»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
SmokeMachine m: { CATCH { default {"HERE"} }; ::("bla") } 01:04
camelia No such symbol 'bla'
in block <unit> at <tmp> line 1
SmokeMachine should it die? 01:05
jnthn m: try { CATCH { default {"HERE"} }; ::("bla") } 01:10
camelia ( no output )
jnthn m: { use fatal; CATCH { default {"HERE"} }; ::("bla") }
camelia ( no output )
jnthn If you don't use one of those two, the Failure will escape the block (since it's the block's return value) and thus explode in the program mainline, which is not covered 01:11
(try implies `use fatal` inside of the block it's covering)
SmokeMachine m: { use fatal; CATCH { default {say "HERE"} }; ::("bla") } 01:14
camelia HERE
SmokeMachine thanks!
skids you can also sink it if "use fatal" is too much. 01:20
{ CATCH { default {"HERE".say} }; my $a = 42; my $b; sink $b = ::("\$a"); $b }().perl.say
evalable6 42
skids m: { CATCH { default {"HERE".say} }; my $a = 42; my $b; sink $b = ::("blah"); $b }().perl.say 01:21
camelia HERE
Nil
skids m: { CATCH { default {"HERE".say} }; my $a = 42; my $b; $b = ::("blah"); $b }().perl.say
camelia Failure.new(exception => X::NoSuchSymbol.new(symbol => "blah"), backtrace => Backtrace.new)
buggable New CPAN upload: HTML-Canvas-0.0.3.tar.gz by WARRINGD cpan.metacpan.org/authors/id/W/WA/...0.3.tar.gz 01:44
Kaiepi m: my $ver := '2.0.0'; my ($maj, $min, $patch) := $ver ~~ /(\d+)\.(\d+)\.(\d+)/; say "$maj $min $patch" 05:45
camelia Unexpected named argument '' passed
in block <unit> at <tmp> line 1
Kaiepi m: my $maj := '2.0.0' ~~ /(\d+)\.\d+\.\d+/; say $maj.trim 05:46
camelia 2.0.0
Kaiepi wait why am i trying to use a regex when i can just do .split 05:47
Zoffix m: dd Version.new: '2.0.0' 05:49
camelia v2.0.0
Zoffix m: dd Version.new('2.0.0').parts 05:50
camelia (2, 0, 0)
Zoffix m: dd '2.0.0'.comb: /\d+/ 05:51
camelia ("2", "0", "0")
Zoffix m: use nqp; subset VLAN of Int where .so; subset Meows of VLAN; my $param := -> Meows $ where .so {}.signature.params.head; dd [.elems, $_] with nqp::getattr($param, Parameter, '@!post_constraints') 05:55
camelia [2, (Meows, -> ;; $_ { #`(Block|64841888) ... })]
Zoffix Is there a way to get more @!post_constraints? More than 2? 05:56
Kaiepi m: Version.new: 'foo' 06:02
camelia ( no output )
Kaiepi m: my ($a, $b, $c) := Version.new: 'foo'
camelia Cannot unpack or Capture `vfoo`.
To create a Capture, add parentheses: \(...)
If unpacking in a signature, perhaps you needlessly used parentheses? -> ($x) {} vs. -> $x {}
or missed `:` in signature unpacking? -> &c:(Int) {}
in block <unit> …
Kaiepi m: my ($a, $b, $c) := dd 'foo'.comb: /\d+/ 06:03
camelia ()
Too few positionals passed to '<unit>'; expected 3 arguments but got 0
in block <unit> at <tmp> line 1
Kaiepi how well does exception handling perform in perl 6? 06:04
Zoffix Don't know what you mean... "well" based on which parameters? 06:05
Kaiepi i mean in node for instance, try/catch makes the entire function it's used in unoptimizable
last i used it at least
Zoffix m: my ($a, $b, $c) = Version.new('2.3.42').parts; dd [$a, $b, $c] 06:06
camelia [2, 3, 42]
Zoffix m: my ($a, $b, $c) = '2.3.42'.comb: /\d+/; dd [$a, $b, $c]
camelia ["2", "3", "42"]
Zoffix Kaiepi: that sounds a lot like premature optimization 06:07
Worrying about that stuff, I mean.
Kaiepi it is for what i'm doing atm 06:09
Zoffix premature optimizing? :) 06:10
Kaiepi yeah
i'm wondering in case i run into some case where i might need to know in the future
Kaiepi t 06:31
Zoffix .tell tyil you have several bugs in your Appsixt module. Variables inside «» gets split on spaces within their values so, say, spaces in paths will break the path into separate args (`my $source = 'bar ber'.IO; dd «foo {$source.IO.path}»;`): gist.github.com/Whateverable/576bc...e13e4a7f26 07:18
yoleaux Zoffix: I'll pass your message to tyil.
perlawhirl hi perlers 07:39
perlawhirl i been playing around with different combinatoric sequences 07:40
i have an issue with one of my functions that generates the next sequency by modifying positional elements in an Array (or List) 07:41
and the gather/take, when evaluated eagerly, only gives me the last entry
gist.github.com/0racle/6be868d7ef3...9b24387299
this is golfed
but how do i work around this?
as mentioned, this is golfed... it's the rotating of positional elements (@list[*]) that does it... if i simply did `@list .= rotate` it doesn't happen 07:42
AlexDaniel .tell Zoffix re run, «» and notifying users about their bugs: I feel encouraged by your attempt to work on the issue on case by case basis… Lately I feel like it's easier to give up :) irclog.perlgeek.de/perl6/2018-01-24#i_15732558 07:44
yoleaux AlexDaniel: I'll pass your message to Zoffix.
perlawhirl but in my function i'm reversing a portion of the list (eg. `@list[$k..$n] .= reverse`) and i can't seem to make it work right when consuming the seq eagerly
AlexDaniel perlawhirl: this is so weird :D 07:55
AlexDaniel m: sub rots($list is copy) { gather for ^$list { $list[*] .= rotate; take $list; } }; my @l = ^3; say rots(@l).list; say rots(@l).Array; 07:56
camelia ([1 2 0] [2 0 1] [0 1 2])
[[0 1 2] [0 1 2] [0 1 2]]
AlexDaniel that has to be a bug 07:57
AlexDaniel m: sub rots($list is copy) { gather for ^$list { $list[*] .= rotate; take $list; } }; my @l = <0 1 2>; say rots(@l); say eager rots(@l); 07:58
camelia ([1 2 0] [2 0 1] [0 1 2])
([0 1 2] [0 1 2] [0 1 2])
Zoffix perlawhirl: you're sticking the same Array into all the cells 08:00
m: sub rots($list is copy) { gather for ^$list { $list[*] .= rotate; take $list.List; } }; my @l = <0 1 2>; say rots(@l); say eager rots(@l); 08:01
camelia ((1 2 0) (2 0 1) (0 1 2))
((1 2 0) (2 0 1) (0 1 2))
Zoffix m: sub rots($list is copy) { gather for ^$list { $list[*] .= rotate; take $list.clone; } }; my @l = <0 1 2>; say rots(@l); say eager rots(@l); 08:02
camelia ([1 2 0] [2 0 1] [0 1 2])
([1 2 0] [2 0 1] [0 1 2])
titsuki Is this a rakudo bug? REPL cannot handle binding properly: github.com/hankache/perl6intro/issues/168
Zoffix titsuki: looks like it 08:03
titsuki OK. then I'll tell him so... 08:04
Zoffix m: my @a = ^3; my @b; @a.=rotate; @b[0] = @a; @a.=rotate; @b[1] = @a; @a.=rotate; @b[2] = @a; say @b
camelia [[0 1 2] [0 1 2] [0 1 2]]
Zoffix perlawhirl: ^ this is what you're doing. The difference between the two cases is in the first one each item gets rendered into a string before the next one gets generated 08:05
Zoffix m: my @a = ^3; my @b; @a.=rotate; my $answer = '['; @b[0] = @a; @a.=rotate; $answer ~ @a; @b[1] = @a; @a.=rotate; $answer ~ @a; @b[2] = @a; $answer ~ @a; say "$answer]" 08:05
camelia WARNINGS for <tmp>:
[]
Useless use of "~" in expression "$answer ~ @a" in sink context (lines 1, 1, 1)
Zoffix m: my @a = ^3; my @b; @a.=rotate; my $answer = '['; @b[0] = @a; @a.=rotate; $answer ~= @a; @b[1] = @a; @a.=rotate; $answer ~= @a; @b[2] = @a; $answer ~= @a; say "$answer]" 08:06
camelia [2 0 10 1 20 1 2]
AlexDaniel aaaaaaaaaaaaaaaaaaaaaaaaaaaahhhhhh
Zoffix++ 08:07
Zoffix m: my @a = ^3; my @b; @a.=rotate; @b[0] = @a; my $answer = '[' ~ [email@hidden.address] @a.=rotate; $answer ~= [email@hidden.address] @b[1] = @a; @a.=rotate; @b[2] = @a; $answer ~= [email@hidden.address] say "$answer]"; say @a 08:09
camelia [[1, 2, 0],[2, 0, 1],[0, 1, 2]]
[0 1 2]
Zoffix m: my @a = ^3; my @b; @a.=rotate; @b[0] = @a; my $answer = '[' ~ [email@hidden.address] @a.=rotate; $answer ~= [email@hidden.address] @b[1] = @a; @a.=rotate; @b[2] = @a; $answer ~= [email@hidden.address] say "$answer]"; say @b
camelia [[1, 2, 0],[2, 0, 1],[0, 1, 2]]
[[0 1 2] [0 1 2] [0 1 2]]
Zoffix perlawhirl: ^ like that :)
$answer is the $r and @b is @r
titsuki: beat you to it. 08:13
titsuki Zoffix: Thanks! 08:14
perlawhirl Zoffix: I understand why your code does what it does, but don't immediately see why the gather/take acts this way 08:28
just so i understand it, when eagerly evaluating the gather/take, does it do the equivalent of your `@b[n] = @a` until the gather ends, then returns `@b`
and if that's the case... whats the best way around it, given that i'm generating a sequence by modifying a list in place 08:29
hrm... so a "quick fix" for my actual code is rather than `take @list[^$k]`, I can do `@a = @list[^$k]; take @a.clone`. 08:33
feels a bit messy, but helps illuminate the issue for me
moritz try @list.head($k) 08:35
Zoffix perlawhirl: no, @b.STORE pulls each of the items from the Seq created by gather/take and sticks into a new element in @b, one at a time. Each time it generates a new element, *all previous elements* of @b are too updated, since they all contain the same object. But by the time you **print** the eager version, the Seq was already fully pulled and all of the cells on @b contain the same @a Array that it's in its 08:42
last rotation. In the lazy version, the Seq is pulled **while the Str for printing is made** so despite the same Array being used for everything, by the time it changes on next iteration a Str snapshot of it was already taken
perlawhirl: possibly-helpful articles: perl6advent.wordpress.com/2017/12/02/ rakudo.party/post/Perl-6-Seqs-Drug...ock-n-Roll and rakudo.party/post/Perl-6-Seqs-Drug...ll--Part-2 08:43
Zoffix m: sub rots($list) { ^$list .map: { $list[flat ++$ ..^$list, ^ ++$ ] } }; my @l = ^3; my $r = rots(@l); say $r; my @r = rots(@l); say @r; 08:49
camelia ((1 2 0) (2 0 1) (0 1 2))
[(1 2 0) (2 0 1) (0 1 2)]
perlawhirl When I ran into the issue the first thing I did was re-read your articles on Seq's :D 08:50
Zoffix m: sub rots { .rotate, *.rotate … {++$ ≥ $_} with $^list }; my @l = ^3; my $r = rots(@l); say $r; my @r = rots(@l); say @r; 08:56
camelia ([1 2 0] [2 0 1] [0 1 2])
[[1 2 0] [2 0 1] [0 1 2]]
perlawhirl Thanks Zoffix. Will dig a bit deeper into this after I put the kids to bed. 08:58
Zoffix \o
Zoffix Actually, I got one more: 09:04
m: my &rots = { [.=rotate] xx $_ }; my @l = ^3; my $r = rots(@l); say $r; my @r = rots(@l); say @r;
camelia ([1 2 0] [2 0 1] [0 1 2])
[[1 2 0] [2 0 1] [0 1 2]]
Zoffix :)
perlawhirl: a lot less messy :)
perlawhirl So... I can't use macros in modules? macro works fine in a standalone script 10:52
but using it inside a function in a module, got "Missing serialize REPR function for REPR MVMContext (BOOTContext)"
I'm cool with that, given macros are experimental... just confirming this is expected behavior? 10:53
AlexDaniel greppable6: macro 10:56
greppable6 AlexDaniel, 2897 lines, 63 modules: gist.github.com/4daf1cc8cd6aa24895...ba175ba2b5 10:57
AlexDaniel greppable6: macros
greppable6 AlexDaniel, 137 lines, 35 modules: gist.github.com/4046709a6ed9751b44...03dd39200d 10:58
AlexDaniel greppable6: use experimental \:macros\;
greppable6 AlexDaniel, 7 lines, 3 modules: gist.github.com/c3c9652a784ff1c083...e66cefe9d3
AlexDaniel perlawhirl: can't comment on the issue but maybe take a look at these modules as they seem to work 10:59
perlawhirl thank... one more lil prob i just stumbled on... possibly related to the recent conversion of all subs to multi?? 11:10
i have a sub in my module that takes a "Callable", ie: foo($iter, &func = &infix:<+>) { ... }
in my tests it's failing on 'foo(@data, &max)' and it's complaining that " expected Callable but got Sub+{is-pure}" 11:11
this worked a few days ago, i'm fairly certain 11:12
perlawhirl m: sub foo($bar, &func) { func($bar) }; foo([^3], &max); 11:14
camelia Type check failed in binding to parameter '&func'; expected Callable but got Sub+{is-pure} (Sub+{is-pure}.new)
in sub foo at <tmp> line 1
in block <unit> at <tmp> line 1
perlawhirl bisectable6: sub foo($bar, &func) { func($bar) }; foo([^3], &max); 11:15
bisectable6 perlawhirl, Bisecting by exit code (old=2015.12 new=235d3f1). Old exit code: 0
perlawhirl, bisect log: gist.github.com/66a1b39ae7c4adacf5...2be5ebb3df
perlawhirl, (2018-02-20) github.com/rakudo/rakudo/commit/55...73aba83013
perlawhirl yeah, it's the commit that made &max a multi. 11:16
perlawhirl m: my &máximo = &max; 11:27
camelia Type check failed in assignment to &máximo; expected Callable but got Sub+{is-pure} (Sub+{is-pure}.new)
in block <unit> at <tmp> line 1
perlawhirl This seems to be a pretty bad regression 11:28
Additionally, it made it into the compiler release 11:32
AlexDaniel perlawhirl: what release? 11:36
plz don't give me a heart attack when it's not needed :) 11:37
c: 2018.02.1,HEAD sub foo($bar, &func) { func($bar) }; foo([^3], &max); 11:38
committable6 AlexDaniel, ¦2018.02.1: «» ¦HEAD(235d3f1): «Type check failed in binding to parameter '&func'; expected Callable but got Sub+{is-pure} (Sub+{is-pure}.new)␤ in sub foo at /tmp/lYR8EVamam line 1␤ in block <unit> at /tmp/lYR8EVamam line 1␤␤ «exit code = 1»»
AlexDaniel 6c: sub foo($bar, &func) { func($bar) }; foo([^3], &max);
committable6 AlexDaniel, gist.github.com/57c2cd5d31a9940305...aacd47e0d2
AlexDaniel perlawhirl: iirc that regression was somewhat “expected”, which is why we waited for the release to happen before trying to switch stuff to multis :) 11:40
perlawhirl: but please submit a ticket with this particular case 11:41
perlawhirl What I find odd (or I just don't understand) is I can define my own pure sub, export it from a module, and it doesn't suffer the same issue. It identifies (.WHAT) as a Sub+{is-pure}, but i can assign it to a &callable just fine
ok.. will submit GH issue
AlexDaniel thanks!
Geth doc: a881041297 | (JJ Merelo)++ | doc/Type/Any.pod6
Fixes the last "return Any"

By looking at the actual implementation. Revised also other "returns empty list". It actually returns an empty list in many cases, so it's OK. I haven't seen any other remainint, so provisionally closes #1731. If anyone finds a particular description that's wrong, I would rather have an specific open issue for that.		 13:50
synopsebot Link: doc.perl6.org/type/Any
Manifest0 hi. I think i found a bug on a error message produced by perl6 15:54
This is the code: paste.opensuse.org/92568701
and this is the error message it produces: paste.opensuse.org/33633715
:-) 15:56
timotimo looks like you have a quoted string around it, that sometimes happens 15:59
i don't see any opening quotes in the paste, though
oh, haha
no, you have to put a space after "sub"
timotimo otherwise you're calling a sub named &sub and trying to pass $t to it, which is not defined 15:59
timotimo m: a => sub($t){ say $t } 16:02
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$t' is not declared
at <tmp>:1
------> 3a => sub(7⏏5$t){ say $t }
timotimo m: a => sub ($t){ say $t }
camelia WARNINGS for <tmp>:
Useless use of "a => sub ($t){ say $t }" in sink context (line 1)
timotimo m: say a => sub ($t){ say $t }
camelia Unexpected named argument 'a' passed
in block <unit> at <tmp> line 1
timotimo m: say (a => sub ($t){ say $t })
camelia a => sub ($t) { #`(Sub|41873912) ... }
timotimo Manifest0: does that help?
timotimo we can probably detect when you're accidentally calling a sub called "sub" rather than declaring an anonymous sub 16:03
Manifest0 timotimo: that's it. A space was missing! 16:39
Manifest0 so without a space is a function call and with is an anonymous call? That's going to be confusing... 16:40
That explains the error message 16:41
timotimo aye, just like when you write "if($foo = 10)" that's not an if statement, that's a call to a sub called "if"
in that case you don't have the "i thought i was declaring that variable?" situation you had with "sub($t)" 16:42
Manifest0 timotimo: thanks for the explanation :-) 16:44
timotimo no prob
jgoff Manifest0: It's one of the reasons I use parentheses sparingly. 16:47
Manifest0 DrForr: yeah. I guess i'll have to do the same. I just started learning. I still to get experience in p6 16:49
jdv79 m: use Test;like(IO::Path.new("."),/\./) 17:20
camelia IO::Path is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in method new at src/RESTRICTED.setting line 32
in block <unit> at <tmp> line 1
jdv79 nice
anyway, the docs for Test::like say i coerces so what gives? 17:21
"Marks a test as passed if the $value, when coerced to a string, matches the $expected-regex. "
IO::Path::Str exists
Juerd Either the documentation or the implementation is wrong 17:22
It does not coerce. It even handles non-Str separately. 17:23
jdv79 i see that. its not in S24 so i guess its up for grabs 17:26
but timotimo - why did you make that change? the docs look correct before your change
timotimo i did?
jdv79 github.com/rakudo/rakudo/commit/ba...c9903bf1c3 17:27
maybe that should have been more like if it can't stringify then kick?
timotimo hm, i think the problem you mentioned was in the code before i changed things 17:28
like the commit message mentions, i had test code like "create an array, then call `like` on the different entries"
and it used to bring down the whole test suite if there wasn't an entry at one of the locations
jdv79 it used to be just: $got ~~ $expected
which would have wrked 17:29
timotimo ah, i see
feel free to turn it into Str() then
jdv79 ok
jdv79 shouldn't 'make t/spec/S24-testing/1-basic.t' work? 18:11
skids breaks zef while --force-installing zef with it. :-( 18:13
smarty 3 18:47
rindolf timotimo: here? 19:06
Geth doc: 0e59d162a9 | (Will "Coke" Coleda)++ | doc/Type/Hash.pod6
fix typos
 19:19
synopsebot Link: doc.perl6.org/type/Hash
Kaiepi is there something i'm doing wrong here? this is the only method in the class that's crashing for me, even though the native sub it uses works fine gist.github.com/Kaiepi/3f837104f66...4499181198 19:26
timotimo rindolf: briefly in between pieces of cooking 19:33
going to cut some veggies right now, but feel free to ask something
rindolf timotimo: any luck optimising my program? 19:34
Kaiepi never mind, looking at the coredump it's the same problem that's breaking some other native stuff on my system 19:43
MasterDuke rindolf: i asked him the same question yesterday, he had some info irclog.perlgeek.de/perl6-dev/2018-...i_15853760 19:44
timotimo rindolf: did you read backscroll? 19:45
ah, thanks MasterDuke
note, however, that i do get a different result after i made that change :(
so clearly some result in the middle passed the 64bit range
rindolf timotimo: was I addressed and online? 19:47
timotimo possibly not online 19:52
moritz we have public logs 19:54
MasterDuke rindolf: you said the equivalent code in perl5 ran much faster? 19:58
rindolf MasterDuke: yes 19:59
MasterDuke: about 20s here
MasterDuke: pypy runs it in under a second
MasterDuke hm. wouldn't perl5 also have problems if intermediate values are getting bigger than 64bit? 20:00
rindolf MasterDuke: see github.com/shlomif/project-euler/t...-euler/618 20:01
MasterDuke: i think so
MasterDuke huh. i guess they don't get quite big enough to actually break things 20:04
El_Che rindolf: do people use pypy nowadays? Most pythonistas I know don't 20:05
El_Che and google staying on 2.7 regular python and moving stuff to go instead of 3+pypy... 20:07
rindolf El_Che: i do for project euler 20:08
El_Che rindolf: if you don't use a lot of libs it's indeed a great candidate
rindolf MasterDuke: i am getting the same result from pypy3 and from p5
MasterDuke i think timotimo has contributed to pypy in the past 20:09
rindolf MasterDuke: on a core i3 machine
El_Che: yes
timotimo a tiny bit 20:10
pypy is fantastic
El_Che timotimo: does it fix the GIL thing? 20:11
timotimo: how is it that fast btw? regular python is also C 20:12
sounds interesting
rindolf El_Che: pypy has jit and stuff like that 20:13
El_Che a 50x speedup by jit? nice 20:14
rindolf timotimo: El_Che : MasterDuke : bye , gotta reboot 20:19
rindolf hi all 20:34
rindolf timotimo: there are intermediate results as output 20:39
MasterDuke man, a profile shows 231s spent in Int's infix:<%> 20:47
rindolf MasterDuke: with my program? 20:55
MasterDuke yeah 21:04
jeromemathias nice ! 21:06
MasterDuke there are 117598512 calls to it... 21:10
rindolf MasterDuke: ah 21:15
hankache hello #perl6 21:16
rindolf MasterDuke: can you shortcircuit it to int64 if it is within bounds? 21:17
hankache: hi
hankache hi rindolf
rindolf hankache: how are you?
hankache m: my $a := 3; $a = 4; say $a
camelia Cannot assign to an immutable value
in block <unit> at <tmp> line 1
hankache rindolf: all is well and you? 21:18
hankache m: my $a := 3; $a := 4; say $a 21:18
camelia 4
hankache is this sound? ^^^^
rindolf hankache: i got some pull-reqs for pysol fc 21:19
hankache rindolf: nice! 21:21
rindolf hankache: and i worked on some euler problems 21:22
buggable New CPAN upload: CommonMark-0.0.3.tgz by JGOFF cpan.metacpan.org/authors/id/J/JG/...-0.0.3.tgz 21:24
timotimo rindolf: what do you mean "there are intermediate results as output"? 21:28
rindolf timotimo: github.com/shlomif/project-euler/b...v1.pl#L574 21:30
timotimo: oh wait
timotimo: seems like it is just for progress
timotimo yeah, the interesting parts are: how big will $p * $m[$m - $p] get? 21:31
Geth ecosystem: drforr++ created pull request #386:
Add perl6-CommonMark
timotimo i don't expect it to get past 64bit, tbh
the numbers in that array are all modulo'd to $BASE 21:32
so i don't know where my code goes wrong
rindolf timotimo: p < 50k and the d < 1e9+7 21:32
timotimo m: say (50_000 * 1e9).msb 21:33
camelia No such method 'msb' for invocant of type 'Num'
in block <unit> at <tmp> line 1
timotimo m: say (50_000 * 1e9).lsb
camelia No such method 'lsb' for invocant of type 'Num'
in block <unit> at <tmp> line 1
timotimo oh, must be Int'ed
m: say (50_000 * 1e9).Int.msb
camelia 45
timotimo that'd get past the 32bit where our big integer math gets around allocating actual big int structures, but not outside of where int64 would be enough 21:34
timotimo or did i misunderstand? 21:37
rindolf 64-bit is over 1e18 21:38
m: say 1 << 63
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of << to do left shift; in Perl 6 please use +< or ~<
at <tmp>:1
------> 3say 1 <<7⏏5 63
rindolf m: say 1 +< 63
camelia 9223372036854775808
Geth doc: 55f79aba96 | (Zoffix Znet)++ | doc/Type/IO/CatHandle.pod6
Document IO::CatHandle.handles

Closes github.com/rakudo/rakudo/issues/1546 R#1546 Spec: github.com/perl6/roast/commit/1a89a1e8f3 Impl: github.com/rakudo/rakudo/commit/d5baa036fa		 21:42
synopsebot Link: doc.perl6.org/type/IO::CatHandle
synopsebot R#1546 [closed]: github.com/rakudo/rakudo/issues/1546 [IO] Implement IO::CatHandle.handles
jeromemathias perl6 c'est de la barre ! 21:44
rindolf jeromemathias: hi 21:58
MasterDuke huh, making all variables ints, including the arrays, makes it slower 22:11
rindolf MasterDuke: that sucks 22:12
MasterDuke not by much, but i would have expected some sort of speedup instead 22:14
rindolf MasterDuke: perhaps we can use bit vectors 22:15
like vec() in p5
timotimo MasterDuke: that's what i was writing about, though; split the expression with the two array accesses into multiple lines, each assigning into a native int var, then it'll be faster 22:17
MasterDuke wait, this is weird. why does my profile of the unmodified code show 45641 calls to calc_S?
i just stuck 'int ' before every variable declaration 22:18
timotimo yeah, it looked like that somehow incurs a whole lot of boxing and unboxing or something 22:20
MasterDuke must be, it was still using the Int, Int candidate for infix:<%> 22:21
pmurias if there are precompiled files in a path passed to -I rakudo should also find them? 22:29
MasterDuke and how does postcircumfix:<[ ]> have so much exclusive time when all it does is call AT-POS?
timotimo fwiw, in the fastest version i had, it was somehow still using the exact same line number for infix:<%> but even with the same number of calls it was significantly faster 22:30
so .. ?!
pmurias github.com/perl6/roast/blob/master...ion.t#L169 - this test is broken as we just don't find the module this way 22:37
pmurias (so if the interning works or not is not even tested) 22:38
Zoffix pmurias, looks like just `'.pm.' ~ $precomp-ext` should be changed to simply `'.pm6'` here github.com/perl6/roast/blob/master...ion.t#L131 doesn't it? 23:28
Zoffix oh, nm, there's code that does stuff to it 23:29
Filed as S#391 23:32
synopsebot S#391 [open]: github.com/perl6/roast/issues/391 Broken todo-ed precomp interning test
MasterDuke timotimo: it's definitely using the Int,Int infix:<%> candidate. i added a note() to the native candidate and it's never getting printed 23:56

Powered by Cro and the Raku® Programming Language.

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

Thank you!