤» 12:20
moritz_ so it's not only the arity
masak includes that in the ticket 12:21
moritz_: $makefile.say("\tinstall -D $file ~/.perl6/$file");
that -D option is -d on Mac OS X.
moritz_ and -d means something completely different on linux 12:22
-d, --directory
treat all arguments as directory names; create all components of
the specified directories
hilarity ensured
masak I see two options: 12:23
(1) mkdir -p
(2) detect OS or, more specifically, the way `install` works.
moritz_ (1) is probably more straight forward
masak any objections to (1)?
ciphertext colomon: speaking of series, i had a few comments and questions the other day, because I had been working on a new patch (to replace the one i submitted that used a special 'scmp' operator) 12:24
phenny ciphertext: 18 Jun 10:28Z <colomon> tell ciphertext I've paused for the moment (on series), but I still have a list of things that need doing and a notion that this is a pretty good time to be doing them.
12:26 bluescreen joined, bluescreen is now known as Guest89026
colomon ciphertext: yes? 12:28
masak rakudo: say "abcdefabc".rindex("b")
p6eval rakudo 11cbd4: OUTPUT«Method 'rindex' not found for invocant of class 'Str'␤ in main program body at line 11:/tmp/bKU9XHf9bg␤»
masak alpha: say "abcdefabc".rindex("b")
p6eval alpha 30e0ed: OUTPUT«7␤»
masak submits rakudobug
ciphertext colomon: ok, first, i had a question about character series. Namely, how should something like 'A'...'DD' behave? should it be the same as 'A'...'D' or 'A'...'Z','AA'...'DD'? 12:32
colomon it's almost the same as 'A'...'D'
ciphertext colomon: 'almost'?
colomon well, is the same, I suppose.
ciphertext ok
colomon it's 'A'...^'DD' that would be different from 'A'...^'D' 12:33
rakudo: say 'A' ... 'DD' 12:34
p6eval rakudo 11cbd4: OUTPUT«ABCD␤»
12:34 bphillips joined
ciphertext second, i was going to point out that the current version doesn't support closures on the rhs (to determine endpoint). I actually have this done in my version (github.com/ciphertext/rakudo/blob/m....pm#L310), but there are a lot of other differences too. 12:35
colomon there is a difference internally ... 'A'...'D' ends because the series hits 'D', while 'A'...'DD' ends because the next element ('E') is on the opposite side of 'DD' than 'D' is. 12:36
masak \o/ first installed project with pls! 12:38
colomon \o/ 12:39
slavik o/
12:40 aCiD2 joined, masonkramer left, masonkramer joined
TimToady colomon: I suspect that for strings, anything other than exact match is going to be un-dwimmy 12:41
rakudo: say 'a' ... 'zzz' 12:42
p6eval rakudo 11cbd4:
..OUTPUT«abcdefghijklmnopqrstuvwxyzaaabacadaeafagahaiajakalamanaoapaqarasatauavawaxayazbabbbcbdbebfbgbhbibjbkblbmbnbobpbqbrbsbtbubvbwbxbybzcacbcccdcecfcgchcicjckclcmcncocpcqcrcsctcucvcwcxcyczdadbdcdddedfdgdhdidjdkdldmdndodpdqdrdsdtdudvdwdxdydzeaebecedeeefegeheiejekelemeneoepeqereseteuevewex…
moritz_ rakudo: say 26**3
p6eval rakudo 11cbd4: OUTPUT«17576␤»
slavik perl6: say 2**64
p6eval pugs: OUTPUT«18446744073709551616␤»
..elf 31421, rakudo 11cbd4: OUTPUT«1.84467440737096e+19␤»
slavik grr 12:43
TimToady commuting &
slavik TimToady: perfect person to ask :) 12:44
moritz_ slavik: at least two implementations agree (though they are both wrong)
slavik moritz_: they are not 'wrong' but I would expect automatic bignum
moritz_: or do you mean wrong in the sense of the 'not following spec' 12:45
moritz_ right
wrong == different answer than specced
rakudo: say (2**64).WHAT
p6eval rakudo 11cbd4: OUTPUT«Num()␤»
slavik moritz_: in rakudo's case, this has to do with parrot?
moritz_ slavik: that's one way to look at it. If parrot supported native bigints, rakudo would support it too 12:46
slavik docs.parrot.org/parrot/1.9.0/html/s...m.pmc.html
moritz_ there are at least two caveats with this 12:47
three, I think
slavik do tell
:)
moritz_ 1) bignum != bigint
slavik agreed 12:48
colomon ciphertext: my quick impression looking at your version is that you've implemented some nice new stuff and broken some stuff at the same time. Is yours passing all the tests? (afk for a bit, getting ready to commute)
moritz_ 2) it's not "native" in the sense that an integer register can hold a bignum (or bigint, if it exists) PMC
3) it's only available when people have GMP installed, and parrot's configure.pl detects it
ciphertext colomon: i don't think it passes everything, no... i haven't had a chance to work on it much in the last few days. 12:49
12:49 redicaps left
slavik moritz_: gmp? 12:49
ahh 12:50
gnu bignum?
moritz_ slavik: the math library that parrot uses under the hood
slavik so if I have libgmp-dev installed, I should get bignum in perl6 and 2**64 in rakudo will output '18446744073709551616' ?
moritz_ nope; because of caveat 2) 12:51
parrot has this model of 'vtables'
and there's one to coerce to integer
12:52 pmurias left
moritz_ called get_integer, iirc 12:52
and it *has* to return a native parrot int
which is broken design, if you consider bigints
because "12244(many digits more)234234".Int boils down to a vtable call
(it doesn't have to for Perl 6 objects, but if it doesn't, HLL interop is in danger) 12:53
slavik parrotbug?
moritz_ so coercions to int or num prevent us from just using big numbers
slavik: parrot design bug, if you want 12:54
slavik ouch :(
moritz_: any idea if they are working on a fix?
12:54 redicaps joined
moritz_ I don't think it's high priorty for most parrot hackers 12:54
slavik :(
12:56 redicaps left
colomon ciphertext: will you have some availability for a long-distance hackathon over the weekend? 13:02
I think we can finish ... if we put our heads together.
moritz_ masak: I've tried to play with tcurtis' optimization framework, and promptly hit two show-stopper bugs (first one is already fixed)
ciphertext probably... what time(s)?
moritz_ it seems every project needs a user to keep 'em honest
13:03 JimmyZ joined
masak moritz_: aye. 13:03
13:03 Guest10194 left
colomon ciphertext: what timezone are you in? 13:05
ciphertext colomon: u.s. central (GMT-6) 13:07
colomon I'm eastern. 13:08
maybe late morning or early afternoon?
13:08 stepnem left 13:09 JimmyZ left
colomon may have to play it by ear around my son's napping schedule. :) 13:09
13:09 stepnem joined
ciphertext that should work. 13:09
colomon in the meantime, we should try to think of test cases. :) 13:10
moritz_ speaking of test cases
colomon remember that series doesn't have to work on just Strs and numbers. ;)
moritz_ I noticed that somebody re-vamped some of the series tests
and changed file names
somebody needs to adapt t/spectest.data accordingly at some point 13:11
I'll try to do that tonight if nobody beats me to it
colomon looks like files were actually added? 13:12
moritz_ added, and series-simple is gone it seems
colomon oh, right, that's just the .rakudo ghost in my ls
pmichaud good morning, #perl6 13:13
colomon I've gotta get a move on here, and my rakudo is out-of-date and dirty, anyway. but maybe later today....
o/
pmichaud: heading your way...
pmichaud colomon: we're in the hospitality suite, 2nd floor
colomon be there in about 30 13:14
afk 13:15
13:15 colomon left, felliott left
moritz_ pmichaud: did you have a parrot BOF/metting/whatever? if yes, how did it go? 13:15
13:16 cjk101011 joined 13:17 cjk101010 left
masak morning, pm. 13:17
ok, I have now pushed a pls proof-of-concept. github.com/masak/proto/tree/pls 13:18
moritz_ and morning, right :-)
masak I'd be very happy to hear if people get the same results using it as I do.
pmichaud moritz_: there was a room reserved for parrot bof, and we had a couple of meetings there 13:19
mainly it was a meeting to brainstorm suggestions for parrot based on rakudo's experiences 13:20
masak to help me test it (1) check out/update the 'pls' branch from the 'proto' repo. (2) optionally run 'make' on it (3) run './proof-of-concept' and see JSON::Tiny fetch, build and test. the tests fail. (4) run './proof-of-concept --force'; the tests should run again, and then JSON::Tiny should install into ~/.perl6/lib.
moritz_ so you fetch it even though it's bundled? 13:21
masak "it"?
I don't understand the question.
moritz_ JSON::Tiny
masak ah, yes. sure.
because pls makes no guarantees as to having the latest JSON::Tiny. 13:22
it has the one it has.
13:22 stepnem left
moritz_ I'll try as soon as my rakudo buid finishes 13:22
masak cool. thanks. 13:23
kolibrie TimToady: I found out that I do not have to leave for the airport until 11:00. Do you have time if I stop by for fifteen minutes some time before then to discuss the flow diagram from my talk?
masak the output is very verbose and confused right now. but that's because it's a proof-of-concept, not a finished product :)
moritz_ env PERL6LIB='/Users/masak/gwork/proto/lib' perl6 --target=pir --output=lib/JSON/Tiny.pir lib/JSON/Tiny.pm 13:25
===SORRY!===
Unable to find module 'JSON::Tiny::Actions' in the @*INC directories.
masak shouldn't have committed Makefile
moritz_ does the Makefile properly declare dependencies?
masak remove it, ufo, and make again :)
mathw o/ 13:26
moritz_ masak: ./proof-of-concept has a hardcoded perl6 path
masak no, Makefiles don't need to consider deps, because deps are already installed when the project tries to build.
dalek kudo: 18682c8 | pmichaud++ | build/PARROT_REVISION:
Bump PARROT_REVISION.
masak moritz_: oh!
masak checks
moritz_ masak: I mean dependencies inside the makefile, so that parallel make works
masak moritz_: the sheband one?
moritz_: I couldn't get 'env' to work here. 13:27
moritz_: I wouldn't know how to "declare dependencies" within a Makefile.
moritz_ what did you try? #!/usr/bin/envperl6 ?
masak: foo: bar
masak: only builds foo if bar is already built
mathw masak: When you can spare a moment, it's time for Form.pm to be abducted
masak mathw: sure. I can spare a moment. are there any questions? 13:28
moritz_: hm.
moritz_: that would complicate the ufo script a bit, but otherwise it'd work. 13:29
moritz_ masak: proof-of-concept reports that json is already installed ... which it is not
masak moritz_: sounds like a bug.
moritz_ though I have an old project.state lying around
masak moritz_: did it create a poc-projects.state file for you?
that shouldn't matter.
moritz_ yes
$ cat poc-projects.state
{ "json" : { "state" : "installed" } }
masak huh.
moritz_: did the proof-of-concept script at least appear to install JSON::Tiny? 13:30
mathw masak: actually no, it seems to have worked perfectly
moritz_ masak: nope
masak mathw: \o/
moritz_: /o\
moritz_: what went wrong?
mathw \o/
mathw goes back to Haskell 13:31
moritz_ masak: nopaste.snit.ch/21536
masak moritz_: rm poc-projects.state
try again.
moritz_ rms, and tries again
masak that's another file I shouldn't have pushed... 13:32
moritz_ masak: nopaste.snit.ch/21537
13:32 Woody2143 left
masak I got that too at one point. 13:33
didn't investigate, but it seemed a "cd" in qx[] persisted to the next &run or something...
moritz_: does cache/json/Makefile look OK?
13:34 Intensity joined
moritz_ masak: yes, looks like a typical ufo generated Makefile 13:34
masak moritz_: and cd-ing into cache/json and doing 'make' works, too?
moritz_ yes 13:35
masak if you enter a debug line `say "cd $cwd && make" above line 234 in proof-of-concept, what does it say? 13:36
er, s:2nd/"/";`/
moritz_ should I first remove poc-projects.state again?
masak not necessary.
it should be at 'fetched' now, which only saves some time. 13:37
moritz_ cd /home/moritz/src/proto/cache/json && make 13:38
now it gets to the failing test
masak huh.
so it didn't work the time it created the Makefile, but it worked the time after?
ay ay ay.
then there is indeed some cwd interference which I haven't patched. 13:39
moritz_ got to the test failure, reported status "built"
13:39 plobsing joined
masak that's correct. 13:39
now try ./proof-of-concept --force
moritz_ and installed it with --force
masak \o/
thanks.
moritz_ how do I uninstall now?
:-)
\o/ indeed
masak not sure uninstall is even theoretically possible. 13:40
because of dependencies.
13:40 skids left
masak and because what constitutes 'install' is up to the Makefile of the project. 13:40
moritz_ right
masak I'll be happy if I get 'update' working :)
the semantics of 'update' are only strange, not impossible.
moritz_ that's why linux distribution package managers install into local dirs first, built archives from that and install them under the control of the package manager 13:41
masak they're "pretend for a moment that the project is absent, try to install it, and then reset the state to "installed" regardless of what happened'.
moritz_ e.g. dpkg-buildpackage installs into debian/$somewhere/ first, and creates a .deb from the contents of debian/$somewhere 13:42
masak moritz_: I guess one could simply 'rm' all modules from ~/.perl6/lib that occur in the project's lib dir, and then remove the project from the cache, too. 13:43
TimToady kolibrie: I'm just sitting here in the hostility room.
masak that could work.
kolibrie TimToady: okay, I'll be over in a bit
TimToady it's back on the 2nd floor 13:44
moritz_ hostility?
nice name for a room
like, 101 in 1984
masak "Hostility is Friendliness"
moritz_ :-) 13:45
TimToady *hospitality
Su-Shee "hostility is fiendliness". easily corrected. ;) 13:47
moritz_ finds in funny that in English "host" can both mean somebody who lets you stay in your hous, and "army" (which sometimes drives you out of your hous)
13:48 [Coke] joined
sbp "cleave" is another autoantonym 13:57
I think there are some others too. oh, "sanction"
masak it's a wonder we manage to communicate at all.
sbp most of the time we don't
masak :) 13:58
PerlJam shared human experience does wonders for communication
13:58 Guest23195 joined
PerlJam (we're really going to have problems when the aliens show up though) 13:58
moritz_ finds that he uses .pairs much more often that .kv to loop over hashes (and arrays, for that matter) 13:59
s/that/than/
masak make: *** No targets. Stop. 14:00
g'ah! I wish I knew why it says that.
and then the next time, it works.
OH!
moritz_ try to add a call to pwd before make
masak tries closing the file handle
14:01 colomon joined
masak moritz_: I had tried that already. it showed the proto dir. 14:01
now it works!
moritz_ \o\
masak /o/ 14:02
pmichaud
.oO(those look an awful lot like TIE fighters to me...)
14:02 [Coke] left
TimToady >ö< <-- X-wing 14:02
moritz_ pmichaud: don't know if you had your "Cool Perl 6" talk yet... comb instead of split would be more idiomatic
wouldn't that be |o| for the TIE fighters? :-) 14:03
pmichaud moritz_: I couldn't figure out how to do it with comb at the time (and was in a rush)
PerlJam (o) for Vader's
14:03 rv2733 joined
moritz_ pmichaud: ok 14:03
pmichaud sure, |o| is the TIE fighter when it's coming straight at you. \o\ looks like a tie-fighter as it's making a swooping pass at you.
masak :) 14:04
moritz_ relatvisticly distorted TIE figher :-)
colomon |o|
pmichaud
.oO( touchdown? )
 14:05
PerlJam Next thing you know we're got ascii animation of the entire star wars end-of-movie fight scene
s/we're/we've/
TimToady )O( # TIE fighter in a convex mirror 14:06
PerlJam www.asciimation.co.nz/
masak :)
14:07 rlb3 joined
moritz_ )O( |O| (O) |O| resonance! 14:07
14:07 rlb3 left 14:08 rlb3_ joined 14:11 rlb3 joined 14:12 ajs joined
masak now that the proof of concept works, it would be fun to see if Emmentaler can finally be built. 14:14
14:15 rlb3_ left
dalek kudo: bf5bbd4 | pmichaud++ | src/core/ (2 files):
Turn Array.exists into List.exists, write it a bit cleaner.
 14:15
kudo: 7e3ddca | pmichaud++ | src/core/EnumMap.pm:
Redo EnumMap.exists .
colomon rakudo: my @a = 'a'...'f'; say (@a[2..*]).munch(20) 14:17
p6eval rakudo 11cbd4: ( no output )
moritz_ that looks rakudobuggy 14:18
colomon it is, I was trying to do it as an example for pmichaud++
(I talked in person for the other bits of the conversation... freaky!)
moritz_ woah there. Don't get used to it! 14:19
14:19 felliott joined
colomon moritz_: don't worry, I've got a few hours more. :) 14:20
moritz_ :-)
cono munch == batch ?
colomon cono: yes
moritz_ munch = destructive batch 14:21
colomon batch is going to be history.
I think?
moritz_ yes
cono it's destruct aray? 14:22
moritz_ yes, just like ship
erm
shift
cono @a[2..3].munch(2) - like splice in perl5 ?
moritz_ don't know if it works on array slices 14:23
splice exists too
.munch(5) is calling .shift 5 times
cono ah, I got it, thanks :)
batch - just copy? 14:24
moritz_ yes 14:25
.batch($x) was the same as .[^$x]
pugssvn r31422 | colomon++ | Intial version of new trig test support code.
cono "batch is going to be history" it means that it will be removed in next versions? 14:26
colomon that was my understanding. but I could be wrong.
moritz_ and remoed from the spec
cono :(
colomon cono: as far as I know, .batch(N) is exactly the same as .[^N] 14:28
cono ah, that's why 14:29
understand
colomon batch was defined on iterators, and we try not to expose iterators to the user now. 14:30
14:31 spinclad joined
masak both my talks were accepted for YAPC::EU. I didn't expect that :) conferences.yapceurope.org/ye2010/talk/2896 conferences.yapceurope.org/ye2010/talk/2898 14:31
colomon rakudo: my @a = 'a' ... 'z'; say @a[*/2]
p6eval rakudo 11cbd4: OUTPUT«n␤»
masak maybe I'll even have to finish one of them before the conference this time. 14:32
arnsholt Congrats. Wish I could attend them 14:33
moritz_ should try to sneak in a lightning talk at least
14:33 thebird left
arnsholt Especially the parsing one sounds cool 14:33
14:33 thebird joined
masak arnsholt: :) 14:33
pmichaud masak: looks like you got more talks accepted than I did. :)
masak arnsholt: still need to implement some of the things for that talk.
arnsholt =D
Now you have incentive ;) 14:34
masak pmichaud: huh. :)
arnsholt: exactly!
pmichaud gee, it lookslike one of my talk submissions got lost
hmmmmm
oh well, I'm happy enough with just one. :) 14:35
14:35 thebird left 14:39 thebird joined, lest_away is now known as lestrrat
pmichaud confirms his talk. 14:39
14:39 plobsing left 14:41 lestrrat is now known as lest_away 14:43 lest_away is now known as lestrrat, ash_ joined 14:44 estrabd_afk is now known as estrabd
masak std: my (@xxx, @yyy, @zzz); for @xxx; @yyy; @zzz -> $x; $y; $z { ... } 14:44
p6eval std 31422: OUTPUT«===SORRY!===␤Missing block at /tmp/0Ied7MayNn line 1:␤------> my (@xxx, @yyy, @zzz); for @xxx⏏; @yyy; @zzz -> $x; $y; $z { ... }␤ expecting any of:␤ desigilname␤ twigil␤Parse failed␤FAILED 00:01 114m␤»
masak thought so 14:45
14:45 alester joined, rlb3 left 14:46 lestrrat is now known as lest_away 14:50 [particle] left
TimToady masak: that is very ancient p6 14:52
14:52 cygx joined
masak TimToady: it's from A04. :) 14:52
just checking that STD is saner than that nowadays. 14:53
moritz_ where A stands for Ancient :-)
pmichaud worries about the notion of "ancient p6".
masak and S stands for "Sane" :P
moritz_ pmichaud: why?
pmichaud it's been around long enough that we can have things that are "ancient?
"
14:54 [particle] joined
moritz_ pmichaud: "ancient" is always relative. Talking about human history has a different notion of "ancient" than talking about creation of the universe 14:54
masak pmichaud: I don't think there's something negative in the fact that Perl 6 has evolved during the past 10 years. 14:55
colomon btw, try one at integrating the new str2num code into Rakudo:
pmichaud I forgot my smiley again.
colomon > say 3.14159_26535_89793_23846_26433_83279_50288 3.925520848
ooo, that paste doesn't work right here.
> say 3.14159_26535_89793_23846_26433_83279_50288
3.925520848
rakudo: say 3.14159_26535_89793_23846_26433_83279_50288
p6eval rakudo 7e3ddc: OUTPUT«===SORRY!===␤PAST::Compiler can't compile node of type BigInt␤»
moritz_ rakudo: say (3.92 - 3.14) / 3.14 14:56
p6eval rakudo 7e3ddc: OUTPUT«0.248407643312102␤»
moritz_ well, 25% error...
masak moritz_: if you arrive to Pisa on the 1st, make sure to see "Gioco del Ponte"! wikitravel.org/en/Pisa#Do
14:57 Woody2143 joined
colomon take 2: 14:57
> say 3.14159_26535_89793_23846_26433_83279_5028 3.14159265358979
pmichaud rather than get a bigger slice, colomon++ is trying to make a bigger pie. :-)
colomon 3.14159265358979
masak moritz_: oh wait. "last sunday of June". nvm.
would have been fun to see, though. :)
moritz_ masak: pity. But thanks for the link anyway
cygx hi 14:58
moritz_ hi cygx
masak moritz_: it's the first link I find that says that there's more than the tower to see :P
cygx moritz_: any thoughts on github.com/moritz/try.rakudo.org/ra...ter/DESIGN
masak cygx! \o/
moritz_ cygx: I've only glanced over it; looks sane so far 14:59
cygx: I'll have to defer detailed reading for a few hours due to external constraints 15:00
cygx moritz_: ;)
moritz_: that's ok, as I had to defer implementing anything of that due to my lacking knowledge of perl5 15:02
15:02 [particle] left, charsbar joined
jnthn pmichaud: Only one of mine got accepted to, fwiw. 15:02
*TOO
cygx moritz_: I'll push what I did so far to the repository; would be nice if someone could look at it and tell me if I did something completely retarded 15:03
moritz_ cygx: if you have any Perl 5 questions, don't hesitate to ask in private message, or in #perlde on irc.perl.org if you speak German
cygx: sure, will look at it (but also deferred to later tonight)
15:04 lest_away is now known as lestrrat
cygx moritz_: will do; nothing urgent comes to mind, though - I'm good so far 15:05
15:05 [particle] joined 15:06 meppl joined 15:07 lestrrat is now known as lest_away, rtward joined
ash_ you could cheat save yourself the db if you just keep the connection to the browser open 15:07
ash_ see comet, server push, long polling, hidden iframe 15:08
cygx ash_: there are a lot of things that can go wrong with that approach: timeouts, proxies,...
moritz_ accidentally closed browser windows
etc.
I'm find with a db
it makes for a more traditional and easier to understand design 15:09
ash_ i'd image a closed browser would result in a new session anyway, unless your doing the sessions on remote ip's
cygx ash_: actually, long polls are used in case of blocking on rakudo output
ash_ (which can be bad for people behind a router)
moritz_ cygx: any preference for a db?
(so that I can know what to set up) 15:10
mysql, postgres, sqlite should all be fine
and easy to install
cygx moritz_: not really - Postgres would most likely be the most robust solution
moritz_ ok
15:12 skids joined
colomon started spectest, now have a bit of time to work on $work. :) 15:13
ash_ if you need any help cygx i have a bit of experience with the web, i'd be more than happy to help out if you want help 15:14
colomon [particle]: o/ 15:15
[particle] hola, colomon
pmichaud \o/ particle++
moritz_ so much for $work, he? :-)
colomon ;) 15:16
[particle] howdy, pmichaud
glad we had that chat yesterday
pmichaud same here :)
[particle] i'm downloading kubuntu 10.04 now, i'm gonna try developing there
colomon moritz_: it's so weird actually knowing people on the channel IRL. :) 15:17
guess I'm going to have to try to make it to Europe one of these days.
pmichaud colomon: who are you calling weird?!? ;-)
moritz_ colomon: right, i fealt the same after the Copenhagen Perl 6 hackathon with jnthn, masak, arnsholt, baest and others 15:18
masak it's good for morale to meet at least once a year :)
moritz_ indeed
masak makes it so much real.
cygx ash_: I did some web-development as well, but not for several years and not in perl, so your help is appreciated 15:19
moritz_ also for the climate
pmichaud looks like we'll have a Pisa hackathon on Aug 3.
masak and for the in-jokes :)
moritz_ it feels so much easier to yell at somebody you've never seen
pmichaud: that's great
colomon moritz_: :)
cygx ash_: I'd like to give developing the frontend a shot, but I'll need some time to familiarize myself with perl; I'll remember to start bugging you when think start going wrong ,)
masak moritz_: or, flip-wise, it's harder to be empathic with someone you haven't.
moritz_ masak: right 15:20
masak s/\./seen./
cygx *think --> things ups ;)
[particle] has to make a yapc::eu some day
moritz_ [particle]: where some day == Aug 2nd 2010 :-) 15:21
[particle] ah, if only....
pmichaud we should probably plan for a Rakudo Star BOF on Aug 4, too. 15:22
ash_ cygx: alright, if your going to do the front end with javascript, for an interactive terminal, i'd suggest jquery, it will make your life a lot easier
moritz_ pmichaud: +1
www.perlmonks.org/?node_id=846326 I wonder how much of that is YAPC::NA inspired :-) 15:23
colomon I was just explaining to pmichaud and TimToady that I'm taking one more vacation on my own next month, and then I figure I don't get to go anywhere without my family for at least 11 months.
masak likes that pmichaud++'s talk "Rakudo Star -- A usable Perl 6 release" is filed under the "Core Perl" track 15:24
15:25 patch joined
cxreg was poking around trying to make FakeDBD::Pg. even if someone else does it first it's fun 15:26
15:26 kensanata joined
moritz_ cxreg: which route are you going? direct NCI, or use parrot's existing bindings? 15:27
cxreg: if you have a github account (or create one), I'm sure we can arrange commit access to FakeDBI for you 15:28
hugme: list projects 15:29
hugme moritz_: I know about Math-Model, Math-RungeKutta, book, fakedbi, gge, hugme, ilbot, json, november, nqp-rx, nqpbook, perl6-examples, perl6-wtop, process-cmd-args, proto, pugs, rmp, star, svg-matchdumper, svg-plot, tardis, temporal-flux-perl6syn, try.rakudo.org, tufte, web, yapsi
moritz_ yes, fakedbi is in there
15:29 moritz_ sets mode: +v hugme 15:32 Trashlord left
cygx ash_: I'll do the server-side part of the frontend first; until that's done, the client-side code currently in the repository should be sufficient 15:33
ash_: once the client-server communication has be worked out, switching to a Jquery solution might be a good idea to add some 'polish' 15:34
cxreg moritz_: NCI for now, i havent looked at parrot's bindings
moritz_ cxreg: if you're interested, here are the tests for them: svn.parrot.org/parrot/trunk/t/library/pg.t 15:35
ash_ cygx: have you thought about doing it in perl6? (just curious) 15:38
15:38 kensanata left
cygx ash_: I have, but I think the first iteration should stay in perl5 15:42
mberends that epic tennis match has resumed at Wimbledon 15:43
moritz_ mberends: continuation based tennis match :-)
I guess it'll be over rather quickly
mberends lol
11 hours, 68-68 in the final set 15:44
masak wow.
ash_ an 11 hour tennis match? o.0 thats crazy...
cygx ash_: the codebase shouldn't become too large and is fairly modular, so migrattion to p6 is definitely an option 15:45
moritz_ my experience from table tennis is that if long-time tied games are interrupted for longer, they are usually decided pretty quickly after the interruption
jnthn is watching Slovakia trying to hold on to a lead over Italy
That'd be a lovely upset. :-)
moritz_ indeed 15:46
beating the world champion
ash_ cygx: i was thinking it might nice to see if rakudo was up to the task more to find out where its not up to the task so we can fix it :P
real world examples of rakudo are always good test cases
colomon the problem with work is std::cout << "got here" << endl; instead of say "got here";
examples++
$work, I mean.
moritz_ ash_: It's pretty obvious to me that rakudo has many weaknesses in the areas of IO and IPC 15:47
ash_: I don't know if a real-life example is really needed to prove that
jnthn: your time to say \o/ :-)
*turn 15:48
ash_ yeah, but you could do most of the interactive console without IPC (i think) and the IO issues don't seem like a show stopper, it would have to have a slightly different design that doesn't involve threads, but i think it could work 15:49
moritz_ I don't think the current design involves threads... or does it? if so, where?
ash_ blocking io causes a second thread in the strategy cygx described 15:50
cygx moritz_: the controller has to listen to both the frontent and rakudo 15:51
*frontend
moritz_ cygx: I'd prefer async IO over threads there, but in the end it's the implementor's choice
TimToady std: #`〝 comment 〞 15:52
p6eval std 31422: OUTPUT«Use of uninitialized value in vec at CursorBase.pm line 548.␤ok 00:01 108m␤»
TimToady sorear: ^^
cygx moritz_: async io is fine with me; the current design still has some 'issues' anyway
colomon TDD works much better if your tests are correct...
wait! I need tests for the tests!
moritz_ colomon: :-)
TimToady sorear: I'll let you fix that one, since it's in the guts of the DFA
moritz_ calls it a day 15:53
colomon oh wait, the test was correct, it's the code that's wrong. 15:54
pugssvn r31423 | lwall++ | [tryfile] find uniprops in . 15:55
ash_ who tests the tests?
:P
moritz_ and who tests the tests that test the tests?
colomon who tests the testman?
moritz_ proposes a meta-circular testing scheme where the tests test themselves 15:56
colomon "1 of 1 tests succeeded"
\o
\o/
see moritz_ -- $work!
pugssvn r31424 | lwall++ | [STDeco] remember $?LONGNAME of module; use that to form name of .syml file
r31424 | (this fixes the bug where it puts Test::Util's .syml in the wrong dir)
16:01 justatheory joined
TimToady rakudo: say '␍'.ord 16:02
16:02 Wolfman2000 left
p6eval rakudo 7e3ddc: OUTPUT«9229␤» 16:02
TimToady darn
rakudo: say '␤'.ord
p6eval rakudo 7e3ddc: OUTPUT«10␤»
ash_ is that right?
colomon ash_: yes 16:03
ash_ i thought ␤ was a unicode newline
TimToady I was hoping p6eval would also translate ␍
colomon oh, I dunno.
ash_ ␤
oh is its p6eval thats translating it? 16:04
16:04 cdarroch joined, cdarroch left, cdarroch joined
masak yes. 16:04
TimToady rakudo: say ' 16:05
p6eval rakudo 7e3ddc: OUTPUT«===SORRY!===␤Confused at line 11, near "say '\x{16}"␤»
TimToady
jnthn moritz_: Yes, \o/ indeed! Not so sure that Sunday will give me anything to \o/ about though. ;-)
pugssvn r31425 | colomon++ | [t/spec] Fudge hyper.t to work in Rakudo again. 16:06
pmichaud jnthn: ping 16:07
16:07 ashleydev joined
jnthn pmichaud: pong 16:07
pmichaud are you aware of anything that is currently using WHENCE in master?
jnthn pmichaud: Some of the BUILD code may be (more) 16:08
pmichaud: However, we don't have a way to make WHENCEs yet
pmichaud: So we're not relying it even if it's there
pmichaud so I suspect it's not presently using it.
okay, that's what I wanted to know.
jnthn *on it
Yeah, afaik you can just rip out whatever's there and do it how you like
pugssvn r31426 | colomon++ | [t/spec] Fudge zip.t to work in Rakudo again.
jnthn Without any adverse effects
pmichaud I may (temporarily) remove the BUILD code (that's the piece I was looking at) 16:09
jnthn pmichaud: Well, if you leave in the non-WHENCE bits, then yes
Should be fine.
pmichaud right, that's what I meant.
jnthn OK :-)
16:11 Ross joined
colomon rakudo: say ~(42, 24, * % * ... 0) 16:12
p6eval rakudo 7e3ddc: OUTPUT«42 24 18 6 0␤»
cognominal go for a movie not taxing my brain before switching to full masak mode. 16:13
*goes
masak cognominal: enjoy!
colomon rakudo: say ~(42, 24, * % * ... 0)[*-2] 16:14
p6eval rakudo 7e3ddc: OUTPUT«6␤»
cognominal A-Team abeilt dubbed. French suburbia... :( As it mattered for that kind of movie.
colomon rakudo: say ~(42, 23, * % * ... 0)[*-2] 16:15
p6eval rakudo 7e3ddc: OUTPUT«1␤»
colomon rakudo: say ~(42, 23, * % * ... 0)
p6eval rakudo 7e3ddc: OUTPUT«42 23 19 4 3 1 0␤»
colomon rakudo: say ~(42, 1, * % * ... 0)
p6eval rakudo 7e3ddc: OUTPUT«42 1 0␤»
colomon I like.
16:16 kst joined, jaldhar left, cjk101011 left
colomon www.perlmonks.org/?node_id=772778 16:16
cognominal Anyway I think I have found my problem 16:17
> my $tree = $_=["01"]; my @path=[0,1]; for @path.kv -> $i, $offset { when Array { die "offset $offset at position $i in path {@path.perl} for tree {$tree.perl} too large" if $offset >= +@($_); $_ = $_[$offset] ; }} 16:18
pugssvn r31427 | colomon++ | [t/spec] Refudge for Rakudo, and add a couple of tests which are variants on failing tests.
cognominal offset 0 1 at position 0 in path [[0, 1]] for tree ["01"] too large
TimToady sorear: also, the lexer doesn't seem to count ^M as whitespace
cognominal it seems that rakudo does not like $_ changed in a when. Needs to golf it further.
colomon rakudo: say ~(24, 42, * % * ... 0) 16:19
p6eval rakudo 7e3ddc: OUTPUT«24 42 24 18 6 0␤»
cognominal don't know it is a bug or a misuse.
afk&
colomon rakudo: say ~(1, 42, * % * ... 0)
p6eval rakudo 7e3ddc: OUTPUT«1 42 1 0␤»
colomon moritz_++ 16:20
masak swimming & 16:23
16:23 masak left
colomon rakudo: (1, 1/2, 1/4, ... 0).[^5].map({.perl}).join(', ').say 16:26
16:26 Ross left
p6eval rakudo 7e3ddc: ( no output ) 16:26
16:27 Ross joined
colomon rakudo: (1, 1/2, 1/4, ... 0).[^5].say 16:28
p6eval rakudo 7e3ddc: ( no output )
colomon rakudo: (1, 1/2, 1/4 ... 0).[^5].say
p6eval rakudo 7e3ddc: OUTPUT«10.50.250.1250.0625␤»
pmichaud jnthn: ping 16:29
colomon std: (1, 1/2, 1/4, ... 0) 16:30
p6eval std 31422: OUTPUT«ok 00:01 110m␤»
colomon rakudo: (1, { 1 / ((1 / $_) + 1) } ... 0).[^5] 16:33
p6eval rakudo 7e3ddc: ( no output )
16:33 aCiD2 left
moritz_ doesn't know how he deserved colomon's last ++ 16:33
colomon perlmonks link above. 16:34
appears you came up with (42, 24, * % * ... 0)[-2] for finding the GCD 16:35
TimToady just impressed Damian with it. :)
[*-2], I mean. 16:36
PerlJam colomon++ that's cool
moritz_ colomon: ah, very old stuff 16:37
colomon but still very worth a ++ for you. :)
16:38 gbacon joined
ash_ std: my @a = 1; say @a[-1]; 16:40
p6eval std 31427: OUTPUT«===SORRY!===␤Unsupported use of [-1] subscript to access final element; in Perl 6 please use [*-1] at /tmp/cwFRpYMiJz line 1:␤------> my @a = 1; say @a[-1]⏏;␤Parse failed␤FAILED 00:01 114m␤»
jnthn pmichaud: pong 16:42
16:42 mberends left
pmichaud jnthn: I have some questions about 'is copy' 16:43
16:43 Ross left
pmichaud jnthn: when doing something like sub abc(@foo is copy) { ... } 16:43
currently we create an ObjectRef to point to the cloned array. Any particular need/reason for the ObjectRef, as opposed to binding the array directly? 16:44
jnthn pmichaud: It may well date back to a pervious model of doing things. 16:46
er, previous
pmichaud: oh 16:47
Maybe constraints
But...no probably not.
pmichaud: I'd say make it look how things in the lexpad normally look for arrays.
So if that involves getting rid of the ObjectRef that's probably fine. 16:48
TimToady rakudo: say ~(1,3,5...3)
p6eval rakudo 7e3ddc: ( no output )
TimToady S03:1937 16:50
16:51 Chillance joined
pmichaud jnthn: okay, and it's okay if I make the scalar "is copy" look like a scalar would? ;-) 16:53
jnthn erm
Probably
The test suite will call you on it if not. ;-)
Or masak. ;-) 16:54
pugssvn r31428 | lwall++ | [STD] worry about extra comma before series operator
moritz_ TimToady++ # that's something I often got wrong
jnthn pmichaud: I think there's some ticket about is copy and hashes also. 16:55
pmichaud jnthn: okay.
jnthn Or something along those lines
pmichaud I'm likely to refactor that a bit (nothing huge)
jnthn I suspect the whole is copy thing is out of date with regard to current state.
pmichaud right 16:56
jnthn I haven't updated it at all
So unless you did, it'll be out of date.
pmichaud and since I'm cleaning up variables, whence, etc.
it seems like a good time to address
jnthn Aye, go for it.
Should be easy enough.
colomon rakudo: say ~(1, 1, -> *@a { [+] @a } ... 20000) 16:57
p6eval rakudo 7e3ddc: OUTPUT«1 1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384␤»
16:58 tomdbs98 joined
colomon TimToady: see? ;) 16:58
16:59 tomdbs98 left 17:00 macdaddy joined
moritz_ now that's evil stuff! 17:00
17:00 macdaddy is now known as Guest73864 17:01 envi^home left 17:03 Ross joined
colomon rakudo: say ~(1, -1/2, 1/4 ... 0).[^5] 17:05
p6eval rakudo 7e3ddc: OUTPUT«1 -0.5 0.25 Any() Any()␤»
TimToady rakudo: say ~(2,3,5 ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n } } ... 100) 17:06
p6eval rakudo 7e3ddc: OUTPUT«===SORRY!===␤Confused at line 11, near "say ~(2,3,"␤»
17:06 [Coke] joined
colomon rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n } } ... 100) 17:06
TimToady : say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n } } ... 100)
p6eval rakudo 7e3ddc: OUTPUT«Divide by zero␤ in 'infix:</>' at line 3580:CORE.setting␤ in 'infix:</>' at line 3245:CORE.setting␤ in 'infix:<%>' at line 3250:CORE.setting␤ in 'negate' at line 1␤ in <anon> at line 1␤ in 'Any::join' at line 718:CORE.setting␤ in 'List::Str' at line 2373:CORE.setting␤
..in ma…
TimToady : say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n }; $n } ... 100) 17:09
pugssvn r31429 | colomon++ | [t/spec] Fix typos, refudge for Rakudo.
TimToady rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n }; $n } ... 100) 17:10
p6eval rakudo 7e3ddc: ( no output )
TimToady rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n !% any(@a) { ++$n }; $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50␤»
17:11 meppl left, rv2733 left
TimToady rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1]; repeat while $n % any(@a) { ++$n }; $n } ... 50) 17:11
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 30␤»
17:12 xuser joined
PerlJam TimToady: don't you want an $n++ before the while loop? 17:12
TimToady rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n++ while $n !% any(@a); $n } ... 50)
PerlJam or that.
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49␤»
TimToady PerlJam: it's a repeat, so it's supposed to do the ++$n first 17:13
colomon rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n++ while [||] ($n <<!%<< @a); $n } ... 50) 17:15
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤»
17:15 thebird left
colomon rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n++ while [||] ($n <<!%<< @a); $n } ... 100) 17:15
p6eval rakudo 7e3ddc: 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␤»
colomon rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n += 2 while [||] ($n <<!%<< @a); $n } ... 100) 17:16
p6eval rakudo 7e3ddc: 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␤»
TimToady rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n++ while $n !% all(@a); $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤» 17:17
moritz_ rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(@a); $n } ... 50) 17:19
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤»
pmichaud rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(@a); $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤»
pmichaud rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2..sqrt(@a)); $n } ... 50)
17:19 pyrimidine joined
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 9 11 13 15 17 19 23 25 29 31 35 37 41 43 47 49␤» 17:20
pmichaud rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2..sqrt(@a[*-1])); $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 9 11 13 17 19 23 25 29 31 37 41 43 47 49␤»
moritz_ sqrt(@) is probably a dangerous assumptioin 17:21
pmichaud moritz_: yes
moritz_ but I don't know enough number theory to prove it :-)
pmichaud it is dangerous :)
rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(@a.grep( * < sqrt($n)); $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 11␤»
pmichaud rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(@a.grep( * < sqrt($n))); $n } ... 50) 17:22
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 9 11 13 17 19 23 25 29 31 37 41 43 47 49␤»
pmichaud rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(@a.grep( * <= sqrt($n))); $n } ... 50)
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤» 17:23
jnthn afk for a while, noms and beers
moritz_ who did the md5 implementation via parrot recently? cosimo? or cognominal? (just remember it was a ^co guy :-) 17:25
ah, cosimo++
just found it on github 17:26
17:26 stepnem joined 17:30 xuser left 17:35 patrickas joined 17:37 perlygatekeeper joined
sorear good morning #perl6 17:39
moritz_ oh hai sorear 17:40
patrickas good evening #perl6 17:41
:-)
moritz_ hi patrickas
pugssvn r31430 | moritz++ | [docs/feather] remove some outdated links 17:43
patrickas hello moritz_ what's new in perl6land ?
moritz_ patrickas: lots of people have fun at YAPC::NA
patrickas so I heard :-) and YAPC::EU is getting close too ! 17:44
moritz_ patrickas: I've tried to play with the new PAST optimization framework - but no luck so far :/
found two bugs in a row 17:45
patrickas :-)
speaking of optimizations
rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2..sqrt($n)); $n } ... 50);
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤»
pugssvn r31431 | moritz++ | [docs/feather] remove another broken link 17:46
patrickas seems to be strangly faster than pmichaud's (using @a.grep( * <= sqrt($n)) ) ... I suppose grep is really not optimized for such tasks
maybe for big enough numbers the grepped version may be faster again 17:47
moritz_ yes, might be different
PerlJam patrickas: pm's version does O(N) sub calls per iteration, whereas 2..sqrt($n) does not 17:50
patrickas PerlJam I don't think so ...
moritz_ PerlJam: define N :-)
PerlJam N==+@a 17:51
patrickas pmichaud's version is cleaning all items form @a that are bigger than sqrt($n) before working on them
so effectively it is O(sqrt(n)) too ... it is just the grep that is adding a small constant tot he time 17:52
PerlJam patrickas: The only difference between the two is one has 2..sqrt($n) and the other has @a.grep(* <= sqrt($n)), right? 17:55
patrickas right
moritz_ the real interesting question is how big sqrt($n) is in comparison to @a 17:57
and how that changes for later values in the sequence
TimToady you could use 2,(3,5...sqrt($n)) to at least avoid dividing all the even numbers
PerlJam moritz_: yeah, at some point the calls to the anonymous sub (and possibly the recomputation of sqrt($n)) in the grep version will no longer swamp the time it takes to generate 2..sqrt($n) 17:59
18:01 bphillips left
patrickas TimToady: I must me missing something :-( 18:02
raudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...sqrt($n))); $n } ... 50); #Is that what you mean ?
TimToady yes
patrickas rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...sqrt($n))); $n } ... 50);
p6eval rakudo 7e3ddc: ( no output ) 18:03
patrickas it is taking so much time I have not been able to calculate it for n = 10 yet :-)
rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...sqrt($n))); $n } ... 5);
p6eval rakudo 7e3ddc: OUTPUT«2 3 5␤»
patrickas rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...sqrt($n))); $n } ... 10);
p6eval rakudo 7e3ddc: ( no output ) 18:04
patrickas rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...sqrt($n))); $n } ... 7);
p6eval rakudo 7e3ddc: ( no output )
PerlJam throw in a little memoization and it should get much quicker :)
TimToady only if hashing two arguments and looking them up is faster than a simple modulus 18:05
patrickas It's shocking that the ... operator is so slow! or is it something else I am not getting ? 18:07
pmichaud it's not ... that is slow here, it's all() and *@a and stuff like that. 18:08
moritz_ patrickas: I guess it's because ... has to handle so many different cases
patrickas pmichaud: but the only difference bwteen TimToady's version and mine is that his uses ... to get the odd numbers while mine uses .. to get them all 18:09
moritz_: Yes of course it should be slower but I would have expected some kind of a constant factor
it times out while coputing the primes less smaller than 7 ... I would say there is a bug 18:10
PerlJam patrickas: you don't know that it's not a constant :)
patrickas maybe the ... is never termination ? because the end of the range is a real ?
PerlJam (a really big constant)
I presume that all(2,3,5...sqrt($n)) forces the complete evaluation of the sequence? 18:12
18:12 justatheory left
PerlJam er, with the proper syntax 18:12
pugssvn r31432 | sorear++ | [Cursor] Re-special-case \r in \h; Perl/Blank doesn't cover it after all 18:13
TimToady rakudo: say ~(2,3,5...50)
p6eval rakudo 7e3ddc: OUTPUT«Unable to figure out pattern of series␤ in main program body at line 1␤»
TimToady rakudo: say ~(2,(3,5...50))
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49␤»
TimToady that why the extra parens
patrickas hehhe got it 18:14
rakudo: say (2,(3,5...sqrt(5)));
p6eval rakudo 7e3ddc:
..OUTPUT«235791113151719212325272931333537394143454749515355575961636567697173757779818385878991939597991011031051071091111131151171191211231251271291311331351371391411431451471491511531551571591611631651671691711731751771791811831851871891911931951971992012032052072092112132152172192212232…
PerlJam patrickas: nice
patrickas it is never terminating :-)
colomon known bug. 18:15
patrickas colomon now you tell me :-)
TimToady we were discussing it earlier :)
colomon :)
I was hacking rather than IRCing. :)
patrickas hehe colomon++ someone's gotta do the work 18:17
slavik wait ...
rakudo: say 2**64
p6eval rakudo 7e3ddc: OUTPUT«1.84467440737096e+19␤»
slavik aww :(
patrickas and TimToady-- for suggesting an "optimization" using a known bug being discussed earlier and confusing me :-)
pugssvn r31433 | sorear++ | [Cursor] Unused Unicode blocks should be initialized to "", not undef
colomon slavik: I'm figuring on taking a stab at proper bigints after R*
sorear TimToady: both bugs you told me in scrollback are now fixed.
slavik colomon: I will buy you a beer
colomon sorear++ 18:18
slavik or vodka
vodka is better imo
18:18 rindolf joined
rindolf Hi all. 18:18
colomon I prefer scotch
or cider
slavik eww
PerlJam patrickas: you can always do 3,5...sqrt($n).Int 18:19
sorear Hello rindolf 18:20
rindolf sorear: hi. 18:21
TimToady or .ceiling
colomon patrickas: 2, 3, 5 ... sqrt(50) should work fine, but if the end point is less than the last LHS value, you'll have trouble.
patrickas PerlJam TimToady nope :(
colomon you are right
PerlJam I guess .Int truncates 18:22
patrickas it's the fact that it is smaller than 5 not that it's a real
PerlJam ah
colomon sure, all Reals are well-ordered. ;) 18:24
patrickas rakudo: say max(4,6);
PerlJam yeah, but that doesn't mean it wasn't the source of a bug :)
p6eval rakudo 7e3ddc: OUTPUT«Unable to handle non-closure Ordering yet␤ in 'Any::max' at line 1362:CORE.setting␤ in 'max' at line 1560:CORE.setting␤ in main program body at line 11:/tmp/j0xtf8Q_mV␤»
colomon patrickas: you want 18:25
rakudo: say [max] 4, 6;
or
p6eval rakudo 7e3ddc: OUTPUT«6␤»
colomon rakudo: say 4 max 6
p6eval rakudo 7e3ddc: OUTPUT«6␤»
patrickas oh it's an operator!
TimToady $x max= $y
patrickas i wouldn't have guessed from the error message!
colomon the sub version of max wants a closure 18:26
sorear I think I'm going to add to Sprixel/VICIL something like FORTH's CODE
colomon FORTH!
sorear a routine declarator, with lower-than-Perl6-level code in it 18:27
colomon rakudo: say ([minmax] 2, 4, 5, 2, 3).perl
p6eval rakudo 7e3ddc: OUTPUT«2..5␤»
sorear this will make it possible to define the Scalar, Parcel, and Capture classes - they can't be defined in Perl 6 because the binder needs them 18:28
colomon rakudo: say ([minmax] 2, 4, 5, 2, 3).bounds.perl
p6eval rakudo 7e3ddc: OUTPUT«(2, 5)␤»
sorear thoughts? 18:29
patrickas rakudo: say ~(2,3,5, ->*@a { my $n = @a[*-1] + 2; $n+=2 while $n !% all(2,(3,5...(5 max sqrt($n)))); $n } ... 50); #So this seems to be fastest 18:30
p6eval rakudo 7e3ddc: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47␤»
patrickas BTW is there a Benchmark.pm6 yet ?
18:31 justatheory joined
PerlJam patrickas: I started github.com/perlpilot/benchmark 18:31
patrickas: there may be others.
It looks like the spec says to return Nil in the 3,5...4 case, but for the prime problem, it looks like returning 3,5 would be more useful. 18:33
colomon I think it's supposed to return just 3 for 3, 5 ... 4 18:35
?
I'm sure I understand how 3, 5 would be more useful. so you don't have to say 5 max sqrt?
PerlJam yes, I was just thinking that. 18:36
18:36 [particle] left
PerlJam rakudo: say 3,5...5 max sqrt(100); 18:38
p6eval rakudo 7e3ddc: OUTPUT«3579␤»
PerlJam interesting
colomon I think making the left hand the pattern and the right hand the limit works better. so that 3, 5 ... 4 would be 3 18:39
PerlJam colomon: S03:1937ish says ''if a supplied limit value is on the "wrong" side of the first of those values, Nil is returned''
colomon "first" 18:40
this is the second.
PerlJam ah, yes.
I'm a little slow today
colomon no worries.
and I just spent an hour discussing this stuff with pmichaud and TimToady, because the tests had changed and I was trying to figure out if they were still correct. 18:41
PerlJam if the LHS is a list, the series op only cares about the last 2 values in the list right? 18:42
18:43 spinclad left 18:45 gbacon left
PerlJam colomon: I mean ... if 3,5...4 returns 3, what should this do? 1,3,5,7,9...4 18:46
PerlJam reads the spec
oh, yes. 1,3 is what I would expect and that's what the spec says 18:47
TimToady++ 18:48
18:48 [particle] joined
colomon dunno if that works yet, mind you. 18:48
rakudo: say 1, 3, 5, 7, 9 ... 4 18:49
p6eval rakudo 7e3ddc:
..OUTPUT«135791113151719212325272931333537394143454749515355575961636567697173757779818385878991939597991011031051071091111131151171191211231251271291311331351371391411431451471491511531551571591611631651671691711731751771791811831851871891911931951971992012032052072092112132152172192212232…
18:49 patspam joined
PerlJam looks like not :) 18:49
colomon so my new decimal number code is 8x slower than the old code.
guess it's still too slow to incorporate.
must optimize....
slavik colomon: run it on top of ghc :P 18:50
PerlJam rakudo: my @a = 13; 1,3...@a 18:51
p6eval rakudo 7e3ddc: ( no output ) 18:52
PerlJam rakudo: my @a = 13; say 1,3...@a
p6eval rakudo 7e3ddc: OUTPUT«135791113␤»
18:52 gbacon joined
moritz_ it runs up to 13? 18:52
that's... kinda weird
PerlJam rakudo: my @a = 1,3,5,7; my @b = 25..35; say ~(@a..@b); 18:57
18:57 rtward_ joined
p6eval rakudo 7e3ddc: ( no output ) 18:57
18:57 rtward_ left
PerlJam rakudo: my @a = 1,3,5,7; my @b = 25..35; say ~(@a...@b); 18:58
p6eval rakudo 7e3ddc: OUTPUT«1 3 5 7 9 11 13 15 17 19 21 23 25 26 27 28 29 30 31 32 33 34 35␤»
19:00 rtward left 19:04 cono left 19:06 cono joined
diakopter shell out to perl 19:07
19:07 Mowah left 19:09 ashleydev left
patrickas is 1,1...2 supposed to return Nil? 19:11
19:11 rtward joined, araujo left
pmichaud patrickas: what would the increment be in that case? I'm guessing it acts like an increment of zero. 19:12
colomon increment of zero, yes
moritz_ so infinite number of 1s?
patrickas currently it returns an infinite lis of 1 19:13
but
colomon moritz_: that's my understanding.
patrickas I could argue that 2 is on the wrong side :-)
"if a supplied limit value is on the "wrong" side of the first of those values, Nil is returned"
19:14 Ross left
patrickas since both sides .succ and .prec are the wrong sides :-) 19:14
1,1...1 would be an infinite list though 19:15
not a theoretical question .... I am asking to see if I should fix it or not :-)
colomon no, it would be just 1
19:16 araujo joined, TiMBuS left, Fozo joined 19:17 Ross joined
moritz_ does anybody want to do next week's challenge? 19:18
I'm running out of ideas
pmichaud question:
my $b = [1,2,3]; sub abc($x is copy) { $x[0] = 4; }; abc($b);
does $b change?
moritz_ I fear it does
pmichaud my $b = [1,2,3]; sub abc($x is copy) { $x[0] = 4; }; abc($b); say $b.perl;
(I'm asking what the spec says should happen, not what rakudo currently implements) 19:19
moritz_ my answer was about what the spec says
or how I think I understand the spec
19:20 meppl joined
patrickas colomon: which part of the spec is that? 1,1...1 is 1? 19:20
19:20 justatheory left
pmichaud (colomon just left) 19:22
19:22 colomon left
PerlJam patrickas: "If any value in the series is C<eqv> to the limit value, the series terminates, including that final limit value." 19:23
patrickas: S03:1812 19:24
patrickas PerlJam: Thanks
19:24 lue joined
lue Hello! 19:24
Fozo hmmmm 19:33
pugssvn r31434 | moritz++ | [feather/sccript] be more specific about where to find t/ 19:34
Fozo HeLlO
moritz_ hi
19:39 tylerni6 joined, justatheory joined 19:42 tylerni7 left
pugssvn r31435 | moritz++ | [util] remove now unneeded logic from version_h.pl; replace it by somthing rather simple 19:43
19:44 gbacon left 19:45 stepnem left 19:54 patrickas left
pugssvn r31436 | moritz++ | [util] be more consistent with argument passing 19:58
moritz_ I'm currently trying to run the generation of syn documents under a different user name 20:00
it's... not amusing
info: 990 smartlinks found and 990 broken in 820 test files (227 test files had no links).
so it seems to find not a single test file. Urks. 20:01
lue That's odd. 20:04
20:05 Psyche^ joined 20:08 whiteknight joined
pugssvn r31437 | moritz++ | [Text::SmartLinks] do not die when one file cannot be parsed 20:08
r31437 |
r31437 | Since the default is to recursively find all .pod and .pm files, it kinda
r31437 | makes sense to only warn if one of them is invalid, instead of dying horribly
r31437 | and dragging everything else down.
20:09 Su-Shee left, Patterner left, Psyche^ is now known as Patterner
pugssvn r31438 | moritz++ | [Text::SmartLinks] moritz-- cannot write Perl 5 code anymore 20:09
moritz_ of course that doesn't fix the generation at all :( 20:10
20:12 rindolf left
TimToady std: 1,2,3, ... 10 20:13
p6eval std 31436: OUTPUT«Potential difficulties:␤ Comma found before apparent series operator; please remove comma (or put parens␤ around the ... listop, or use 'fail' instead of ...) at /tmp/sWXCl1l3VJ line 1:␤------> 1,2,3,⏏ ... 10␤ok 00:01 113m␤»
PerlJam TimToady++ 20:14
20:15 nimiezko joined
moritz_ hum, the problem seems to be elsewhere 20:16
it includes the .t files inline in the specs 20:17
but still reports the smartlinks as broken
20:21 Guest73864 left, ashleydev joined 20:24 rtward left
pugssvn r31439 | moritz++ | [feather/script] synopsis htmls are now generated by the perlcabal user 20:25
20:26 Guest89026 left
moritz_ if anybody notices some rendering errors, 404s or so on perlcabal.org/syn/, please ping me 20:27
20:30 patspam left
pmichaud afk for a while # commuting 20:34
arnsholt moritz_: The most important one for me (S05) works at least, so I'm not complaining =) 20:35
20:37 patspam joined
sorear moritz_: I wonder how hard it would be to get all the perlcabal .pm6 files and fragments run through STD_hilite 20:37
I debitrotted that with that in mind...
20:38 Sanitoeter left 20:40 nimiezko left
pugssvn r31440 | lwall++ | [STD] also collect $?LONGNAME on module Foo {} form, if first declaration 20:44
20:47 gbacon joined
pugssvn r31441 | lwall++ | [t/spec] fix CR on two test files, add svn props 20:49
r31442 | lwall++ | [S03] qualify misleading assertion re 1,2,3...$n when $n < 3 20:55
20:55 cygx left 20:56 spinclad joined
sorear TimToady: What are we up to on snaptests? 20:57
20:57 patspam left
TimToady just running one right; should know in about 15 minutes, but I'm hoping for ~96% 20:58
*right now
21:01 lest_away is now known as lestrrat 21:05 Chillance left 21:07 pyrimidine left
sorear After my A $x; my B $y; $x := $y; what values can I legally assign to $x? $y? 21:09
21:11 Guest23195 left
TimToady we're at 95.76% now 21:14
21:23 gbacon left 21:25 macdaddy joined 21:26 macdaddy is now known as Guest5815 21:31 Ross left 21:36 kensanata joined 21:44 Sanitoeter joined 21:45 lichtkind joined
lue afk 21:45
lichtkind greetings 21:46
is end of june still the rakudo target? 21:53
21:55 tylerni6 is now known as tylerni7, tylerni7 left, tylerni7 joined
ash_ which raukudo target? 21:56
lichtkind target date 21:57
for star
ash_ to be released "by july 29" 21:59
rakudo.org/node/73
lichtkind thank you very much 22:00
22:05 cj left, skangas left 22:06 cj joined 22:08 d4l3k_ joined 22:09 dalek left, d4l3k_ is now known as dalek 22:24 ash_ left 22:26 estrabd is now known as estrabd_afk 22:33 Sanitoeter left 22:35 quietfanatic joined
TimToady quietfanatic: o/ 22:35
quietfanatic hey
so I was wondering
rakudo: class A {has $.x = rand}; say A.new.x
p6eval rakudo 7e3ddc: OUTPUT«Could not find sub &rand␤ in <anon> at line 11:/tmp/rlUBmjp7Ec␤ in main program body at line 11:/tmp/rlUBmjp7Ec␤»
quietfanatic Why don't we have rand during build time?
22:35 Sanitoeter joined
TimToady beats me 22:36
rakudo: say rand
p6eval rakudo 7e3ddc: OUTPUT«0.692584724255966␤»
TimToady perhaps it's the same lexical scope problem in class blocks we've always had 22:37
if rand is in CORE
rakudo: my $a = 42; class A { has $.x = $a; }; say A.new.x
p6eval rakudo 7e3ddc: OUTPUT«Lexical '$a' not found␤ in <anon> at line 11:/tmp/XIEFY1zH5p␤ in main program body at line 11:/tmp/XIEFY1zH5p␤»
TimToady that one
quietfanatic ah, I see
There doesn't seem to be any workaround save for making a custom BUILD 22:38
TimToady rakudo: class A { has $.x = CORE::rand; }; say A.new.x
p6eval rakudo 7e3ddc: OUTPUT«Can not find sub CORE::rand␤ in <anon> at line 1␤ in main program body at line 11:/tmp/ZIawOLSCLn␤»
quietfanatic Or passing a parameter to new instead 22:39
TimToady rakudo: our sub foo { rand }; class A { has $.x = foo; }; say A.new.x 22:40
p6eval rakudo 7e3ddc: OUTPUT«0.70248685786234␤»
TimToady there's a workaround
pmichaud dhmmmmm
I'm a little surprised that plain rand didn't work.
yes, it's likely the same bug. But this one is also more easily fixable.
quietfanatic I thought I tried that but I guess I didn't
22:49 Patterner left, pragma_ is now known as tits 22:50 tits is now known as pragma_ 22:53 Psyche^ joined, Psyche^ is now known as Patterner
cxreg en.wikipedia.org/wiki/Guillemets#Pr..._Languages 23:10
23:10 saaki left
cxreg that's either wrong or outdated, right? 23:10
TimToady «» is more like shell quoting, but their example isn't exactly incorrrect 23:12
23:14 kensanata left, rv2733 joined 23:15 saaki joined 23:23 felliott left 23:25 skids left 23:29 ash_ joined 23:35 justatheory left 23:39 azert0x left 23:41 Patterner left 23:42 rv2733 left 23:50 rgrau_ joined 23:51 Fozo left 23:53 justatheory joined 23:55 Psyche^ joined 23:56 Psyche^ is now known as Patterner 23:59 jrtayloriv joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!