»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.
Geth doc: d5abcebc15 | (Zoffix Znet)++ | 3 files
[io grant] Write docs for all spurt routines

  - Reflect changes in &spurt
  - Document all the args in IO::Path.spurt
  - Document newly-added IO::Handle.spurt
Impl: github.com/rakudo/rakudo/commit/099512b20c Tests: github.com/perl6/roast/commit/79ff022881		 00:37
u-ou Not sure if this is a bug or not, but when I try this script I get broken and glued together lines in my IO::Socket::Async's Supply. Docs say "Control characters (such as \n) always serve as grapheme boundaries, so any text-based protocols that use newlines or null bytes as terminators will not need special consideration." script here: pastebin.com/NHbLfqhQ 03:02
geekosaur u-ou, were you expecting lines to transit the network as lines? 03:07
u-ou I'm not sure what it's supposed to do but that line from docs sounds as though it's meant to 03:08
geekosaur the network does not know of anything but octets. if you are not using UDP, routers and kernels will arbitrarily segment and reassemble packets; there is no guarantee the receiver will see anything but the same *stream*
how that stream is reported to you will vary
u-ou ok
u-ou so I have to take care of it myself 03:08
geekosaur grapheme boundaries are not related to anything but how it looks for combining characters in a grapheme 03:09
yes
u-ou ok
geekosaur this is true of any network protocol. yours mayhappen to be line based, but this is not a universal
u-ou ohhhh
geekosaur presumably someone could do a layer that reassembled lines and passed them on as such (just make sure that's always what you need; some protocols have text-like commands but the data following them might be length delimited binary) 03:11
u-ou yeah, I'm just wondering what is the right way to do it 03:12
geekosaur I want to say another Supply which your receiver feeds reassembled lines 03:13
u-ou *nod* 03:14
geekosaur s/feeds/emits/
llfourn anyone have an idea what it means when rakudo gives me a bogus line number in the backtrace. 03:21
Too few positionals passed; expected 2 arguments but got 1
in method stage2 at /Users/llfourn/src/Spit/lib/Spit/SAST.pm6 (Spit::SAST) line 1 03:22
(obviously that method isn't on line 1)
u-ou hmm. I want to split a string but keep the separator as part of the returned string (at the end) 03:40
no wait. I don't need to. nvm. 03:41
memori How different is perl 5 from perl 6? What major changes in syntax (if any) might cause problems in scripts? 06:42
moritz Perl 6 is newly designed from scratch, but inspired by Perl 5 07:01
so it's quite different. You can't throw Perl 5 scrits at a Perl 6 compiler and expect them to run 07:02
samcv it is not back compatible
but it is so much fun :)
TEttinger there's inline stuff also 07:03
moritz right, you can load the perl 5 interpreter in Rakudo, thus using Perl 5 code in Perl 6 (or vice versa)
TEttinger it's neat and seems to be widely useful 07:06
tane howdy 07:16
is there anyway to return a pair of array from a sub routine? i.e. my (@a, @b) = some-sub; ?
masak m: my (@a, @b) := foo; say @a; say @b; sub foo { return [1, 2], [3, 4] } 07:22
camelia [1 2]
[3 4]
masak morning, #perl6
m: my (@a, @b) := sub { return [1, 2], [3, 4] }(); say @a; say @b 07:23
camelia [1 2]
[3 4]
masak that latter one is grammatically correct -- and I don't have to name the sub -- but it kind of goes against my "visual pill" instincts
tane masak, thanks 07:28
timotimo u-ou: there is a method on supplies that will split things into lines for you 11:14
u-ou: docs.perl6.org/type/Supply#method_lines
geekosaur: you may also be interested in this
nadim Moning #perl6, can someone tell me if "also" is something more than allowing inheritence (class, role) to be defined in the body of a class rather than the class header? 11:37
timotimo nothing else to my knowledge
moritz same here; except maybe it allows some weird stuff that would otherwise be hard 11:38
timotimo stuff that derives from itself? :P
moritz m: class A { class B { }; also is B }
camelia ( no output )
moritz m: class A { class B { }; also is B }; say A ~~ A::B
camelia True
Voldenet why ಠ_ಠ
timotimo neat, you can make that private and it'll be super strange and useless but cool!
Voldenet oh welp, it'd work for some trees
nadim "super strange and useless but cool", weirdly, I like that 11:39
Voldenet "I'll use this as my 'good practice'" :)
leont My rakudobrew doesn't want to rebuild rakudo :-/ 11:42
«Command failed (status 256): /home/leont/perl5/perlbrew/perls/perl-5.24.0/bin/perl Configure.pl --optimize --prefix=/home/leont/.rakudobrew/moar-nom/install --make-install»
moritz any output above that? 11:43
leont paste.scsys.co.uk/557779
timotimo leont: heads up, i'm hoping i can add some better error messages to YAMLish; got any hints for me? like what kinds of code changes you'd like or not like to see? 11:44
leont: that error is mea culpa. i rm'd the libtommath folder and reincarnated it as a git submodule and you have to rm -rf it manually to make stuff work again :)
moritz right, if you rm -rf /home/leont/.rakudobrew/moar-nom/nqp/MoarVM you should be fine 11:45
leont YAMLish needs a major refactor first. It really needs an AST
timotimo leont: i don't have that much time :D
leont I had this hunch a rm -rf would be involved in fixing this :-p
timotimo also: holy crap, a whole lot of stuff you'd intuitively think would be invalid ... is actually valid and not quite what you'd expect
moritz yes, YAML is "fun" like that
leont Yeah, I've learned a lot of things about YAML that I didn't want to know. 11:46
timotimo i can definitely put in an error when a map has a mis-indented key
but not when a sequence does
however, when a sequence has a mis-indented key, YAMLish rejects that outright as a parse fail, which it shouldn't
so that's also something i'd like to "fix"
leont My favorite exampe is putting a block scoped thingie inside an flowing thingie 11:47
timotimo no, i don't want to see that. ever.
timotimo are you looking to implement something like YAMLReference? like, a streaming parser for tokens/"events" and an interpreter for that to create the objects or proper errors from it? 11:49
leont I don't thing it has to be streaming 11:51
YAML documents are rarely large enough to benefit from that
timotimo fair enough
the difference is just whether the list is consumed right away, or stored until the document has ended
timotimo though also, it'll potentially change what errors you'll get 11:52
timotimo you won't get errors from the consumer rejecting the token stream, unless the producer reached the end without complaining 11:53
timotimo but i'm not sure if there are cases like that 11:53
leont I don't think of it as a token stream, mainly because Grammars are recursive-decent 11:54
timotimo that's fair, i guess
leont Even if all other YAML parsers do work that way 11:55
timotimo i didn't dive down far enough to get a clear understanding of the whole, rather than just individual pieces
leont I have a moar--j=2/ folder, that looks like a mistake to me 12:00
timotimo :D 12:03
yeah, you don't give j two dashes
leont TAP.pm kind of requires a Dual-life workflow 12:16
timotimo taps some of his cards to harness the power of their magic
leont Editing it on both ends is problematic and not keeping them in sync is problematic.
*is even more
timotimo you mean how tap::harness fails with many test suites of perl6 modules? 12:17
or have i missed something additional?
leont I've just synced them
timotimo i think i've missed something; sync tap::harness and what?
leont It's in rakudo, and it's in separate repository, but it's the same code 12:18
timotimo oh, i didn't realize
leont Possibly the easy solution is to not have it in the ecosystem, given that it has been in rakudo for a year now. Not sure how far zefs backward compatibility goes. 12:21
leont Hmmm, doesn't Test.pm6 have a skip-all option? 12:25
timotimo skip-rest 12:27
+ exit 0
or something
leont That is not the same 12:29
perl -MTest::More -e 'plan skip_all => "Reason"'
timotimo oh, ok 12:30
andrzejku hi people
leont Solved it with a print and exit, but it is kind of awkward 12:31
timotimo a bit, yeah
lizmat leont: PR's to Test.pm6 are welcome :-)
andrzejku hi people would you like to update this greate intro 13:02
document for Native Call
in C++
lizmat andrzejku: I think hankache takes PR's :-) 13:03
andrzejku lizmat: yes yes 13:04
leont Yeah, adding it to Test should be trivial, just surprised no one felt they needed it 13:09
timotimo andrzejku: hankache wants the intro to only be an intro, not everything 13:16
andrzejku: that's also why there is no mention of "is rw", or repr('CStruct')
lizmat is C++ support explained on docs.perl6.org ? 13:31
docs.perl6.org/language/nativecall#C++_Support # guess it is 13:32
andrzejku: perhaps you could improve that ?
lizmat and perhaps the intro could also link to that section ? 13:33
timotimo it already links to language/nativecall, doesn't it? that starts with a nice TOC that has "C++ support" in it. in bold, even! :) 13:34
leont I doubt it's possible to support any of the more interesting parts of C++ (as that requires the C++ compile time) 13:35
timotimo right, can't make templates work unless they are already instantiated in the way you want them to 13:35
moritz which is a real PITA even when using C++ libraries in C++ 13:44
it means you always have to keep the full source code o fthe libraries around 13:45
lizmat which, atm, is not unlike Perl 5 or Perl 6 for that matter :-) 13:47
for Perl 6 we will be able to do better in the future :-)
timotimo not the full source code, just the header files! 13:54
except of course that people write their whole projects in header files instead of source files!
moritz well, if you write only headers into the header files, and give somebody an object file and the headers, they can't instantiate templates with unknown combinations of parameters 14:11
moritz guesses he missed the irony in timotimo++'s statement :-)
pmurias hi 14:18
lizmat pmurias o/ 14:19
Geth doc: ee9455a05d | (Zoffix Znet)++ | doc/Language/testing.pod6
Document `skip-all` option for Test::plan()

Rakudo impl: github.com/rakudo/rakudo/commit/14b6d5f360 Tests: github.com/perl6/roast/commit/680970a2e5		 16:01
Geth doc: 9a993275ec | (Brad Gilbert)++ | doc/Type/IO.pod6
Typo slurp -> spurt
 17:27
doc: 5c9a788f6e | (Zoffix Znet)++ | doc/Language/testing.pod6
Improve docs for plan :skip-all, bail-out, and skip-rest

Mentions the other two in each of the sections and clarify a bit when to use which.
Geth doc: bd491f13b8 | (Zoffix Znet)++ | t/pod-htmlify.t
Reword test to work in earlier rakudos

IO::Handle.spurt is not yet available in any of the released rakudos. Fixes #1264		 17:47
Geth doc: 951f5f6c5a | (Zoffix Znet)++ | t/pod-htmlify.t
Fix broken test

tempfile returns a Str not IO::Path for $filename		 17:51
leont \o/ (skip-all) 18:21
andrzejku hi 18:52
:)
raschipi oi 18:54
masak y0 18:56
geekosaur œ 18:56
masak ℌℑ 19:00
El_Che too "whaat" for utf8. I am still in shock Ubuntu is dropping unity 19:13
:)
raschipi Unity, Mir, Snappy... 19:17
raschipi GNOME won the desktop wars. Soon SuSE will have to switch too, to avoid being the weird one. 19:18
masak Ubuntu is droppoing Unity? hadn't heard, but that sounds like good news from where I'm sitting. 19:20
dropping*
El_Che masak: I suffered through the first buggy releases and I end up liking it :) 19:21
just installed gnome on this laptop
:)
t.co/3Lu0WHO8k5 19:23
The announcement
raschipi All because of language bindings. 19:25
El_Che gtk3? 19:26
raschipi Qt bindings are few an far between and they suck everywhere but on C++ or Python. 19:27
El_Che ah, unity 8
not 7 19:28
geekosaur stays with mate
El_Che needs to relearn quite a few of keyb shortcuts it seems 19:30
AlexDaniel wow, that was unexpected 19:35
wait, how are they going to integrate their ads in GNOME exactly? 19:36
El_Che the amazon thing has been gone for a long time 19:36
leont Gnome Shell's awesome side is its extensibility, if they would want to they easily could 19:37
El_Che I switched this "irc + firefox" machine first (and 2gb 9y old ultrabook) 19:38
and gnome3 does not feel slow 19:39
that's a win
raschipi Amazon integration comes disbled by default for a long time. And it wouldn't be hard to write an amazon search provider for GNOME. 19:42
AlexDaniel thinking about it, Canonical has done a lot to make sure I would never recommend Ubuntu to anyone 19:46
… only to revert it all back
leont I loved Ubuntu when it was an up-to-date Debian, but got annoyed by the QA and now I'm on Arch 19:47
samcv i use arch too 19:48
AlexDaniel debian unstable works fine too :P
raschipi It sure does. 19:49
samcv i have debian unstable on my server
arch on desktop/laptops 19:50
AlexDaniel :O
well, wouldn't really do that on a server… :)
samcv well. unstable is not unstable, it's just. the version numbers change
stable refers to the versions not changing
which i am totally fine with
though it also refers to the other kind of stability too, but. i want the latest versions 19:51
raschipi I don't put unstable on servers because I like to activate automatic security upgrades. 19:52
And unstable doesn't have those.
AlexDaniel well, you can upgrade it automatically anyway 19:55
and then check your server every day to see if it is actually running…
raschipi My clients would kill me...
samcv i guess that's fair enough raschipi
raschipi It's good to avoid new features unless you can sell them. 19:58
[Coke] yawns. 20:03
MasterDuke arch on desktop, kubuntu on laptop 20:07
El_Che is there a demand for arch packages (like the one I provide for centos/debian/ubuntu/fedora? 20:14
I mean rakudo pkgs 20:15
raschipi I think they already have the newest version in AUR 20:16
El_Che ah nice
(then no use case there) 20:17
MasterDuke yeah, they're usually updated only a couple days after a release
El_Che impressive
I'll try to add alpine soon
raschipi Debian also updates within a few days when they're not frozen. 20:19
El_Che I ran testing and unstable in the early 2000 :) 20:20
El_Che was glad when ubuntu came out: frozen recent packages 20:21
on the other hand, I learnt a lot about Linux in those days when X (Xf86) was often broken on unstable :)
Voldenet El_Che: do you maintain any repo with debian/* files, I'm playing around with moar's code so I'd be glad not to just rm -rf /opt/rakudo 20:50
but to use apt instead
El_Che Voldenet: no a repo, just dpkg -i 20:53
it upgrades the pkg
El_Che Voldenet: github.com/nxadm/rakudo-pkg/releases 20:54
Voldenet: you could run the container that create the package and chaneg stuff there
Voldenet oh, docker 20:55
I didn't notice there was a source there, thanks.
AlexDaniel covers his ears 21:06
Geth doc: bdb7ab42af | (Zoffix Znet)++ | doc/Programs/00-running.pod6
Explain the option more clearly

Never know which one it is when looking for env vars and end up clicking the wrong thing :/		 21:24
SmokeMachine m: sub bla⸨:$key, :$value⸩ {say "$key => $value"}; bla "a" => 42 22:13
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3sub bla7⏏5⸨:$key, :$value⸩ {say "$key => $value"};
expecting any of:
new name to be defined

Powered by Cro and the Raku® Programming Language.

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

Thank you!