»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.
00:00 chenryn left 00:07 cdg left 00:08 cdg joined
TEttinger .u SELFIE 00:10
yoleaux No characters found
00:10 RabidGravy left 00:12 cdg left 00:17 skingsbu left 00:21 Ven left
geekosaur not till the next unicode update :p 00:21
00:22 BenGoldberg joined, Zoffix left 00:23 BenGoldberg left, Zoffix joined 00:24 BenGoldberg joined 00:29 BenGoldberg left 00:30 BenGoldberg joined 00:32 BenGoldberg left 00:33 BenGoldberg joined 00:38 BenGoldberg left 00:40 BenGoldberg joined 00:42 BenGoldberg left 00:43 BenGoldberg joined 00:44 BenGoldberg left 00:45 BenGoldberg joined
skink Is there an official way to make arbitrary syscalls yet? 00:53
00:56 BenGoldberg left 00:57 yqt left, BenGoldberg joined 00:59 kurahaupo left
BenGoldberg skink, You can call arbitrary library functions using NativeCall. 01:03
skink BenGoldberg, It's not in a library file :) 01:04
01:04 BenGoldberg left 01:05 Zoffix__ joined 01:06 BenGoldberg joined 01:07 Zoffix__ is now known as sexy-coder-girl
skink BenGoldberg, It's not in a library file :) 01:09
01:09 Zoffix left 01:10 sexy-coder-girl is now known as Zoffix, Zoffix left, Zoffix joined
BenGoldberg How would you perform this syscall from a C program? 01:10
skink man7.org/linux/man-pages/man2/getrandom.2.html 01:11
01:11 kurahaupo joined
geekosaur syscalls can't really be invoked portably 01:12
01:12 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg 01:16 kurahaupo_ joined, kurahaupo` joined 01:17 espadrine left, BenGoldberg left 01:18 kurahaupo left 01:19 BenGoldberg joined
skink geekosaur, I already separated OS-specific functionality into separate modules 01:19
01:20 kurahaupo_ left
geekosaur man7.org/linux/man-pages/man2/syscall.2.html 01:25
01:28 huggable left
skink Disgusting :) 01:28
01:28 huggable joined, kurahaupo` left 01:32 _4d47 left 01:33 Zoffix left 01:34 TimToady left 01:35 Zoffix joined, BenGoldberg left, jdv79 left, ggoebel115 left, Celelibi left, jdv79 joined 01:36 remmie left, ggoebel115 joined, BenGoldberg joined 01:37 Praise left, pierrot_ joined, BenGoldberg left 01:39 BenGoldberg joined, Praise joined, Praise left, Praise joined, BenGoldberg left 01:40 TimToady joined 01:41 BenGoldberg joined, Celelibi joined, pierrot left, skink left, tbrowder joined
tbrowder back again, ref pod and docs 01:42
01:42 BenGoldberg left 01:43 BenGoldberg joined, araujo_ left 01:46 ilbot3 left, remmie joined
tbrowder pmurias: current p6 pod is not handled completely, e.g., tables. after loking at some other markup langs again, I think, given the present state of things, p6 pod (and tools) is okay but needs more attention. I have just submitted an e-mail to the perl6 users presenting my simple take on my impression of p6 pod and some possible projects. I would appreciate any commennts on that thread. 01:47
01:47 BenGoldberg left, ilbot3 joined
tbrowder bye all 01:48
01:48 tbrowder left, kurahaupo joined, BenGoldberg joined 01:49 BenGoldberg left 01:50 BenGoldberg joined 01:51 molaf left
Zoffix Any response to this FB thread? www.facebook.com/groups/1595443877...%22R%22%7D 01:55
"Looks like C++ is not supported yet. Is that correct?" "Yep, thats what I was referring to. It did not work for me. May be it is work in progress?"
01:55 BenGoldberg left 01:57 BenGoldberg joined, kanishka left 01:59 ssotka left 02:04 molaf joined 02:05 jack_rabbit joined, kurahaupo left, noganex joined 02:08 noganex_ left 02:11 kid51 left
tailgate are there examples of how to make a class/static method in a perl6 class? 02:12
02:12 dimon__ joined
dimon__ hey, what's the difference between "unit module Aaa;" and "unit package Aaa;" ? 02:13
geekosaur tailgate, our method foo(...) { ... }
02:19 BenGoldberg left 02:20 dimon__ left, BenGoldberg joined 02:21 dimon__ joined
dimon__ hey, what's the difference between "unit module Aaa;" and "unit package Aaa;" ? 02:21
02:21 BenGoldberg left 02:22 BenGoldberg joined 02:23 pierre_ joined
Zoffix Goood question. 02:25
The speculation on the subject seems a bit outdated: "package Foo; ... This form is illegal in a Perl 6 file. If you wish to have a file-scoped package, either use the brace form or declare it with the module keyword instead." 02:26
(and we now have `unit` keywords)
02:29 pierre_ left
dimon__ Zoffix: I see. 02:29
And if I have file1.pm6 and file2.pm6 in my library MyComp::MyLib, what should be the 1st on those file?
unit module MyComp::MyLib;
unit module MyComp::File1Class;
assuming that
I define those classes as 02:30
within those 2 modules
02:30 BenGoldberg left
dimon__ correction: class MyComp::MyLib::File1Class { .... } 02:30
class MyComp::MyLib::File2Class { .... }
sorry for the typos 02:31
02:31 BenGoldberg joined 02:32 jack_rabbit left 02:35 BenGoldberg left 02:36 BenGoldberg joined 02:47 BenGoldberg left 02:48 BenGoldberg joined 02:53 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg 02:55 dimon__ left 02:56 BenGoldberg left 02:58 BenGoldberg joined 02:59 BenGoldberg left 03:00 BenGoldberg joined 03:01 |2701 left 03:08 cdg joined 03:12 cdg left 03:15 BenGoldberg left 03:16 BenGoldberg joined 03:17 BenGoldberg left 03:18 BenGoldberg joined 03:19 BenGoldberg left 03:21 BenGoldberg joined 03:22 ssotka joined 03:27 BenGoldberg left, dimon_ joined
dimon_ when I'm using a file-class declaration "unit class MyModule::Class1", how can I write "does RoleX" for Class1? 03:28
03:28 BenGoldberg joined
timotimo m: unit class FooBar; also does Positional; 03:29
camelia ( no output )
dimon_ timotimo: and if I need to export the module where "Positional" is defined? say, it's defined in MyModule 03:31
then this won't compile also does Positional; 03:32
timotimo export? you mean import?
you can put "use MyModule" in front
dimon_ neither this also does MyModule2::Positional;
I mean, MyModule2, not Module for Positional
timotimo i'm not sure how you mean that
dimon_ I mean, this won't compile "; also does MyModule2::Positional;" 03:33
timotimo so why not put "use MyModule2" in between?
03:33 khw left
timotimo oh, you mean you want Positional to be available as a member of the unit class' package? 03:33
dimon_ tnx
timotimo you'd have to do something like "my \Positional is export = Positional" 03:34
or something like that
dimon_ nope, I meant how to "use" MyModule2::Positional
before "also does MyModule2::Positional"
timotimo the thing after "use" is the file you want rakudo to look for in the filesystem
so if you want to get MyModule2::Positional, you'd usually "use MyModule2" 03:35
03:36 BenGoldberg left 03:37 BenGoldberg joined
dimon_ tnx 03:38
03:38 dimon_ left, BenGoldberg left 03:40 BenGoldberg joined, dimon_ joined, pierre_ joined
dimon_ by the way, if my library is available in "panda", and if I update its source code at github, will it be updated in panda automatically? 03:41
03:41 BenGoldberg left
dimon_ does panda use code from github of a package? 03:41
timotimo panda always pulls the code directly from github, yeah
dimon_ ok
03:41 dimon_ left
timotimo we will have something different in the future that's more similar to cpan and PAUSE 03:41
03:42 BenGoldberg joined 03:43 BenGoldberg left 03:45 BenGoldberg joined 03:48 BenGoldberg left 03:50 BenGoldberg joined 03:51 BenGoldberg left 03:52 BenGoldberg joined 03:53 pierre_ left 03:54 BenGoldberg left 03:55 BenGoldberg joined 03:56 BenGoldberg left 03:57 BenGoldberg joined 03:58 BenGoldberg left 04:05 ssotka left 04:06 BenGoldberg joined 04:07 BenGoldberg left 04:08 BenGoldberg joined 04:10 BenGoldberg left, mtj_ left 04:11 BenGoldberg joined 04:14 BenGoldberg left, mtj_ joined 04:15 BenGoldberg joined 04:17 BenGoldberg left 04:18 BenGoldberg joined 04:27 BenGoldberg left 04:32 BenGoldberg joined 04:39 dmc00 left 04:48 BenGoldberg left 04:50 Cabanossi left 04:51 Cabanossi joined 04:53 espadrine joined 04:54 BenGoldberg joined, pierre_ joined 04:58 BenGoldberg left 05:00 pierre_ left
dalek line-Perl5: 0dbc189 | (Zoffix Znet)++ | README.md:
Newer version of P5 available now
line-Perl5: 9d70afd | lizmat++ | README.md:
Merge pull request #62 from zoffixznet/patch-1

Newer version of P5 available now
05:03 wamba joined, BenGoldberg joined 05:04 mr-foobar left 05:05 BenGoldberg left, dmc00 joined 05:07 dmc00 left 05:10 BenGoldberg joined 05:14 dmc00 joined, CIAvash joined 05:16 BenGoldberg left 05:21 dmc00 left 05:22 BenGoldberg joined 05:23 BenGoldberg left 05:28 huggable left, huggable joined 05:29 BenGoldberg joined 05:31 BenGoldberg left 05:36 BenGoldberg joined 05:39 BenGoldberg left 05:44 BenGoldberg joined 05:47 kaare_ joined 05:49 BenGoldberg left 06:01 domidumont joined 06:06 domidumont left 06:07 domidumont joined 06:21 spider-mario joined 06:23 grondilu joined 06:26 lizmat joined 06:27 espadrine left 06:32 kaare_ left 06:42 telex left 06:44 telex joined, wamba left 06:53 lizmat left 06:59 jack_rabbit joined 07:07 CIAvash left 07:16 molaf left 07:18 CIAvash joined 07:57 firstdayonthejob joined 08:10 RabidGravy joined 08:15 darutoko joined 08:16 kaare_ joined 08:20 Ven joined 08:30 grassass left 08:36 rindolf joined 08:38 grassass joined 08:41 pierre_ joined, sortiz left 08:44 Emeric joined 09:03 pierre__ joined, pierre_ left, Ven left 09:06 cpage_ joined
masak good almost-noon, #perl6 09:21
09:22 espadrine joined, buharin joined
RabidGravy our cat decided it was time for us to get up an hour and a half ago 09:22
buharin hiho
09:27 lue joined 09:28 huggable left, huggable joined, espadrine left, ShimmerFairy left, lue is now known as ShimmerFairy
grondilu just adding a 'BEGIN say "hi";' in one of my modules make it fail. Go figure. 09:33
can't reproduce it with a one liner though. 09:34
m: module A { our @a is export = rand xx * }; import A; say "ok";
camelia rakudo-moar ad8265: OUTPUT«ok␤»
grondilu m: module A { our @a is export = rand xx * }; BEGIN say "hi"; import A; say "ok";
camelia rakudo-moar ad8265: OUTPUT«hi␤ok␤»
grondilu something like that
it's for my clifford module (github.com/grondilu/clifford). I was trying to do some precomputation with a BEGIN block but even the simplest BEGIN generates: ==SORRY==Variable '@e' is not declared 09:36
also after this error removing the BEGIN say "hi"; is not enough. I have to manually remove the .precomp directory. 09:39
very LTA
RabidGravy grondilu, iirc doing IO in a module that may be precompiled causes badness
grondilu aren't all module susceptible to be precompiled? 09:40
RabidGravy that is if the IO will happen at compile time
grondilu oh
hang on
grondilu tries with 'BEGIN sleep 1;' 09:41
nope, BEGIN sleep 1 causes the same error and requires manual remove of .precomp just the same. 09:42
grondilu wonders if BEGIN {} would just be ignored. 09:43
no, it's not ignored and it causes the error as well. 09:44
09:45 CIAvash left 09:47 buharin left, cpage_ left 09:49 CIAvash joined 09:50 zakharyas joined
grondilu :q 09:50
but seriously I have hard times figuring why adding 'BEGIN {}' in a module could make any difference whatsoever. 09:54
09:54 TEttinger left, skingsbu joined
dalek rl6-most-wanted: b49b9f9 | titsuki++ | most-wanted/modules.md:
Fix a Text-Tabs-Wrap link
rl6-most-wanted: 53ac1e0 | titsuki++ | most-wanted/modules.md:
Merge pull request #31 from titsuki/fix-not-working-link

Fix a Text-Tabs-Wrap link
masak grondilu: I'd be quite interested to see a golf of what you're observing.
RabidGravy yeah I deeply suspect there is something else going on
llfourn writing to stdout during compile time used to be a problem but I thought it was more or less fixed now 09:57
09:57 kaare_ left
llfourn I always use BEGIN note "recompiling"; for that reason 09:58
grondilu masak: so would I. I've tried to make one but I failed. 10:01
I could try harder I guess.
10:01 pierre__ left 10:02 pierre_ joined
grondilu last time I tried I gave up when I realized the code I was executing was not the code I had typed because of the .precomp issue. 10:02
m: BEGIN say "compiled on " + now 10:08
camelia rakudo-moar ad8265: OUTPUT«5===SORRY!5=== Error while compiling /tmp/GRX4YX4MEy␤An exception occurred while evaluating a BEGIN␤at /tmp/GRX4YX4MEy:1␤Exception details:␤ Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5compile…»
10:08 pmurias joined
grondilu silly me 10:08
m: BEGIN say "compiled on " ~ now
camelia rakudo-moar ad8265: OUTPUT«compiled on Instant:1463306963.554297␤»
grondilu ^I guess I could use somethink like that to check that my code is up-to-date or something. 10:09
dogbert17 good day #perl6 10:10
pmurias dogbert17: hi
dogbert17 pmurias: hi 10:11
noob question of the day, what does this syntax mean: my class X::Seq::Consumed { ... } # from src/core/Seq.pm
grondilu it's the standard way of making an exception IIRC 10:12
10:12 cpage_ joined
dogbert17 shouldn't ther be an 'is Exception' in there somewhere? 10:13
grondilu I think it was not considered necessary 10:14
there is no Exception class in Perl 6 10:15
m: say Exception
camelia rakudo-moar ad8265: OUTPUT«(Exception)␤»
grondilu oh there is
my bad
I don't know then. 10:16
RabidGravy dogbert17, that is a "forward declaration" 10:18
bartolin dogbert17: I think that is a stub definition for that class, which get fully defined in src/core/Exception (and there you get 'is Exception' as you expected) 10:19
RabidGravy it's actually defined in src/core/Exception.pm
dogbert17 RabidGravy++ bartolin++ many thanks
grondilu++ as well 10:20
RabidGravy it can do it like that (in separate files) because of the way that the CORE setting is compiled
dogbert17 aha, the reason for asking is that I want to try to fix github.com/perl6/doc/issues/515 10:21
hadn't seen the syntax before and thought, for some reason it meant 'not yet implemented' 10:22
RabidGravy yeah I was going to respond to that, I think the time is very soon that the exceptions should have their own section in the docs, on the grounds that if they are *all* documented it will swamp all the other types
dogbert17, it means "not implemented here" :) 10:23
or "yet"
dogbert17 RabidGravy: :) 10:24
and I'm guilty as charged :), have added several 'missing' exceptions to the docs in the last few days 10:25
RabidGravy that is to say that if by the time the whole thing is finished compiling there isn't a complete definition a "he following packages were stubbed but not defined" will be thrown
arnsholt In user code, you'd end up using that syntax if you have two types that depend on each other
10:26 araujo joined
arnsholt Like "class B {...}; class A { has B $.b; }; class B { has A $.a }" 10:26
bartolin dogbert17: you can read more about stub declarations in the speculations: design.perl6.org/S06.html#Stub_declarations
arnsholt (And no, that can't be split into two separate files)
dogbert17 thx all, very interesting discussion 10:27
10:27 aries_liuxueyang left 10:29 aries_liuxueyang joined
RabidGravy the thing with "document all the exceptions" is that there are greater than 250 defined in src/core/Exception.pm which is roughly more than the rest of the types 10:30
so a) the types page becomes rather large and b) they are mostly not all that interesting except in the context of *where* they are thrown 10:31
so having them in a separate section would prevent a) and allow the linking from the documentation for the things that throw them 10:33
dogbert17 RabidGravy: maybe I should hold off then until we have a solution ? 10:34
RabidGravy I'd carry on to be honest 10:35
dogbert17 RabidGravy: ok will do :) 10:36
RabidGravy I'd say that people agreeing will take forever or until someone just does it and it's better that things are documented, it can be re-arranged later
dogbert17 RabidGravy: how about =SUBTITLE 'Error due to trying to reuse a consumed sequence' is that intelligible or just bad english? 10:39
10:44 buharin joined 10:45 Actualeyes joined 10:47 kaare_ joined
gfldex the optimisations of the last 2 weeks have shaved off 13% execution time of my pod renderer 10:55
you totally improved my productivity!
11:01 Roamer` left 11:02 fildon joined
RabidGravy yay! 11:10
dalek c: 9728090 | (Jan-Olof Hendig)++ | / (2 files):
Added docs for X::Seq::Consumed
RabidGravy yay!
dogbert17 :)
yesterday an odd thing happened when writing some docs, after having run htmlify.p6 my changes weren't there, i.e. in the generated html files. Had to nuke the 'precompiled' directory in order to make it work! 11:15
teatime waves g'morning.
dogbert17 o/ teatime 11:16
11:17 kid51 joined 11:18 kaare_ left 11:19 [TuxCM] joined 11:23 zakharyas left 11:27 pierre_ left 11:32 mr-foobar joined 11:36 nightfrog left 11:40 grassass left
llfourn hmm I have to keep deleting .precomp because I get "Missing or wrong version of dependency" often. Seems to have only started today. 11:46
I will see if I can bisect it 11:47
11:50 kurahaupo joined
llfourn seems to happen on 2016.03 as well o.o 11:56
RabidGravy :-\ 11:57
llfourn not using git so hard to know what I did to trigger it.
12:00 snarkyboojum joined 12:01 snarkyboojum left 12:04 mr-foobar left
Zoffix I was getting that when I installed a module and stuff depending on it and then got a local copy of the module to hack on. So basically the installed, precompiled modules were precompiled against the "old" installed module, not my local copy. 12:06
llfourn hmmm that interesting because a few of the compunits in what I'm working on do have the same name as some that are installed 12:07
Zoffix And I reported it: rt.perl.org/Ticket/Display.html?id=128088
12:08 snarkyboojum joined 12:12 pierre_ joined
tailgate if I have ((a -1) (e -5) (d -4) (c -3) (b -2)) how do I make it a hash of the form a => -1 e => -5 etc? 12:12
12:13 mr-foobar joined
llfourn tailgate: I'm not sure what you have there but try .hash 12:13
if it's a list of lists which is what it looks like you may have to flatten them first
12:13 _4d47 joined
llfourn m: ( <a,b>,<c d> ).perl.say 12:14
camelia rakudo-moar ad8265: OUTPUT«("a,b", ("c", "d"))␤»
llfourn ...
m: ( <a b>,<c d> ).perl.say
camelia rakudo-moar ad8265: OUTPUT«(("a", "b"), ("c", "d"))␤»
llfourn m: ( <a b>,<c d> ).hash.say
camelia rakudo-moar ad8265: OUTPUT«{a b => (c d)}␤»
llfourn m: ( <a b>,<c d> ).map(|*).hash.say 12:15
camelia rakudo-moar ad8265: OUTPUT«{a => b, c => d}␤»
llfourn ^ seems to do the trick
m: hash( (<a b>,<c d>) ).say 12:16
camelia rakudo-moar ad8265: OUTPUT«{a => b, c => d}␤»
llfourn ^ better
tailgate ah, I was using .Hash 12:17
llfourn m: %( (<a b>,<c d>) ).say # also works I think
camelia rakudo-moar ad8265: OUTPUT«{a b => (c d)}␤»
llfourn m: %( |(<a b>,<c d>) ).say # or this perhaps?
camelia rakudo-moar ad8265: OUTPUT«{a b => (c d)}␤»
llfourn ..nevermind 12:18
.Hash should work too I think
m: ( <a b>,<c d> ).map(|*).Hash.say
camelia rakudo-moar ad8265: OUTPUT«{a => b, c => d}␤»
llfourn at least there it does
12:19 snarkyboojum left 12:20 CIAvash left, [Tux] joined
tailgate what do | and |* mean? 12:21
llfourn | means turn the thing after it into a Slip ( an auto-flattening list )
* in this case means $_, which means each item in the map
so .map(|*) is short for .map({ |$_ }) 12:22
the problem is without it the .Hash gets two Lists as arguments where as what you want it four elements which will then pair up as key-values. 12:23
s/want it/want is/
12:33 pierre_ left 12:35 [TuxCM] left 12:46 mr-foobar left
tailgate thanks llfourn 12:47
llfourn np :)
12:47 mr-foobar joined 12:52 pierre_ joined 12:56 kanishka joined 13:00 Ven joined
Zoffix (╯°□°)╯︵ ┻━┻ 13:06
zef fails to install Inline::Perl5 with a non-descript error... tried with panda and it fails while trying to install prereqs because they're already installed -_- 13:07
timotimo hey zoffzoff
Zoffix hey
13:11 kaare_ joined
Zoffix Oh, pffftt... I forgot to perlbrew switch to proper perl, that's why Inline::Perl5 is failing. panda++ (actually showing the error) 13:13
13:14 CIAvash joined
teatime I am curious about how exactly Inline::Perl5 works / is implemented. 13:15
Zoffix I think it uses NativeCall to call into Perl5's C lib that handles XS.
RabidGravy yeah, it's something like that 13:18
teatime cool. 13:21
timotimo it might be easier to understand how Inline::Python works. i bet the python API is simpler than XS is :P
RabidGravy well the P5 "embedding" API is quite well documented perldoc.perl.org/perlembed.html 13:22
13:24 domidumont left 13:28 huggable left, huggable joined 13:29 Zoffix left 13:30 Zoffix joined
Zoffix What's a good monospace font that has good unicode support and doesn't look like shit? 13:30
RabidGravy just in case anyone is on a "search and document" mission for exceptions I've created github.com/perl6/doc/issues/517 to cover off the rest of them as there are loads
teatime Zoffix: PragmataPro, but it's not free, and actually is kindof expensive IMO. 13:31
Zoffix :(
I like free
13:31 kaare_ left
teatime Zoffix: Are you more interested in various scripts, or in Latin (and maybe Greek, Cryllic) + *symbols* ? 13:31
timotimo and emoji!
Zoffix Well, I'm interested in not seeing boxes with numbers in them: ¯\_(ツ)_/¯(╯°□°)╯︵ ┻━┻ 13:32
nine I think the basic parts of Inline::Perl5 and Inline::Python are equally easy to read
teatime Zoffix: Is this (primarily) for use in a terminal emulator (like xterm or rxvt-unicode)? 13:33
Zoffix This is what I see with the above: i.imgur.com/0G0HF0l.png
timotimo Zoffix: so no support for those boxed numbers emoji?
Zoffix I have 'Monospace' font that seems to have better support than Deja Vu, but it looks atrocious. 13:34
timotimo as a kid i loved "fixedsys" 13:35
but unicode wasn't so much a thing back then :P
teatime if you were on Linux, "monospace" would be a, like, "virtual" font, and inside the (terribly over-complicated) fontconfig config files (which you can customize via iirc ~/.fonts.conf), there would be defined a primary font and then a series of fallback fonts used to implement it. but I have nfc how that works on windows. 13:37
(it wouldn't surprise me if, for once, the Windows way to accomplish the same thing was actually more sane than the Linux way)
mst oh gods fonts 13:38
timotimo fonts are a Mighty Sucky Thing, aren't they, mst? :) 13:40
geekosaur I would suspect windows does much the same but through the registry and using uuids
13:42 Su-Shee joined
Su-Shee good afternoon. who's the person to talk to about the buildscript of rakudo-star? 13:42
Zoffix \o 13:43
Su-Shee Zoffix: is that a "it's me"? :)
Zoffix That's a hello :)
Su-Shee Zoffix: :) 13:44
Zoffix I dunno if there's one person. Just ask the question, I guess.
There's also #perl6-release, I believe.
Su-Shee Zoffix: too long, I just wanted to prepare said person(s) for a well not quite bug report, but for some buildscript cleanup thing mr shee is on the way of submitting to github...
CIAvash Zoffix: In my editor, it's falling back to DroidSansFallback to show those characters 13:45
Su-Shee Zoffix: or, rather not too long "something with DESTDIR support is broken" - but what exactly is too long :))
13:45 jameslenz joined
Zoffix :) 13:45
mst Su-Shee: heh, DESTDIR fixes are what got my first patch in 13:46
Su-Shee he tried to fix stuff, but after one fix the next problem came up so he stopped ;)
mst also didn't somebody create a perl6-dev channel for rakudo hacking?
timotimo we have #p6dev
mst oh for fuck's sake that's why I couldn't find it
can people please try and get the naming right? 13:47
mst sobs
timotimo right, it deviates from the ... :D
oh my
geekosaur iirc the release script was on a "this quick hack worked for me, if it breaks for you then you get to keep the pieces" basis
Zoffix +1 I too couldn't find it the first time for the same reason
RabidGravy there there, have a ponie
Su-Shee mst: I'm not quite up2date, I essentially just politely wanted to warn $person or if it is somebody I know, directly point person A to person B. ;)
13:47 Zoffix left
mst Su-Shee: oh, yeah, totally 13:48
Su-Shee geekosaur: well it DOES build things, but it has quirks making it hard to make proper packages with the use of DESTDIR. that's all.
13:48 Zoffix joined
Su-Shee I only know because I just made mr shee shorten the 5km long report. ;) 13:49
mst lol
timotimo Zoffix: perl6.party runs on rakudo? it seems surprisingly snappy compared to last we spoke
mst I don't think it does anymore
there was a memory leak, so he rewrote everything
Su-Shee mst: yeah well he was in the middle of explaining what pkgsrc is/does and putting in footnotes and that's never a good sign ;)
timotimo dang. 13:50
mst Su-Shee: probably not strictly necessary, we do actually know what DESTDIR is, it just doesn't always get exercised as often as it should
Zoffix timotimo, yeah huge memory leaks (irclog.perlgeek.de/perl6/2016-05-08#i_12449184), I saved the P6 version in a separate branch and rewrote the whole thing with Mojolicious. 13:51
Su-Shee mst: I'll just let him submit his stuff, what do I care if netbsd has a proper rakudo package build process :))
Zoffix Well, that and the fact that you can't have async stuff with Bailador. The in-browser code samples could take up to 20 seconds to run 13:52
timotimo Zoffix: were you still evaling the templates every time a page was requested?
13:52 Ven left
Zoffix CIAvash, the one download I found tells me "Doesn't look like a valid font file" when I try to install it on Windows :( 13:53
mst Su-Shee: do tell me once the PR's in, I'm something of a netbsd/pkgsrc fan and I'd quite like to keep an eye on it
Zoffix timotimo, yeah. Though first I had is cached trait on sub that gave me rendered templates. But it's not just that app that leaks. huggable leaks. Its restarter script leaks 13:54
timotimo mhm
Zoffix At least leaked at that time. I've not re-tried after the fixes you pushed to MoarVM
mst I'd've rewritten the restarter script in p5
timotimo the fixes i made were almost exclusively to fix --full-cleanup to give better results, fwiw 13:55
13:55 [Tux] left
Zoffix Ah 13:55
timotimo not for regular running
13:55 _mg_ joined
timotimo so you could use valgrind more helpfully 13:55
13:55 kurahaupo left
timotimo but i do agree we're not doing extremely well with memory usage at the moment 13:55
i guess i'll check out huggable and look if i find something obvious in the heap explorer 13:57
CIAvash Zoffix: I'm using this package for Arch Linux, www.archlinux.org/packages/communi...ttf-droid/
Zoffix It seems the leakage happens when she's addressed and generates a response
timotimo oh lord 13:58
cloned huggable, "panda installdeps ."
CIAvash Zoffix: Not sure it helps, www.archlinux.org/packages/communi.../download/
timotimo ==> huggable depends on IRC::Client, IRC::Client, huggable, IRC::Client::Plugin::Factoid, IRC::Client
who sees the problem here ..
13:59 JustinHitla joined
Su-Shee mst: oh, it seems the problem is already reported at least twice. 13:59
tadzik I do :)
timotimo huggable depends on ...Plugin::Hug, which is provided by huggable
Zoffix huh
tadzik it's fixed in a branch, fwiw
wait what
13:59 kurahaupo joined
Su-Shee mst: github.com/rakudo/star/issues/65 and the link to the "Open" thing "Rakudo Star: Installation of modules doesn't respect DESTDIR" 13:59
timotimo IRC::Client::Plugin::Hug is in its depends section
Zoffix Oh, I confused provides with depends :)
timotimo hah
RabidGravy doh
mst Zoffix: you always were a little backwards 14:00
Zoffix Fixed...
timotimo xiffoz
JustinHitla I have some text, and I want to get all words from it that contains only specific lettes, "a b c d e f g h i j k l m n" so words that fit are "leaf, leen" but not "leave, noun" 14:01
Zoffix JustinHitla, .words.grep: /<[tehletters]>/
timotimo so you'll write a regex like / << <[a b c d e f g h i j k l m n]> >> /
or that
kurahaupo "leave" would appear to contain both a and e.
JustinHitla so I don't even need perl ? I can use grep tool ?
Zoffix heh 14:02
JustinHitla, I guess... but .grep is a method in Perl 6
JustinHitla kurahaupo: yes, a and e are allowed
kurahaupo JustinHitla: do you mean "starts with"?
JustinHitla no, contains any of "allowed" letters
Zoffix m: say 'The stuff with words'.words.grep: /<[wu]>/
camelia rakudo-moar ad8265: OUTPUT«(stuff with words)␤»
JustinHitla I mean "consist" of
kurahaupo JustinHitla: so explain your "not" example 14:03
timotimo i forgot to put in a + after the <[ ... ]> in my regex
JustinHitla kurahaupo: other way around, if a word contains at least one letter not in the allowed list it doesn't fit
Zoffix for "consists of" just negate the match to exclude words that have letters you don't own
timotimo kurahaupo: leave contains v, which is not in the list of allowed letters
teatime pokes Zoffix
kurahaupo oh, comprises only those letters. just search for anything else and exclude the word
14:04 pierre_ left
kurahaupo like ! <[^a-n>] 14:04
timotimo no, we spell that <-[a-n]>
Zoffix: what's with the "use lib" in bin/huggable.p6? 14:05
kurahaupo grep -v '[o-z]' on the command line 14:06
timotimo oh
Zoffix timotimo, I was hacking on IRC::Client
timotimo won't "use lib 'lib'" in a bin/foo.p6 actually point to bin/lib? 14:07
JustinHitla kurahaupo: but first I need to split the text word by word then
Zoffix timotimo, I start it with perl6 bin/foo.p6
14:07 pierre_ joined
Zoffix timotimo, actually IRC::Client has a bug that might prevent you joining: github.com/zoffixznet/perl6-IRC-Cl...nt.pm6#L31 14:08
Sometiems messages get split onto multiple lines and that MOTD end message never gets reporter
This new keyboard is starting to get on my nerves :(
I don't know why there are such giant gaps between the keys 14:09
14:09 Su-Shee left
Zoffix (and keys themselves tiny) 14:09
timotimo hmm
JustinHitla Zoffix: you can always look at keyboard when typing
Zoffix watches timotimo descend into piles of half-arsed Zoffix-ware :) 14:10
timotimo the heap dump profile segfaults because it's stumbling over a somehow-invalid object
Zoffix JustinHitla, no thanks :)
RabidGravy over-rated 14:11
timotimo wonder how long it'll take my huggable to join my channel when run under valgrind
14:11 _mg_ left
timotimo aha! 14:12
14:12 _4d47 left
timotimo i now see it sloooowly process each individual message :D 14:12
Zoffix :) 14:15
timotimo it's about 1/2 into the initial welcome stuff
Zoffix How are you gonna kill it? 14:16
timotimo ctrl-c should be enough
i actually forgot --full-cleanup
that's not smart
Zoffix Won't doing that make valgrind show everything as leaked?
Zoffix loves the new single-key emoticons :D 14:17
timotimo i was hoping it'd show what or how the object got b0rked. but maybe i'll actually perl6-valgrind-m --profile=heap to get the crash reproduced guaranteed
kurahaupo JustinHitla: If you already have your words in a Perl (version 6) data structure, then this is the channel to ask. If they're in a file somewhere, ask #bash. If you don't know which version of Perl you're using, you're probably in the wrong place
timotimo the thing is, it's probably not the kind of leak valgrind would show that happens in the huggable case
it's probably objects being kept alive by something we've not thought about clearing out properly 14:18
for example, when spesh analyzes a piece of code, it'll log a bunch of values. if that piece of code doesn't run often enough to be fully specialized, that record will stick around until the program exits, keeping alive all those intermediate values that are otherwise uninteresting