»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
[Coke] pushed another set of updates for QAST on partcl-nqp/nqp2, and is now getting this error: 02:14
Error while compiling block : Error while compiling op : hash: can't use null as key 02:15
MikeFair_ [Coke]: how's that error coming? 03:34
sorear random, certainly unoriginal, thought on number theory: 05:42
TimToady ... 05:45
sorear given a putative number p of the form 4k+1, for all x, let y = x^2, z = y^2, now z^((p+1)/4) = x^(p+1) = y mod p, so square roots are easy iff p is prime (Blum-Goldwasser) 05:47
now imagine turning this into a primality test: take a random number, try to take a square root, fail if the putative root is wrong 05:48
if this test gives a false positive, it has found a square root of a random number modulo a composite
but we know (Rabin) that that is exactly as hard as factoring.
so a single iteration of this test will only give the wrong answer, if the test could have factored the number 05:49
if factoring is hard on average for classical algorithms, this algorithm fails with neglible probability on random inputs
I've seen a lot of primality testers using the generalized Riemann hypothesis 05:55
I can't be the first one to think of primality testing connected to standard cryptographic assumptions
MikeFair_ you're not, but those who are interested in the field tend to flock together or not talk about their work. :) In fact I think several movies even make references to such things (sneakers comes to mind, not that many people actually understood what the match the professor was explaining that chip actually did) 06:29
Anyone here aware of Java's JSR223 Scripting Engine Framework API? 06:30
I just got made aware of it today, and think it could be relevant to the Perl6 on Java / Parrot on Java discussion 06:32
sorear MikeFair_: nah, what's really interesting here is the faint glimmer of a connection between the GRH and P vs NP. Pity I'm nowhere near good enough to follow up on it. 06:34
MikeFair_ sorear: I hear that pain. :) 06:35
Man, some people have done a lot of very cool stuff out there 06:38
sorear o/ 06:47
FROGGS good morning 06:48
jnthn o/ 06:50
wow, short backlog
diakopter :) 06:51
sorear o/ 06:52
jnthn commuts 07:01
FROGGS jnthn: latest rakudo complains about the \ in \$arr here: multi method at_pos(::?CLASS:D \$arr: $pos) is rw { # NativeCall.pm#L149 07:03
should the \ be removed?
"Obsolete use of | or \ with sigil on param $arr"
just a warning though
tadzik good morning #perl6 07:14
FROGGS morning tadzik 07:15
GlitchMr FROGGS: it is \arr if I remember correctly 07:31
FROGGS: perlcabal.org/syn/S06.html#Parcel_binding 07:32
GlitchMr perl6: class Mumbler { method mumble { say "Mumble!" } };(-> \_ = OUTER::<_> { _.mumble })(Mumbler.new()) 07:35
p6eval rakudo ef228d: OUTPUT«Mumble!␤»
..niecza v21-15-gf226bf6: OUTPUT«===SORRY!===␤␤Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?) at /tmp/cTelw01ND9 line 1:␤------> { say "Mumble!" } };(-> \_ = OUTER::<_> ⏏{ _.mumble })(Mumbler.new())␤␤P…
GlitchMr std: class Mumbler { method mumble { say "Mumble!" } };(-> \_ = OUTER::<_> { _.mumble })(Mumbler.new())
p6eval std 77327a4: OUTPUT«Use of uninitialized value $first in string eq at STD.pm line 65741.␤Use of uninitialized value $first in string lt at STD.pm line 65746.␤===SORRY!===␤Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?) … 07:36
GlitchMr std: class Mumbler { method mumble { say "Mumble!" } }; (-> \$_ = OUTER::<$_> { $_.mumble })(Mumbler.new())
p6eval std 77327a4: OUTPUT«===SORRY!===␤Unexpected block in infix position (two terms in a row, or previous statement missing semicolon?) at /tmp/mEPr__Wf87 line 1:␤------> ay "Mumble!" } }; (-> \$_ = OUTER::<$_> ⏏{ $_.mumble })(Mumbler.new())␤ expecting…
kresike hello all you happy perl6 people 08:30
sorear FROGGS: what GlitchMr says, you're actually supposed to remove the sigil and leave the \ 08:36
sleep&
GlitchMr Interesting how sleep program on my PC requires time argument 08:43
masak morning, #perl6 08:47
masak MissingSemicolon: welcome :) 11:45
JavaScript programmer? :P
MissingSemicolon lol, no, just tried to think of the cleverest screen name I could. Didn't get very far with that, obviously. :-) 11:46
masak it's a fine name.
MissingSemicolon Thanks. :-) So is this the place to bring my Perl 6 programming questions, or is it more geared for p6 developers? 11:47
masak rn: say "Hey $_, Rakudo and Niecza greet you!" given <Missing Semicolon>.join
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«Hey MissingSemicolon, Rakudo and Niecza greet you!␤» 11:48
masak MissingSemicolon: right now it's a place for both.
MissingSemicolon: Perl 6 is still small-and-cozy enough to have all conversations in the same cahnnel ;)
MissingSemicolon yes, I'm hoping it'll become big-and-cozy sometime soon. 11:49
masak working on that.
MissingSemicolon I've been a Perl programmer since the stone age, and I like a lot of things about Perl 6. I'm now very embarrassingly stumped on an extremely simple problem, though. 11:50
masak bring it.
MissingSemicolon To get the index of a substring, is it supposed to be $haystack.index(' needle ')? 11:51
I'm trying to skip over lines that contain a substring, and I'm doing that with: next if $line.index('thing_i_do_not_want') == -1; 11:52
MissingSemicolon But with that I get: Parameter '' requires an instance, but a type object was passed 11:53
in method Bridge at src/gen/CORE.setting:2786
in sub infix:<==> at src/gen/CORE.setting:2709
in block at ./exercise1a.p6:10
Su-Shee MissingSemicolon: why not use grep?
masak MissingSemicolon: probably $line was never assigned to.
MissingSemicolon: seems it is undefined, whatever the reason.
r: my $line; say $line.index('thing_i_do_not_want') == -1
p6eval rakudo ef228d: OUTPUT«No such method 'index' for invocant of type 'Any'␤ in block at /tmp/szPuwtVKy5:1␤␤»
masak r: my $line = Str; say $line.index('thing_i_do_not_want') == -1
p6eval rakudo ef228d: OUTPUT«use of uninitialized value of type Str in string context in block at /tmp/jLzGprxlTz:1␤␤Parameter '' requires an instance, but a type object was passed␤ in method Bridge at src/gen/CORE.setting:2814␤ in sub infix:<==> at src/gen/CORE.setting:2737␤ in block at…
masak that's the error, I think. 11:54
MissingSemicolon OK, but if I use 11:55
MissingSemicolon … hang on a sec. Something strange :-) 11:56
MissingSemicolon rakudo: my $s = 'foobar'; say "OK" if $s.index('baz') == -1; 12:19
p6eval rakudo ef228d: OUTPUT«Parameter '' requires an instance, but a type object was passed␤ in method Bridge at src/gen/CORE.setting:2814␤ in sub infix:<==> at src/gen/CORE.setting:2737␤ in block at /tmp/Rysur7bGGQ:1␤␤»
MissingSemicolon There's a minimal example of what I'm seeing.
Am I doing something fundamentally stupid? (quite possible!) 12:20
jnthn r: my $s = 'foobar'; say $s.index('baz).perl 12:29
p6eval rakudo ef228d: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 2, near "'baz).perl"␤»
jnthn grr
r: my $s = 'foobar'; say $s.index('baz').perl
p6eval rakudo ef228d: OUTPUT«Int␤»
jnthn Yeah, so I think index in Perl 6 doesn't return -1
Instead it returns an undefiend value
So you probably want 12:30
rakudo: my $s = 'foobar'; say "OK" unless defined $s.index('baz')
p6eval rakudo ef228d: OUTPUT«OK␤»
jnthn Like that
MissingSemicolon Ok, that makes sense. 12:33
Thanks! 12:34
[Coke] MikeFair_: I went to bed right after posting, so it's still there. ;) 12:46
[Coke] finds a bunch of :pasttypes that should be :ops 12:56
it would be helpful tracking this down if it gave me a line number on the Actions.pm when it died rather than saying only: "Error while compiling block" 12:58
jnthn if it says that, you're way past the actions
PerlJam "past the actions" ... that's almost punny :) 13:02
masak you're way qast the actions. 13:04
[Coke] jnthn: ah, excellent. where am I? ;)
jnthn You're compiling the QAST down to PIR
What *is* the error you get?
[Coke] Error while compiling block : Error while compiling op : hash: can't use null as key 13:05
(that after running "./partcl" and entering any command.
jnthn ah
you maybe didn't set the op
nqp: use QAST; QAST::Op.new( :pasttype('list') ) 13:06
p6eval nqp: OUTPUT«Method 'pasttype' not found for invocant of class 'QAST::Op'␤current instr.: '' pc 1053 (src/stage2/QASTNode.pir:464) (src/stage2/QASTNode.nqp:54)␤»
jnthn It should do that though
Like, as soon as you instantiate it
So I'm surprised you're getting as far as you are
[Coke] found a few more :pirops... 13:08
[Coke] changed, same error. 13:08
[Coke] gist.github.com/3761375 - jnthn: various types of ops I'm using. 13:12
I didn't see docs on the actual ops available, these are leftover or updated (incorrectly?) from the past version 13:13
masak oh! closures are moving code at runtime. macros are moving code at compile time. 13:20
PerlJam "moving code"? 13:21
[Coke] jnthn: so, there are some Ops that don't have an :op() - is that legal? 13:22
(some have only a :name() - should those have :op<call> also?
masak PerlJam: closures are first-class function values. so you can put it in your bag and steal away with it in the night. 13:24
sirrobert Can I export an enum? 13:25
how do I reference it in the outer namespace?
outer=remote 13:26
sirrobert figured it out... enum Foo is export <X Y Z>; elsewhere: Foo::X 13:28
colomon phenny: tell TimToady I'm having a hard time getting comfortable with the name is-prime, because it feels like it promises something it cannot deliver. Should it be is-likely-prime? is-known-composite? 13:29
phenny colomon: I'll pass that on when TimToady is around.
flussence reading the backscroll gave me an idea: a lot of cryptic errors/warnings come from unexpected undefinedness - maybe it'd be useful to show datatypes (Mu, Any) in the backtrace 13:30
arnsholt Sounds like a good idea to me 13:31
masak sirrobert: um. I think X should also be visible as-is. 13:32
PerlJam colomon: ha! irclog.perlgeek.de/perl6/2012-09-20#i_6004806
masak sirrobert: dunno if implemented, though.
colomon PerlJam: glad to see I'm not alone. :)
PerlJam colomon: however, later ... irclog.perlgeek.de/perl6/2012-09-20#i_6004983 13:33
sirrobert masak: X wasn't visible, but it was visible scoped
err namespaced
masak sirrobert: yeah. I'm thinking of Bool::True and True as an example. 13:34
sirrobert I wonder why that works and this doesn't?
masak because Bool isn't a real enum? 13:37
as implemented, I mean.
colomon PerlJam: I guess I'm guilty as charged as being wary of probabilistic algorithms. but citing 4 ** -100 is cheating -- default error rate is 4 ** -10, no?
sirrobert fair enough =)
masak I dunno how close Niecza got with this, but I know Rakudo didn't get all the way there, and neither of the main implementors were happy about Bool being an enum. 13:38
r: say (class A {}).HOW.WHAT
p6eval rakudo ef228d: OUTPUT«Perl6::Metamodel::ClassHOW()␤»
masak r: say (enum E <a b>).HOW.WHAT
p6eval rakudo ef228d: OUTPUT«Perl6::Metamodel::EnumHOW()␤»
masak r: say Bool.HOW.WHAT 13:39
p6eval rakudo ef228d: OUTPUT«Perl6::Metamodel::ClassHOW()␤»
masak yeah.
sirrobert masak: I still haven't gotten a feel for how I should think about "HOW" 13:43
masak hehe.
I don't blame you. :) 13:44
I remember staring dumbfounded at au++'s MOP diagram back in the day, thinking "this is all very nice, but do I really need a metaphysics for my objects?"
sirrobert heh
arnsholt That'd be a great title for a Perl 6 parallel to The art of meta-object protocol. "The metaphysics of objects" =D 13:45
masak well, think of it this way: somewhere in Rakudo there is code to tell Perl 6 how to lay out attributes, how to do method lookup, how inheritance works.
having a MOP in the first place means you expose those details to a greater or smaller extent. 13:46
jnthn [Coke]: er, that sounds bad
[Coke]: yes, they totaly should have :op<call>
sirrobert masak: makes sense
masak the .HOW (and .^) is how you start talking about that API.
sirrobert: turns out OO is a great fit for implementing a MOP :) 13:47
which is when you get into the whole metacircularity whirlpool. 13:48
r: say (class A {}).WHAT
p6eval rakudo ef228d: OUTPUT«A()␤»
masak r: say (class A {}).HOW.WHAT
p6eval rakudo ef228d: OUTPUT«Perl6::Metamodel::ClassHOW()␤»
masak r: say (class A {}).HOW.HOW.WHAT
p6eval rakudo ef228d: OUTPUT«No such method 'gist' for invocant of type 'NQPClassHOW'␤ in sub say at src/gen/CORE.setting:7173␤ in block at /tmp/flhmjZl_D7:1␤␤»
masak r: say (class A {}).HOW.HOW.HOW.WHAT
p6eval rakudo ef228d: OUTPUT«No such method 'gist' for invocant of type 'KnowHOW'␤ in sub say at src/gen/CORE.setting:7173␤ in block at /tmp/k5Nhxk5MbJ:1␤␤»
masak r: say (class A {}).HOW.HOW.HOW.HOW.WHAT
p6eval rakudo ef228d: OUTPUT«No such method 'gist' for invocant of type 'KnowHOW'␤ in sub say at src/gen/CORE.setting:7173␤ in block at /tmp/RFQnPolDpx:1␤␤»
masak it's KnowHOW's all the way down from there.
arnsholt These bits are relevant to what makes Zavolaj tick, incidentally
masak arnsholt: because of reprs, I presume. 13:49
arnsholt Because repr, yeah
[Coke] nqp: use QAST; new QAST::Op( :name<eek> ); 13:50
p6eval nqp: OUTPUT«Confused at line 2, near "new QAST::"␤current instr.: 'panic' pc 19999 (src/stage2/gen/NQPHLL.pir:7315) (src/stage2/gen/NQPHLL.pm:324)␤»
[Coke] nqp: use QAST; QAST::Op.new( :name<eek> );
p6eval nqp: ( no output )
jnthn Maybe we should make that complain somehow. 13:51
[Coke] jnthn: so, should that barf, then?
barfing sooner than later is boss.
jnthn [Coke]: It'd certainly be more helpful if it did
[Coke] jnthn: now I'm getting Error while compiling block : Error while compiling op call: Can only use get_how on a SixModelObject 13:55
so, a naive translation is not working. I'll have to do more than a global search and replace, so that means: the weekend.
(unless someone beats me to it. ;)
jnthn [Coke]: That may mean there's a PAST node in the tree 13:56
Or an undef or a literal
Note that literals must be an IVal, NVal or SVal now
Not put directly as a child
sirrobert is there an operator (or method?) to compare if two hashes have the same keys and values? 14:47
I only need shallow comparison, but deep is cool too
tadzik Test.pm does that in is_deeply 14:48
sirrobert great, thanks (using it in a test) 14:49
masak eqv should do that.
rn: say { foo => 1, bar => 2 } eqv { bar => 2, foo => 1 }
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«True␤»
sirrobert rn: say { foo => 1, bar => 2 } eqv { bar => 2, foo => 5 }
p6eval rakudo ef228d: OUTPUT«False␤»
..niecza v21-15-gf226bf6: OUTPUT«True␤»
masak submits nieczaissue 14:50
sirrobert sad path =)
masak sirrobert++
colomon rn: say { foo => 1, bar => 2 } cmp { bar => 2, foo => 5 } 14:55
p6eval niecza v21-15-gf226bf6: OUTPUT«Same␤»
..rakudo ef228d: OUTPUT«Decrease␤»
colomon actual bug is in cmp, I think
masak adds that to the issue 14:56
colomon wishes he had TextMate on Linux.
colomon .... errr, how does one determine Hash cmp Hash? does the question even make sense? 15:01
jnthn
.oO( maybe if you've had enough hash... )
 15:02
I'm not sure increase vs decrease does though
sirrobert if: class A { method foo {"Hallo!"} }; my $mtd = 'foo'; 15:06
arnsholt I'd argue that Hash cmp Hash doesn't really make sense
sirrobert how do I do: say A.new.$mtd;
masak I don't really see the problem.
at the first differing key (sorted lexographically), there's bound to be a well-defined ordering.
masak sirrobert: A.new."$mtd"() 15:06
sirrobert ah, thanks (was missing the () in my tries) 15:07
arnsholt Hmm. masak's idea might work
masak++ # More cleverer than me =)
I'm not sure if it'd be a useful ordering over hashes, but it's workable 15:08
It depends on there being an ordering over the keys though, which might not hold for non-Str keys
sisar masak++ #release 15:09
masak arnsholt: troo. it feels a little weird to base any feature of a hash on ordering of its keys. 15:12
arnsholt I guess. I'm not sure we'd really want to make two hashes comparable just for the sake of making them comparable 15:19
kresike bye folks 15:20
flussence I can't say I've ever wanted a built in feature for that in any language 15:22
GlitchMr rn: say { bar => 2, foo => 5} cmp { foo => 1, bar => 2 }
p6eval rakudo ef228d: OUTPUT«Increase␤»
..niecza v21-15-gf226bf6: OUTPUT«Same␤»
GlitchMr Wait... why those hashes can be compared? Let me guess, string conversion? 15:23
Hmmm...
rn: say { bar => 2, foo => 5} cmp { foo => '5', bar => 5.0 }
p6eval rakudo ef228d: OUTPUT«Increase␤»
..niecza v21-15-gf226bf6: OUTPUT«Same␤»
GlitchMr rn: say { bar => 2, foo => 5} cmp { foo => '5', bar => 2.0 }
p6eval rakudo ef228d: OUTPUT«Increase␤» 15:24
..niecza v21-15-gf226bf6: OUTPUT«Same␤»
GlitchMr rn: say { bar => 2, foo => 5} cmp { bar => 2.0, foo => '5'}
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«Same␤»
GlitchMr rn: say { bar => 2, foo => 5} eqv { bar => 2.0, foo => '5'}
p6eval rakudo ef228d: OUTPUT«False␤»
..niecza v21-15-gf226bf6: OUTPUT«True␤»
TimToady colomon: wikipedia claims: If n is composite then the Miller–Rabin primality test declares n probably prime with a probability at most 4 ** -k. On the other hand, the Solovay–Strassen primality test declares n probably prime with a probability at most 2 ** -k. 15:25
phenny TimToady: 13:29Z <colomon> tell TimToady I'm having a hard time getting comfortable with the name is-prime, because it feels like it promises something it cannot deliver. Should it be is-likely-prime? is-known-composite?
TimToady and we're defaulting to 100 tests
so maybe you're thinking of SS instead of MR?
colomon no, I'm just not comfortable with naming the "is very likely prime" function is-prime. :) 15:26
GlitchMr When it makes invalid results anyways?
TimToady anyway, as for the name, if the probability is much larger that a cosmic ray is going to flip a bit in your memory, then all our functions should have "likely" in the name 15:27
GlitchMr Also, low primes could be hardcoded? 15:28
TimToady and are, in libtommath
so I'd feel uncomfortable saying "is likely prime" when we know for sure :) 15:29
modulo cosmic rays 15:30
TimToady has been running his longest left truncating prime program in niecza again for more than 24 clock hours (4500 cpu minutes) just to see if it blows up again in the same place, of if the first blowup was a cosmic ray :) 15:33
GlitchMr That's waste of CPU, but whatever 15:33
TimToady hopes that when is-prime gets into rakudo it can run this faster, and maybe more correctly, since my last rakudo run with a Perl 6 MR test was producing some incorrect results 15:34
GlitchMr Using IRC is waste of CPU too
TimToady GlitchMr: not a waste at all, it heats up my house quite efficiently
GlitchMr oh, ok
huf small house or huuuuge irc?
TimToady with 100% efficiency, in fact
not the irc, that wastes some heat outside the house
the ltp algo 15:35
colomon back when I was single (and in Michigan, mind you), I didn't notice that my heat wasn't working one winter for about a month because my computers were generating enough heat to keep me comfortable. 15:36
huf ... what 15:37
i cant even imagine that
benabik I can. In my college dorms in upstate NY, we regularly had the window open in the middle of winter to handle computer heat. 15:37
huf i've already had a morning when i considered turning the heat on 15:38
and it's only september
TimToady a 1 in one novemdecillion, six hundred six octodecillion, nine hundred thirty eight septendecillion, forty four sexdecillion, two hundred fifty eight quindecillion, nine hundred ninety quattuordecillion, two hundred seventy five tredecillion, five hundred forty one duodecillion, nine hundred sixty two undecillion, ninety two decillion, three hundred forty one nonillion, one hundred sixty two octillion, six hundred two septillion, five hundred twenty two sexti
colomon benabik: of course, one of my buddies had the same issue in his dorm back in the mid-80s. they just kept the dorm's heat way too high. :) 15:39
sirrobert I often kept a box fan blowing cold (snowy-cold) air into my dorm
sirrobert I had hot-water pipes running through three of my walls in the dorm =) (well, two walls and the floor) 15:40
arnsholt TimToady: Do you happen to know what pieces are missing to get is-prime from tommath in Rakudo? 15:41
TimToady moritz++ already has a patch, I believe
arnsholt Ah, cool. Then I won't duplicate his efforts
TimToady didn't apply it before release though, sanely enough... 15:42
arnsholt Probably not a bad idea
TimToady is not always in favor of sanity :)
colomon rn: my $a = set 4, 5, 6; say $a.keys.sort 15:43
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«4 5 6␤»
skids pfeh. Obviously we need to develop CPUs that have a heating CoP over 1. But then we'd have to find a way to connect them to the outdoor air or groundwater.. 15:44
TimToady we should invent CPUs that run faster when you put a blowtorch on them... 15:44
sirrobert in HS I had a DX2-66mhz. The fan fell off the CPU, so it kept slowing and stopping after (fairly short) intervals. my brother and I ended up keeping ice cube trays in the freezer and putting ice in zip-lock bags to lay on the CPU to keep it cool =) 15:47
colomon rn: my $a = set <a b c>; say $a.keys.sort 15:48
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«a b c␤»
GlitchMr sirrobert: makes sense :-) 15:49
But isn't "ice cube" water?
sirrobert that's why we used the plastic bags =) 15:49
(and put a paper-towel around the bag, but that was just an implementation detail... I don't think they ever got wet
colomon 's happily productive morning has been derailed by trying to figure out how to make Niecza's cmp handle hashes. 16:00
prammer 16:11
TimToady yup, mono crashed again at about the same place. apparently arbitrary precision integers are not so arbitrary. or it's just a bug that shows up when integers start to get arbitrary... 18:05
TimToady now waits patiently(?) for is-prime to show up in rakudo 18:07
diakopter TimToady: it might need some refactoring har har 18:26
diakopter TimToady: incorrect results are incorrect 18:28
TimToady well, one reason I'm waiting is that the current is-prime written in Perl 6 may not work right under rakudo; my program is consistently getting the wrong answer for every base 8 or higher
on rakudo, that is
niecza works up till the blowup at base 18 18:29
so if I replace is-prime, that might show where the bug is
as a first divide and conquer step
TimToady wonders what is different between base 7 and base 8 18:30
in every case it stops too soon, so it must be calling something composite that is actually prime, or it's not generating all the possibilities to begin with 18:31
diakopter maybe it's related to one of the integer overflow bugs 18:33
TimToady could well be 18:34
I could probably match against the correct answers and see which pass loses the truncated substring of the correct answer 18:38
diakopter how do you know the correct answers? 18:38
TimToady and in fact that technique might isolate to a single calculation
rosettacode.org/wiki/Talk:Find_larg...given_base 18:39
those answer check for all the cases I've been able to check under niecza
*swers
diakopter Microsoft's MS-PL BigInteger.cs++ 18:40
TimToady which is everything up to base 23 except for 18
TimToady is just a bit chagrined that ruby can calculate the base 18 case and Perl 6 can't yet 18:41
the bug could also be that correct answer is in the finalists but the sort isn't sorting it to the front correctly 18:42
[Coke] sees that java calls is-prime isProbablePrime(). 18:43
TimToady the finalists do not contain the correct answer, so it's not being generated 18:45
TimToady [Coke]: and we should take our design decisions from Java because they got it all right, eh? :) 18:45
[Coke] refuses to respond to the trollbait. 18:48
diakopter but it's so enticing 18:50
masak evenin', #perl6 19:26
colomon \o
sorear good * #perl6 19:42
[Coke] hio. 19:44
colomon o/ 19:53
colomon sorear: any idea why adding gist.github.com/3763534 to infix:<cmp> makes it blow sky-high when you try to compare two hashes? error message is "System.InvalidOperationException: Comparison threw an exception. ---> Numeric failed to return core numeric type" 19:55
[Coke] n: my %hash; say +%hash; 19:59
p6eval niecza v21-15-gf226bf6: OUTPUT«0␤»
[Coke] I presume cmp is cheating. :)
colomon [Coke]: if you look at the gist, you'll see my attempt to support comparing hashes. but it doesn't work. :( 20:00
sorear n: say 1 cmp 3
p6eval niecza v21-15-gf226bf6: OUTPUT«Increase␤»
colomon I think the error is happening in the sort call, but I've no clue why. 20:01
sorear colomon: if you get to the end of the for loop in the gist, what does cmp return?
colomon errr.... good question
adding a return and trying again. 20:02
sorear a return Same; I hope
colomon based on the full error message I don't think that's the issue, but it certainly was wrong the way I had it.
sorear n: say Nil.Numeric 20:03
p6eval niecza v21-15-gf226bf6: OUTPUT«0␤»
colomon n: say set("foo", "bar").keys.sort 20:03
p6eval niecza v21-15-gf226bf6: OUTPUT«bar foo␤»
colomon if I do that with my cmp-hash patch in, I get the error
sorear n: say set("foo","bar").keys.WHAT 20:04
p6eval niecza v21-15-gf226bf6: OUTPUT«List()␤»
sorear n: say set("foo","bar").keys>>.WHAT
p6eval niecza v21-15-gf226bf6: OUTPUT«Str() Str()␤»
sorear out of curiousity could you gist the output of "git diff"? I'd like to see your addition in context 20:05
sorear n: say 0 + Nil 20:05
p6eval niecza v21-15-gf226bf6: OUTPUT«0␤»
sorear n: say 0 + any("foo","bar") 20:06
p6eval niecza v21-15-gf226bf6: OUTPUT«Unhandled exception: Cannot parse number: foo␤ at /home/p6eval/niecza/lib/CORE.setting line 1435 (die @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3525 (ANON @ 11) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3527 (NumSyntax.str2num @ 5) ␤ at …
sorear n: say 0 + set("foo","bar")
p6eval niecza v21-15-gf226bf6: OUTPUT«2␤»
colomon gist.github.com/3763594 20:07
masak rn: say 0 + set("foo", "foo") 20:08
p6eval rakudo ef228d, niecza v21-15-gf226bf6: OUTPUT«1␤»
sorear n: say set(<foo bar>).sort 20:10
p6eval niecza v21-15-gf226bf6: OUTPUT«bar foo␤»
sorear n: say set(<foo bar>).keys.sort
p6eval niecza v21-15-gf226bf6: OUTPUT«bar foo␤»
sorear n: say set(<1 2>).keys.sort 20:11
p6eval niecza v21-15-gf226bf6: OUTPUT«1 2␤»
sorear n: say "foo" ~~ Hash
p6eval niecza v21-15-gf226bf6: OUTPUT«False␤»
moritz \o 20:14
sorear o/ moritz
colomon: I just applied this loc ally and it breaks for me 20:15
... and it just stopped breaking 20:16
colomon ? 20:18
sorear colomon: it's consistant. THe "Numeric failed to return core numeric type" occurs if and only if the setting was recompiled in the same run of niecza
if you rerun your test without changing the setting, it will work. 20:19
sorear colomon: I can reproduce this without your patch: 20:26
$ echo >> lib/CORE.setting
$ mono run/Niecza.exe --obj-dir=obj -e 'say ["foo", "bar"].sort'
[auto-compiling setting]
[done]
Unhandled exception: System.InvalidOperationException: Comparison threw an exception. ---> Numeric failed to return core numeric type
colomon it's not my fault! \o/ 20:27
unless, of course, it's because of something I did earlier. ;)
moritz bisect can tell, no? 20:29
btw I have a local rakudo path for expmod and is-prime, haven't gotten around to test it 20:32
will do tomorrow, and then push
jnthn moritz: yay :) 20:36
[Coke] I bet if you stick it somewhere, larry will test it. he seems pretty eager. 20:37
... almost hyper
moritz oh well, it's after the release, I can afford to break nom for half a day :-)
jnthn but racing for it would be out of order :P
dalek p: c458a08 | moritz++ | src/ (2 files):
make Rabin-Miller prime check available as nqp::isprime_I

also special-case 1 not to be a prime		 20:38
jnthn moritz: I can run a spectest
jnthn has the next two weeks off o/ 20:39
sorear moritz: bisect requires a good revision
sorear this has been broken for much longer than I thought and I'm having trouble finding a revision where it worked 20:39
dalek kudo/nom: 2e33656 | moritz++ | / (3 files):
add expmod and is-prime
 20:40
masak moritz++
masak moritz: is it because of libtommath that mp_prime_is_prime reports 1 as prime? 20:41
moritz: would they consider it a bug if we told them?
moritz masak: I don't know, which is why I haven't done it yet 20:42
masak: it claims to implement the rabin-miller prime test, not to check if something is a prime 20:43
jnthn has kicked of a build/spectest
sorear found a good rev. kicking off a bisect 20:44
masak \o/ 20:45
manual or with a script?
sorear script 20:46
sorear remarkably, this is the first time I have used git bisect 20:50
moritz you usually just don't regress stuff :-)
$ ./perl6 -e 'say (1..100).grep: *.is-prime' 20:56
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
looks better than before :-)
Tene why .is-prime instead of just .prime I wonder ?
sorear .prime-p
moritz Tene: maybe to not confuse it with "find the $nth prime"? 20:57
Tene Hmm.
moritz PrimeQ # in Mathematica
sorear my $funky = 42 but Prime;
std: 1 =< 2 21:00
p6eval std 77327a4: OUTPUT«===SORRY!===␤Unable to parse quote words at /tmp/0GmRNw2zTA line 1:␤------> 1 =⏏< 2␤Couldn't find final '>'; gave up at /tmp/0GmRNw2zTA line 1 (EOF):␤------> 1 =< 2⏏<EOL>␤Parse failed␤FAILED 00:00 42m␤»…
moritz sorear: I was about to do that too :-)
eval: 1 =< 2
buubot_backup moritz: ERROR: Unterminated <> operator at (eval 20) line 1.
sorear 13:58 <@LeoNerd> Whomever is responsible for eval: 2 =< 4 giving Reversed <= operator at (eval 1830) line 1. ERROR: Unterminated <> operator at (eval 1830) line 1. I will buy you a beer :) 21:01
moritz -> sleep
[Coke] moritz++ 21:02
"sleep well, and dream of large numbers."
sorear bisect blames 3d9f4b463b37ad13de9addc6e793bc3372af846b 21:03
masak smiles at the "sleep well and dream of X" meme being picked up 21:04
sorear I'll poke more at this later
dayjob&
masak .oO( sorear has a *job* now!? ) 21:05
clearly I haven't been keeping up.
colomon me neither!
jnthn moritz: spectest looks good :) 21:08
[Coke] O_o 21:10
[Coke] Meme? Why didn't you list among our assets in the first place? 21:13
masak ducks "why didn't you list among our assets" and gets the weirdest webcomic 21:16
[Coke] list *that 21:17
[Coke] masak: it's from the princess bride. 21:19
masak oh, princ... right.
I need to re-watch that movie :)
[Coke] masak: which is also where your meme is from, btw. :P 21:20
at least in my version of your meme.
masak so it seems. 21:24
I can honestly say that I wasn't thinking of the movie when I started using the meme on the channel. 21:25
sorear
.oO( "my name is S.J.Gould" )
masak but I'm not too disappointed of it containing prior art -- it contains a lot of nice stuff :)
PerlJam does that mean that people are going to start saying "as you wish" rather than give hugs?
[Coke] Good night, #perl6. I'll most likely kill you in the morning. 21:26
masak PerlJam: ;)
anybody want a peanut? 21:27
popl This is how the world ends.
PerlJam popl: But I haven't been Dread Pirate Roberts yet 21:28
popl s/r/w/g 21:29
PerlJam: I always hear it in Andre's voice.
Excuse me, Mr. the Giant. 21:30
PerlJam :-)
raiph phenny, tell moritz: there's a typo in github.com/moritz/ilbot/blob/maste...e/day.tmpl -- "<TMPL_UNLESS NAME=IS_SUMAMRY>" tried to fix and pull request via github but there's something going on that seems to break when i try. 22:05
phenny raiph: I'll pass that on when moritz is around.
masak 'night, #perl6 22:45
timotimo i made a code to calculate the intersection of two lines and i failed horribly :\ 23:28
shachaf timotimo: ? 23:30
timotimo i dunno what went wrong, but the coordinates of the intersections are on neither of the lines 23:33
and i didn't come up with proper tooling to see in what way it went wrong :|
raiph timotimo: are you using jnthn's debugger yet? 23:41
timotimo oh! OH! 23:43
should have totally done that
still doesn't help me see what's happened ;) 23:48
it's probably my formula that's wrong or something

Powered by Cro and the Raku® Programming Language.

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

Thank you!