»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:12 Psyche^ joined 00:15 kid51 left, Psyche^_ left 00:20 Rotwang1 left
TimToady .pong 00:24
00:38 jnap left 00:59 xenoterracide left, xenoterracide_ joined 01:04 xenoterracide_ left 01:05 Ben_Goldberg joined 01:07 BenGoldberg left 01:10 Ben_Goldberg left, FROGGS_ joined 01:14 FROGGS left 01:20 berekuk left, dhaivat joined 01:23 berekuk joined 01:25 berekuk left 01:27 berekuk joined
[Coke] updated gist.github.com/coke/5879701 01:27
01:27 daniel-s_ joined 01:29 berekuk left 01:36 SamuraiJack_ joined 01:44 EvanTeitelman left 01:45 EvanTeitelman joined, xinming left, EvanTeitelman left 01:47 xinming joined, abnorman left 01:52 SamuraiJack_ left 01:53 danaj left 01:56 xenoterracide_ joined 01:59 dayangkun left 02:07 xenoterracide_ left 02:09 danaj joined 02:13 BenGoldberg joined 02:15 PacoAir left 02:19 abnorman joined 02:37 dhaivat left 02:41 BenGoldberg left
[Coke] .tell lizmat S02-names/is_default.rakudo.jvm aborted 114 test(s) 02:46
yoleaux [Coke]: I'll pass your message to lizmat.
02:48 thou left 02:56 abnorman left
[Coke] here's a rakudo-jvm bug: IO::Spec::Win32.abs2rel('/a'); - throws an error about not finding 'match' 03:09
rakudo: IO::Spec::Win32.abs2rel('/a');
camelia rakudo afe36d: ( no output )
[Coke] rakudo: IO::Spec::Win32.abs2rel('/a')
camelia rakudo afe36d: ( no output )
[Coke] (whoops)
benabik r: IO::Spec::Win32.abs2rel('/a').say
camelia rakudo afe36d: OUTPUT«..\..\a␤»
[Coke] r: say "that's about {33+19} tests more..." 03:17
camelia rakudo afe36d: OUTPUT«that's about 52 tests more...␤»
[Coke] zzz 03:20
03:23 fridim_ joined 03:24 dhaivat joined 03:27 jaldhar joined 03:32 mauke left 03:33 mauke joined
lue rn: sub infix:<IFSO OTHERWISE>($a, $b, $c) { $a ?? $b !! $c }; say 1 IFSO 2 OTHERWISE 3 03:33
camelia niecza v24-90-gd827fa8: OUTPUT«Unhandled exception: No value for parameter '$c' in 'infix:<IFSO OTHERWISE>'␤ at /tmp/XXgQDasMgd line 0 (infix:<IFSO OTHERWISE> @ 1) ␤ at /tmp/XXgQDasMgd line 1 (mainline @ 4) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4579 (ANON @ 3) ␤ at /home/p6…
..rakudo afe36d: OUTPUT«===SORRY!=== Error while compiling /tmp/PJVGEK_bWq␤Two terms in a row␤at /tmp/PJVGEK_bWq:1␤------> >($a, $b, $c) { $a ?? $b !! $c }; say 1 ⏏IFSO 2 OTHERWISE 3␤ expecting any of:␤ postfix␤ infix stopper␤ i…
lue :/
03:33 preflex left
benabik Trinary isn't supported that wawy. 03:35
03:35 preflex joined, ChanServ sets mode: +v preflex
benabik ?? !! has a custom parse in the grammar 03:35
lue S02:3250 suggests differently. 03:36
synopsebot Link: perlcabal.org/syn/S02.html#line_3250
lue (although I'd instead vouch for an incircumfix category, as opposed to overloading infix with that ternary notion) 03:37
benabik Erm. That is how the _name_ is parsed. 03:38
But Perl6::Grammar doesn't translate that name to a trinary op. 03:39
lue Not that I expected it to work, but that S02 line feels incredibly misleading. 03:40
benabik Eventually you might be able to do something weird like infix:<<A B>> is parsed(/'A' <expr> 'B'/)
It might be better to use a (post)circumfix:<< >> as an example instead. 03:41
lue circumfix would be the easiest example. I think I'll change it. 03:42
03:45 abnorman joined, abnorman left
dalek ecs: ccc6176 | lue++ | S02-bits.pod:
[S02] Replace misleading example

Specifically, infix:<?? !!> wrongly implies that you can implement ternary operators in this fashion, when it reality it's much more complex a task. This example was thus replaced with a situation where multi-element operator names makes sense (circumfix:«<< >>»).
benabik Hm. I hadn't actually expected you to use circumflex:«<< >>», but it works. :-D
lue++ # finding and removing confusion 03:47
lue I was going to do circumfix:<( )> or similar but I think part of the point of that example was to demonstrate the possibility of multi-char multi-element names (thus emphasize the need for the separating space in <> form). 03:48
benabik fair
03:57 dhaivat left, dhaivat joined 04:09 jaldhar left 04:10 xenoterracide_ joined 04:11 jaldhar joined 04:15 bluescreen10 left 04:20 fridim_ left 04:25 dhaivat left 04:26 cooper joined, cooper left, cooper joined 04:32 dhaivat joined 04:41 telex left 04:42 telex joined 04:47 dhaivat left 04:52 arlinius left
TimToady umm, << >> is not a circumfix 04:57
benabik Details. 04:58
TimToady it's a quoter
it's not a circumfix unless it takes an EXPR in the middle
05:01 dhaivat joined 05:09 skids left 05:14 dhaivat left 05:15 kaare_ joined 05:21 arlinius joined 05:24 aindilis joined 06:16 berekuk joined
dalek : 61eb362 | (Tobias Leich)++ | / (2 files):
return subcaptures if applicable
06:34 berekuk left 06:40 cbk left 06:55 moritz_ joined 06:58 LordVorp joined, slavik1 joined 06:59 moritz left 07:00 zakalwe_ joined, renormalist joined 07:01 ugexe_ joined, kst` joined 07:02 gabriel joined, kshannon_ joined 07:04 logie left, segomos_ joined, TimToady_ joined, cooper_ joined 07:05 risou_awy_ joined 07:06 silug_ joined 07:11 cooper left, risou_awy left, kst left, slavik left, silug left, cooper_ is now known as cooper, cooper left, cooper joined, domidumont joined 07:14 domidumont left 07:17 Timbus joined, BinGOs joined 07:28 domidumont joined 07:31 benabik left 07:33 domidumont left, domidumont joined 07:37 domidumont left 07:40 dmol joined 07:41 FROGGS_ is now known as FROGGS
FROGGS jnthn / masak: I'd like to create a (perl5-ish) nqp::pack and nqp::unpack like we did for sprintf, would that be okay? 07:43
08:02 denis_boyun joined 08:07 denis_boyun left, dayangkun joined
moritz_ \o 08:08
08:08 moritz_ is now known as moritz
FROGGS moin moritz 08:08
08:23 denis_boyun joined 08:42 PacoAir joined
lizmat morning all 08:55
yoleaux 02:46Z <[Coke]> lizmat: S02-names/is_default.rakudo.jvm aborted 114 test(s)
lizmat .tell [Coke]: I have no idea: 114 is *all* tests. on parakudo it says: Files=1, Tests=114, 3 wallclock secs ( 0.03 usr 0.00 sys + 2.75 cusr 0.17 csys = 2.95 CPU) 08:56
yoleaux lizmat: What kind of a name is "[Coke]:"?!
lizmat .tell [Coke] I have no idea: 114 is *all* tests. on parakudo it says: Files=1, Tests=114, 3 wallclock secs ( 0.03 usr 0.00 sys + 2.75 cusr 0.17 csys = 2.95 CPU)
yoleaux lizmat: I'll pass your message to [Coke].
08:56 teslos joined 09:03 arlinius left
lizmat FROGGS: fwiw, that seems like an excellent ide 09:14
FROGGS yeah, because v5 contains an improved pack/unpack, and it should be faster when it was in nqp 09:15
lizmat would v5 be able to use that as well then? 09:16
FROGGS it would... btw, nqp::sprintf gave v5 about 1000 passing tests 09:17
masak good antenoon, #perl6 09:25
lizmat masak /o
masak /o\ 09:27
09:27 teslos left
jnthn morning, #perl6 09:27
masak .oO( at least I'm here before jnthn... ) :P 09:28
lizmat hopes jnthn got some serialize_tc inspiration...
jnthn took advantage of being able to sleep in, given he has to get up and teach all 5 days next week... 09:29
lizmat oooh wow 09:30
masak has it easy, only has to teach 4 days next week ;) 09:34
jnthn It's almost like it's autumn :P
masak .oO( and boy do we "fall" for that, ever time ) :P 09:35
jnthn *groan*
jnthn springs to the coffee machine
masak ...and we're fresh out of season-related puns. 09:36
lizmat nuts, we'll soon have nuts!
09:36 berekuk joined
masak no, that's a squirrel-related pun at best :P 09:37
arnsholt Definitely seasonal, I'd say =p 09:39
09:40 kaare_ left
masak .oO( fair enough. let's pick nuts, not pick nits ) 09:40
arnsholt We could also go nuts 09:42
But maybe not bananas
arnsholt ponders when bananas are actually in season 09:43
.oO( bananaphobe... )
arnsholt Gee, thanks. There goes that day's worth of music listening ^_^ 09:44
jnthn bwahaha
.oO( what would a movie called "Bananarama 3D" be like )
arnsholt And from that song, I remembered spotify:track:0PCyUmv2yn4h82GOZ6SKNT. Yet another hit from the early days of my internet career 09:46
masak arnsholt: I have no idea how to visit that resource specifier. 09:48
arnsholt Put it into your spotify client's search field (less the period, which is sentence final) 09:49
Although to be fair, the proper proper link is www.albinoblacksheep.com/flash/french 09:50
09:50 arlinius joined
.oO( Let it be Rick Astley, let it be Rick ... )
masak arnsholt: oh, that one :)
arnsholt Oh no, way before Rick =)
And much more hilarious (although possibly not to lizmat)
jnthn :) 09:52
Yes, seen that one some time ago :)
masak no, listening to deformed lyrics of a song in your own language is hopeless. 09:53
arnsholt Also, relistening to Bananaphone, there are some truly #perl6-worthy puns in there
masak the whole thing is ruined by your brain helpfully telling you what the song is all about.
arnsholt "It's no baloney, it ain't a phoney!"
09:54 Rotwang1 joined
masak .oO( I sausage that one coming miles away ) 09:54
lizmat :-)
masak it has a nice rhythm to it. 09:56
(Bananaphone, not French)
*lol* "it's a phone with appeal!"
that one snuck up on me. and it feels like a really "pure" pun, for some reason. 09:57
like, you'd create the entire song around it, just to get that pun in at the end.
lizmat some books start with the final line being written first :-) 09:58
arnsholt It's even in there twice. At the beginning there's "I've got a feeling, so appealing" as well =D
masak that one doesn't trigger it for me, even though I see the connection. 09:59
the nice thing about the second use is that you can hear the sentence two ways.
10:01 kaare_ joined 10:03 hummeleBop joined
lizmat wonders whether she should wait for jnthn to fix the serialize_tc issue, or start on something else 10:04
jnthn lizmat: Probably the latter, it will need quite some golfing... 10:05
lizmat okidoki 10:06
let me know if I can help you with the glof
jnthn lizmat: Well, most helpful would be to have an example of it just in terms of a role and a BEGIN-time mixin of it to some class. 10:07
(If it can be done that way.)
lizmat maybe strip the CORE_SOURCES list and get a smaller sample?
jnthn Well, no, I think it can be reproduced just in a single file 10:08
lizmat but still during building of setting, right ?
jnthn Without mentioning Hash at all, just creating something with a similar structure and using a BEGIN method that does a mixin
Because we managed to reproduce it last night separately.
lizmat did we?
jnthn Yes, you had a short file that used a typed hash that wouldn't pre-compile... 10:09
Which suggests any usage of one triggers it
lizmat I just added that class to Bag.pm (which was otherwise clean)
jnthn Oh...I thought you'd done it as a separate file? 10:10
Outside of setting...
lizmat no, sorry
jnthn Ah
May still be worth checking if that explodes though...
OK, then it's harder.
lizmat r: class A { my Int %h{Any} } 10:11
camelia rakudo afe36d: ( no output )
lizmat doesn't bomb
I'm not sure what you're getting at?
jnthn Yes, but that's not serializing.
Would need to be in a module, then pre-compile the module 10:12
lizmat perl6 --target=pir --stagestats --output=test.pir -e 'class A { my Int %h{Any} }' test # works 10:13
jnthn ah
So the usage has to be within the setting...
Grr 10:14
lizmat the generated .pir files clearly states: 10:17
push $P5004, "!UNIT_MARKER"
push $P5004, "Hash[Int,Any]"
push $P5004, "TValue"
push $P5004, "TKey"
$ perl6 -Mtest -e 1 # also works 10:18
even with -I. :-) 10:19
Teratogen pirrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr 10:21
10:21 ssutch left
masak Teratogen: :) 10:27
Teratogen masak what is up!?
masak Teratogen: clouds. feather.perl6.nl. the positive y axis. 10:29
Teratogen actually Buckminster Fuller made the observation that really there is no up or down, there is just in and out
masak the study of the Zermelo-Fraenkel axioms has made me draw the same conclusion. 10:30
jnthn But...but...I go in the elevator which goes up or down, then I got out...
lizmat haw can you be sure the building didn't move down and up ? 10:31
jnthn Well, some of the tallest buildings I've been in were in earthquake zones... 10:32
.oO( or is that "been up" :P )
masak r: say sgn(4) 10:36
camelia rakudo afe36d: OUTPUT«===SORRY!=== Error while compiling /tmp/2oZ5kXrBH4␤Undeclared routine:␤ sgn used at line 1. Did you mean '&sign', '&sin'?␤␤»
masak r: say sign(4)
camelia rakudo afe36d: OUTPUT«1␤»
Teratogen jnthn, when you get into an elevator and go "up" you are actually going "out" from the earth. 10:40
masak also, if you look at "up" upside-down, it says "dn"
jnthn :P 10:41
Teratogen well in upside down is ni, like the Knights who say Ni 10:42
masak no "in" upside-down is "uı" 10:44
10:44 domidumont joined
jnthn masak: fail, you want a COMBINING DOT BELOW on that ı :P 10:44
Teratogen we want... a SHRUBBERY! 10:45
10:45 denysoniqeu is now known as denysonique
masak jnthn: aye :) 10:48
lizmat r: say ~Mu 10:58
camelia rakudo afe36d: OUTPUT«Nominal type check failed for parameter ''; expected Any but got Mu instead␤ in sub prefix:<~> at src/gen/CORE.setting:1368␤ in block at /tmp/9zUTGWsOVJ:1␤␤»
lizmat since Mu has .Stringy, and prefix ~ calls .Stringy, any objections against making prefix ~ take Mu ? 10:59
masak no objections.
lizmat spectesting 11:00
masak though the obvious question is how it interacts with Junction
lizmat maybe an extra candidate then ?
jnthn -1. This feels like a slipperly slope. If we do this to ~, how many more operators will we do it to? 11:01
lizmat it's just that there is a failing spectest for ~Mu
so maybe the spectest should go ?
masak maybe. 11:02
why is it we have a Mu type again?
lizmat to allow for more cow puns
but only if you change them :-)
masak .oO( bullshit puns ) 11:03
lizmat hmmm… actually, slightly above in the test file it says: 11:04
# jnthn and pmichaud believe the next test is incorrect
# is +Mu, 0, "+ context forcer works (8)";
masak heh.
jnthn Mu is outside of Any to allow dispatch fail-over to trigger stuff like auto-threading.
masak right, also known as "we put junction semantics in the type system".
jnthn Right.
Or rather, use the type system to flag up the case.
That's a general Perl 6 thing, though. 11:05
masak troo
jnthn (Pushing decision making into the type system)
lizmat so: ~Mu should never work, ergo we need a spectest that it fails ?
otoh, what are .Str and .Stringy doing in src/core/Mu then? 11:06
shouldn;t they be in Any then ?
jnthn Yes, should have a test that it fails...
masak there's been a bit of back and forth about this in the backlog over the years, I recall... 11:07
lizmat and what to think about methods in Mu.pm that don't have a Mu signature ?
masak I think it was differences like this that led to something like .gist 11:08
lizmat r: say Mu.Str
camelia rakudo afe36d: OUTPUT«use of uninitialized value of type Mu in string context in block at /tmp/a6ez5MmVGq:1␤␤␤»
lizmat r: say Mu.Stringy
camelia rakudo afe36d: OUTPUT«use of uninitialized value of type Mu in string context in block at /tmp/f6CZNG45VI:1␤␤␤»
jnthn I don't think "~ should auto-thread" automatically means "Mu shouldn't know how to .Str". 11:09
lizmat ok
jnthn They're at lesat somewhat orthogonal.
lizmat I'll just add tests for +Mu and ~Mu failing 11:10
r: say +Mu
camelia rakudo afe36d: OUTPUT«Nominal type check failed for parameter ''; expected Any but got Mu instead␤ in sub prefix:<+> at src/gen/CORE.setting:3500␤ in block at /tmp/AgVUdmylxX:1␤␤»
lizmat r: say +Any
camelia rakudo afe36d: OUTPUT«use of uninitialized value of type Any in numeric context in block at /tmp/uBKOARA6P4:1␤␤0␤»
11:11 Faxmachinen joined
Faxmachinen Hi folks 11:11
Is first always eager?
jnthn I guess it has to reify up until it finds the first thing that matches. 11:12
lizmat first is lazy
there are spectests for that
jnthn r: (gather { for 1..20 { .say; take $_ } }).first(3) 11:13
camelia rakudo afe36d: OUTPUT«1␤2␤3␤»
jnthn Looks like it's behaving as I described. :)
Faxmachinen r: (Date.new(1900, 1, 1) ... Date.new(10000, 1, 1)).first({ .day-of-week == 7 }) 11:16
camelia rakudo afe36d: ( no output )
Faxmachinen Is this a recent change maybe? On my local Rakudo Whatever that line takes like ten thousand years to evaluate 11:18
This is perl6 version 2013.05 built on parrot 5.3.0 revision 0 11:19
jnthn Faxmachinen: I seem to recall fixes to first in the not too distant past
FROGGS maybe it is doing in in real-time?
jnthn ah 11:20
New in 2013.07
+ List.first is now lazy
Faxmachinen Thanks :)
jnthn (from docs/ChangeLog)
11:24 xenoterracide_ left 11:27 denis_boyun left 11:29 rindolf joined 11:32 jaldhar left 11:33 jaldhar joined, denis_boyun joined
dalek ast: 4ea3b29 | (Elizabeth Mattijsen)++ | S03- (8 files):
Some skip/todo fudges and some new tests
lizmat running some errands&
11:48 Rotwang1 left
arnsholt Weeeeirdness 11:48
To the gdb-mobile! 11:49
Parrot Y U return 0x1???? 11:52
11:54 jaldhar left 11:55 Rotwang1 joined 12:00 Rotwang1 left 12:01 jaldhar joined 12:12 cognominal joined 12:16 PacoAir left, Rotwang1 joined 12:34 xenoterracide_ joined 12:38 mtk left 12:39 Rotwang1 left 12:42 mtk joined
flussence r: say (Date.new(1900, 1, 1) ... Date.new(10000, 1, 1)).first({ .day-of-week == 7 }) 12:52
camelia rakudo afe36d: OUTPUT«1900-01-07␤»
12:54 xenoterracide_ left, Rotwang1 joined
lizmat just realizes today is the 5th palindrome date of the year: 31-8-13 12:55
masak r: say (Date.new(1900, 1, 1) ... *).first({ .day-of-week == 7 })
camelia rakudo afe36d: OUTPUT«1900-01-07␤»
masak r: sub palindrome { $^s eq $s.flip }; .say when &palindrome for (Date.new(2013, 1, 1) ... Date.new(2013, 12, 31)).map({ "{.day}-{.month}-{.year.substr(2)}" }) 12:58
camelia rakudo afe36d: OUTPUT«31-1-13␤31-3-13␤31-5-13␤31-7-13␤31-8-13␤»
masak \o/
masak ♥ Perl 6 12:59
12:59 benabik joined
masak of course, that .substr(2) isn't Y10K-compatible. my bad. 13:00
r: sub palindrome { $^s eq $s.flip }; .say when &palindrome for (Date.new(2013, 1, 1) ... Date.new(2013, 12, 31)).map({ "{.day}-{.month}-{.year % 100}" })
camelia rakudo afe36d: OUTPUT«31-1-13␤31-3-13␤31-5-13␤31-7-13␤31-8-13␤»
masak r: say "{.day}-{.month}-{.year % 100}" if .month.chars == 1 && .day == 31 for Date.new(2013, 1, 1) ... Date.new(2013, 12, 31) 13:02
camelia rakudo afe36d: OUTPUT«31-1-13␤31-3-13␤31-5-13␤31-7-13␤31-8-13␤»
masak also, we won't have a palindrome date of that kind until 2021. 13:03
so today is indeed very special.
13:09 lichtkind joined 13:12 berekuk left 13:14 berekuk joined 13:17 teslos joined
lizmat woolfy came up with 4-11-14 13:32
colomon woolfy++
just have to ignore the dashes
woolfy I can come up with more: 5-11-15, 6-11-16, 7-11-17, 8-11-18, 9-11-19 13:33
masak yes, ignoring the dashes produces more solutions, indeed.
woolfy :-)
I am happy to ignore dashes.
timotimo masak++ #t3 reviews and t4 reviews in a little bit :D
woolfy 02-11-20 13:34
masak I've been spending parts of the day implementing t3 myself.
to find out things about optimal solutions.
woolfy: now you're adding zeros :)
lizmat there were rulez?
.oO( Miep Cijfer )
woolfy To make palindromes look pretty, dashes are not necessary and zeroes can be.
13:39 domidumont left 13:42 zakharyas joined 13:44 Rotwang1 is now known as Rotwang, Rotwang left, Rotwang joined
lizmat r: my @one=1,2; say @one .. 3 13:44
camelia rakudo afe36d: OUTPUT«[1, 2]..3␤»
lizmat r: my @one=1,2; say ~(@one .. 3)
camelia rakudo afe36d: OUTPUT«No such method 'succ' for invocant of type 'Array'␤ in sub postfix:<++> at src/gen/CORE.setting:2217␤ in sub postfix:<++> at src/gen/CORE.setting:2216␤ in method reify at src/gen/CORE.setting:6649␤ in method reify at src/gen/CORE.setting:6779␤ in method reify …
lizmat r: my @one=1,2; .say for @one .. 3 13:45
camelia rakudo afe36d: OUTPUT«No such method 'succ' for invocant of type 'Array'␤ in sub postfix:<++> at src/gen/CORE.setting:2217␤ in sub postfix:<++> at src/gen/CORE.setting:2216␤ in method reify at src/gen/CORE.setting:6649␤ in method reify at src/gen/CORE.setting:6779␤ in method reify …
lizmat hmmmm...
# Test that the operands are forced to scalar context 13:46
## From pmichaud 2006-06-30: These tests may be incorrect.
## C<@one> in ## item context returns an Array, not a number
## -- use C< +@one > to get the number of elements. So, we
## need to either declare that there's a version of infix:<..>
## that coerces its arguments to numeric context, or we can
## remove these tests from the suite.
# RT #58018
# RT #76950
Time to remove these tests ?
r: my @a=1,2; say for 0 .. @a 13:47
camelia rakudo afe36d: OUTPUT«␤␤␤»
lizmat r: my @a=1,2; .say for 0 .. @a 13:48
camelia rakudo afe36d: OUTPUT«0␤1␤2␤»
13:49 fridim_ joined
masak lizmat: +1 on removing. 13:51
lizmat finding out that 0 .. @a works
made me change perspective: if that works, @a .. 100 should also work 13:52
masak I think infix:<..> favors the rhs in determining the type. 13:54
I may misremember on that one, though.
13:54 dhaivat joined, PacoAir joined
Faxmachinen r: sub foo(Int @a) { return [+] @a }; say foo([1,2,3]); 13:58
camelia rakudo afe36d: OUTPUT«Nominal type check failed for parameter '@a'; expected Positional but got Array instead␤ in sub foo at /tmp/7qXCh4ORUS:1␤ in block at /tmp/7qXCh4ORUS:1␤␤»
lizmat RT #58018 13:59
jnthn Crappy error reporting, but it's correct to reject it..
lizmat was hoping for an RT bot to provide a link :-)
jnthn r: sub foo(Int @a) { return [+] @a }; say foo(Array[Int].new(1,2,3));
camelia rakudo afe36d: OUTPUT«6␤»
Faxmachinen Huh, forgot about the Class[Whatever] syntax 14:00
The error message is perhaps not ideal though
jnthn Yes, agree 14:01
14:05 dhaivat left 14:08 dhaivat joined
dalek kudo/nom: 2d5abca | moritz++ | tools/build/NQP_REVISION:
bump NQP revision
lizmat running spectest with this while cycling& 14:17
14:22 dhaivat left 14:31 xinming left
FROGGS perl6 -e 'use v5; $x = "" =~ /./; { use 6.0.0; say $x.WHAT }' 14:45
I love that way to debug things :o)
JimmyZ :P
14:48 cognominal__ joined 14:49 dhaivat joined 14:50 teslos left 14:51 xinming joined 14:52 cognominal left, berekuk left 14:57 berekuk joined
arnsholt O(n!) > O(2^n), isn't it? (If you'll forgive the flagrant abuse of notation) 14:58
14:59 raiph joined
masak arnsholt: yeah. 15:00
arnsholt: because the left side keeps multiplying bigger factors, as the right side keeps adding the same factor. 15:01
arnsholt Right, right
15:13 denis_boyun left 15:15 ugexe_ is now known as ugexe
moritz n! is more like N^N 15:25
erm, n^n
timotimo nân? 15:26
15:26 zakharyas left, TimToady_ is now known as TimToady
masak cf en.wikipedia.org/wiki/Stirling%27s...roximation 15:27
moritz tadzik: fwiw the '==> Fetching panda' step in bootstrapping panda seems awefully slow 15:28
timotimo on jvm already? ;) 15:30
15:33 colomon left
moritz no 15:33
masak: I've used that a lot in statistical physics 15:34
15:35 denisboyun joined
dalek Iish: 99608bb | moritz++ | t/10-mysql.t:
[mysql] use the documented test crendetials
timotimo did arnsholt fix all the stuff needed for you to do your dbi stuff? and stuff? 15:37
moritz timotimo: dbiish mostly seemed to work even with zavolaj broken-ish 15:38
timotimo: I'm now testing it more thoroughly though
timotimo cool :)
moritz postgres also works \o/ 15:39
15:39 ggoebel joined, ggoebel2 left
dalek Iish: 51cfd5a | moritz++ | lib/DBDish/ (2 files):
avoid deprecation warnings in bless
15:42 logie joined
timotimo i wish i knew enough about plack and websockets and such so that i can build a tiny web framework on rakudo-jvm 15:43
wait, not plack, PSGI 15:44
moritz read up on it :-) 15:45
15:49 jaldhar left, EvanTeitelman joined 15:52 hummeleBop left 15:55 colomon joined
timotimo hm, http::easy already has something for psgi, maybe it can be adpted to handle async, too? 15:56
lizmat is back and saw that the spectest is green after moritz++ upgrade of Parrot version 16:01
colomon as of last night, we're back to NativeCall failing and all dependent modules failing with it. 16:03
timotimo so it seems like there's a listen socket that speaks the psgi protocol and it accepts one connection per incoming request to the "outer" web server and speaks a specific protocol over that connection 16:05
lizmat colomon: could this be Bag/Set related ???
timotimo that means it should be rather simple to plop an async in there and have it work
colomon lizmat: doubt that very greatly
timotimo i hope :)
sadly supernovus isn't idling around here 16:06
colomon I'd be dumbfounded if NativeCall is using Bag or Set.
timotimo intermittent failure brought to light by some environment change?
colomon " 16:09
Cannot unbox argument to '' as a native str
t/02-simple-args.t 16:10
timotimo i thought that was aded? 16:11
colomon wrong computer, bother 16:14
16:14 ajr joined 16:15 ajr is now known as Guest49411
colomon no, that's still the error I get with an *almost* up-to-date Rakudo. (Doesn't have moritz++'s NQP bump.) 16:16
next step is to try with that, I guess 16:17
16:18 Faxmachinen left 16:20 berekuk left 16:23 berekuk joined 16:25 berekuk left 16:31 Rotwang left 16:32 fridim_ left 16:36 ssutch joined, Guest49411 is now known as ajr_
timotimo i read a commit today or yesterday about something unboxing to native str 16:37
FROGGS I think there was a patch about Mu and Stringy 16:39
nwc10 colomon: are you able to use git bisect to work out which commit? 16:40
colomon don't have time right now 16:41
16:42 daniel-s_ left
colomon is cooking lunch to take to the library, then making a bassoon check stop, then heading out to pick up two months' worth of meat from a farm, and needs to get back before the football game starts. :) 16:42
nwc10 mmm, meat.
16:42 daniel-s_ joined 16:47 pecastro_ is now known as pecastro, daniel-s_ left 16:48 daniel-s_ joined
masak another round of horse jokes? :/ 16:49
nwc10 I think that we've managed to reign them in
TimToady it's the long tail 16:50
nwc10 we're no longer saddled with bad puns
masak I see we're back on track
TimToady still jockeying for top punster?
jnthn I thought my position at the top was stable... :P 16:51
TimToady you're just stalling
16:51 logie left 16:52 berekuk joined 16:53 btyler joined 17:09 denisboyun left 17:11 daniel-s_ left
masak whaddya take me for, a foal? 17:12
17:12 daniel-s_ joined
geekosaur calf-baked ideas? 17:13
masak the cow puns were earlier today. 17:15