»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.
00:02 espadrine left 00:04 spider-mario left 00:19 cfedde left 00:20 cfedde joined 00:23 cschwenz left, protium left, flussence left 00:26 flussence joined 00:28 shadowpaste left 00:29 shadowpaste joined 00:32 protium joined 00:38 SCHAAP137 joined, protium left 00:39 protium joined 00:50 BenGoldberg joined, protium left 00:51 Herby_ left, protium joined 00:52 yeahnoob joined, protium left 00:54 protium joined
BenGoldberg m: { my sub foo { 42 } }; say foo; 00:57
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Lo96KfyYq3␤Undeclared routine:␤ foo used at line 1␤␤»
BenGoldberg m: { my sub foo { 42 }; say foo };
camelia rakudo-moar 770d10: OUTPUT«42␤»
BenGoldberg m: { say my sub foo { 42 }; }; 00:58
camelia rakudo-moar 770d10: OUTPUT«sub foo () { #`(Sub|64228656) ... }␤»
BenGoldberg m: my $secret = { my sub foo { 42 }; }; say $secret();
camelia rakudo-moar 770d10: OUTPUT«sub foo () { #`(Sub|57335376) ... }␤»
01:00 AlexDaniel left
dalek osystem: 8a34595 | (brad clawsie)++ | META.list:
add WebService::AWS::V4
 01:02
osystem: 3c9f418 | (Zoffix Znet)++ | META.list:
Merge pull request #127 from bradclawsie/master

add WebService::AWS::V4
BenGoldberg m: my $secret = { anon Int sub { 42 }; }; say $secret(); 01:03
camelia rakudo-moar 770d10: OUTPUT«sub ( --> Int) { #`(Sub+{Callable[Int]}|62201728) ... }␤»
BenGoldberg m: my $secret = anon Int sub { 42 }; say $secret(); 01:04
camelia rakudo-moar 770d10: OUTPUT«42␤»
01:18 SCHAAP137 left 01:31 mcmillhj joined 01:33 adu joined
dalek kudo-star-daily: 3227cc6 | coke++ | log/ (6 files):
today (automated commit)
 01:36
kudo-star-daily: 3491280 | coke++ | log/ (6 files):
today (automated commit)
kudo-star-daily: 3c084de | coke++ | log/ (6 files):
today (automated commit)
rl6-roast-data: d2d8319 | coke++ | / (6 files):
today (automated commit)
rl6-roast-data: b66f6e4 | coke++ | / (7 files):
today (automated commit)
01:36 BenGoldberg left 01:39 _nadim left 01:44 Actualeyes joined 01:51 vendethiel joined 01:52 adhoc left 02:09 leont left, BenGoldberg joined 02:11 xpen joined 02:14 vendethiel left 02:16 Cabanossi left 02:19 vendethiel joined 02:24 pierre-vigier joined 02:32 Cabanossi joined 02:34 kid51 joined 02:42 vendethiel left 02:49 pierre-vigier left 02:50 linuxuser9000 joined 02:53 Calibellus joined 02:54 yqt left 02:57 Calibellus left 03:00 Calibellus joined 03:01 Calibellus left 03:05 pierre-vigier joined 03:10 kid51 left 03:13 psy_ left 03:15 psy_ joined, psy_ left 03:16 psy_ joined 03:17 Calibellus joined 03:26 protium left 03:35 linuxuser9000 left 03:38 MadcapJake left, MadcapJake joined 03:42 BenGoldberg left 03:43 Calibellus left 03:49 noganex_ joined 03:52 noganex left 04:02 linuxuser9000 joined 04:03 TEttinger joined
MadcapJake any suggestions on how i could keep a linecount while parsing with a grammar? 04:17
04:18 gregf_ left, gregf_ joined
llfourn MadcapJake: I'm not sure, if there is a builtin way, btu maybe just increment a counter at linebreaks? 04:24
skids MadCapJake: increment a dynamic in the newline rule, assuming you never backtrack it.
llfourn that sounds even cleverer :)
MadcapJake cool, that's good! thanks! I was leaning towards having a `token line { <useful> | <gibberish> }` but i think making a newline rule will be a lot easier! 04:26
04:28 pierre-vigier left 04:30 n0tjack joined 04:35 n0tjack left 04:48 pierre-vigier joined 04:57 revhippie left 05:06 pierre-vigier left 05:09 pierre-vigier joined 05:14 marmay_ joined 05:16 psy_ left, khw left 05:38 n0tjack joined 05:40 marmay_ left 05:42 n0tjack left, skids left 05:49 Cabanossi left 05:50 Cabanossi joined 05:59 pierre-vigier left 06:09 pierre-vigier joined 06:13 jme` left 06:30 darutoko joined 06:31 adu left 06:41 sjoshi joined 06:46 sjoshi left 06:47 sjoshi joined
dalek osystem: d4c384a | (brad clawsie)++ | META.list:
apologies, fixed package repo name
 06:47
osystem: 315f659 | sylvarant++ | META.list:
Merge pull request #128 from bradclawsie/master

apologies, fixed package repo name
06:47 ChoHag left
06:48 salva left 07:06 geraud left 07:09 linuxuser9000 left 07:10 ChoHag joined 07:14 vendethiel joined 07:15 quester joined 07:16 pierre-vigier left, bjz joined 07:17 pierre-vigier joined 07:21 bjz left
MadcapJake what's a quick way to remove the first character of a string? 07:23
I tried comb.shift.join, but apparently you can't do that on Sequences
I tried comb.eager.shift.join but apparently you can't do that to immutable lists :(
maybe comb[1..*]? 07:24
07:24 FROGGS joined
MadcapJake that did it, thanks! 07:25
ugexe $str.substr(1) 07:26
MadcapJake haha that works too :D
I was wondering if anyone was reading me talk to myself xD
weird, it's still printing the `\` must not actually be there or something :S 07:29
cognominal when creating a postcircumfix op, how to tell that the operand should not be a list. I seem to remember there is a way to tell the loosest precedence accepted but I can't seem to find it
MadcapJake cognominal: maybe this: doc.perl6.org/language/functions#Coercion_Types 07:30
07:30 [Sno] left
cognominal nope, I am messing with nqp. thx anyway 07:32
on the other hand, I may have seen that feature used in rakudo because it is written in nqp 07:34
07:36 Actualeyes left 07:37 firstdayonthejob joined, Actualeyes joined, vendethiel left 07:39 salva joined 07:41 bjz joined
[Tux] csv-ip5xs 50000 18.300 12.374 07:47
test 50000 23.677 23.010
test-t 50000 12.662 11.739
csv-parser 50000 52.584 -0.576
07:49 abraxxa joined 07:54 _mg_ joined 07:59 n0tjack joined 08:00 RabidGravy joined
MadcapJake cognominal: maybe this: doc.perl6.org/language/functions#Coercion_Types 08:01
oops sorry, trying to access terminal history :P 08:02
08:03 n0tjack left 08:05 quester left 08:07 pierre-vigier left
masak morning, #perl6 08:07
08:08 pierre-vigier joined 08:09 [Sno] joined
moritz \o masak, * 08:09
RabidGravy erp!
08:11 CIAvash joined, vendethiel joined, _nadim joined 08:15 bjz_ joined 08:16 bjz left, zakharyas joined 08:26 pierre-vigier left, pierre-vigier joined 08:30 uruwi left 08:32 uruwi joined
cognominal Venus. I mean morning * 08:33
08:34 vendethiel left 08:35 frankjh joined
RabidGravy Boo! 08:39
08:39 firstdayonthejob left 08:41 cc9989 joined
RabidGravy so yesterday, I stated my intention to make Perl 6 versions of the RabbitMQ tutorials, today I realise that I am going to spend most of the time adding functionality to Net::AMQP 08:42
A WHOLE FARM OF HIRSUTE YAKS 08:43
moritz the yak is strong in this one 08:47
masak good ♀, cognominal 08:49
08:52 Actualeyes left 08:54 dakkar_ joined
masak Unicode has made my life so much richer. before Unicode 6.0.0, I didn't even know cats could cry tears of joy. 09:00
or pout. 09:02
09:03 lizmat joined
lizmat wonders why there still isn't a Rakudo *, a P6W or a decision on how to work with versioning 09:04
09:04 lizmat left 09:06 pierre-v_ joined, pierre-vigier left, Actualeyes joined
RabidGravy I think I need a thing that is like a collection of promises such that when one is Kept or Broken it is removed and can be treated as a single promise that can be kept or broken 09:09
masak RabidGravy: not quite clear to me what you mean, but check out .allof and .anyof on Promise 09:10
09:10 brrt joined
cognominal messing with the NQP grammar : github.com/cognominal/nqp/blob/bra...ti-rules.t 09:11
RabidGravy the reason for this is that something like Net::AMQP may have a bunch of promises that are waiting on a '*-ok' message from the server, but if the connection craps out it may for instaance get a close-ok but none of the others 09:12
soOooooo potentially a program could just be left hanging on a Promise that will never be Kept 09:13
09:14 cschwenz joined
RabidGravy masak, oh I know *how* to make it, I just want someone else to do it :) 09:14
cschwenz o/ #perl6 :-)
brrt \o cschwenz 09:15
09:17 nakiro joined 09:22 leont joined
masak RabidGravy: ok. I thought either of those two might be what you were looking for. 09:24
09:28 avs_ left
RabidGravy oh yeah, it would almost certainly involve anyof :) It's basically a loop round anyof(@bunch-o-primises).then({ # remove done ones from @bunch-o-promises}; 09:29
09:30 n0tjack joined
moritz there should be a primitive that turns a bunch of promises into a stream, based on their completion/breakage 09:31
FROGGS cant we .grep them? 09:33
09:34 n0tjack left
RabidGravy well you can do something like Supply.merge(@promises.map({ $_.Supply })).grep(...) 09:36
which is fun
moritz but that doesn't give you the list of pending promises, right? 09:40
09:40 vendethiel joined
RabidGravy oh I see, no that just gives the Kept ones 09:41
09:42 yeahnoob left
moritz maybe it's enough 09:43
RabidGravy m: say (Promise.new, Promise.new.keep, Promise.new).grep({ $_.status ~~ Planned}) 09:44
camelia rakudo-moar 770d10: OUTPUT«Use of Nil in numeric context in block at /tmp/4k9xA3o6KX line 1␤(Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned) Nil Promise.new(scheduler => Th…»
RabidGravy ok, that's special
but something like that
oh yeah, the Promise.new.keep no return a Promise 09:47
brrt masak: nice blog post (as usual) 09:48
RabidGravy m: my @ps = (Promise.new, Promise.new, Promise.new); @ps[1].keep; say @ps.grep({ $_.status ~~ Planned})
camelia rakudo-moar 770d10: OUTPUT«(Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned) Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_h…»
RabidGravy there
09:49 rindolf joined
masak brrt: thanks 09:50
09:51 abaugher left
moritz m: m: my @ps = (Promise.new, Promise.new, Promise.new); @ps[1].keep; say @ps.classify(*.status).perl 09:52
camelia rakudo-moar 770d10: OUTPUT«(my Any %{Any} = PromiseStatus::Planned => $[Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned), Promise.new(scheduler => ThreadPoolScheduler.new(initia…»
09:52 abaugher joined
moritz m: m: my @ps = (Promise.new, Promise.new, Promise.new); @ps[1].keep; say @ps.classify(*.status).keys 09:52
camelia rakudo-moar 770d10: OUTPUT«(Planned Kept)␤»
09:53 kjs_ joined
RabidGravy Ooh classify doesn't stringify the keys 09:55
ah, Hash[Any, Any] 09:56
09:58 g4 joined, g4 left, g4 joined 10:01 vendethiel left 10:04 espadrine joined, cognominal left 10:10 brrt left 10:12 ely-se joined
ely-se moroning 10:12
oops :P 10:13
Is it possible to query errno after a NativeCall call?
Currently I have "fail 'cannot create pipe' if $status == -1;" but I'd like to fail with something more descriptive. 10:14
RabidGravy yes, it's doc.perl6.org/language/nativecall#E..._variables 10:19
10:23 TEttinger left
ely-se That's quite scarce and low-level. What if I have a different libc version? Does it respect errno being thread-local? What about __errno_location? 10:24
There is a reason libc implementations define errno as a macro that expands to __errno_location. Does cglobal treat errno as a special name? 10:25
or rather, expands to (*__errno_location()) 10:26
timotimo i don't think it does; it just uses that dlopen mechanism
10:26 leont left 10:28 joydo joined
moritz ely-se: if you want to take care of such abstractions, the most reliable way is to write your own C function my_errno() or so that returns errno, and which you can use reliably in NativeCall 10:29
ely-se: if you think such a thing is universally helpful, maybe submit a pull request that adds it to NativeCall? 10:30
ely-se I think I will use this: gist.github.com/rightfold/a6cd608544912f91e401
10:31 adhoc_ joined
RabidGravy the example in that doc is not quite up-to-date with what cglobal actually does, it uses "guess_library_name" so you can pass it anything that you can pass the native trait 10:32
including Str
ely-se moritz: I like the way Go handles errno (golang.org/cmd/cgo/). Maybe some trait that enables a similar mechanism would be nice.
10:33 brrt joined
ely-se to also automatically wrap it in an exception object 10:33
RabidGravy i.e. cglobal(Str, "errno", Int) should work
timotimo except if the libc has errno as a macro
in which case, you're out of luck
RabidGravy indeed
in which case one would need to compile a tiny helper I guess 10:34
10:34 lnrdo joined
timotimo ely-se: you can build a proxy for the errno in your code 10:34
ely-se timotimo: solution: gist.github.com/rightfold/a6cd608544912f91e401 but fragile if the VM decides to do some I/O in between.
timotimo alternatively, a term for errno would also work
10:35 pmurias joined
ely-se it has to be atomic 10:35
timotimo hm. i don't think the VM would decide to do I/O by itself. __errno_location is supposed to be thread-safe, yeah?
ely-se yes. that's the point of __errno_location. it uses thread-local storage
timotimo right
ely-se I'll make the current solution into a library and use that. Thank you all.
10:39 max___ joined
RabidGravy good good 10:39
pmurias for binding c++ libraries just wrapping everything in c functions and using NativeCall is the only way? 10:42
timotimo no 10:43
we implement name mangling for methods
max___ Hi, I've installed the Rakudo Xmas release only just recently - up until then I'd been playing with some older version. Now I wonder what happened to Parcels...they
don't seem to be there anymore - can anybody confirm that? 10:44
moritz max___: they were remove with the Great List Refactor
so gone since 2015.09
max___ thanks...good to know!
10:46 virtualsue joined
max___ while we'r at it...I can't access the perl6maven.com site anymore - I've been using it quite a bit. Anybody seeing similar problems? 10:46
moritz isup.me/perl6maven.com "It's not just you! perl6maven.com looks down from here." 10:47
max___: you could shoot an email to [email@hidden.address] who runs that
timotimo "gateway time-out" 10:48
stmuk_ perl6maven is restarted on a cron according to his YAPC talk 10:50
apparently due to the language it uses :)
10:54 lnrdo left 10:58 lnrdo joined 11:03 brrt left
ely-se RabidGravy: done: github.com/rightfold/NativeCall-Errno :) 11:04
11:05 pierre-v_ left 11:06 Roamer` left 11:07 ely-se left 11:15 ely-se joined, _nadim left
RabidGravy coolio 11:17
11:18 vendethiel joined
RabidGravy .seen retupmoca 11:22
yoleaux I saw retupmoca 12 Dec 2015 19:56Z in #perl6: <retupmoca> .tell nine renaming dlls works fine if libfoo.dll is standalone. But in the case of openssl: ssleay32.dll depends on libeay32.dll. I load libeay32.dll (works great!), then I try to load ssleay32.dll which tries to find or load a dll named 'libeay32.dll'. Since I didn't load it with that filename, ssleay32 won't load because it can't find libeay32.dll
11:25 ely-se left
max___ just to let you know - Gabor Szabo just restarted his perl6maven.... 11:26
pmurias timotimo: does that work with templates and inline methods? 11:31
11:31 virtualsue left
timotimo i don't think we can do templates, no. not without a compiler. 11:32
11:33 xpen left 11:34 xpen joined
pmurias is there any hope for transparent "read the header", use the compiler, maybe offer a clunky interface style interop? 11:34
being able to parse a c++ header and generate a c wrapper+native call binding would be really awesome 11:39
11:40 brrt joined 11:42 CIAvash left, vendethiel left
timotimo we have gptrixie by skarsnik which generates NC bindings from a header file 11:44
11:44 llfourn left 11:45 vendethiel joined
RabidGravy but a sort of damn clever Inline::C that just takes the header and a library and creates the subs on the fly might be cool 11:46
11:47 lnrdo left
RabidGravy infact I guess you could sorta kinda do that now if you took the output of gptrixie and EVAL it 11:47
brrt why don't we just write a c compiler in perl6 11:48
moritz because ugly
brrt is grinning widely at the thought
RabidGravy but don't let use stop you ;-)
us
brrt moritz: do you manage pl6anet.org? 11:50
moritz brrt: no, itz does
or does itz have a new nick? I can't remember
RabidGravy stmuk mostly nowadays 11:51
or some variant therepf
moritz stmuk_: please add up-to-date contact information (your current nick name, possibly email address) to pl6anet.org. Thanks!
RabidGravy and goats, we need more goats 11:52
timotimo is a javascript dev here? how do nodejs version numbers work? i have 0.10.36 locally, but apparently there's already 4.2.4 LTS and 5.4.1 stable ... o_O 11:54
11:56 perlawhirl joined
pmurias timotimo: they changed the version scheme after a fork took over 11:56
timotimo i see it now 11:57
fair enough
stmuk_ moritz: done 12:00
brrt: is the expired cert? I'll be on the case shortly :) 12:01
brrt no, that's not what it is about
i was wondering if we could make it so that only my posts categorised perl6 would get syndicated 12:02
so i can safely publish stuff that is not of interest to the perl6 community without spamming everyones feed reader
stmuk_ I really would need a perl6 specific RSS feed 12:03
unless you can 'tag' non-perl 6 and I put a special case hack in
brrt hmmm 12:04
let me see if that can be arranged
12:06 joydo left, vendethiel left
arnsholt brrt: I believe some Lisp machines had C compilers that compiled to Lisp or somethiing like that =) 12:08
brrt cool :-)
stmuk_: url would be brrt-to-the-future.blogspot.com/fee...lt/-/perl6 12:09
stmuk_ ok it will be updated shortly
brrt thanks :-)
max___ quit 12:10
12:12 Skarsnik joined 12:13 kjs_ left
Skarsnik brrt, hello. I tried on 64 bits and my code still segfault on parse x) 12:15
12:15 virtualsue joined 12:17 pierre-vigier joined 12:20 frankjh left 12:29 telex left 12:30 vendethiel joined, pmurias left, telex joined, pmurias joined 12:37 kid51 joined 12:41 llfourn joined 12:42 virtualsue left, pierre-vigier left 12:47 llfourn left 12:48 lucasb joined, lnrdo joined
brrt oh... great 12:48
:-(
can you run perl6-gdb-m and give us a backtrace?
12:49 pierre-vigier joined
Skarsnik gist.github.com/Skarsnik/ef3c2613aec884b2202f 12:50
12:50 kid51 left 12:51 vendethiel left 12:52 virtualsue joined
Skarsnik you can find the code on github.com/Skarsnik/gptrixie the reftype branch x) 12:52
12:53 xpen left 12:56 ely-se joined 12:59 virtualsue left 13:00 virtualsue joined
timotimo well, a C compiler is just a different QAST code-gen that also tells the jit to output its machine code stuff into a .o file, right? :P :P :P 13:02
"basically" :D
Skarsnik x) 13:04
RabidGravy "just a simple matter of programming"
pmurias timotimo: we could have a C to moar bytecode compiler 13:06
13:07 virtualsue left
timotimo if we simply map the mov instruction to a moar bytecode, we can translate any C program to moar bytecode 13:07
pmurias brrt: re write a c compiler, I want to be able to bind c++ libraries
brrt: I have a lot of things I would rather do than write a c++ compiler 13:08
brrt :-)
13:09 domidumont joined
RabidGravy binding c++ libraries works 13:09
brrt well, darn, skarsnik, that looks ugly
Skarsnik it's when I added github.com/Skarsnik/gptrixie/blob/...ss.pm6#L73 that is start segfaulting. but even removing it did not make it work x) 13:11
brrt no, it's a GC issue 13:12
i wonder how
dalek osystem: 79ba90b | rightfold++ | META.list:
Add NativeCall::Errno to ecosystem

See github.com/rightfold/NativeCall-Errno		 13:14
ely-se I oh thought it would create a pull request and not instantly commit it to the main repo. :v 13:15
13:15 vendethiel joined
timotimo it's totally fine 13:15
moritz ely-se: no harm done
ely-se ok :p
timotimo i don't mind direct commits to the ecosystem repo
in fact, i've added my most recent modules like that, too
13:16 lokien_ joined
ely-se I do use .t6 not .t for my tests. Is that problematic? 13:16
13:16 cschwenz left
ely-se I think Panda won't pick them up. 13:16
moritz ely-se: yes, that's problematic
13:17 max___ left
ely-se I do it because my editor otherwise thinks it's Perl 5. 13:17
But I'll change that tehn
moritz teach your editor to take a 'use v6;' as an indicator for Perl 6 code, maybe? 13:18
13:18 kjs_ joined 13:19 musiKk joined 13:20 sufrostico joined, sufrostico left
pmurias Skarsnik: having something like gptrixie for C++ to generate things like an OpenCV binding would be awesome 13:21
Skarsnik well it should not be hard to add c++ support to it
I am pretty sure gccxml give me class information 13:22
what is the name of opencv c++ lib? 13:23
13:24 lnrdo left
ely-se moritz: ok, I changed it :0 13:24
:)
pmurias Skarsnik: you mean the name of the .so? 13:26
pierre-vigier is there any possibility to make a binding between a list elements, and a hash element? like my @array = $r; my %h = 'key' => $r; 13:27
Skarsnik I mean libopencv-dev is the right lib? not the C one?
pierre-vigier and if i do %h<key> = 12; then my array first element = 12 ?
13:28 lucasb left
pmurias Skarsnik: it to seems to be 13:29
FROGGS m: my $r = 42; my @a; @a[0] := $r; my %h; %h<key> := $r; say @a; say %h; $r = 111; say @a; say %h # pierre-vigier 13:30
camelia rakudo-moar 770d10: OUTPUT«[42]␤key => 42␤[111]␤key => 111␤»
FROGGS m: my $r = 42; my @a; @a[0] := $r; my %h; %h<key> := $r; say @a; say %h; %h<key> = 111; say @a; say %h
camelia rakudo-moar 770d10: OUTPUT«[42]␤key => 42␤[111]␤key => 111␤»
13:30 lnrdo joined
Skarsnik I get output with class and stuff with gccxml 13:30
so it's doable x) 13:31
pierre-vigier thanks,
strnage i had that message: 13:32
Cannot use bind operator with this left-hand side
it might come from somethign else
oh, i was using a list, not an array 13:33
so it ws immutable
FROGGS aye
13:34 iH2O joined 13:36 vendethiel left 13:37 _mg_ left
dalek blets: 87b312f | lichtkind++ | docs/tablet-3-variables.txt:
preparing the next changes
 13:38
13:38 musiKk left 13:39 _mg_ joined, vendethiel joined, domidumont1 joined 13:40 domidumont left 13:43 llfourn joined
ely-se When I write "#| C<foo> bar" before my sub, Pod::To::HTML doesn't generate "<code>foo</code> bar" but "&lt;code&gt; bar". is that a bug in Pod::To::HTML or am I using #| wrong? 13:44
Skarsnik pod parser is "weak" 13:45
not your fault ^^
ely-se ok 13:46
13:47 iH2O left 13:48 llfourn left 13:51 molaf joined 13:52 domidumont joined 13:53 Roamer` joined 13:55 domidumont1 left, sufrostico joined 14:00 vendethiel left 14:01 brrt left 14:07 vendethiel joined 14:10 cdg joined 14:11 lnrdo left 14:12 skids joined 14:15 _Vi left 14:19 n0tjack joined
pmurias is zef the panda replacement? 14:20
timotimo i wouldn't call it "replacement" per se 14:21
but it fills the same niche
pmurias panda's slowness is annoying 14:25
ely-se what is slow about Panda?
it's always been fast for me
timotimo you seem to have low expectations, then 14:26
pmurias the json parsing is slow
time panda help takes 21seconds on my machine 14:27
timotimo but it already uses json::fast! 14:28
what more do you want ;(
(yes, json::fast is also very slow)
14:29 __HOO joined 14:30 vendethiel left 14:31 adu joined
pmurias timotimo: why don't we have a fast json parser? 14:31
ely-se pmurias: only 1.47s on mine 14:32
Skarsnik rakudo is slow x) 14:35
gfldex if i overwrite a method in a subclass (that does exactly the same thing) that is used by some operator, should that method call have a performance impact? 14:36
pmurias would it be possible to parse json faster using nativecall? 14:37
ely-se gfldex: it might, because the base method may have been JIT-compiled already whereas the override might not have 14:38
Skarsnik pmurias, well it's sad that everytime we want perf we use a libc xD 14:39
ely-se pmurias: if the bottleneck is with the actual parsing and not with data structure construction, then I'm pretty sure that's possible
gfldex pmurias: the individual json files are quite small. So the overhead might outweigh the benefit
Skarsnik It annoy me a lot that it take more than a 1 second to parse xml file that are less that 1Mb 14:40
ely-se ini > JSON :)
14:41 llfourn joined, adu left
ugexe its slow because it uses a grammar 14:42
14:43 rindolf left
ugexe parsing speed can be increased by only parsing the parts that need to be parsed 14:44
gfldex can i precomp a script?
ugexe if i dont access $json<some-key2> it doesnt need to parse the possibly 100 levels under "some-key2" : 14:45
you can precomp a script but you still need a perl6 entry point
ely-se ugexe: interesting, but how will you do that? 14:46
14:46 llfourn left
ugexe originally (and probably in the future again) bin/ scripts got precompiled, and the wrapper was the perl6 entry point 14:46
ely-se brace balancing won't work; braces may appear in string literals that you have to parse
pmurias Skarsnik: in perl5 a lot of the speed critical things are in XS
ugexe ely-se: there was an article on hacker news a few months ago about the "worlds fastest json parser". i forget the language it was in even, but that was how it was done
it was in D 14:47
news.ycombinator.com/item?id=10430951 # some explanations here as well 14:49
14:50 perlawhirl left, prammer joined
ely-se ugexe: apparently it accepts invalid JSON 14:50
that's not good for a general-purpose JSON parser 14:51
ugexe its a json parser yes, not a json validator
14:52 mohae left
Skarsnik ok that not funny 10-20 sec to parse a 650kb xml file 14:56
14:57 pierre-vigier left
ugexe if you think thats bad try `to-json()`ing a large structure 14:57
14:57 __HOO left
Skarsnik funny how a simple c++ headers imply the inclusion of 47 files just by using std::string x) 14:59
jnthn If only there were as many people able to work on performance as there were people willing to carp about it. :P
timotimo pmurias: we have a NativeCall-based json parser already
15:00 sftf joined
ugexe i certainly tried, but couldnt do any better than timotimo/tony-o 15:00
timotimo only for a json parser? 15:01
why not try to make rakudo itself or moarvm faster? :) :)
join us at the internals side. it's fun! i promise*! 15:02
* not an actual promise
15:02 kaare_ joined, sufrostico left
Skarsnik do you have cookies? 15:02
timotimo no, but we have a CCookie REPR 15:03
15:03 g4 left 15:04 sufrostico joined
ugexe because i dont have what it takes to properly benchmark/optimize/repeat 15:04
15:04 CIAvash joined
ugexe it ends up into just optimize/repeat, and then eventually confusion 15:04
15:04 adu joined
Skarsnik should I start working on my c++ support in gptrixie on spend time binding libxml2 to have xml parsing perf xD 15:05
15:05 raiph joined, _Vi joined
arnsholt timotimo: Do you know how the NativeCall JSON parser compares to the pure Perl 6 ones? 15:05
timotimo i never actually ran it
arnsholt From memory the overhead in NativeCall is non-trivial, sadly
timotimo it's probably between 10000000x and 100000000000000000000000000000000x faster
FROGGS Skarsnik: do we want to join our libxml2 effort?
15:06 adu left
[Coke] comes back online, and is forced to skip backscroll here. 15:06
timotimo hm
yoleaux 17 Jan 2016 20:27Z <bartolin> [Coke]: looks like you have about 100 processes (some of them a few days old) running on hack. no real problem so far, just got some notifications (we're monitoring the number of procs on hack)
Skarsnik FROGGS, well I probably don't want the whole libxml2, just having a dom from a xmlfile and convert it to p6 XML object
15:07 colomon left, colomon joined
[Coke] bartolin: fixed the symptom, will prevent the problem from happening again.... but this means that spectests are hanging on hack, which is bad. 15:07
15:07 lnrdo joined
FROGGS Skarsnik: and I basically "just" want to make XML::Compile work, but I dont have enough tuits for anything these days :/ 15:08
Skarsnik XML::Compile?
15:09 adrusi left
FROGGS Skarsnik: a P5 module that loads a wsdl, compiles it to Perl 5 code, so you call subroutines and it does the right things 15:09
15:10 pierre-vigier joined, Praise left, sjoshi left
FROGGS "Where other Perl modules (like SOAP::WSDL) help you using these schemas (often with a lot of run-time XPath searches), XML::Compile takes a different approach: instead of run-time processing of the specification, it will first compile the expected structure into a pure Perl CODE reference, and then use that to process the data as often as needed." 15:11
bbl 15:13
15:13 FROGGS left, adrusi joined
Skarsnik gah I wish there was a module C { class are export are rw; } to put commons traits to the class in the same module xD 15:15
timotimo you could build a module with an EXPORTHOW "xclass" and "xstruct" that automatically apply "is export" and "is export is rw" and treat the rest of the stuff as classes 15:18
15:19 lokien_ left 15:22 pmurias left
ely-se I see people in labcoats. 15:22
Skarsnik hm 15:24
timotimo ely-se: the scary kind or the nice kind? 15:26
15:28 jvcakg joined, pmurias joined, jvcakg left
ely-se timotimo: yes 15:29
pmurias Skarsnik: for what I need from opencv I should be fine with just manually extending the hand written binding
ely-se There's a lab on the other side of the canal here.
timotimo ah
15:30 jvcakg joined
Skarsnik hm, what can prevent a class to be composable? 15:30
ely-se use of global mutable state 15:31
jvcakg is perl 6 totally different from perl 5?
DrForr One easy way to find out :)
Skarsnik Na, I mean I have an error telling me that one calss it not composable
ely-se jvcakg: it has some similarities but there's little chance your existing Perl 5 code will work without being completely rewritten
jvcakg ah i see 15:32
DrForr But you can use Inline::Perl5 in order to still use it.
ely-se Going from Perl 5 to Perl 6 is not like going from Perl 4 to Perl 5 but more like going from C++ to C#
jvcakg I see your point 15:33
Skarsnik m: class A { has $.a}; class B does A { has $.b}; class C does B {has $.c};
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Hr2P3_FWwG␤A is not composable, so B cannot compose it␤at /tmp/Hr2P3_FWwG:1␤»
15:33 pmurias left
Skarsnik hm, I need to use is and not does? 15:33
jvcakg so people like me that are starting to learn Perl "6" will benefit 15:34
timotimo Skarsnik: well, classes can't be applied as if they were roles 15:35
Skarsnik Oh is the right thing
jvcakg good thing I didn't learn Perl 5
timotimo you can use roles as if they were classes because they can, at any point, be turned into a class by composing them into an empty class. that's called "punning"
jvcakg: ah! welcome to the club :) 15:36
ely-se Learning Perl 5 may still be useful.
Skarsnik It's confusing that is the same keywoerd that for trait
timotimo i went straight from python to perl6
jvcakg thank you
ely-se Especially if you want to use a Perl in production right now.
jnthn Skarsnik: does is also a trait modifier :)
Skarsnik m: class A { has $.a}; class B does A { has $.b}; class C is B {has $.c}; say C.^attributes;
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/TP2m8h9n_U␤A is not composable, so B cannot compose it␤at /tmp/TP2m8h9n_U:1␤»
jvcakg is there a book in Perl 6?
DrForr At least a few people here are already using perl6 in "production", whatever that means. 15:37
jvcakg or there all documented in web?
timotimo we don't have a perl6 book yet, but a bunch of docs online, yeah
jvcakg ok
interesting
timotimo perl6.org/downloads/ - i recommend all of the "introductory material" links on the right side of this page
ely-se I have had too much trouble with using experimental software in production. I don't want to make that mistake again. 15:38
timotimo i'm certainly hopeful our stability stuff works as well as it sounds 15:40
15:40 _Vi left
Skarsnik well aside from my leak issue and my weird sigseg it's stable? x) 15:40
15:42 CIAvash left, llfourn joined
gfldex are there any guidelines on how to do versioning of modules? 15:42
Skarsnik nop 15:44
15:46 _Vi joined 15:47 llfourn left
timotimo do you ever wonder if things would have been way better if instead of "Perl 6" we had called it "PerlScript"? 15:47
15:47 PerlJam left
timotimo hm. or perhaps ECMAPerl 15:48
15:48 perl6newbee joined
jvcakg any website examples written in Perl 6? 15:48
moritz jvcakg: doc.perl6.org
jvcakg thank you
15:48 kjs_ left
ely-se timotimo: Oystear 15:49
Perl ~ Pearl hence Oystear ~ Oyster
timotimo excellent 15:50
jvcakg moritz, not that. a finish example, someone mention made a site base on Perl 6 15:52
Skarsnik doc is generated with perl6 code
jvcakg ok
Skarsnik but if you mean something hosting dynamic content in perl6, well
jvcakg hehe
15:52 virtualsue joined
jvcakg waiting.. 15:53
moritz perl6maven.com/
jvcakg appreciate it
15:54 nakiro left, xpen joined
ely-se I was working on one recently. Current status is ABANDONED 15:54
but it's proprietary anyway :P
15:56 _mg_ left
Skarsnik it's annoying this error does not generate the backtrace: Cannot unbox a type object 15:58
15:59 vendethiel joined 16:00 xpen left
moritz Skarsnik: have you submitted a bug report already? 16:01
16:02 sena_kun joined, lnrdo left 16:03 PerlJam joined 16:04 lnrdo joined
Skarsnik duh C allow for anonymous struct? :( 16:05
16:07 frankjh joined 16:08 mohae joined
PerlJam timotimo: re perlscript ... it would have been worse IMHO 16:08
timotimo Skarsnik: only new-ish C allows that, i think 16:09
or, wait
i think i'm thinking of anonymous union 16:10
Skarsnik pthread as it apparently
geekosaur struct { ... } is fine, yes
PerlJam C has had anonymous structs for as long as I can remember.
geekosaur back to K&R days
PerlJam Though, you usually typedef them or something so that you can do more useful things with them.
16:11 pierre-vigier left
geekosaur yes. there was inconsistency in the old days, some compilers would consider identical anonymous structs to be type compatible and others wouldn't 16:11
that is, given `struct { whatever} a; struct { same thing } b;` different compilers would or would not allow `a = b` 16:12
although if we have to care aout what pre-ANSI C compilers do, we may have already lost 16:13
ely-se timotimo: you are confusing C with C++. in C++ it is a very new feature (C++11), except for unions
C has had them since longer
geekosaur yeh, C++ made it go away because `struct foo` and `foo` are the same thing
and then brought it back\
16:14 domidumont left
timotimo ely-se: what now. unions or structs? 16:15
ely-se timotimo: in pre-C++11 C++, only unions could be anon. since C++11, structs can be too 16:16
timotimo OK
Skarsnik hm interesting calling $o<key>:exists on an object that does AT-KEYS does not work 16:17
moritz Skarsnik: it needs EXISTS-KEY too
16:17 jme` joined
Skarsnik hm, I should add it to XML them 16:17
frankjh Hi, I have a question wrt to types and units: I want to refactor some code, but if I move the fuction nonce() called from github.com/soundart/perl6-tweetnac..._box.t#L12 which is currently defined here github.com/soundart/perl6-tweetnac....pm6#L187, into another unit module then I get an error from line 13 of 03-crypto_box.t: Type check failed in binding 16:19
So what is the difference between the two types and why does it depend on the unit where the nonce function is defined? 16:20
16:20 cc9989 left 16:21 lokien_ joined, vendethiel left
timotimo <list type="ordered" start="4" tight="true" delimiter="period" sourcepos="21:1-22:0"> 16:22
sorry
cat stepped on laptop's trackpad ...
ely-se punish the cat 16:23
16:23 ely-se left
stmuk_ cat = entropy for key generation 16:23
16:25 kjs_ joined
gfldex frankjh: sub crypto_box (Blob $buf!, CArray[int8] $nonce!, CArray[int8] $pk!, CArray[int8] $sk!) 16:27
that's what you call
native shaped arrays and error messages are not the best friends right now
Skarsnik ? 16:28
frankjh gfldex: uh I still do not understand.?
Skarsnik CArray are not shapped array 16:29
gfldex indeed, that error message part still stands
RabidGravy CArray type constraints are somewhat wonky
16:29 vendethiel joined
RabidGravy it did work at some point, then stopped working somewhere between september and december 16:30
gfldex frankjh: try to add the same type to 'my $nonce = nonce();' and see if that points you to the righ line/file
16:33 _mg_ joined, nemo left
stmuk_ is there any "cross installer" (eg. panda and zef) way (API whatever) of querying what's installed from the ecosystem and where? 16:33
frankjh If I modify line12 to my CArray[int8] $nonce = nonce(); and add use NativeCall, Then I get the same Error from the same call of crypto_bob(...) 16:35
16:35 rudi_s left, xdg joined 16:40 nemo joined, nemo is now known as Guest53072
[Coke] (entropy generation) I just had a bug trying to launch a coldfusion container due to not enough entropy. I may need to borrow your cat. 16:41
16:42 khw joined, Guest53072 is now known as nemo 16:43 llfourn joined 16:45 rudi_s joined 16:48 llfourn left 16:49 stmuk left, stmuk_ is now known as stmuk, salva left 16:51 vendethiel left
frankjh RabidGravy: Is above a known Bug? And is NativeCall::Types::CArray[int8].new a type or something I can call? 16:51
RabidGravy It is known, the message isn't very nice and you just want to relax the typeconstraints, you still need to create shaped CArrays 16:53
16:55 sena_kun left
raiph m: my @d; @d.append: 1,2; @d.push: 3, 4; @d.append: [5,6]; @d.push: [7,8]; dd @d 16:55
camelia rakudo-moar 770d10: OUTPUT«Array @d = [1, 2, 3, 4, 5, 6, [7, 8]]␤»
raiph ^^ doc.perl6.org/routine/append says "The difference to method push is that with an array or list argument, append appends several elements (one for each array or list element), whereas push appends just one element."
^^ Do folk agree the doc is wrong and I should s:g/array or// ?
(I'll backlog later) 16:56
16:56 raiph left, musiKk joined
RabidGravy oh well, clearly not that interested 16:57
moritz RabidGravy: @d.push: 3, 4 ; is a call with two arguments, not with a list argument
RabidGravy I know
moritz sorry, meant raiph
m: my @d = 1, 2; @d.push: (3, 4); say @d.perl 16:58
camelia rakudo-moar 770d10: OUTPUT«[1, 2, (3, 4)]␤»
RabidGravy I was going to say that what the description in the doc exactly matches the result
16:59 raiph joined
flussence [Coke]: you may find `haveged` a bit safer than a cat in production there :) 17:00
raiph moritz, RabidGravy: I meant s:g/or list//
moritz raiph: I figured. And the answer is: the docs are correct as is. See my example above. 17:01
frankjh RabidGravy: I am trying to read what a shaped array is... How do I relax my type constraint? 17:02
17:02 rindolf joined
RabidGravy CArray[uint8] is shaped, CArrat is relaxed 17:02
frankjh RabidGravy: Ah ok..
17:04 musiKk left, kjs_ left 17:05 llfourn joined
RabidGravy raiph, @p.push: 1,2 is neither an array or a list, it is multiple arguments 17:05
17:06 vendethiel joined
RabidGravy (1,2) is a list; [1,2] is an array 17:06
stmuk hmmm shouldn't the panda "state" file contain the version of module installed?
frankjh RabidGravy: I relaxed the type constraints and it compiles and tests are passing again. Thank you! 17:09
raiph RabidGravy, moritz: Makes sense, thanks for your steady patience.
RabidGravy no worries both :)
17:10 raiph left 17:11 sufrostico left 17:19 andrewalker left 17:23 SCHAAP137 joined 17:27 MageAlpha joined 17:29 vendethiel left, zakharyas left
MageAlpha 'p6: say 3;' 17:32
p6: say 3
camelia rakudo-moar 770d10: OUTPUT«3␤»
[Coke] tries to read Shimmerfairy's "calmer" rant. 17:33
Nope. Ah well. 17:35
17:35 breinbaas left, larion left 17:38 dakkar_ left
kaare_ [Coke]: Where? 17:39
17:39 _mg_ left
[Coke] I'm not linking to it. 17:39
it was related to rt.perl.org/Ticket/Display.html?id=127108
17:41 psy_ joined, psy_ left, Actualeyes left 17:42 psy_ joined
mspo [Coke]: link? 17:50
[Coke] mspo: 12:39 < [Coke]> I'm not linking to it. 17:51
b2gills Can someone look at github.com/rakudo/rakudo/pull/686 I'm waiting on it to push a Palindromic Code Golf 17:53
17:54 lichtkind joined
moritz b2gills: I'm trying to find out why travis complains about this 17:54
without success so far
17:54 vendethiel joined
timotimo on my end, travis says rakudo/rakudo isn't an active repository >_> 17:55
b2gills I think GitHub itself might have been too slow when it was pulling the code????
17:56 MageAlpha left
b2gills I could force push a newer version ( that is only the time will change) 17:56
17:56 xpen joined
b2gills Which I think will cause it to try it again 17:56
I already force pushed the associated ROAST changes ( I forgot two `;` ) 17:57
moritz timotimo: yes, same here 17:58
mspo found it
moritz did anybody deactivate travis for rakudo? 17:59
17:59 ab6tract joined
ab6tract o/ #perl6 17:59
b2gills If you are planning on just clicking the accept button I will force push an iteration with newer timestamps 18:01
[Coke] anyone know if Grammar::Parsefail::Exceptions is in a module somewhere?
18:01 xpen left, _mg_ joined
stmuk maybe its a "provides" you could try grepping the panda.json 18:02
18:03 autarch1 joined
[Coke] ah, this user declared the dep in a readme instead of in a deps file. 18:03
stmuk github.com/ShimmerFairy/grammar-pa...META6.json 18:04
ab6tract masak: it probably already occurred to you, but `subset PlannedPromise where { $_.status ~~ Planned }` has a really nice feel to it :)
stmuk its there
autarch1 how can I fork a running p6 program? there doesn't seem to be a fork builtin and the Proc class seems to be for spawning another program
18:04 firstdayonthejob joined
stmuk shouldn't it be X::Grammar::Parsefail :P 18:04
18:04 captain-adequate joined
ab6tract autarch1: what's your use case? 18:06
autarch1 ab6tract: I want to test some IPC code so I'd like to fork in my test code
I could write two separate scripts and have one call the other but it'd be nice to keep this all in one file
ab6tract autarch1: ah, i see
mspo [Coke]: ~/perl6 is unclean :) 18:07
err. ~/.perl6
RabidGravy sub fork() returns int32 is native { * }
autarch1 I think that threads are preferred over forks in p6 but in this particular case I'd like to test the code both ways
also, sometimes forking is better since it gives you better isolation between interpreters 18:08
ab6tract autarch1: yeah, in general there hasn't been call for fork afaict
autarch1 that seems odd
but if it's not implemented I guess that's my answer ;)
ab6tract i wonder if RabidGravy's proposal would work 18:09
PerlJam It should
RabidGravy jt does
PerlJam (unless you've got a libc without fork :)
And I think that's part of the reason why it hasn't been implemented.
(It was easy enough if ever anyone wanted it) 18:10
dalek osystem: 674191d | (Zoffix Znet)++ | spec.pod:
Delete "spec.pod" from the repo. It"s outdated and the README already mentions Test::META
ab6tract can someone on Linux try installing Audio::Sndfile ?
I'm curious whether my troubles are OS X related 18:11
RabidGravy what's the matter with it?
dalek osystem: 6380513 | (Zoffix Znet)++ | META.list:
Math::Matrix now uses META6.json; not META.info
RabidGravy I will check in a minute
ab6tract I get 42 fails in 010-read.t
18:11 sftf left
mspo what if you want to daemonize? 18:12
RabidGravy ah, I took the todo out
PerlJam ab6tract: trying now.
ab6tract mspo: then your code could just look like: my $daemon = supply { ... }; await $daemon;
18:13 gregf_ left
PerlJam ab6tract: All tests successful. 18:13
ab6tract mspo: unless i'm mistaken, which happens frequently enough :)
mspo ab6tract: I mean without forking
unless supply does a lot more than I think
RabidGravy ab6tract, it's an os/x specific rakudo bug, there's an RT 18:14
ab6tract RabidGravy: nooooooo
:
:(
there is a lack of NativeCall hackers running OS X :(
18:15 gregf_ joined
ab6tract my example doesn't fork. but if you mean without using any extra threads, you could use loop { ... } ? 18:15
ab6tract has never written a daemon, just imagines them as processes that don't exit until they get a message to do so 18:16
flussence there's nothing magical about a daemon, except maybe the amount of effort needed to get the @&$%ing thing to behave in a proper environment when it insists on backgrounding 18:17
18:17 vendethiel left
mspo ab6tract: to get into the background you need to fork, sedsid, cd /, umask, close some fd's 18:17
18:17 abraxxa left
arnsholt That's more or less it. Except there's a bunch of OS-specific crap you need to deal with to run properly in the background 18:18
18:18 spider-mario joined
mspo on a bsd system you can just call daemon(3) 18:18
but on linux daemon(3) is broken or something
nfi how to run a windows process
autarch1 is serialization in p6 supposed to be done by implement .perl methods for one's objects? 18:19
ab6tract autarch1: the default should do it for you 18:20
autarch1 ok, that works
18:21 _Vi left, vendethiel joined
ab6tract thanks for the primer on daemon architecture. 20 years on *nix and never bothered to do more than start or stop them :) 18:22
mspo ab6tract: api.metacpan.org/source/AJDIXON/Da...emonise.pm 18:23
fwiw
18:24 espadrine left
mspo I should give it a try 18:24
RabidGravy ab6tract, this is the RT for that BTW rt.perl.org/Ticket/Display.html?id=125408
18:24 abraxxa joined
RabidGravy but I did infact take the TODOs on the afflicted tests 18:26
ab6tract m: class A { has A $.a; has $.b }; my $a = A.new(a => A.new( b => 'bye!'), b => 'brooklyn'); say $a.perl.EVAL.b 18:27
camelia rakudo-moar 770d10: OUTPUT«brooklyn␤»
ab6tract mspo ^^
erm, sorry, autarch1 ^^
awwaiid what is a name for the concept of classes-as-special-values, such as Inf, NaN, and maybe True, False?
18:28 llfourn left
PerlJam Type object? 18:28
oh, no, not that
perigrin They're not that yeah.
18:28 abraxxa left 18:29 raiph joined
awwaiid "special-values" is what came to mind 18:29
ab6tract perigrin: they aren't type objects?
awwaiid also Nil
perigrin awwaiid: I've never heard of a single name other than "special-values"
awwaiid well they ARE type objects, but their usage is more than that
PerlJam yeah
masak [Coke]: if I may just hold two contradictory ideas for a while, I think (a) ShimmerFairy is valuable to #perl6 and losing her like this would make us all poorer, (b) that second post was still very tone-deaf, in ways that none of us managed to convey to ShimmerFairy without making her feel like crap
ab6tract right, i just thought of type objects as the overall genre
18:29 lokien_ left
RabidGravy ab6tract, if you have an application for Audio::Sndfile on OSX I could stick the todo back in just for that platform 18:29
18:30 autarch1 left
ab6tract in which case 'special types' is more evocataive than 'special values' 18:30
awwaiid ab6tract: I want to have a name for them in documentation. Heck maybe even a list of them.
ab6tract RabidGravy: that would be great :)
awwaiid ab6tract: the reason I like special-values more than special-types is that they are being returned as and treated as values 18:31
RabidGravy it works fine for non-floating point data iirc
(makes a change from fixing Net::AMQP to be able to do the RabbitMQ tutorial examples :)
Skarsnik hm, what can I use to count subset of an array/hash? 18:32
map then elems?
18:33 FROGGS joined
ab6tract awwaiid: maybe we could call them 'value-types' ? 18:33
18:33 mindos left
[Coke] masak: my solution is to find out what the actual bug is. 18:34
which I think I just did.
wow, that took... 15 minutes. :P
18:34 mindos joined, ggherdov left, corbyhaas left, jnap left, kipd_ joined, corbyhaas_ joined 18:35 parisba left, ugexe left
[Coke] Ok. more of a toolchain question, and this is something that definitely would have been impacted by CURL?I 18:35
github.com/ShimmerFairy/SUPERNOVA/...test.p6#L6
18:35 kipd left, kipd_ is now known as kipd
[Coke] there's a 'use Grammar;' line there. I think the expectation is that that is going to load github.com/ShimmerFairy/SUPERNOVA/...rammar.pm6 18:35
18:36 robinsmidsrod left
masak makes sense, no? 18:36
[Coke] if I put a BEGIN { say "GOAT THERE" } in that file, nothing. if I change that use to an EVALFILE, I get the note. Is it possible that she's inadvertently loading an internal Grammar ?
awwaiid indeed
18:36 ugexe joined
masak m: use Grammar 18:37
camelia rakudo-moar 770d10: OUTPUT«===SORRY!===␤Grammar is a builtin type. You can use it without loading a module.␤»
masak [Coke]: indeed
looks like a name collision of some kind
18:37 robinsmidsrod joined 18:38 ggherdov_ joined
hoelzro I think that if all of the supernova stuff were prefixed with SUPERNOVA:: as a namespace, it would probably work, but that's LTA 18:38
RabidGravy ab6tract, are you sure it was the t/010-read.t rather than the 020-write.t that was failing? the former never had any todos in it
18:38 cbk__ left
flussence
.oO( not to mention all-caps generally means "internal, don't touch" )
 18:38
awwaiid ab6tract: thanks for the input, I'll noodle a bit. will be s/// relaceable either way and I guess there's no standard yet
18:39 jnap joined, parisba joined, parisba is now known as Guest70706
[Coke] also, the Grammar.pm6 file has a grammar called Pod6::Grammar in it, which is weird to me. I'd expet that to live in Pod6/Grammar.pm6 18:39
FROGGS [Coke]: aye 18:40
ab6tract RabidGravy: yup. 'seekable' returns False instead of True, and a lot of others 18:41
seems to happen for multiple formats
i wonder if i'm missing some compilation flags in my install of libsndfile?
awwaiid: sounds good :) 18:42
[Coke] ah, changing the BEGIN { say ... to a straight up "die" does show it's calling that file. 18:44
18:45 marmay_ joined
RabidGravy ab6tract, oh, there were never any todos in that file, this is going to be one of those "won't get fixed because I can't test" things 18:45
ab6tract, can you post the verbose output of the failing test files, then I can at least see if there is anything I can do without being able to reproduce 18:49
18:49 domidumont joined
RabidGravy that is post it as an issue on the Audio::Sndfile GH 18:49
ab6tract RabidGravy: verbose would be just running the test file with perl6 instead of through the test harness? 18:50
or with --ll-debug
RabidGravy yeah just running the test files without the harness will do fune
fine
18:55 virtualsue left 19:08 lnrdo left 19:10 jme` left 19:14 jme` joined 19:19 silug left
[Coke] .tell shimmerfairy to see latest comment on RT #127108 that will give you a workaround so you can keep moving. 19:21
yoleaux [Coke]: I'll pass your message to shimmerfairy.
[Coke] it's a precompilation bug. Like many of the others that have been reported.
19:24 llfourn joined 19:26 _nadim joined, ely-se joined 19:29 llfourn left, AlexDaniel joined 19:32 cdg left
AlexDaniel o/ 19:32
19:32 perl6newbee left 19:33 _mg_ left, musiKk joined
Skarsnik I think I will never understand how to use grep/map on stuff like hash xD 19:33
AlexDaniel Skarsnik: do it over keys? 19:34
b2gills moritz: currently permutations(1) and permuations(0) don't work at all, so I don't know how it would expose a bug that wouldn't have already been exposed
star-m: say permutations(1).perl; say permutations(0).perl # I think it is a bug that this produced ().Seq because +permutations(0) should be 1
m: say permutations(1)
m: say permutations(0)
camelia star-m 2015.09: OUTPUT«((0,),)␤().Seq␤»
rakudo-moar 770d10: OUTPUT«Index out of range. Is: -1, should be in 0..Inf␤ in block <unit> at /tmp/VvcE1t9ZY_ line 1␤␤Actually thrown at:␤ in block <unit> at /tmp/VvcE1t9ZY_ line 1␤␤»
rakudo-moar 770d10: OUTPUT«Constraint type check failed for parameter '$n'␤ in block <unit> at /tmp/SstxroKePy line 1␤␤»
19:34 raiph left, _Vi joined
ely-se How can I conditionally define a subroutine at compile-time? 19:35
19:36 raiph joined
Skarsnik m: my %h = a => 1, b => 2; my @t = %h.values.grep{ %% 2}; say @t; 19:36
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ZxPrf6GW7M␤Two terms in a row␤at /tmp/ZxPrf6GW7M:1␤------> 3=> 1, b => 2; my @t = %h.values.grep{ %%7⏏5 2}; say @t;␤ expecting any of:␤ infix␤ infix stopper␤ statemen…»
Skarsnik m: my %h = a => 1, b => 2; my @t = %h.values.grep{ * %% 2}; say @t;
camelia rakudo-moar 770d10: OUTPUT«Cannot call grep(Seq: ); none of these signatures match:␤ ($: Bool:D $t, *%_)␤ ($: Mu $t, *%_)␤ in block <unit> at /tmp/rLhgx828jD line 1␤␤»
Skarsnik I really don't get it xD 19:37
AlexDaniel m: my %h = a => 1, b => 2; my @t = %h.values.grep: * %% 2; say @t;
camelia rakudo-moar 770d10: OUTPUT«[2]␤»
AlexDaniel m: my %h = a => 1, b => 2; my @t = %h.values.grep: { $_ %% 2 }; say @t;
camelia rakudo-moar 770d10: OUTPUT«[2]␤»
AlexDaniel m: my %h = a => 1, b => 2; my @t = %h.values.grep(* %% 2); say @t; 19:38
camelia rakudo-moar 770d10: OUTPUT«[2]␤»
Skarsnik hm
how I can remove these result from the hash? 19:39
19:39 psy_ left
AlexDaniel m: my %h = a => 1, b => 2; -> $a, $b { %h{$a} :delete if $b == 2 } for %h.kv; say %h 19:41
camelia rakudo-moar 770d10: OUTPUT«a => 1␤»
AlexDaniel m: my %h = a => 1, b => 2; { %h{$^a} :delete if $^b == 2 } for %h.kv; say %h
camelia rakudo-moar 770d10: OUTPUT«a => 1␤»
AlexDaniel Skarsnik: is it good enough? ↑ 19:43
awwaiid I find state-changing adverbs on things that don't look like verbs (such as non-state-changing hash value lookups) to be... interesting. Like veribifying a noun kinda.
19:44 psy_ joined
gfldex awwaiid: we do that in german all the time. It works well. 19:44
awwaiid gfldex: which, verbifying nouns? we do that in english as well; I'm just not sure this is quite that 19:46
gfldex the order feels a little odd 19:48
i would say :delete {$^a} in %h. Both in english and in german. In german however I would be allowed to change the order, it doesn't got strict SPO 19:49
AlexDaniel m: :<x> 19:51
camelia rakudo-moar 770d10: OUTPUT«WARNINGS for /tmp/XbevNzTxXa:␤Useless use of constant value x in sink context (line 1)␤»
AlexDaniel m: say :<x>
camelia rakudo-moar 770d10: OUTPUT«x␤»
AlexDaniel m: say ::<x>
camelia rakudo-moar 770d10: OUTPUT«Nil␤»
AlexDaniel m: say :::<x>
camelia rakudo-moar 770d10: OUTPUT«===SORRY!===␤Could not locate compile-time value for symbol :<x>␤»
AlexDaniel m: say ::::<x>
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/eJ5lull58K␤Name component may not be null␤at /tmp/eJ5lull58K:1␤------> 3say ::7⏏5::<x>␤»
19:54 sufrostico joined
awwaiid reading rakudo:src/core/Hash.pm, what does 'is raw' on methods mean? mmm... looks like that basically does 'is raw' on each of the params, which looks like it means it uses binding instead of assignment 19:54
19:57 xpen joined
awwaiid AlexDaniel: what is :<x>? 19:57
AlexDaniel awwaiid: I don't know
awwaiid the rest kinda make sense
FROGGS m: say (:<x>)
camelia rakudo-moar 770d10: OUTPUT«x␤»
vendethiel m: say (:<x>).perl #:) 19:58
camelia rakudo-moar 770d10: OUTPUT«"x"␤»
gfldex awwaiid: it's binding
awwaiid oh, like :("x") ?
AlexDaniel binding what? 19:59
gfldex m: sub ($a is raw, \b){}
camelia ( no output )
awwaiid m: say :("x").perl
camelia rakudo-moar 770d10: OUTPUT«:(Str $ where { ... })␤»
stmuk does anyone know of an ecosystem module with good POD which isn't DBIish?
awwaiid oh, gfldex was referring to my 'is raw' comment
gfldex :() <-- signature literal
and a big monkey ofc
awwaiid gfldex: yeah, but :<x> doesn't seem to be shorthand for :("x") 20:00
20:00 Ven joined
Skarsnik m: my %h; say %h<a>.elems # why 1? 20:00
camelia rakudo-moar 770d10: OUTPUT«1␤»
gfldex m: say (:<x>).WHAT
camelia rakudo-moar 770d10: OUTPUT«(Str)␤»
Skarsnik should it be 0? x) 20:01
awwaiid examines AST output
gfldex m: my %h; say %h<a>:exists; 20:02
camelia rakudo-moar 770d10: OUTPUT«False␤»
20:02 xpen left
gfldex m: my %h; say %h<a>:exists; say %h.elems; 20:02
camelia rakudo-moar 770d10: OUTPUT«False␤0␤»
awwaiid m: %h<x>.WHAT.say
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/0Q_gDoP_jx␤Variable '%h' is not declared␤at /tmp/0Q_gDoP_jx:1␤------> 3<BOL>7⏏5%h<x>.WHAT.say␤»
gfldex m: my %h; say %h<a>:exists; say %h.elems; say %h<a>; say %h.elems;
camelia rakudo-moar 770d10: OUTPUT«False␤0␤(Any)␤0␤»
awwaiid m: my %h ; %h<x>.WHAT.say
camelia rakudo-moar 770d10: OUTPUT«(Any)␤»
gfldex m: my %h; say %h<a>:exists; say %h.elems; say %h<a>; say %h.elems; say %h<a>.elems; 20:03
camelia rakudo-moar 770d10: OUTPUT«False␤0␤(Any)␤0␤1␤»
awwaiid Skarsnik: %h<a> returns Any. One of them :)
RabidGravy stmuk, all of mine have *comprehensive* pod I'm not sure about *good*
Skarsnik Oh make sense
geekosaur hm, could that :<x> be colliding with a unfortunate degenerate case of base-literals?
gfldex %h<a> is a list of Pair with one element. The value of that Pair happens to be undefined/false
geekosaur m: say :0<x>
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/K6184Wfysf␤Radix 0 out of range (allowed: 2..36)␤at /tmp/K6184Wfysf:1␤------> 3say :0<x>7⏏5<EOL>␤»
dalek c: 28e6394 | (Steve Mynott)++ | bin/p6doc:
teach p6doc about the existance of panda installed precomp ecosystem modules. Unlikely to work with zef
geekosaur maybe not 20:04
probably have to look at the grammar to see if it's falling through the cracks or something
RabidGravy stmuk++ # nice one, there's an actual point in putting POD in them now :)
gfldex awwaiid: where did you find that :<x> ? 20:05
20:05 psy_ left
stmuk I'm not very happy with how that works! There probably should be proper APIs. But I figure partly working hack is better than not working at all 20:06
awwaiid gfldex: AlexDaniel was playing with it 20:07
AlexDaniel gfldex: yeah, that's a typical result of my brain fart :)
awwaiid AlexDaniel: gfldex: I tried to suss it out looking at the AST of :<x> vs "x", but haven't gained any insight: diff -u <(perl6 --target=AST -e 'say "x"') <(perl6 --target=AST -e 'say :<x>') 20:08
20:08 psy_ joined
moritz awwaiid: the main difference seems to be that "x" is an allomorphic str/Str thingy, while :<x> is a pure Str 20:11
awwaiid moritz: ... interesting. Is there another way to make a pure Str?
moritz awwaiid: presumably through Str.new 20:12
20:14 darutoko left
ely-se Is the current way to make public constants to make them subroutines? our \blah = 1; is not yet supported by rakudo 20:16
Skarsnik ? 20:17
awwaiid moritz: looks legit... so is this a feature or implementation side-effect?
jnthn ely-se: constant blah = 1; # is our-scoped by default 20:18
Skarsnik m: module A { constant PIKO is export = "hello" }; import A; say PIKO;
camelia rakudo-moar 770d10: OUTPUT«hello␤»
ely-se jnthn: oh I see
I couldn't find "constant"
awwaiid ely-se: might want to use :: instead of .
20:20 Calibellus joined
moritz awwaiid: the :<x> not being allomorphic looks like an oversight 20:20
ely-se: or "our constant" if you want to be explicit 20:21
awwaiid moritz: but is it legit that :<x> is more or less "x", and is :<x> shorthand for something in this case? Or in other words, do I get to add another thing to my colon collection (thelackthereof.org/Perl6_Colons)? 20:25
20:25 llfourn joined
ely-se what is "Koalatee" in the module listing? 20:25
jnthn: thanks, that works :)
awwaiid moritz: and if it is a lovely additon to my colon collection, does the construct have a name? 20:26
20:27 Praise joined, Praise left, Praise joined 20:28 telex left
awwaiid ely-se: github.com/perl6/modules.perl6.org...oalatee.pm is the code for it -- in other contexts it is a somewhat-subjective score hopefully correlated to the module authors doing good stuff (subjective is why it isn't spelled "quality") 20:28
ely-se oh ok
moritz awwaiid: no idea
ely-se ah, very trivial stuff I see 20:29
moritz awwaiid: looks like the degenrate form of a colon pair to me
20:30 marmay_ left
nemo Heh. unc0rr just linked me to some code he wrote for our haskell game server that I thought might entertain you perl6 folks w/ your unicode-in-the-language 20:30
hg.hedgewars.org/hedgewars/file/tip...Glicko2.hs
20:30 llfourn left
awwaiid imagines a slang where latex equations are valid 20:32
ely-se I've always used Unicode in my code, for ASCII is a subset of Unicode. 20:33
20:33 grondilu left
awwaiid ely-se++ # profound 20:33
nemo heh
20:33 breinbaas joined
huf pretty much everyone ever has used unicode 20:33
moritz awwaiid: the Mathematica language has something very close, and some other scientific language whose name I forgot too
huf since the point of unicode is to contain everything
(and it's not an encoding)
moritz "One code to rule them all" 20:34
20:34 dfcarpenterak joined, grondilu joined
ely-se Next step is to figure out how to run an extra build step with META.info. 20:36
moritz ely-se: look at Inline::* modules as examples? 20:37
ely-se Hmm. I'm not sure I want it.
It's the numeric values of the <errno.h> constants.
They never change so I could just as well not have them generated.
20:38 telex joined, frankjh left
El_Che ely-se: famous last words :) 20:38
ely-se I vaguely remember those constants being on doc.perl6.org/ about a year ago 20:39
But they aren't there anymore.
arnsholt They're constants, until suddenly they're not anymore
(If your code moves to a different OS, f'rinstance)
ely-se POSIX error numbers will never change. It would break too much software.
They are also not OS-specific. POSIX defines them.
arnsholt Aren't there OS-specific extensions too, though? 20:40
ely-se I don't support them.
arnsholt Well, that solves it, yeah =)
ely-se looks like this now: github.com/rightfold/NativeCall-Er...m6#L14-L92
20:41 zakharyas joined 20:45 domidumont left
ely-se manually copied from generator output 20:45
Skarsnik what the hell is short unsigned int 20:49
20:49 kaare_ left 20:51 lucasb joined
geekosaur C lets you permute those 20:51
ely-se volatile const auto short unsigned int x = 0; 20:53
Skarsnik great I have wchar_t what the hell is this dumb header
ely-se wchar_t is an abomination.
lucasb wtfchar 20:54
nemo ely-se: amen
arnsholt Skarsnik: "Wide character"
From ).
some functions and attributes of the gc module behave in a slightly different way: for example, gc.enable and gc.disable are supported, but instead of enabling and disabling the GC, they just enable and disable the execution of finalizers.
Whoops. Mispaste
Damn tiny laptop keyboard
ely-se doesn't matter; still hilarious
nemo arnsholt: back from when people thought 16 bits would be enough for all chars - didn't count on CJK being quiiiite as big as it was
poor naive western designers
Skarsnik I am writing fundamental C type to perl6 table
arnsholt Yeah, what nemo said
ely-se 8 bit is enough 20:55
people should just learn English
Skarsnik so wchar_t will be... fund this
arnsholt From the days before Unicode was fully figured out
moritz ely-se: 7 bit is enough!
nemo ely-se: I feel sorry for all the poor kids in this area learning chinese as a second language. they are all feeling super confident at learning the accents and all that, then someone turns the script loose on them and crushes their tiny spirits
20:55 dfcarpenterak left
ely-se hmm, no equivalent for !eqv :( 20:55
nemo Funny essay. www.pinyin.info/readings/texts/moser.html
ely-se teach them Perl 6 20:56
arnsholt (Without reading it) is that "Why Chinese is so damn hard"? =)
nemo arnsholt: indeedy
'I couldn't remember how to write the character 嚔, as in da penti 打喷嚔 "to sneeze". I asked my three friends how to write the character, and to my surprise, all three of them simply shrugged in sheepish embarrassment'
'three Ph.D. students in the Chinese Department at Peking University, all native Chinese' 20:57
arnsholt I've been told the Japanese have a term for that: wopurobaka (literally "word processor stupid") for the phenomenon where people can recognize characters when reading, but not produce them
Skarsnik hm, what can I do with complex of C99 ? 20:58
alpha123 tried to learn chinese in grade school, spoken isnt bad but writing killed him
ely-se Skarsnik: what do you mean?
nemo arnsholt: at least the set of kanji is a lot smaller in japanese, and semi-avoidable while writing
alpha123 Skarsnik: complex number arithmetic and a few transcendental functions
Skarsnik how to translate the complex float to a perl6 type
ely-se complex numbers are useful in a variety of disciplines
20:59 lnrdo joined
Skarsnik I assume a complex float is just a fancy float? 20:59
alpha123 arnsholt: it's a pretty common phenominon in japan, kids aren't learning stroke order anymore (they just type hiragana or katakana and phones etc complete the kanji)
arnsholt Skarsnik: On a first pass, I'd mark them as TODO, and see if people have good ideas =)
alpha123 Skarsnik: if you don't know what it is, you probably don't care
nemo arnsholt: really, the only utility to kanji in japanese seems similar to variant spellings of same sound in french, to help in disambiguation a bit
20:59 [Sno] left
nemo although in french it is probably just due to sounds fading off over time 20:59
ely-se Skarsnik: a complex float is typically implemented as a pair of floats
Skarsnik I am writing something that produce Nativecall code from c header
nemo arnsholt: and I guess to look pretty
arnsholt Yeah, French spelling is a confluence of insanely archaic orthography and some annoying sound changes 21:00
Skarsnik I don't care the use I have of these type xD
ely-se struct complex { float real; float imag; };
arnsholt Much the same is true for English, really
alpha123 nemo: kanji actually mean a lot more than that, for example they sometimes convey respect and formality
Skarsnik ok, I will stick to returning NYI or something like that x)
mspo wouldn't a french word be written in katakana or romanji?
alpha123 yeah
nemo arnsholt: english is a whole different beast... a frankenstein stiched from german, french and a ton of other languages
[Coke] m: say complex;
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Jr7p08l6vD␤Undeclared routine:␤ complex used at line 1␤␤»
21:00 psy_ left
alpha123 nemo: but at least english verbs are great 21:01
[Coke] m: say Complex;
camelia rakudo-moar 770d10: OUTPUT«(Complex)␤»
ely-se You should give Finnish a try.
That's a weird language.
arnsholt That's vocabulary though. Completely different =)
alpha123 still firmly believes every language other than english and mandarin got verbs wrong
nemo my russian friends claim russian is a pretty rational language
ely-se Dutch is hte best language.
21:01 psy_ joined
alpha123 ely-se: screw dutch, messed up our english spelling 21:01
;)
nemo and a couple of my russian speaking french friends (fair # of french speak russian for some reason, and vice versa) agree 21:02
arnsholt nemo: Eh. Similar claims have been made for (off the top of my head): French, German, Latin and Greek
nemo perhaps due to a history of mutual invasions
arnsholt It's all bollocks =)
nemo arnsholt: latin is just hilarious
alpha123 I only know a bit of russian but it seems nice. At least they spell mostly phonetically
nemo arnsholt: every time a short latin phrase comes up online on some latin forum, and someone asks for the "correct" text, you get a forum thread like 20 posts long
alpha123 lol
arnsholt Latin isn't too bad, TBH. The morphological system is really quite regular, due to it being radically restructured relatively recently 21:03
nemo I ran into this when looking up that latin from that space horror movie
the one that is kinda like warhammer
21:03 revhippie joined
nemo Event Horizon 21:03
latindiscussion.com/forum/latin/lib...eris.7499/ 21:04
I stand corrected. only 18 responses
diakopter spoiler alert: there is blood 21:05
nemo alpha123: eh. I like the flexibility of tenses in other languages. more expressive
arnsholt *tutemet is the correct form of the pronoun, if you care =p
nemo alpha123: there have been times when the limitations of english have kind of annoyed me
alpha123 nemo: english has a bunch of tenses and moods and aspects too, they're just not used much and we don't have 20 versions of each verb because of it
for as weird as english is, at least we have non-gendered nouns and comparatively simple verb forms 21:06
21:06 lokien_ joined
geekosaur there is no such thing as a regular language. if it gets used, it sprouts irregular "conveniences" for the frequent speaker. 21:06
nemo alpha123: well, the number of tenses in latin-derived languages is pretty insane, but most of 'em aren't used, and the ones that are used are pretty easy to remember I think
geekosaur: lojban/
?
alpha123 geekosaur: HTML?
;D
nemo I considered trying to teach my kid lojban but was 1) too lazy and 2) told it would be child abuse 21:07
geekosaur I shoul hae qualified that with "natural" laguage. lojban etc. don't really count since they are artifically structured
alpha123 nemo: lol
that would be interesting at least
geekosaur (also anyone who thinks HTML is regular should try to parse actual web pages >.> )
alpha123 geekosaur: that was rather tongue-in-cheek, referencing the stackoverflow HTML IS NOT A REGULAR LANGUAGE post 21:08
geekosaur there's a reason the most popular html parser is called "tag soup"
ely-se it's offensive towards soup
soup is unlike HTML: good
alpha123 the html5 parsing algorithm is a minor disaster
nemo alpha123: did you ever see the regex on stackoverflow for HTML?
alpha123 nemo: No, was it PCRE?
because standard regexes can't do html 21:09
nemo alpha123: stackoverflow.com/a/5233151 ♥
well. that's just XML
but came from an XHTML question on stackoverflow 21:10
so there's some overlap
and still funny
geekosaur surprised nobody added CTHULHOID TENTACLES to unicode just for the benefit of that post
alpha123 lol
ely-se In Perl it's easier
just embed a block that calls into an HTML parser
alpha123 ely-se: well perl regexes aren't regular anyway
ely-se neither are .NET's
alpha123 what even are perl regexes, LR(0)?
[Coke] (;,;) 21:11
nemo alpha123: pastebin.com/HUmPzP5M the pastebin unpack from that post in case you're feeling lazy or TLDR
diakopter [Coke]: XD
[Coke] /|\(;,;)/|\ . o O (zzz)
alpha123 nemo: bWVuLCBpIHNwZWFrIGJhc2U2NCBuYXRpdmVseQ==
nemo alpha123: heh. I piped that to xxd in case you were being evil 21:12
but, yeh, perl -e 'print unpack "B*","hi"' - old silly forum favourite 21:13
alpha123 perl6 grammars are really awesome, even though PEG is not really my favorite algorithm
Skarsnik libxml2 is fun... gist.github.com/Skarsnik/c4b3bd674fbb407d9d0e 1068 functions xD 21:14
nemo hm. does that line of perl5 still work in perl6?
m: print unpack "B*","hi"
camelia rakudo-moar 770d10: OUTPUT«5===SORRY!5=== Error while compiling /tmp/gW0xGUtMZH␤Undeclared routine:␤ unpack used at line 1. Did you mean 'pack'?␤␤»
nemo oh well
21:15 cdg joined, lucasb left
alpha123 I hope something like grammars becomes more standard in languages in the future. My own language (which is a bit like an illegitimate child of Perl 5 and Haskell) ~~stole~~ adopted the idea 21:16
AlexDaniel actually, Russian is just insane. You might like it if only you don't mind that the same thought expressed in English is twice shorter than in Russian… 21:18
21:18 maslan joined
alpha123 AlexDaniel: English is pretty efficient in general, but 2x seems like a stretch 21:18
AlexDaniel and that its alphabet barely fits into your keyboard layout
21:20 gtodd left
AlexDaniel alpha123: sure, I am exaggerating a bit 21:21
alpha123 also russian has a vastly more expressive array of profanity than english :p
21:22 dfcarpenterak joined, geraud joined, sno joined 21:23 dfcarpenterak left, dfcarpenterak joined 21:24 llfourn joined
gfldex nemo: Date: Wed Dec 23 22:31:59 2015 +0100 21:24
Make pack and unpack experimental
21:25 _mg_ joined
AlexDaniel alpha123: yes, what an advantage. Not sure where would you use it though… 21:25
gfldex russians use that advantage at every opportunity
21:28 molaf left
gfldex i used to play eve-online quite a lot. One day I had a russian cussing at me for 15 minutes. At least 25 individual insults. I was impressed. 21:28
21:29 llfourn left
nemo AlexDaniel: I understand where russian truly shines, is profanity 21:30
en.wikipedia.org/wiki/Mat_%28Russi...ofanity%29
"Mat has thousands of variations but ultimately centers on four pillars"
21:31 FROGGS left
nemo AlexDaniel: www.russki-mat.net/page.php?l=RuEn&a=%D0%95 21:32
21:42 YP-QMUL-W joined 21:44 psy_ left
Skarsnik right side of 'and' or '&&' is never evaluated if the left is false? 21:47
gfldex they short circuit 21:48
Skarsnik because I have
say $t.ref-type ~~ FundamentalType;
return 'Str' if ($t.ref-type ~~ FundamentalType and $t.ref-type.name eq 'char') ||
and I get Method 'name' not found for invocant of class 'GPT::Class::PointerType' on the second line 21:49
and it say false
So I am really confused now x) 21:50
21:54 colomon left, nadim joined, cognominal joined 21:55 _nadim left
Skarsnik and obviously I only have this bug on a huge ass file 21:58
21:58 xpen joined 22:00 skids left 22:01 leont joined, prammer left
dalek c: ac1285c | (Steve Mynott)++ | bin/p6doc:
-l now lists panda installed modules as pod candidates
 22:01
22:03 xpen left, Ven left 22:04 espadrine joined 22:07 Guest70706 is now known as parisba__, Skarsnik left, AlexDaniel left
timotimo you may want .^name on that? 22:11
or does FundamentalType have an actual name method? 22:12
i didn't see skarsnik leave :\ 22:14
22:15 rindolf left 22:18 colomon joined 22:21 llfourn joined 22:28 _mg_ left 22:29 zengargoyle left, pmurias joined
pmurias re koalatee it seems to be broken as it gives a travis score for a module with unknown travis status 22:30
gfldex there may be caching involved 22:31
pmurias I think it would be great to either document it or just remove it
22:31 AlexDaniel joined
AlexDaniel nemo: what a good list 22:31
pmurias by document it I mean is that the module.perl6.org page doesn't mention what it is
22:32 yqt joined
RabidGravy right no more fixing software for the day 22:32
22:33 zakharyas left, zengargoyle joined 22:34 musiKk left
AlexDaniel nemo: of course there are much more. People come up with these words on the fly, so honestly I haven't thought that someone will attempt to create a list like this… 22:34
22:35 ely-se left 22:36 n0tjack left 22:37 RabidGravy left, n0tjack joined
dalek kudo-star-daily: a142780 | coke++ | log/ (5 files):
today (automated commit)
 22:38
22:45 n0tjack left
nemo AlexDaniel: I heard some russian artist is making like a 6 volume encyclopedia set of mat 22:48
kinda like as an art piece I guess
ah. 12 volumes 22:49
'The Russian linguist (and protest art guru) Alexey Plutser-Sarno is working on a dictionary of Russian curses.'
www.quora.com/Which-are-the-best-l...o-swear-in (from here)
mspo can I use rakudo directly and then moar to execute directly? 22:50
like a traditional compile/run cycle
alpha123 mspo: why
mspo alpha123: I think I'd prefer it
flussence that's how the core setting is done, I don't know how well it'd work for other code
mspo alpha123: especially if I can eventually create compiled things to deploy 22:51
alpha123: which, for example, also package their dependencies
alpha123 mspo: I don't think MoarVM bytecode is portable across platforms or Moar versions 22:52
mspo that's fine
alpha123: I'm not a fan of artifacts getting autocreated here and there 22:53
pyc files are bad news :)
llfourn mspo: perl6 --target=mbc --output=foo.moarvm foo.p6 # worked last time I tried it
though rakudo is doing this for you and putting it into .precomp 22:54
mspo ~/.perl6/.precomp ? 22:55
llfourn mspo: depends. If you are using a lib like perl6 -Ilib foo.p6 it will put it in lib/.precomp 22:56
it won't compile foo.p6 -- just the dependencies in lib
precompile*
if you have an installed module it will probably be precompiled in .perl6 22:57
22:58 colomon left
pmurias mspo: you are a fan of Makefiles? or just having them autocreate in a single place? 23:00
23:01 AW3i left 23:03 [Tux] left
awwaiid PolyConf call for papers is open (polyglot programming conference in Poland) -- eventil.com/events/polyconf-16 . Would be great to have some p6 representation :) 23:04
23:08 lichtkind left
b2gills .ask jnthn wouldn't Seq.Numeric always be the same as Seq.elems? 23:10
yoleaux b2gills: I'll pass your message to jnthn.
23:11 leont left 23:15 [Tux] joined 23:17 skids joined 23:21 cpage_ left
AlexDaniel nemo: and sure enough we have some profanity in perl6: github.com/Skarsnik/acme-wtf 23:26
23:26 vendethiel left
AlexDaniel nemo: though it would have been nicer if it accepted all sorts of “fuck” words as “die” alternatives 23:28
23:28 kid51 joined
pmurias I get a an error while doing rakudobrew build-panda 23:31
it seems to have disappeared 23:32
nemo AlexDaniel: I'm disappointed it isn't licensed under the DWTFYW
AlexDaniel nemo: WTFPL you mean? 23:33
nemo oh right
nemo forgot the name
23:33 maslan left 23:35 YP-QMUL-W left 23:43 TEttinger joined 23:45 leont joined
pmurias ahh, an old version of shell command was causing the error 23:48
23:49 xdoctor joined 23:51 nadim left 23:54 colomon joined 23:55 n0tjack joined 23:59 xpen joined, lokien_ left

Powered by Cro and the Raku® Programming Language.

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

Thank you!