Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2015-02-04

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:03 jack_rabbit joined #perl6
00:06 [Coke] pmichaud: I'm focusing on RT/impl tickets, not github/spec tickets. We have too many to say "all of these are required to fix before our RC release."
00:06 [Coke] (but someone needs to own those github tickets, as well)
00:15 kaare_ joined #perl6
00:23 adu joined #perl6
00:32 apv_ joined #perl6
00:33 apv_ Perl6 n00b enjoying rakudo brew. Anyone active right now who might indulge me on some p5->p6 questions?
00:34 pmichaud I'm here.
00:34 apv_ (also IRC n00b, please forgive any accidental /monkeybiz.)
00:34 tony-o also here
00:34 apv_ Hey there.
00:35 apv_ __FILE__ - I use this idiom with File::Spec and friends to let a script determine its lib. What's the right, or a good, idiom for p6.
00:35 apv_ For example, with Path::Class: my $self_file = file( File::Spec->rel2abs(__FILE__) );
00:36 pmichaud m: say $?FILE
00:36 camelia rakudo-moar 5b4907: OUTPUT«/tmp/2i1w2ozcPf␤»
00:36 pmichaud $?FILE is what you're looking for, I think.
00:36 apv_ Ah, nice. Thanks.
00:37 pmichaud http://doc.perl6.org/language/vari​ables#Compile-time_%22constants%22
00:37 apv_ Better still.
00:41 apv_ That is a constant and not an object I take it? What kind of idiom or package to get the File::Spec->rel2abs(...) behavior? I have panda installed and working fine.
00:42 apv_ (Is sri doing a mojolicious for p6?)
00:42 pmichaud m: say ($?FILE).WHAT
00:42 camelia rakudo-moar 5b4907: OUTPUT«(Str)␤»
00:42 pmichaud m: say $?FILE
00:43 camelia rakudo-moar 5b4907: OUTPUT«/tmp/XoCZgRl8af␤»
00:43 Mouq methinks that could probably be an IO
00:43 pmichaud yes, that's what I was checking.
00:43 Mouq Although perl6 -e'say $?FILE' => '-e'
00:44 pmichaud m: say IO::Path.new($?FILE).abspath
00:44 camelia rakudo-moar 5b4907: OUTPUT«IO::Path is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:1␤  in method new at src/RESTRICTED.setting:32␤  in block <unit> at /tmp/KT_gvEBueu:1␤␤»
00:44 pmichaud hmmm, weird.
00:44 apv_ https://github.com/labster/p6-IO-Path​-More/blob/master/lib/IO/Path/More.pm interesting in there…
00:45 Mouq m: say $?FILE.IO.abspath
00:45 camelia rakudo-moar 5b4907: OUTPUT«/tmp/tcagcZsaKo␤»
00:45 pmichaud Anyway, IO::Path might be the answer you're looking for
00:45 apv_ Fabulous! IO.abspath works for me.
00:45 apv_ I figured there would be something like that.
00:46 apv_ How's interest these days? I haven't played with p6 much and it's been quite awhile. rakudobrew/panda made it interesting and productive to jump back in. Quite exciting so far.
00:46 pmichaud Interest is way up.  :)
00:47 apv_ Great.
00:47 pmichaud On Sunday we had this:  https://fosdem.org/2015/sched​ule/event/get_ready_to_party/
00:50 apv_ Is there anything like mojolicious in the works? I saw Bailador already.
00:50 apv_ Oh, of course. I forgot the talk already happened.
00:50 pmichaud I'm not familiar with the latest modules / ecosystem items, so someone else would have to chime in on that.
00:50 _sri apv_: once perl6 is production ready (my definition of it)
00:51 Mouq I feel that both #120916 and #122613 are misinterpretations of what `...` means in the given context... Perl 6 really has a lot of different ways to say "Yeah, whatever," and they're all stuffed into the spellings "*" or "..."
00:51 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=120916
00:51 apv_ _sri++
00:52 apv_ Anyone have a good tut/advent entry to read for multi methods/subs?
00:52 Mouq apv_: https://github.com/tadzik/Template-Mojo/ is a recent addition
00:55 apv_ Goals might help so I'm not XY'ing everyone. I want to do request method based dispatch with the method corresponding to URI. Just messing around now. So it would be multi user GET multi user PUT, etc.
00:58 pmichaud apv_: this one is a bit old, but may help:  https://perl6advent.wordpress.com/2011/12/24/​day-24-subs-are-always-better-in-multi-ples/
01:02 pmichaud I'm afk again, be back tomorrow
01:10 b2gills apv_: I've thought about having a module that you would use by defining `multi method GET ( 'user', %params, *% ){ ... }; multi method GET ( /<[0-9]>+/, %params, *% ){ ... }` etc
01:16 TimToady well, in current syntax would have to be Str where /<[0..9]>+/
01:22 Klaudioh joined #perl6
01:24 apv_ Thanks much, all. I have to go too. I'll experiment with that stuff.
01:25 TimToady well, actually, needs ^ and $ too...
01:25 flussence There was that one thing that routed URIs by just throwing them straight into a grammar... sounds crazy but it's a pretty cool idea. Is it a module? I can't find it...
01:26 TimToady what you probably don't want to do is Foo."$insecure"()
02:07 yeahnoob joined #perl6
02:16 yeahnoob joined #perl6
02:24 colomon joined #perl6
03:10 chenryn joined #perl6
03:14 skids joined #perl6
03:17 espadrine joined #perl6
03:19 Mouq joined #perl6
03:23 chenryn joined #perl6
03:29 noganex joined #perl6
03:55 Mouq joined #perl6
04:12 anaeem1 joined #perl6
04:14 kjs_ joined #perl6
04:22 telex joined #perl6
04:29 kaleem joined #perl6
04:35 stux|RC-only joined #perl6
04:39 nige joined #perl6
04:55 silvest joined #perl6
04:57 chenryn joined #perl6
05:34 kjs_ joined #perl6
05:44 silvest joined #perl6
05:51 TimToady Perl 6 Ant, Perl 6 Bee, Perl 6 Cicada, Perl 6 Dragonfly, Perl 6 Earwig, Perl 6 Fruitfly, Perl 6 Gnat or so
05:54 TimToady just an idea to differentiate standards from implementations
05:55 TimToady feel free to bikeshed systems of names (trees, flowers, musical instruments, authors)
05:55 TimToady but insects seemed appropriate for some reason :)
05:56 TimToady especially since they have 6 legs :)
05:58 TimToady this would completely decouple standards from years and numbers, but be very easy to determine order
06:00 TimToady probably need something shinier than "ant" and "bee"
06:00 TimToady and dragonfly is a bit long
06:03 telex joined #perl6
06:04 JimmyZ Perl 6 Duke
06:04 TimToady Perl 6 Half
06:05 TimToady Perl 6 Grand
06:06 TimToady Perl 6 Final :)
06:06 JimmyZ well, Per 6 Duke will be good to the first implementation :P
06:07 JimmyZ and Perl 6 Forever
06:07 adu joined #perl6
06:10 TimToady Perl 6 Arsenic, Boron, Chlorine, Dubnium, ,,,
06:15 flussence could play along with the perl 5 dinosaur theme... it'd be kinda fitting to name p6 after dragons :)
06:19 alini joined #perl6
06:24 TimToady zzz &
06:32 firefish5000 joined #perl6
06:38 raydiak adu: ping
06:39 adu hey
06:39 b2gills I think the name could be from a name generator
06:39 raydiak hey!  I have something stating to make a zmq.pm6
06:39 raydiak but cast (3?) has brokenness
06:39 adu uh oh
06:40 raydiak if I parse w/cdump it works with just a few warnings, but if I use C::Parser and then .parse (so it uses cast instead of cdump's default 'nil' or whatever)...then errors
06:41 b2gills I know there was a name generator used for the Parallel Realities game named Project Starfighter
06:42 raydiak adu: so I have my script hacked to parse the raw parse tree from C::Parse::Grammer...but I'd really like to use that C::AST you've worked so hard on :)
06:42 adu raydiak: oh, so does it pass with the grammar but fail with the AST?
06:43 kaleem joined #perl6
06:43 raydiak yes, the grammar does work now! very awesome, I actually have a working model for zmq now
06:43 adu nice
06:44 raydiak though I have differences of opinion about how pointers are handled :)
06:44 raydiak in the grammar I mean...on the cast level may be different, idk yet
06:45 adu raydiak: cast level?
06:45 adu what is cast 3?
06:45 adu oh C::AST
06:45 adu duh
06:46 raydiak I mean idk how you handle pointers with c::ast since I'm not using it yet, I'm just talking about the ouput from the grammar itself
06:46 * raydiak thinks of * as part of the type more than part of the name
06:47 chenryn_ joined #perl6
06:47 adu raydiak: that's true, that's exactly what I was talking about semantic AST vs syntactic AST
06:48 raydiak ah, I see...I have a couple places where I have to check in my script for .<pointer> specially because it isn't automatically part of the type where I do all the other checks
06:48 dalek doc: 0aeeca3 | paultcochrane++ | lib/Language/classtut.pod:
06:48 dalek doc: Minor typographical etc. corrections to classtut.pod
06:48 dalek doc: review: https://github.com/perl6/doc/commit/0aeeca3ec7
06:48 dalek doc: 7a992e3 | paultcochrane++ | lib/Language/classtut.pod:
06:48 dalek doc: Wrap paragraphs in classtut.pod consistently
06:48 dalek doc: review: https://github.com/perl6/doc/commit/7a992e325c
06:48 dalek doc: 8ca4e59 | paultcochrane++ | / (12 files):
06:48 dalek doc: Merge branch 'master' of github.com:perl6/doc
06:48 dalek doc: review: https://github.com/perl6/doc/commit/8ca4e59e33
06:48 raydiak and depending on where it occurs, where you check for that is different (return type on a function vs param to a function)
06:49 adu the syntactic AST for c is decl=(spec+, (ptr, name, designators, inits)), but the semantic AST for c is decl=[(complete type, name, init)]
06:49 raydiak so you probably smooth all that over with c::ast, if I had to guess, but I'm wondering if you wouldn't benefit from that change on the grammar level
06:49 raydiak hm I see
06:50 adu I actually stayed with the syntactic AST, and was planning another layer for conversion for the semantic AST, but I tried to design the AST so that it could encode both layers
06:51 adu so, for example TyKind::pointer_declarator would be converted to TyKind::pointer_type
06:52 raydiak I'm starting to catch up to what you were specifically looking at...sometimes I need a working model, like us thick-headed people do from time to time :)
06:53 adu also, I found an obscure part of the standard that gives nice names for things
06:53 adu so "x[4]" is a declarator
06:53 adu and "[4]" is a designator
06:53 adu so those are the names you see in TyKind
06:54 adu maybe once I get the AST sorted out, I'll write some documentation about it
06:55 raydiak I think im(somewhat)ho, a semantic ast would be best, given that the goal of your module is to separate users of your module from C syntactic concerns
06:55 adu well, if I can do it all in one pass, then you can buy me a beer :)
06:56 raydiak well...okay deal :)
06:57 adu I've heard there are C compilers that have a single pass, but I'm pretty sure the popular ones today don't
07:00 raydiak hm
07:04 FROGGS joined #perl6
07:04 raydiak I'm honestly still wrapping my head around the single-pass parsing thing and how it's different from ordinary approaches...I come to these things from a more mundane web development background, in large part
07:06 adu webdev is fun
07:07 moritz ... except when it's not :-)
07:07 raydiak not profess...what moritz said :)
07:08 adu oh, the IE6 kind?
07:08 adu n/m
07:08 raydiak it's a lot less horrifying than it used to be, kinda...except now you have phones and tablets
07:09 adu yeah, I just use bootstrap for everything
07:09 adu no more headaches
07:09 konsolebox joined #perl6
07:10 Tux__ joined #perl6
07:10 raydiak I'll bet that helps...the main thing about web stuff is just that you have to write one thing that works in so many different "everywhere"s, that you're nearly trying to polyglot, or at least were in the old days of, like you said, IE6
07:10 Rounin joined #perl6
07:11 adu modernizr helps with that
07:12 moritz jquery also helps a lot (for the JS part)
07:13 raydiak yeah jquery is awesome
07:13 adu I only use jquery when I have to
07:14 adu I actually prefer Sizzle
07:15 raydiak but I's been doing web stuff professionally since half a decade before jquery's first release, and did things by hand for several years after still...so I have some special hate for JS :)
07:15 adu because Sizzle is jQuery, minus the dom and plugin and ajax stuff
07:15 fhelmberger joined #perl6
07:15 moritz so not much left, eh? :-)
07:17 raydiak ( or ECMAScript as they like to call it these days, to feel less dirty than the terrible association everyone got with "javascript" after the first 5 or 10 years of it )
07:17 * raydiak has not read about Sizzle...adds it to the lizt
07:20 El_Che ll bet that helps...the main thing about web stuff is just that you have to write one
07:20 El_Che thing that works in so many different "everywhere"s, that you're nearly trying to
07:21 raydiak El_Che: yes exactly! :)
07:28 denis_boyun joined #perl6
07:28 raydiak adu: btw, it doesn't look much like it would with the c::ast errors fixed, but here's how my zmq test looked at this point: https://gist.github.com/ra​ydiak/4af7a883df58886c5763
07:28 adu moritz: https://github.com/jquery/jquery/bl​ob/master/src/selector-sizzle.js#L6
07:29 adu moritz: so Sizzle is about 15% of jQuery
07:30 Alina-malina joined #perl6
07:33 muraiki_ joined #perl6
07:35 adu raydiak: yikes
07:35 adu raydiak: ouch
07:35 coffee` joined #perl6
07:36 adu raydiak: you might also try C::Parser::Utils::get_declarator_name()
07:36 raydiak well, I think I'm supposed to be a C::Parser consumer, not a C::Parser::Grammar consumer, so I didn't worry about it much :)
07:36 raydiak ah, will look at that
07:37 adu https://github.com/andydude/p6-c-parse​r/blob/master/lib/C/AST/Utils.pm6#L23
07:37 raydiak mainly I think the parts you could improve to start, would be the places where I have to call .trim (you can probably just s/token/rule/ in some places), and the pointer thing I already mentioned
07:37 adu oh, right
07:38 rindolf joined #perl6
07:40 raydiak but even more primarily, of course...just the C::AST errors which I assume you can already reproduce since you mentioned you tried zmq on your last pass of fixes and testing
07:40 adu I only tried zmq with the grammar, I haven't tried it with the AST
07:43 raydiak yeah it blew up...the first one was just that "name => " thing I think I brought up before, but something else after that was slightly less obvious but you'll see it whenever you have time to look again
07:44 dalek rakudo/birdless: 0f7a0b9 | moritz++ | src/Perl6/World.nqp:
07:44 dalek rakudo/birdless: Fix error reporting of undeclared symbols in the setting
07:44 dalek rakudo/birdless:
07:44 dalek rakudo/birdless: for ab5tract++
07:44 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/0f7a0b9b2a
07:44 dalek rakudo/birdless: 5916b85 | moritz++ | INSTALL.txt:
07:44 dalek rakudo/birdless: Excise parrot from INSTALL.txt
07:44 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/5916b856a5
07:44 dalek rakudo/birdless: 946c413 | moritz++ | README.md:
07:44 dalek rakudo/birdless: no more parrot in README.md
07:44 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/946c4139e0
07:45 dalek rakudo/birdless: 7d85749 | moritz++ | docs/glossary.pod:
07:45 dalek rakudo/birdless: remove outdated glossary items
07:45 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/7d85749a11
07:45 dalek rakudo/birdless: f67ba35 | moritz++ | / (4 files):
07:45 dalek rakudo/birdless: Remove more parrot legacy
07:45 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/f67ba3571a
07:45 dalek rakudo/birdless: cbd059e | moritz++ | t/spectest.data:
07:45 dalek rakudo/birdless: Get rid of "# icu" markers in t/spectest.data
07:45 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/cbd059e8f0
07:45 dalek rakudo/birdless: e755c6e | moritz++ | t/spectest.data:
07:45 dalek rakudo/birdless: remove # conc markers from t/spectest.data
07:45 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/e755c6ee8f
07:45 FROGGS birdless O.o
07:45 raydiak ...ooh
07:46 FROGGS moritz: what exactly is that about?
07:47 FROGGS I guess I should backlog
07:48 moritz FROGGS: it's a branch with parrot support dropped
07:48 raydiak afk
07:48 adu raydiak: how do I filter out Nil's from a list?
07:48 FROGGS moritz: but the intent to let it catch up is still there?
07:49 moritz FROGGS: I never understood the lagging support / catch up narrative
07:50 moritz FROGGS: things don't catch up by themselves
07:50 raydiak adu: just Nil? maybe .grep(* !=== Nil)
07:50 adu oh
07:50 FROGGS moritz: that's true
07:50 adu I didn't think it would be called grep
07:51 moritz FROGGS: and for parrot support to stay viable, somebody would need to invest quite some energy to fix leaky abstractions in NQP first, so that we'd need fewer #?if's in rakudo's core
07:51 moritz FROGGS: ... and if somebody wants to do that, all the code is still in the git history
07:52 moritz FROGGS: fwiw this branch isn't in any way officially sanctioned or so; so far it's an experiment and an RFC by me
07:52 FROGGS I see
07:53 moritz it's fun to delete code :-)
07:54 FROGGS yeah, I know :o)
07:55 _mg_ joined #perl6
08:01 prime joined #perl6
08:06 raydiak un-afk
08:07 ven moritz: https://github.com/rakudo/rakudo/compare/nom...bi​rdless#diff-15c5b4c6f757ed54da181d4639921d11R296 typo, should probably be "jvm"
08:08 moritz ven: no, we never git-clone the JVM :-)
08:08 ven moritz: I
08:08 ven Day #231: I still hate that keyboard
08:09 ven moritz: Imoritz: I'd say you're on the wrong line, because it says "--backends=vm,moar" here, which seems to be a mistake
08:09 moritz ven: oh, sorry, indded
08:10 dalek rakudo/birdless: e8b36e0 | moritz++ | Configure.pl:
08:10 dalek rakudo/birdless: Fix typo in help message, ven++
08:10 dalek rakudo/birdless: review: https://github.com/rakudo/rakudo/commit/e8b36e0aa5
08:11 darutoko joined #perl6
08:15 trone joined #perl6
08:15 fhelmberger joined #perl6
08:19 [Tux] ping nine
08:21 JimmyZ will nqp will still support parrot?
08:22 JimmyZ s:2nd/will//
08:22 [Tux] https://gist.github.com/Tux/6520a23f118908ae8237 <= Anyone with Inline::Perl5 knowledge. I've now got it working but want more
08:24 [Sno] joined #perl6
08:26 DrForr_ joined #perl6
08:27 * DrForr_ sighs and pokes his nose in.
08:28 andreoss` joined #perl6
08:29 nige joined #perl6
08:29 * raydiak wonders what happened to the cartilage in DrForr_'s nose
08:31 moritz JimmyZ: if/when rakudo drops parrot support, I don't see why nqp should keep it
08:35 raydiak g'night #perl6 o/
08:36 adu raydiak: night
08:36 adu raydiak: wait, did you leave yet?
08:40 virtualsue joined #perl6
08:42 KPTN joined #perl6
08:42 KPTN left #perl6
08:44 FROGGS moritz: I don't see a reason to remove the parrot support in nqp... it would just make it difficult to get it back in
08:45 moritz FROGGS: ... until it becomes a pain to maintain
08:46 FROGGS moritz: aye
08:46 FROGGS but please let it in until then
08:48 Alina-malina joined #perl6
08:48 Alina-malina joined #perl6
09:00 abraxxa joined #perl6
09:01 telex joined #perl6
09:02 salv0 joined #perl6
09:02 coffee` joined #perl6
09:02 pmurias joined #perl6
09:03 zakharyas joined #perl6
09:05 Alina-malina joined #perl6
09:05 dalek doc: 03aeec6 | paultcochrane++ | lib/Language/containers.pod:
09:05 dalek doc: Minor typographical etc. corrections to containers.pod
09:05 dalek doc: review: https://github.com/perl6/doc/commit/03aeec62e0
09:05 dalek doc: aed0041 | paultcochrane++ | lib/Language/containers.pod:
09:05 dalek doc: Wrap paragraphs in containers.pod consistently
09:05 dalek doc: review: https://github.com/perl6/doc/commit/aed00416fe
09:07 dalek nqp-js: 48d088b | (Pawel Murias)++ | src/vm/js/QAST/Compiler.nqp:
09:07 dalek nqp-js: Implement nqp::bindattr_i, nqp::bindattr_n, nqp::bindattr_s.
09:07 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/48d088bb6b
09:07 dalek nqp-js: 7be108b | (Pawel Murias)++ | t/nqp/58-attrs.t:
09:07 dalek nqp-js: Add tests for nqp::bindattr_i, nqp::bindattr_n, nqp::bindattr_s, nqp::bindattr.
09:07 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/7be108b7bb
09:08 pmurias .tell jnthn how do I do pipe stuff to a process in nqp?
09:08 yoleaux pmurias: I'll pass your message to jnthn.
09:10 pdcawley joined #perl6
09:14 masak good antenoon, #perl6
09:15 FROGGS pmurias: grep for openpipe in rakudo or nqp
09:17 pmurias masak: good antenoon
09:17 pmurias how do we portably test openpipe?
09:18 timotimo http://www.heise.de/newsticker/meldung/​Vermutlich-kein-Scherz-Perl-6-koennte-z​u-Weihnachten-erscheinen-2534683.html - just saw this headline
09:18 FROGGS nqp-m: my $fh := nqp::openpipe('ls', nqp::cwd(), nqp::getenvhash(), ''); say(nqp::readallfh($fh)); say(nqp::closefh_i($fh))
09:18 camelia nqp-moarvm: OUTPUT«Perlito␤bin␤evalbot␤evalbot.log␤log␤m​box␤niecza␤p1␤p2␤perl5␤rakudo-inst␤rakudo-in​st-1␤rakudo-inst-2␤rakudo-star-2014.12␤rakud​o-star-2014.12.tar.gz␤rakudo1␤rakudo2␤star␤s​tar-2014.09␤std␤test␤ugexe-was-here␤␤0␤»
09:18 FROGGS pmurias: ^^
09:19 FROGGS pmurias: not sure about testing it portably
09:19 FROGGS that's always tricky
09:22 pmurias is it possible to do two way piping using that?
09:23 dalek doc: 2d0bb11 | paultcochrane++ | lib/Language/variables.pod:
09:23 dalek doc: Rename compile-time "constants" -> compile-time variables
09:23 dalek doc:
09:23 dalek doc: This is in response to #GH46.  In the ticket, it is recommended to remove
09:23 dalek doc: the quotes around the word "constants", since this causes a link to not
09:23 dalek doc: work.  Also, it is recommended that the word "variables" is a better choice.
09:23 dalek doc: Thus "variables" was chosen since the values aren't really constants (which
09:23 dalek doc: is the sense, the quotes were trying to convey before).
09:23 dalek doc: review: https://github.com/perl6/doc/commit/2d0bb116d5
09:23 [Tux] If I do nine++ for makeing Inline::Perl5 PASS, will he become ten?
09:24 telex joined #perl6
09:26 jluis joined #perl6
09:28 espadrine joined #perl6
09:32 woolfy I am personally in favor of Perl 6.0.0 as the name for what is going to be released later this year.  I think it as appropriate a version  name as Perl 5.22.0 is for the newest release for Perl 5 in May.
09:33 woolfy I am not afraid of any anger from people from Perl 5.  I think Perl 5 is strong enough to handle this.  Perl 5.22 will be strong enough.  Any confusion is in my eyes deliberate by people who don't want Perl 6 to be a success.
09:33 woolfy I think naming Perl 6 anything else than Perl 6.0.0 will weaken any potentional success of Perl 6.
09:33 moritz m: $_ = '[Tux]'; $_++; .say
09:33 camelia rakudo-moar 5b4907: OUTPUT«[Tuy]␤»
09:33 woolfy s/potentional/potential/
09:34 woolfy Marketing Perl is difficult enough, calling it something like Perl 6 version 1.0 or Perl6 2015a is making marketing a futile exercise.
09:34 woolfy How are we going to explain "Perl6 1.0" to people.  I think it is quite silly.
09:34 timotimo these comments on the heise article are absolutely worthless
09:35 woolfy And that only to make the Perl 5 people less angry?
09:35 dalek nqp-js: 0f36631 | (Pawel Murias)++ | src/vm/js/ (2 files):
09:35 dalek nqp-js: Add a --beautify option to nqp-js.
09:35 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/0f366311cf
09:35 moritz timotimo: which heise article?
09:35 pmurias FROGGS++ # find the right nqp opcode
09:35 woolfy And I really am not afraid of any smacking by mst or whoever.
09:36 molaf__ joined #perl6
09:36 ab5tract woolfy: fwiw, i  completely agree
09:36 woolfy Explaining Perl 6.0.0 is easy.  It is what it is.  It is beyond "Rakudo Perl 6 beta development whatever".  It is "the first release version of Perl 6."
09:36 timotimo moritz: the one i linked to above, "Vermutlich kein Scherz: Perl 6 könnte zu Weihnachten erscheinen"
09:36 andreoss` how this issue was handled in time of Perl 4?
09:37 DrForr_ The version was bumped because ORA wanted to release a new book.
09:37 moritz DrForr_: wasn't that the 3 -> 4 version change?
09:38 huf what if it's perl 6.0.0 but -v reports it as perl 6, version 0, subversion 0? :D
09:38 woolfy If naming Perl 6.0.0 makes it more difficult for Perl 5 people to hire new people, to get new products, it means they should put more effort in explaining and marketing.
09:38 pmurias woolfy: it's not that easy to explain, Perl 6.0.0 implies it's a new major version of Perl 5
09:38 El_Che woolfy: the explanation "easiness" cuts always one side of the equation. What make easy to explain about 6.0.0 make it pretty awkward explaining for 5. I plan to expend time using both languages instead of explaining the versions numbering.
09:39 woolfy And as Perl 6 people do not recomment naming conventions for Perl 5, I think naming conventions for Perl 6 releases should be up to Larry Wall, Patrick Michaud and others who put a lot of work in Perl 6.
09:39 huf how does acknowledging that the number has become part of the name of the language hurt anyone?
09:39 moritz I don't reall see the advantage of 6.0.0 "first version of Perl 6" vs. Perl 6 2015a "2015 version of Perl 6"
09:39 moritz *really
09:40 f3ew Just name it P6 1.0.0.
09:40 woolfy Perl 6.0.0 does not imply the new major version of Perl 5.  It has nothing to do with Perl 5.  I do not see a 5 in Perl 6.0.0.  Anybody suggesting this, is making a joke out of this.
09:40 huf woolfy: so how does calling it "Perl 6 1.0" hurt perl6 then?
09:41 woolfy "2015 version of Perl 6" is like "the 2015 version of the BMW 3".
09:41 woolfy "Perl 6 1.0" is unpronouncable.  How will we be able to explain this to big companies, universities?
09:42 huf same thing everyone's been saying for years?
09:42 woolfy What the hell would Perl 6 1.0 mean?  Take away some spaces and we have Perl 61.
09:42 huf this is perl 6. not perl 5. this is v1.0 of perl6.
09:42 huf done.
09:42 * masak agrees with woolfy
09:42 moritz woolfy: it's common pratice to use years in language versions, see C and Fortran, for example
09:42 [ptc] maybe Perl 6-1.0?
09:42 masak I'm all for not pissing off fivers, but not at the cost of eternal version confusion.
09:42 andreoss` m: say 6-1.0
09:42 camelia rakudo-moar 5b4907: OUTPUT«5␤»
09:43 FROGGS :P
09:43 El_Che woolfy: personaly, I find 1.0 a *lot* clearer that 6.0.0 (even ignoring the wish to make the 5 and 6 parts of the community happy)
09:43 [ptc] :D
09:43 woolfy Perl 6.0.0 would be the version of 2015 of Perl 6.  Just like Perl 5.22 is the 2015 version of Perl 5.  I don't see any problems with that.
09:43 masak Perl 6 is 5! QED.
09:43 DrForr_ Besides, wasn't there already a change in the documentation to change perl 5 to perl5 to make room for perl 6?
09:43 El_Che I find 6.0.0 confusing as hell
09:43 El_Che for a first release, at least
09:43 osfameron won't naming 6.0.0 force the hand of 5 - e.g. to become Perl 5 v.22 (precisely to clarify that 6.0.0 isn't the next version) ?
09:43 huf i dont actually have strong feelings either way, because it doesnt fucking matter.
09:43 woolfy I find Perl 5.22.0 as clear as Perl 6.0.0.
09:43 huf i just dont like this trend of people making up shit reasons to back up their feelings
09:43 timotimo woolfy: not clear at all? :P
09:44 woolfy When Perl 6 has to be called Perl 6 1.0 2015, I think we should call the next version of Perl 5 "Perl 5 version 22 of 2015" or something silly like that.
09:44 huf as long as it's an actual version number and not some ubuntu-like name that's impossible to remember or compare, i'm fine :)
09:44 Grrrr Perl 6.1.0;  there, everybody's side of the argument accounted for
09:44 huf woolfy: guess what perl5 calls itself.
09:45 huf This is perl 5, version 16, subversion 2 (v5.16.2) # some ancient version i found
09:45 DrForr_ Yes, but perl6 feels almost orthogonal to perl5 to me. So perl5 v22 and perl6 v1.o (bowing to Java for leading the way) feels perfectly natural.
09:45 huf the usual ostrich game of "written 5.16.2, pronounced "perl 5, version 16, yada yada"
09:46 huf DrForr_: yeah, but the actual version number could still be 6.1.0 :D
09:46 woolfy When we use the same naming conventions as Perl 5, we will have Perl 6.0.0 as the very first release of Perl 6.  The first minor update will be Perl 6.0.1, the next one Perl 6.0.2.  The new development version will be Perl 6.1.0.  The next major release of Perl 6 will be Perl 6.1.0.  I have no idea why this is difficult.
09:46 andreoss` This is perl, v5.8.8 built for x86_64-linux
09:46 huf andreoss`: yeah, perl 5 is ready for perl 6 now :)
09:46 huf wasnt back in the stone age
09:46 DrForr_ andreoss`: yes, that was before the documentation change was made.
09:47 DrForr_ I should know, I was listening to the conversation :)
09:47 nwc10 This is perl, version 5.000
09:47 woolfy Modern Perl: 2014 Edition. Updated for 5.18.  http://onyxneon.com/books/mo​dern_perl/mp_cover_full.png
09:47 nwc10 please everyone, also try perl -e 'use v20';
09:47 nwc10 and perl -e 'no 6'
09:47 huf call it perl 6 version 6.0.0
09:47 huf YAAAY
09:47 woolfy Well, everybody is entitled to my opinion.  There you have it.
09:48 huf only one more six needed to summon santa
09:48 dj_goku_ joined #perl6
09:48 El_Che anyway, it's important to find a way to make the perl5 and 6 people happy to stay one community. Emotions run high, when it could be a chance for both sister languages to reposition themselves.
09:48 El_Che confusion will hurt perl5 and 6.
09:49 [ptc] how was version 1 of perl 5 labelled?  was it 5.1.0?  (just interested)
09:49 timotimo i've read all of the ~120 commens on that article and it turns out only two came from people who've apparently already touched perl6 in any sort of fashion at all ... all the others were about perl5, python, php, ruby ...
09:49 nwc10 [ptc]: see above.
09:49 woolfy I will try do do marketing for Perl 6.  And marketing for Perl 6 will go on for many many years.  Calling it Perl 6.0 instead of Perl 6 1.0, instead of Perl 6 2015a, instead of Rakudo 1.0 or Camelia Perl 1.0, I am convinced is the easiest for marketing.
09:50 [ptc] nwc10: yeah, it wasn't clear to me, that's why I was asking
09:50 El_Che woolfy: not if it turns perl5 users off. Perl 5 needs space to grow and  Perl 6 needs perl5's popularity/userbase as a start
09:50 nwc10 [ptc]: I'm not near the machine with perl 4 built on it, so I can't offhand report what it says
09:50 masak woolfy: I agree, but I'll also note that I'm less convinced of the simple truth of that today than I was yesterday.
09:51 nwc10 but it's something stranger than 4.036, I think.
09:51 FROGGS woolfy: there are two different things to put a version stamp on: the spec (=testsuite) and the compilers like rakudo
09:51 [ptc] nwc10: ah, ok.  thanks :-)
09:51 masak FROGGS: pretty sure woolfy is aware of this :)
09:52 nwc10 woolfy: I also agree that there needs to be something bloody clear for marketing. I think that Pm has been thinking about all of this, and is close(r) to coming up with a sane plan.
09:52 FROGGS well, it was just treated equally in her sentence
09:52 nwc10 woolfy: it is unfortunate that he's not in your timezone
09:53 _mg_ joined #perl6
09:53 masak there is lots of interesting backlog from yesterday.
09:54 woolfy Perl 5.18 and Perl 5.20 and Perl 5.22 are perfectly good names for Perl 5 (for the years 2013, 2014 resp. 2015).  So are Perl 6.0, Perl 6.2 and Perl 6.4 (for 2015, 2016 resp. 2017).
09:55 huf btw, are a lot of recruits expected from perl5 land?
09:55 woolfy If anybody is going to look for a job in the future, things will happen just like now for Perl 5: if you are not capable of doing Perl 5.18 but are stick with 5.6, you have less of a chance of finding a Perl job.  Now Perl 6.0 comes in.  At the moment, not many jobs are available for Perl 6.0 capable people.
09:55 huf otherwise none of this even matters much...
09:55 [ptc] huf: one would hope so...
09:55 woolfy But I am quite sure that will change in the future.
09:55 masak huf: we're not pinning our hopes on it, I'd say.
09:55 nine [Tux]: pong
09:55 Humbedooh joined #perl6
09:56 pmurias nqp-m: sub slurpy(*%slurpy) {for %slurpy -> $key, $value {}};slurpy(:pivo<1>)
09:56 camelia nqp-moarvm: OUTPUT«(signal SEGV)»
09:56 pmurias nqp-j: sub slurpy(*%slurpy) {for %slurpy -> $key, $value {}};slurpy(:pivo<1>)
09:56 camelia nqp-jvm: OUTPUT«(signal ABRT)*** Error in `java': double free or corruption (fasttop): 0x00007fc394007c90 ***␤*** Error in `java': double free or corruption (out): 0x00007fc39401a8d0 ***␤*** Error in `java': double free or corruption (!prev): 0x00007fc39401a8d0 ***␤#␤# There…»
09:56 nwc10 huf: I believe the answer is (1) assume "que sera sera" (2) it would not be diplomatic to actively try to undermine any other language
09:56 pmurias nqp-j: say(1)
09:56 camelia nqp-jvm: OUTPUT«(signal ABRT)*** Error in `java': double free or corruption (fasttop): 0x00007fa12c007c40 ***␤»
09:57 huf nwc10: true
09:57 nwc10 campaign on the awesome, not negatively
09:57 moritz pmurias: I don't think -> $key, $value works in NQP
09:57 nwc10 also, please notice, the Raspberry Pi just went multicore, still for $35
09:57 nwc10 can $other-language actually use more than one core?
09:57 pmurias moritz: it segfaults
09:58 masak nwc10: I believe so.
09:58 woolfy FROGGS: I think we are past the "Perl 6 is just the specification".  By now, Perl 6 is a working programming language.   Did the Perl 6 people work 14 years on this to get dictated by other people how they should name their programming language?  Just because  of fear for their anger?  Aw, come on...
09:58 moritz pmurias: that's less than awesome :-)
09:58 nwc10 (not obvious when pypy STM reaches Christmas, and I've no idea whether there's an easy way to get Rubinus for it"
09:58 masak woolfy: I think you misunderestimate what "Perl 6 is just the specification" means...
09:58 andreoss` woolfy: should perl.org, cpan.org and etc. be reclaimed by Perl 6 in the future as well?
09:58 moritz woolfy: fwiw I agree that we shouldn't let other people dictate our versioning scheme
09:59 masak +1
09:59 woolfy masak: no, I think it has been an irritating sentence for years...
09:59 FROGGS woolfy: I am just saying that "Perl 6" is the language like Perl 5 is the language the perl5 interpreter implements
09:59 woolfy andreoss: I have no idea why you say that.
09:59 masak woolfy: that's why I think you've got it on the wrong foot...
09:59 moritz woolfy: but it's not obvious to me that 6.0.0 is indeed best for us, or easiest to market
09:59 pmurias huf: re recruits from perl5, for me the Perl 5 legacy is one of the main reasons I'm working on Perl 6
09:59 btyler but it would be (to put it mildly) inconsiderate to do something openly damaging to perl5, if there's another way.
09:59 woolfy Perl 6 is just Perl.  Like Perl 5 is Perl.  And Perl 4 was/is Perl too.
09:59 moritz woolfy: pmichaud++ wrote something to similar effect in yesterday's backlog
09:59 FROGGS so, "Rakudo Perl 6" implements the "Perl 6" language
10:00 nwc10 FROGGS: whoa, "Perl 5" is not *one* language, by that way of defining it
10:00 nwc10 which version of the interpreter?
10:00 huf pmurias: i suppose. it just feels like a church schism to me
10:00 nwc10 they are all subtly different
10:00 masak which platform and OS of the interpreter?
10:00 btyler I'd very much like to continue being a happy user of all the modern perls -- there are loads of really talented people in both 'camps' that have a ton to contribute in both directions
10:00 woolfy So perl.org is for Perl 4 and for Perl 5 and for Perl 6 and in the future, when Perl 6 has become a success, I hope there will be a Perl 7.0.0.
10:00 huf pmurias: with increasingly less hope of reconciliation. but let's hope
10:01 btyler but it seems to me that there needs to be some awareness here around what the stakes feel like for people really invested in perl5
10:01 nwc10 I rather hope there won't be a need for a 7.0.0
10:01 El_Che Whatever happens, if no positive consensus is found, the python 2/3 drama would look like peanuts compare to this. And we'll all loose.
10:01 masak woolfy: sorear and I have called dibs on Perl 7. we will design it while journeying to the stars in a generation ark.
10:01 btyler indeed El_Che
10:01 FROGGS nwc10: Perl 5 is one language, of course evolution happened and will happen
10:01 huf masak: why not a generational ork?
10:02 huf would be much better.
10:02 pmurias or a generation awk?
10:02 masak you people aren't taking generation arks seriously enough.
10:02 moritz huf: how many good languages do you know that were designed by orks? :-)
10:02 woolfy I don't see why all of this seems to sensitive for people.  With Inline::Perl5, all (almost all) of Perl 5 will run in Perl 6.  nine++ is working on Inline::Perl6 to have all (almost all?) of Perl 6 run in Perl 5.  The two versions are so near for the untrained eye, that they are more than sister languages.
10:02 huf moritz: ... all of them? :D
10:02 huf moritz: hint: the orcs are within us all!
10:03 chenryn_ joined #perl6
10:03 woolfy The term sister languages was just used in the past to keep the Perl 5 people happy.  Which did not work very well, seeing (hearing) all the snide remarks from so many Perl 5 people.
10:03 pmurias moritz: that's not funny, I'll be writing php in < 2 weeks :(
10:03 btyler woolfy: here is some discussion yesterday in #mojo. awesome people who have (and continue to make) awesome things in perl 5, who deserve tons of respect, who are generally excited and positive on perl 6: http://irclog.perlgeek.de/​mojo/2015-02-03#i_10052178
10:04 DrForr_ btyler: Indeed. That means that people doing support and writing new modules for perl5 code will be dealing with perl5.22, 5.24, 5.30, 5.70 and so on. While there may be amazing progress going on, it won't be seen that way from the outside. And participation will suffer just because of the perceived (yet incorrect) lack of progress.
10:04 moritz pmurias: sorry to hear that
10:04 woolfy "Job opening for programmer able to work with Perl 5.16 or newer, preferrably Perl 5.22. "   "This presentation will be about Perl 6.0.".  I think explaining the difference will be easy.
10:05 nwc10 pmurias: sigh, why have you taken a PHP job? :-(
10:05 [ptc] woolfy: that sounds good imo
10:07 DrForr_ "Job opening for perl5.16 or higher" "Wow, a perl6 opening! Cool! ... Oh, crap."
10:07 nine My personal goal is to make this whole naming discussion obsolete. There's a Perl 5 part of Perl and a Perl 6 part of Perl. And when you invoke perl it will be smart enough to figure out which part to use and you are free to mix and match them as you like. Who will then care all that much? Damnit, they can even be released together!
10:07 masak woolfy: "[sister languages meme] did not work very well" -- sorry, that's wrong. it saved us an outright civil war.
10:07 krunen joined #perl6
10:07 nine And the technical part of this can be done _this_ year.
10:07 huf postponed it perhaps :D
10:07 huf drove it underground, certainly.
10:07 El_Che nine:
10:08 masak woolfy: I find the idea that *this* is the moment for Perl 6 to assert itself as no longer believing in the sister languages message ridiculous and disruptive.
10:08 woolfy masak:  I've been called plenty of bad names for supporting Perl 6.  So that "sister languages" did not do me any good.
10:08 moritz huf: I haven't seen many jibes towards Perl 6 from Perl 5 folks in the last one or two years, with the exception from a few known trolls
10:08 El_Che nine: the inline stuff is amazingly cool, but there are many usecases wher /usr/bin/perl is the only option (sysadmin stuff, speed foor cli, etc)
10:08 huf moritz: no, instead it's treated as a nonentity
10:08 huf moritz: which is not better imho
10:08 nine El_Che: and how would unifying the Perls be a problem for that usecase?
10:08 moritz huf: it is
10:08 masak woolfy: I'm sorry you've been called bad names for supporting Perl 6. I didn't say it fixed everything.
10:08 woolfy I think Perl 5.22 and Perl 5.24 and Perl 5.26 and so on will live happily next to Perl 6.0.0 and Perl 6.2 and Perl 6.4 and so on.
10:09 pecastro joined #perl6
10:09 huf moritz: but i'll be the first to say i'm not the most optimistic person :)
10:09 El_Che nine: it won't. It would be fantastic. My point is that it will not cover 100% of use cases.
10:09 dalek perl6.org: b8f7fe0 | moritz++ | / (3 files):
10:09 dalek perl6.org: Replace "spefication" by "language design"
10:09 dalek perl6.org:
10:09 dalek perl6.org: Also remove outdated link
10:09 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/b8f7fe0e15
10:10 huf "this is perl 5.98.4+6.8.2"...
10:10 huf by our powers combined...
10:10 nine El_Che: haven't found one where the theory breaks down. The perl binary does not have to load the Perl 6 part if it detects only Perl 5 code.
10:10 woolfy The message for the world should be: "Perl 5 will exist for a very long time, so new products and new projects can be started anytime."  Any manager that does not understand this, is throwing away precious money.
10:10 nine El_Che: thus it's no different from now except for some lines of code for transparent Perl 6 support.
10:11 masak woolfy: as Perl 6 people we're essentially arguing from a *privileged* position of the higher version number.
10:11 masak woolfy: "Poor people should just accept that we're richer."
10:12 masak woolfy: "I don't understand why women get so upset about gender roles in society."
10:12 huf to which the other side says "rich people have no concept of real life, they just do their weird rich people things"
10:12 woolfy masak: anybody who talks bad of Perl 6 because it would mean that it is better than Perl 5, is in my eyes being nasty.  Having a higher version number does not mean it is better.  It is the next type of Perl.  Perl 6 still has to prove it is better than Perl 5.  Just like Perl 5 had to prove it was better than Perl 4.
10:12 huf except everyone covers it in prettier words so they can pretend they're not passive aggressive about it :D
10:13 El_Che nine: e.g. Linux distro's where only perl5 is installed is a use case.
10:13 nine woolfy: which is absolutely true. But unfortunately laypeople tend to simplify such things.
10:13 DrForr_ This naming by fiat guarantees there will never be a new major release of perl5. I understand that this community sees perl6 as the logical successor (in the perl 5++ sense), there are still lots of people in the perl5 community that would love to release a new shinier perl5 as 6, but that's been taken away by fiat, 15+ years ago.
10:13 woolfy Perl 6 was started to fix several problems in Perl 5.  Perl 5 has fixed some of those problems themselves and can be proud of that.  Perl 6 will have its own problems and will try to fix those.
10:14 masak woolfy: let's assume you're right. even so, you're *still* being unnecessarily bellicose about how you present the p5-p6 relation.
10:14 nwc10 I'm going to be blunt
10:14 nine El_Che: the language detection code does not have to depend on Perl 6 being installed. Just like right now it tells you, that you'd need Perl 6.0.0. Perl 6 can then be loaded dynamically (if installed) just like Inline::Perl6 does.
10:14 nwc10 there are lots of people who want a new Perl 5, but have no FUCKING CLUE what should be in it (that is actually acchievable) because none of them have the skills to actually do it.
10:14 Ugator joined #perl6
10:14 woolfy DrForr: the next major release of Perl 5 will be Perl 5.22.  rjbs++ already gave a very nice talk of that at FOSDEM.  The next one will be in 2016 and will be called 5.24.  What is the problem here?
10:14 * [ptc] has to look up the word bellicose...
10:15 huf DrForr_: most things in the world are named by historical accident, which we eventually just learn to accept ;)
10:15 woolfy I am trying to be very clear (and not bellicose) about not naming Perl 6 as "Perl 6 1.0" or "Camelia Perl 1.0" or "Perl 6 2015a".  Those names are marketing nightmares.  Perl 5 is what it is, and Perl 5.22 is on its way.  That has its own merits.
10:16 [ptc] nwc10: the bar is set very high in Perl 5 to be able to make contributions to it
10:16 [ptc] nwc10: basically all of the low-hanging fruit are pretty high up...
10:16 nwc10 [ptc]: this is a side effect of well, history
10:16 nwc10 pugs went that way too - the low hanging fruit vanished
10:17 DrForr_ huf: Yes, and in that sense the damage has been done. But this issue comes up after every convention, and we *still* bitch about it. Every time.
10:17 [ptc] nwc10: which makes it hard for newcomers to get a toehold and make a contribution
10:17 [ptc] nwc10: just saying...
10:17 huf DrForr_: yes, hence my despair
10:17 woolfy By the way, did you see these beautiful pictures of FOSDEM by Chris Jack?  https://plus.google.com/photos/104598318166​622233830/albums/6111769864082826929?sort=1
10:17 El_Che woolfy: perl 5.22 is a nightmare when a 6.0 is there, according to a marketing logic
10:17 huf DrForr_: the issue itself is almost too small to see compared to the damage it's doing
10:17 El_Che woolfy: that's what I mean by trying to find a positive compromise
10:17 huf or rather, the unending bikeshed around it
10:18 nwc10 [ptc]: there is no good way to fix this.
10:18 El_Che perl5 renaming itself as perl v 22 would be a markering nightmare to perl 6 release (hey it's 16 versions behind!)
10:19 woolfy El_Che: I really don't see that.  Perl 5.22 is the newest version of Perl 5 and will have a tremendous amount of users.  Perl 6 will have Perl 6.0.0 and will have less users (for now).  The naming makes clear which is which.  Perl 6.0 will not be the newest version of Perl 5, and is not meant to replace Perl 5.  Just like Perl 5.22 is not meant to replace 5.20, many people will keep on using Perl 5.20 for a while.
10:19 [ptc] nwc10: I know  :-/
10:19 nwc10 [ptc]: may of the people whom I helped to get to the point of being comitters haven't (obviously) stuck around enough to help the generation beyond
10:19 El_Che so, no compromise == we all loose (certainly those people that like to use and promote both versions)
10:20 DrForr_ So is it fair to say that this debate was lost in the last *millennium* and we should simply deal with the consequences?
10:20 woolfy El_Che: I think this is almost evil.  Not allowing Perl 6 to call itself Perl 6.0 is forcing Perl 6 to give itself a name that will be a marketing disaster for Perl 6.  That way, all the work on Perl 6 will become useless.
10:21 huf DrForr_: yeah. and talking about perl marketing seriously is a bit weird to me too. apart from people already in the know, the marketing of perl for the last 15 years has been ... sad.
10:21 huf DrForr_: i'm sure lots of great people tried, but the "brand" isnt in a good way
10:21 [ptc] nwc10: yeah, I know.  It's a pity really.
10:21 woolfy "Let's call it Camelia 1.0 and get it over with" is my nightmare.  Perl 6 will loose all connection to Perl, the fame of the brand name.  And just like Perl 5 was Perl next to Perl 4, Perl 6 is Perl next to Perl 5.
10:22 btyler woolfy: I don't think that's quite fair. perl 5 has a pretty clear "these are the supported releases" policy. older versions are openly and deliberately left behind
10:22 [ptc] nwc10: I'm guessing we discussed this in Salzburg??  (If I have decifered the nick properly...)
10:22 woolfy I've been doing a lot of sad stuff in the last year.  I've done marketing for Perl 5 and Perl 6.  Nobody ever noticed me.  I was so sad.
10:22 nwc10 It was me in Salzburg
10:22 nwc10 but I can't remmeber what I discussed there
10:22 [ptc] :-)
10:22 nwc10 as I was somewhat sleep deprived, stressed, and $other
10:22 woolfy btyler: people are still using Perl 5.6.
10:23 El_Che the choice is between making shure bot languages florish or both languages confusing the world outside the echochamber and being ignored. I have perl hackers here at work (outside of the community) that really have trouble understanding the perl 5 and 6 thing (6 is the new major release of perl that never came out, right? We'll use whatever old version RHEL supplies)
10:23 [ptc] nwc10: well, we had some good chats :-)
10:23 nwc10 most of those are close to resolving themselves, I'm pleased to say
10:23 huf woolfy: i didnt mean to minimize what you've done, only to say that the brand "perl" usually brings on a million idiots crying saying stupid things. and perhaps one or two people who've actually tried the language.
10:23 [ptc] nwc10: yay!
10:23 btyler woolfy: right, for various reasons. but perlpolicy is pretty clear that "we consider releases older than X to be unsupported and you should stop using them if you can"
10:23 huf woolfy: we all know how perl acquired this image, we can rehash it endlessly, doesnt change a thing.
10:24 woolfy huf: so, if we take into account what millions of idiots say about Perl, if that would be the guiding line of our life with Perl, maybe we should just all roll over and die.  No thank you.
10:24 huf you are the one bringing up marketing, that's exactly about convincing those millions of idiots.
10:24 btyler I just mean that drawing a comparison between perl 5.20/5.22 and perl 5/6 is maybe a bit complicated from that perspective
10:24 woolfy btyler: and at the moment, the oldest version of Perl 5 that is actively supported is Perl 5.8.1.  Which was released a long time ago.
10:25 virtualsue joined #perl6
10:25 woolfy El_Che: I think it is not that difficuly.  Maybe many people do NOT want to understand the difference.
10:25 nine huf: no, we don't have to convince the idiots. We have to convince the silent majority.
10:25 woolfy nine++
10:25 huf is there some way to make a decision about the perl6 version number thing that'll stick and nobody will ever bring it up again? :D
10:26 woolfy huf: no, I am afraid not.
10:26 DrForr_ The problem isn't the perl version, it's the distribution the perl version is running on. Changing a distribution is scary, especially if you don't have a good test suite for your application. And what application has 8taht* robust of a test suite?
10:26 huf god dammit.
10:26 pmurias nwc10: seems to be a decent place to work (they use other languages for their other projects), and if the PHP get too annoying (I haven't tried using it ever before), I'll try to just relocate to a different city and get a perl job
10:26 nine There is a solution!
10:26 nine Share the version number :)
10:26 El_Che woolfy: I am talking about people that like perl a lot as a chainsaw, that don't troll the language, but are not in touch with the community. A lot of (positive) people are in that space.
10:26 DrForr_ huf: It was made last millennium.
10:26 huf let's rename perl5 to camelia!
10:26 alini left #perl6
10:26 nine By unifying the languages on a technical level
10:26 huf and perl6 can be 6.0.0
10:26 nwc10 pmurias: sounds like some good reasons. I hope they turn out to be a nice, sane environment
10:27 tadzik what was the problem with 6.0.0 again?
10:27 nwc10 pmurias: and good luck subverting their language choices :-)
10:27 woolfy El_Che: and for them, nothing changes, they can happily ever after use Perl 5.20, Perl 5.22, Perl 5.24 and so on.  No problem there.
10:27 nine Perl 6.1.3 will support both the language we now know as "Perl 6" and the language we know as "Perl 5"
10:27 DrForr_ nine: perl5 is already getting dropped from distros for having too much crap in core, perl5+6 runtime would probably blow that out of the water.
10:27 El_Che woolfy: it will. They think that perl6 obsoletes 5 and they have zero interest in invest in learning 6.
10:27 woolfy nine: well volunteered.  :-)
10:28 DrForr_ This is also a reason why we're pulling stuff like CGI.pm out. As it is, distributions pull documentation because it's too much space.
10:28 El_Che woolfy: they they will move to something else installed on RHEL.
10:28 nine woolfy: I volunteered for that months ago. Just trying to convince people that all this energy spent in this discussion might actually be wasted :)
10:28 woolfy El_Che: press release: Perl 6 does not obsolete Perl 5.  Please continue to work with Perl 5.  If you have some spare time, try out Perl 6 too and see if you can use it too."
10:28 huf DrForr_: too much space? what. what. what are people running these distros on?
10:28 huf dish towels?
10:28 tadzik from this discussion I liked the link to G+ pictures :P
10:28 tadzik woolfy++
10:28 El_Che woolfy: *I* know that. I am talking about a big part our intended audience.
10:29 huf "perl 5 and perl 6 together turn the perl up to eleven..."
10:29 [ptc] tadzik: agreed, the photos are good :-)  woolfy++
10:29 huf because spinal tap is sooooo popular
10:29 El_Che woolfy: and sadly, version numbering is a very logical choice to draw -in our case- false conclusions
10:29 woolfy nine: I know.  But the discussion still happens.  I've told many people that your work on Inline::Perl 5 for Perl 6 and Inline::Perl 6 for Perl 5 (and Inline::Python is interesting too) will make a bridge between Perl 5 and Perl 6.  Add to that the developments of v5 and we can all live happily together.
10:30 pmurias nwc10: a friend who works there is already running haxe code compiled to js on his custom made js-on-c# interpreter ;)
10:30 El_Che woolfy: nice pics!
10:30 DrForr_ huf: Not what they're running *on*, what they're running *from*. If it can't fit on the CD ISO it's not going to make it into the distro, in lots of cases.
10:30 woolfy El_Che: people will get used to new things.
10:30 woolfy Ow, careful, not my pics!  Pics are from Chris Jack.  Chris++
10:30 huf DrForr_: oh. huh. that's a use-case i havent thought of since ... ever :) but i suppose people still install distros without internet
10:30 El_Che woolfy: I am not convinced of that :(
10:31 alini joined #perl6
10:31 El_Che woolfy: python2/3
10:31 woolfy El_Che: that is Python.  Perl is not Python.
10:31 El_Che woolfy: we should learn from them instead of making fun of them. (And the split here is potentially a *lot* bigger)
10:31 [ptc] the split here is *much* bigger
10:32 woolfy The message must be: Perl 5 has existed for over 20 years, and will exist for many years to come.  Next to Perl 5 we now have Perl 6, which is slightly different, and when you use Perl 6, you can still use all of your Perl 5 code.  Perl 6 will live next to Perl 5.  If you are used to Perl 5, please continue to do so.
10:32 DrForr_ Well, they're already the main group of trolls on the #perl's.
10:32 nine But our split can be healed. Theirs can't. Or at least they refuse to.
10:32 btyler I think you find few python developers who would (given equal library availability) choose 2 over 3. it is almost always "the software I need doesn't run in py3k yet -- when it does I'll move"
10:32 tadzik python doesn't really have that problem, we might. More similar real-world confuson is probably "C/C++" in job offers
10:33 [ptc] btyler: agreed
10:33 woolfy Again, don't underestimate the work on Inline::Perl(5|6) and use v5.  This will mean that both languages can be used simultaneously.
10:33 nwc10 does anyone have updated stats on Python 3 adoption, a year on from https://alexgaynor.net/2013/dec/30/about-python-3/  ?
10:33 [ptc] btyler: I program in Python at $day_job, and that's my opinion :-)
10:33 chenryn_ php would be version 7...
10:33 El_Che tadzik: no one confuses version numbers (or standards) of C and C++
10:33 El_Che tadzik: sadly, in our case...
10:34 [ptc] nwc10: no stats.  The feeling I get is that python3 is an inevitablity
10:34 tadzik just like no one will ever confuse Perl 5 and 6 the moment they look at it
10:34 tadzik I still don't see what the problem is
10:34 [ptc] nwc10: Guido has said that there won't be a py2.8
10:34 DrForr_ huf: Talk with rjbs about the distro problem, I think. At least I think that's who I haerd it from.
10:34 El_Che woolfy: idd. Inline is very exciting. I hoppe the fosdem video is out soon.
10:34 tadzik other than "people will think that 6.0 is a newer version of 5.22", which I think is really, really overdramaticised
10:34 tadzik people who are competent in either will never confuse these
10:35 woolfy El_Che: since I did not see any of the talks at FOSDEM, I hope all videos will be online soon.
10:35 woolfy tadzik: exactly my opinion
10:35 tadzik I *still* don't see what the fuss is about
10:35 nine There are still people who think Perl 6 is a wrapper around Perl 5 written in Haskell
10:35 DrForr_ The people that are competent in one or the other ARE NOT THE CROWD TO WORRY ABOUT.
10:35 woolfy So maybe we will listen to TimToady who said that we will see later this year how things will be called.
10:36 El_Che tadzik: people are newbies before 'comptent'. Perl5 also needs newbies. Newbies thinking perl5 is an old release is bad. IMHO
10:36 woolfy DrForr: we should not worry at all.  Perl is strong.  Perl-people are smart.  We will deal with it.
10:37 tadzik El_Che: Yes. It's also unfixable, and every 3 months that we have this discussion we come to the same conclusion
10:37 DrForr_ Then wny are we yelling?
10:37 DrForr_ *why
10:37 woolfy El_Che: anybody who thinks perl5 is an old release, is not relevant.  At FOSDEM I had talks with people that were convinced that Perl-code looks the same after compression as before, and they never worked with any version of Perl.  Those people are completely irrelevant.
10:37 tadzik DrForr_: that's what I'm trying to establish
10:37 tadzik teaching people the difference doesn't take more than 5 minutes
10:38 Sqirrel joined #perl6
10:38 El_Che tadzik: there isn't always someone around to teach the difference.
10:39 El_Che on a happier note, I need a new phone, are there happy/unhappy jolla users here?
10:39 tadzik there also isn't always someone around to teach people Perl
10:39 tadzik I don't see how this is relevant, or related
10:39 tadzik El_Che: happy
10:39 tadzik see also sergot and sjn :)
10:39 woolfy Anybody who is saying that with Perl 6 all Perl 5 is old, is being evil, and trolling, and lying, and trying to make us angry.  Not a good thing, and it will be remedied by us all, people from both Perl 5 and Perl 6, by telling the truth, which is, Perl is awesome and will be working well for almost an eternity.
10:39 DrForr_ Wait a minute, the people that don't already use perl are irrelevant to the discussion? Then we're inside an echo chamber. Which of course we are, since this is a chat network using a protocol from the 80s which is by now *completely* hidden from the rest of the world.
10:39 El_Che there is a -50€ fosdem discount
10:40 tadzik does that bring it down to 200€? Dayum
10:40 woolfy DrForr: don't you troll on my remark.  That was not what I said.
10:40 El_Che code is FOSDEM2015
10:40 woolfy DrForr: I said: "At FOSDEM I had talks with people that were convinced that Perl-code looks the same after compression as before, and they never worked with any version of Perl.  Those people are completely irrelevant."
10:40 * El_Che off to post offcie to send old phone back to amazon under warranty :)
10:41 woolfy DrForr: please beware the "and" between "being convinced that perl-code..." and "never worked with Perl".
10:42 DrForr_ I don't think I agree that they're irrelevant. If they walk over to the booth purely to badmouth the language and walk away, that's one thing.
10:43 pecastro joined #perl6
10:44 kjs_ joined #perl6
10:44 rurban joined #perl6
10:44 espadrine hating on perl for its sigils is the new hating on lisp for its parentheses
10:45 pyrimidine joined #perl6
10:48 Tux__ nine, https://gist.github.com/Tux/6520a23f118908ae8237
10:48 DrForr_ A different way of looking at that is "Those signs are silly. Explain to me what they mean and I might change my mind."
10:49 anaeem___ joined #perl6
10:49 huf they're almost never *that* open-minded
10:49 jluis joined #perl6
10:50 ab5tract DrForr_: you are giving that crowd too much credit
10:50 huf it's usually "convince me that my snap judgement is incorrect. go on, try hard. dance."
10:50 DrForr_ Online, to be sure. But in your face? That's a different story.
10:50 huf i've encountered this irl too
10:50 ab5tract it's trendy to hate on perl, so trendy people hate on perl. and the programming world is *full* of trendy people
10:51 huf we're all about retroactively inventing reasons to prop up our hasty feels-based judgements
10:51 nine ab5tract: those trendy people are the ones writing the code the next generation of trendy people will hate a language for.
10:51 huf but vim is seriously the best editor because butts.
10:51 btyler also there's a strain of the python community that seems to be built around hating on perl. the last page of the big "programming python" tome is a yoda-style joke about how perl is the dark side
10:51 btyler I went to a python meetup, mentioned I did perl, someone asked what it was, the answer they got was "the old python"
10:52 DrForr_ So... the answer then is to do nothing, because everyone that talks to you already has their mind made up. So we shouldn't bother changing our behavior.
10:52 btyler so..yeah. productive discussion is tough
10:52 ab5tract btyler: yup. but that's very historical. and the perl community fueled that hate
10:52 huf DrForr_: yeap. dehumanize yourself and face to bloodshed.
10:53 dalek nqp-js: 73dcbe1 | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
10:53 dalek nqp-js: Fix bugs in checking for emptiness of iterators.
10:53 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/73dcbe1b68
10:53 dalek nqp-js: 47396cc | (Pawel Murias)++ | t/nqp/68-associative-for.t:
10:53 dalek nqp-js: Test for iterating over a hash manually using nqp::iterator.
10:53 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/47396cc375
10:53 ab5tract because of a relatively natural, "hey this new kid is eating my lunch!" response. (when perl was the giant and python the david)
10:53 nine Tux__: looks like all that's missing is support for passing a file handle to Perl 5 code?
10:53 ab5tract nine: that's why i'm very excited about perl 6. leapfrogging the trends :D
10:53 Tux__ possibly
10:53 huf oh come on, i think most of the perl hate comes from many people having to maintain horribly shitty perl codebases without first learning perl.
10:53 huf that sort of experience tends to shade their perceptions for a long time
10:54 tadzik most of the people haven't ever seen perl
10:54 tadzik they just heard that it's bad
10:54 ab5tract huf: the perl <-> python rivalry is deeper than that
10:54 huf okay, yeah, by now it's tribal wisdom of the internet
10:54 [ptc] sounds like the same kind of hate people have for fortran
10:54 tadzik we all love laughing at cobol, who has ever written it?
10:54 [ptc] it is (at least theoretically) possible to write good fortran code...
10:54 huf we love being mean to the "others" :)
10:54 nine I've seen worse languages than COBOL...
10:55 huf nine: true for me too, since i've never seen cobol :)
10:56 ab5tract [ptc]: there you are, continuing unnecessary downtalking of Fortran
10:57 huf can we stop being so reasonable before we start singing the praises of php?
10:57 nine COBOL does have some good excuses for its quirks. Yes, it tends to be somewhat verbose, but it's also more than 50 years old! PHP on the other hand was just developed by clueless people...
10:57 [ptc] ab5tract: actually, I really like fortran.  It's a great numerical language, and a good one to have in the toolbox
10:57 ab5tract "it is (at least theoretically) possible to write good $language code..."
10:57 ab5tract my point exactly :)
10:57 [ptc] ab5tract: however, I've seen some *awful* code written in fortran....
10:58 huf [ptc]: shocking. most code is awful.
10:58 [ptc] ab5tract: yeah, you have a point :-)
10:59 [ptc] ab5tract: I've also seen some *awesome* code in fortran.  And I think that was because one *didn't* have the tool support when it was written.  Hence it *had* to be clear and well documented etc
10:59 huf but can real programmers write java in fortran?
10:59 huf (i know they can write java in perl5)
11:00 [ptc] huf: well, you can do OOP in Fortran now, so quite possibly
11:00 huf \o/
11:01 jluis joined #perl6
11:01 ab5tract regarding PHP: we either learn from the successes of that language, or we repeat some failures of Perl 5
11:02 ab5tract the number one there being plug-and-play deployment
11:02 DrForr_ Ashton's Law.
11:08 Juerd ab5tract: Oh, you mean like how C, Python, Java, etc, etc, have failed, because they couldn't be used out of the box on shared webhosting providers?
11:08 Juerd ab5tract: What you're saying is only true for that specific niche.
11:09 huf also the success of any language is in large part accident, and has nothing to do with any features or niches or anything
11:11 moritz huf: I'm sure the success of Java is in large part due to massive investment and marketing by Sun
11:12 lizmat feels like it's going to take at least a week to catch up on all of the backlog and related posts
11:12 moritz huf: as well as carefully chosing a narrative ("simple, easy to program") and sticking with for quite long
11:13 huf moritz: hmm. yes, but far fewer people would've been interested in that narrative without c++
11:13 El_Che lizmat: you're lucky, you got in when the bashing was on other languages :)
11:13 huf moritz: i'm not saying it's all complete accidents. just that accidents play a very large part. sure, a coherent message helps.
11:14 Ugator left #perl6
11:16 FROGGS Perl 6 "just" needs to be branded as the 'Next Generation Enterprise Programming Language"... then we can just wipe away Java :D
11:16 FROGGS afk
11:16 huf you cant just afk after dropping *that*
11:16 * FROGGS reads the rants about that in a bit
11:18 chenryn_ joined #perl6
11:19 El_Che FROGGS: Moose already reserved the "post-modern" adjective :)
11:19 denis_boyun joined #perl6
11:19 El_Che (OMG, I *hate* postmodernism :) )
11:21 ab5tract Juerd: no, i am saying that even PHP has things to teach us.
11:22 ab5tract Juerd: and the difficulty of deploying ruby apps has played a part in it's "spiral"
11:23 FROGGS El_Che: the Enterprisy bit is important to get in the Java settled area
11:24 El_Che Java 2 Platform, Enterprise Edition 5 for the win!
11:44 jluis joined #perl6
11:44 dalek rakudo/nom: 756a4ac | lizmat++ | src/core/Array.pm:
11:44 dalek rakudo/nom: [].VAR.name returns something more obviously wrong
11:44 dalek rakudo/nom:
11:44 dalek rakudo/nom: Namely "$ANON_VAR__10".  The "$x" seen originally, was an artefact of the name
11:44 dalek rakudo/nom: used of the variable used to itemize in the circumfix sub.  Hopefully, we can
11:44 dalek rakudo/nom: adapt nqp::p6var to not actually have a .name in the case of variables matching
11:44 dalek rakudo/nom: this template.
11:44 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/756a4acd40
11:46 lizmat FWIW: +1 for putting NativeCall into the core
11:46 nwc10 But COBOL is now more populare than Ruby. I read it on the Internet, it must be true
11:46 nwc10 and my naughty fingers still can't type
11:46 masak I'd like to take this moment to thank all the people who self-identify as predominantly Perl 5 programmers, but who hang out here and occasionally provide their opinions. heartfelt thank you on behalf of the Perl 6 community. it's a way to sanity-check us and be able to see things from across the fence.
11:47 tadzik lizmat: which core? :) I
11:47 tadzik I'd vote for First Core though, along with lib.pm
11:47 nwc10 you can't be me - I'm self identifying as "trouble maker"
11:47 nwc10 s/be/mean/
11:47 nwc10 brain fail
11:47 nwc10 fingers fail
11:47 nwc10 something.
11:47 masak also, PSA: the opposite of "win" is spelled "lose" with one "o". the opposite of "tight" is spelled "loose" with two.
11:47 nwc10 should I get more coffee?
11:47 masak nwc10: or less, it's up to you :P
11:47 tadzik easy to remember because loose is less tight than lose :P
11:47 lizmat tadzik: yes, first core, like lib.pm / Test.pm
11:47 tadzik lizmat: +1
11:47 nwc10 but not fewer coffee? :-)
11:49 moritz nwc10: yes, tiobe publishes funny numbers
11:51 masak they're only funny if you don't take them seriously.
11:51 nwc10 they have this wonderful spike up in the past month for COBOL and some SAP-related language
11:52 nwc10 and a page that doesn't render without JS
11:52 nwc10 and has a JS error on Firefox
11:52 nwc10 competant, they are.
11:54 nwc10 oh, curious, https://github.com/rubinius-x/rubinius-x has been eliminated
11:56 ven (good :P)
11:56 nwc10 I wonder where the source code is now
12:00 pecastro joined #perl6
12:05 kjs_ joined #perl6
12:11 fhelmberger_ joined #perl6
12:12 sjn FROGGS: can you be ready & online at 18:00 tonight to see if we can get the arnsholt's NativeCall session stream working?
12:13 charsbar_ joined #perl6
12:14 Util_ joined #perl6
12:15 moritz_ joined #perl6
12:15 gugod1 joined #perl6
12:15 huf_ joined #perl6
12:15 ugexe_ joined #perl6
12:16 rurban_ joined #perl6
12:17 bonsaikitten joined #perl6
12:19 simcop2387_ joined #perl6
12:20 pmurias when is the NativeCall hackathon planned?
12:20 nwc10 [Coke]: "if ther's a ticket you think must be resolved before we ship" then attach it as a dependant ticket in RT, on the release ticket. RT makes this easy, and it's what Perl 5 has been doing for a while
12:21 camelia joined #perl6
12:21 dakkar joined #perl6
12:23 tadzik pmurias: today
12:24 pmurias what time?
12:24 pmurias tadzik: where is it taking place?
12:24 tadzik pmurias: oslo
12:24 tadzik evening, not sure how late in the evening :)
12:25 tadzik you'll have to ask sjn
12:25 FROGGS sjn: I can
12:25 FROGGS bbl
12:26 sjn pmurias: it's strictly not a hackathon
12:26 sjn it's a presentation + hands-on teaching
12:26 sjn "workshop"
12:27 molaf__ joined #perl6
12:34 ab5tract ven: what's your beef with rubinius-x ?
12:34 arnsholt pmurias: 1800 CET
12:38 El_Che sjn: mmm ubuntu phone or jolla...
12:39 rmgk_ joined #perl6
12:41 sjn El_Che: Jolla's is ok. I like it, but it's far from perfect
12:41 sjn dunno anything about ubuntu's
12:41 sjn Jolla are pretty responsive though
12:42 El_Che sjn: does jolla compare positively to rooted android?
12:43 sjn never had an android
12:43 sjn but if you want root, Jolla's great :)
12:44 sjn uses rpm as a package manager, comes preinstalled with Perl5
12:44 sjn El_Che: tadzik mumbled something about wanting to compile Perl6 on the Jolla phone too
12:44 sjn dunno if that got any traction
12:44 tadzik yeah, I tried
12:45 sjn (it would be a nice way to
12:45 sjn ...nice milestone to reach though)
12:46 * sjn imagines it would be possible with some careful use of swap
12:46 nwc10 tadzik: what/how did you try? I think that it ought to be possible to compile MoarVM natively (on it)
12:46 nwc10 and NQP and then Rakudo elsewhere
12:46 nwc10 and then just ship the Rakudo install over, to the same path
12:46 jluis joined #perl6
12:47 tadzik nwc10: yes, moarvm ran fine
12:47 tadzik nwc10: running any moarvm bytecode on it resulted in SIGBUS iirc
12:47 nwc10 aha, OK, when?
12:47 tadzik well, "any" being nqp
12:47 virtualsue joined #perl6
12:47 tadzik I think about half a year ago
12:47 tadzik I have not tried for a while
12:47 nwc10 if you have time, try again
12:48 tadzik noteworthy: I was building in a VM sdk, not on-device
12:48 tadzik nwc10: will do
12:48 nwc10 The ARM alignment fixes were May last year I think
12:48 nwc10 so it should now work on all ARM
12:48 nwc10 but this isn't tested
12:48 tadzik oh, I did not even have this phone back then
12:48 nwc10 mmm OK
12:48 pmurias Jolla phones?
12:48 tadzik yes
12:48 tadzik but I can try again and squeeze out as much debugging info as I can
12:49 nwc10 that would be useful
12:49 kaare_ joined #perl6
12:49 nwc10 I may be able to give a suggestion pretty quickly if you can get a line number in a C source file
12:50 nwc10 if you can do that, also contents of /proc/cpuinfo from it would be useful
12:53 ab5tract http://blogs.msdn.com/b/dotnet/archive/20​15/02/03/coreclr-is-now-open-source.aspx
12:53 ab5tract so, who's up for porting NQP to the CLR? :)
12:54 nwc10 well volunteered!
12:55 tadzik :)
13:00 alini joined #perl6
13:03 * TimToady woke too early :)
13:04 TimToady but waking up, I thought of more alphabetic naming schemes
13:04 El_Che TimToady: how was the flight?
13:04 TimToady fine
13:05 El_Che great
13:05 _mg_ joined #perl6
13:07 chenryn_ joined #perl6
13:08 TimToady scifi destinations, gems, substances, flying things, utopias, adjectives, particles, and so on
13:09 TimToady course, there's alway the old standby Abel Baker Charlie, etc
13:09 El_Che "Next Generation" :)
13:09 TimToady but we can probably choose our overriding metaphor
13:09 TimToady minerals and gems are solid
13:09 TimToady flying things...fly
13:09 TimToady utopias are dreams
13:10 TimToady adjectives are problematic insofar as ubuntu's short names tend that way
13:10 TimToady fruits, vegetables, spices
13:10 kjs_ joined #perl6
13:10 TimToady there are just lots of ways to adorn the 6a, 6b, 6c idea
13:11 donaldh joined #perl6
13:11 nwc10 Ubuntu's short names also seem a bit long. My opinon currently is that ideally it would (mostly) be a sequence of single words, two sylables, easy to pronounce and spell
13:11 TimToady and going a b c doesn't commit us to a schedule
13:11 TimToady yes, short is also good
13:11 moritz ... and limits us to 26 releases :-)
13:11 nwc10 I suspect that that's important. The spec shouldn't imply any sort of schedule release
13:12 moritz unless we enter the weird realm of non-ASCII collation
13:12 nwc10 whereas the distro (or compiler) releases probably should try to be regular
13:12 nwc10 moritz: but isn't that something that Perl 6 is good at?
13:13 moritz nwc10: but perl 6 users might not be :-)
13:13 nwc10 details!
13:13 moritz (nor are all Perl 6 programmers, yours truly included)
13:16 fhelmberger joined #perl6
13:17 kjs_ joined #perl6
13:23 [ptc] hrm, I'm getting an unhandled exception error when running htmlify.p6 through perl6-debug-m
13:25 [ptc] Here's the complete traceback: http://pastebin.com/q3Cr654M   Any ideas?
13:26 nwc10 [ptc]: no, but where do I get htmlify.p6 from? I have an ASAN build handy - I can quickly tell you if it's a C level bug
13:26 [ptc] nwc10: perl6/doc
13:28 nwc10 or not, as I fail with
13:28 nwc10 Could not find Debugger::UI::CommandLine in any of: /home/nicholas/Perl/doc/lib, /home/nicholas/Sandpit/moar-​san-jit/languages/perl6/lib, /home/nicholas/Sandpit/moa​r-san-jit/languages/perl6
13:29 [ptc] nwc10: you need to install that via panda first
13:29 nwc10 where's the TFM that I should R about that?
13:30 [ptc] urm, dunno.
13:30 nwc10 so much for "quickly" :-)
13:30 [ptc] hehe :-)
13:30 [ptc] nothing goes quickly ;-)
13:31 nwc10 https://github.com/tadzik/panda/
13:31 yeahnoob joined #perl6
13:32 nwc10 note that this is the first thing that I've done above NQP level for, well, quite a while.
13:33 TimToady on the spec vs implementation thing, my current thinking is along the lines of what pmichaud++ mentioned earlier, that the initial implementations of a standard target a draft standard, and then then actual standard is negotiated mostly only among the groups who have first implemented the draft
13:33 tadzik nwc10: I really recommend https://github.com/tadzik/rakudobrew :)
13:33 TimToady as long as we only have rakudo, this negotiation is easier :)
13:33 nwc10 OK, but I already *have* a Rakudo, built with somewhat whacky custom options
13:33 tadzik the nice thing about it is that it also updates your panda-stuff after rakudo update, so you don't have to do it manually
13:33 TimToady it's known to target a draft
13:33 tadzik alrighty then :)
13:34 nwc10 tadzik: ==> Please make sure that /home/nicholas/Sandpit/moar-sa​n-jit/languages/perl6/site/bin is in your PATH
13:34 TimToady but basically, we force convergence among all draft implementors, to the extent possible
13:34 TimToady and we do this every cycle
13:34 nwc10 um, why? In as much as, why can't I just copy or symlink the panda binary out somewhere else, and it know where it was installed?
13:35 nwc10 AAAAAAAAAAAARGH
13:35 nwc10 /usr/bin/env: perl6: No such file or directory
13:35 nwc10 that is not a universal solution
13:35 nwc10 SRLY
13:35 tadzik nwc10: because when you install something with panda that puts stuff there it won't be available for you
13:35 nwc10 ah OK. hmm. OK.
13:35 tadzik p6doc, lwp-request or whatnot
13:36 tadzik the thing about env, yeah, I can see why it's annoying
13:36 [ptc] I've struggled with that too in my setup
13:36 tadzik I may have a bug opened for it for which I can bump the priority
13:36 nwc10 perl6 is not in my PATH. it's not a relocatable binary. It already has its location hardcoded
13:36 tadzik ah, that'd be https://github.com/tadzik/panda/issues/28
13:36 tadzik for which FR<TAB><TAB> says it's fixed in a branch
13:36 tadzik I guess I could cherry-pick that onto master
13:37 nwc10 we have the same $turd with the spread of #!/usr/bin/env in CPAN modules, which breaks the designed hack/feature of EU::MM to rewrite #! lines to work on installation
13:37 [ptc] nwc10: I've got perl6 in a non-standard location and it's in my PATH; the panda stuff works ok
13:37 kjs_ joined #perl6
13:37 [ptc] nwc10: there's a lot of manual stuff required to keep things up to date tho'
13:37 nwc10 in that case, yes, if you're not using EU::MM, /usr/bin/env is probably the best solution
13:37 nwc10 but if you are, it's not.
13:37 [ptc] oh
13:37 tadzik nwc10: one workaround I can think of is 'your-perl6 `which panda` <arguments>'
13:37 nwc10 I can fix PATH
13:37 nwc10 but it's *not* going in .profile
13:39 nwc10 because, like perl 5, I have more than one perl 6 installed
13:39 nwc10 and I have reason to want to use different ones for different things
13:39 tadzik right
13:40 nine Why is perl6 a shell script anyway instead of a binary?
13:40 tadzik so the preffered solution for you would be for panda to rewrite its shebang on installation to reflect that?
13:40 moritz nine: because nobody has taken nwc10++'s patches yet, made them readier, and applied them
13:40 nwc10 tadzik: yes. but I don't know what other assumptions that would break
13:40 TimToady if different versions of something are going to be supported simultaneously, you need a namespace to negotiate that in, basically, and mechanisms for aliasing longnames to shortnames
13:41 nine moritz: where are those patches?
13:41 moritz nine: on the perl6-compiler mailing list
13:41 nwc10 mine: in particular, fixing them require some policy decision about how to handle build tools that are common to NQP and Rakudo
13:41 moritz nine: I'll forward the mail to you if that's any help
13:42 TimToady one can see why microsoft came up with a registry, inadequate though that turned out to be
13:42 nine moritz: I'm definitely curious
13:43 nwc10 [ptc]: OK, I get the same error. So it's not hitting any C level undefined behaviour
13:45 zakharyas joined #perl6
13:45 psch https://github.com/rakudo/rakudo/pull/330 has another patch for a binary target and references the branch
13:45 psch *branches
13:45 [ptc] nwc10: ok, good to know :-)
13:46 [ptc] nwc10: thanks for looking into it!
13:49 dalek star: cd30f37 | moritz++ | modules/rakudo-debugger:
13:49 dalek star: Remove modules/rakudo-debugger
13:49 dalek star:
13:49 dalek star: this was renamed to modules/debugger-ui-commandline
13:49 dalek star: review: https://github.com/rakudo/star/commit/cd30f3790d
13:50 nine Seems like quite a few people would like to see this happen :)
13:51 alpha- how would one get a division to say 50th digit after dot ?
13:51 TimToady FatRat
13:51 moritz m: say FatRat.new(1, 17)
13:51 camelia rakudo-moar 756a4a: OUTPUT«0.058824␤»
13:52 nwc10 nine: I believe that I managed the most dogfood (with a lot of help from FROGGS and jnthn)
13:52 alpha- nice
13:53 moritz it just seems we have no method to stringify with arbitrary precision
13:53 TimToady I'm not sure we actually have a method to print out all 50 digits though
13:53 psch m: say FatRat.new(1, 10 ** 50)
13:53 camelia rakudo-moar 756a4a: OUTPUT«0.0000000000000000000000​0000000000000000000000000001␤»
13:54 psch m: say sprintf("%.50f", FatRat.new(123, 10 ** 52))
13:54 camelia rakudo-moar 756a4a: OUTPUT«0.0000000000000000000000​0000000000000000000000000001␤»
13:54 TimToady .base probably needs a scale optional argument
13:55 * psch might have misunderstood the question
13:55 TimToady well, I suspect there are really two parts to it
13:56 TimToady how to get that precision, and how to output it
13:56 TimToady I was trying to answer both parts in sequence :)
13:57 TimToady but looks like %.50f will work, at least in decimal
13:57 TimToady would be nice to have it in any radix though
13:58 TimToady currently the .base method only takes a $base, and guesses on the precision
13:58 TimToady we should have a way to override that guess
13:59 moritz same for .Str
14:01 TimToady not sure about that one
14:02 * TimToady still thinks coercions should be thought of as 1-to-1
14:02 TimToady and if you can't coerce it reasonably, you should use a different function
14:02 rindolf joined #perl6
14:02 TimToady but I'm not hard and fast in that opinion
14:04 FROGGS joined #perl6
14:04 TimToady but things like coercion types Str(Cool) are not going to want to have optional parameters
14:04 xfix joined #perl6
14:05 TimToady and even if we allowed them, different source types might want entirely different optionsl
14:05 TimToady so I think we should reserver coercion for easy things, not hard things
14:05 alpha- m: say sprintf("%.20f", FatRat.new (22) / FatRat.new (7))
14:05 camelia rakudo-moar 756a4a: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/toggH_WGPdâ�¤Unable to parse expression in argument list; couldn't find final ')' â�¤at /tmp/toggH_WGPd:1â�¤------> [32msay sprintf("%.20f", FatRat.new [33mâ��[31m(22) / FatRat.new (7))[0mâ�¤   …»
14:05 TimToady gah, can't type today
14:06 anaeem1_ joined #perl6
14:07 TimToady m: say sprintf("%.20f", FatRat.new(22) / FatRat.new(7))
14:07 camelia rakudo-moar 756a4a: OUTPUT«3.14285714285714000000␤»
14:07 TimToady m: say sprintf("%.20f", FatRat.new(22) / 7)
14:07 camelia rakudo-moar 756a4a: OUTPUT«3.14285714285714000000␤»
14:07 TimToady it looks a bit like it's going via floaters
14:07 alpha- why cant it find final (
14:07 alpha- in my example
14:07 TimToady no whitespace allowed before parenthesized args in standard p6
14:07 alpha- wow
14:08 alpha- I am so used to just ignore whitespace since it's allowed everywhere in p5
14:08 alpha- ok...
14:09 moritz m: printf '%.20f', FatRat.new(22, 7)
14:09 psch alpha-: https://github.com/FROGGS/p6-Slang-Tuxic exists, although i don't know if it already handles methods
14:09 camelia rakudo-moar 756a4a: OUTPUT«3.14285714285714000000»
14:09 moritz psch: it does
14:09 moritz uhm, are those trailing zeros correct?
14:09 ab5tract alpha-: iiuc, it enables p6 to be a lot cleaner about lists (and function parameters) than p5
14:09 TimToady no
14:09 moritz m: printf '%.50f', FatRat.new(22, 7)
14:09 camelia rakudo-moar 756a4a: OUTPUT«3.142857142857140000000​00000000000000000000000000000»
14:09 TimToady that's why I said it's going via num
14:09 cdc joined #perl6
14:10 moritz looks like it
14:10 TimToady and why .base needs an option, I suspect
14:11 moritz .oO( it's all about the .base )
14:12 * nwc10 is wondering whether it's possible have a naming convention that is puns
14:12 [ptc] all your .base are belong to .us
14:13 ab5tract alpha-: it's not quite the same, but if you want/need whitespace, $x.method\ (@args); should work
14:13 alpha- ew
14:13 ab5tract or Slang::Tuxic :)
14:14 ab5tract alpha-: i'd rather have all the good stuff in p6 than a space between my methods and their args
14:14 TimToady yes, it does reduce ambiguity significantly to do it this way
14:14 TimToady and fits in with "all postfixes exclude whitespace by default"
14:15 TimToady and the other way, you get the p5 "print (3*2), 42" problem
14:15 alpha- might actually be good
14:15 alpha- one less thing to fight about with regard to style and indentation
14:15 ab5tract right, like %hash{} # one of the few places {} doesn't introduce a lexical scope
14:16 TimToady well, we're still not trying to be a boa constrictor language, but we do squeeze in certain strategic spots
14:16 ab5tract if you allow %hash {key}, then you're {...} blocks become less clean
14:16 ab5tract your processing of those blocks become less clean, i mean
14:16 TimToady it defeats the self-clocking features that tend to let us give good error messages
14:19 ab5tract speaking of, jnthn++ # i thought the time you spent on error reporting was very effective, at least afaict from reading the slides
14:19 ab5tract TimToady: thanks for clarifying so succinctly :)
14:22 yeahnoob joined #perl6
14:24 firefish5000 joined #perl6
14:27 [Coke] Perl 6 relationship status: It's complicated
14:28 moritz aye
14:28 nwc10 [Coke]++
14:30 pmichaud good morning, #perl6
14:30 nwc10 good UGT, Pm
14:31 moritz good am, pm
14:32 * pmichaud reads backscroll
14:36 TimToady .oO(Perl 6 Alfred, Batman, Catwoman...)
14:36 tadzik :)
14:37 nwc10 :-)
14:37 dalek star: 13f5f95 | moritz++ | modules/doc:
14:37 dalek star: bump modules/doc version
14:37 dalek star: review: https://github.com/rakudo/star/commit/13f5f950ed
14:41 kaleem joined #perl6
14:45 skids FROGGS,sjn: I uploadedthe business-end of libmhash support toSum
14:46 skids panda will install it but the test fileis nerfed so Star can still install Sum
14:46 skids The integrationwith the Sum core will wait till after the (soon?) Starrelease
14:46 skids But themodule can be used standalone (Sum/libmhash.pm6)
14:46 skids Feel free to show/use it duringthe hackathon. It uses the new Buf stuff.
14:47 masak skids: something wonky with your text
14:47 skids Sorry cut and paste from emacs.
14:47 skids Is it readable?
14:47 tadzik to me it is :)
14:49 ven ab5tract: "I'm changing the language because I don't like to implement some parts of it" is not really going forward. it has a bad name in the ruby community, overall
14:50 skids BTW, I found the sweet spot when funneling data to NativeCall in chunks is 65Kish.  Below that and overhead starts to eat in.
14:50 skids Got within an order of magnitude of C sha1sum binary.
14:51 salva joined #perl6
14:51 arnsholt I think there's a good chunk of overhead we can try to kill in NativeCall itself too
14:52 FROGGS joined #perl6
14:52 ab5tract ven: interesting. i had only heard good things about rubinius, but i hadn't heard of rubinius-x. thanks for explaining
14:53 * PerlJam reads backlog and is glad we're discussing the "Perl 6.0.0" (or whatever) moniker now rather than waiting for closer to christmas
14:58 * skids commute &
15:03 alpha- moniker is good but don't overdo it
15:03 alpha- like ubuntu or fedora
15:04 alpha- beefy miracle srsly
15:04 TimToady we would try to be tasteful
15:09 eiro joined #perl6
15:18 moritz 14.07 "tasty tarpit"
15:20 PerlJam yeah, alliteration is always good too :)
15:23 [Coke] tasty armpit?
15:23 FROGGS sjn: 1800 is in about 1.5 hours, right?
15:23 sjn yes
15:25 lizmat good morning pmichaud
15:26 lizmat do you have an ETA for your post-FOSDEM blogpost ?
15:26 lizmat this regarding inclusion in this week's P6Weekly or not
15:26 pmichaud later today, perhaps?
15:27 lizmat ok, then I'll won't wait for it  :-)
15:27 lizmat (the perhaps clinched it :-)
15:28 pmichaud seems reasonable.
15:29 Kristien joined #perl6
15:30 PerlJam lizmat: but you can put a tease in p6w about it for next time ;)
15:30 lizmat yup
15:33 FROGGS jnthn: I'd like to talk about vmarray problems on jvm, which you might already have a solution for because of nsa and nfg
15:33 adu joined #perl6
15:34 FROGGS jnthn: not now though, more like in the evening(s)
15:34 adu raydiak: ping
15:36 pmichaud lizmat: how quickly would you need it for inclusion in p6weekly?
15:37 lizmat I was aiming for publication at 20:00 (about 2.5 hours from now)
15:37 lizmat 3.5 hours from now
15:37 lizmat 8pm CET
15:38 FROGGS pmichaud++ # 'use nqp;'
15:38 pmichaud I need a 15 min break, perhaps I can whip it out within that period of time
15:38 pmichaud er...
15:38 pmichaud after a 15 minute break, I may be able to write it up within the next 2 hours
15:38 bobv joined #perl6
15:38 nwc10 regular time, Fergie time, or IE download time?
15:38 FROGGS >.<
15:39 lizmat ++pmichaud
15:41 bobv left #perl6
15:41 bobv joined #perl6
15:46 _mg_ joined #perl6
15:48 hhjhj joined #perl6
15:49 cdc perl6:  say lol(<a b> X <1 2>).perl
15:49 camelia rakudo-{parrot,moar} 756a4a: OUTPUT«("a", "1", "a", "2", "b", "1", "b", "2")␤»
15:49 cdc perl6:  say lol(<a b> X <1 2>).lol.perl
15:49 camelia rakudo-{parrot,moar} 756a4a: OUTPUT«((("a", "1"), ("a", "2"), ("b", "1"), ("b", "2")).list.item)␤»
15:49 cdc Hello, it this ^ expected?
15:49 cdc *is
15:50 cdc "say lol(<a b> X <1 2>).perl" comes from S03
15:50 ab5tract perl6:  say lol(<a b> X <1 2>).map: *.perl.say
15:50 camelia rakudo-{parrot,moar} 756a4a: OUTPUT«(("a", "1"), ("a", "2"), ("b", "1"), ("b", "2")).list.item␤True␤»
15:51 TimToady the lol function is probably not behaving right there
15:51 ab5tract so it looks like an interaction with the .perl output in the first example
15:51 TimToady oh, yeah, maybe that
15:51 bobv joined #perl6
15:52 TimToady hopefully one of the crinkles that gets ironed out with GLR
15:53 pmichaud s/hopefully//
15:53 cdc no need to patch S03 then?
15:53 ggoebel111111113 here's a language version number suggestion from the peanut gallery... v6.[timestamp] where timestamp is 3/14/15 9:26:53.589 in whatever format floats the community's boat :-)
15:54 moritz ggoebel111111113: UNIX timestamps for the win, of course
15:55 * PerlJam readies the jetski, ramp, and shark
15:55 alpha- v6.2015.12.25
15:56 pmichaud (use nqp)   Could someone make sure a note gets into the release notes or whatever warning authors that "use nqp;" will be required at some point.  And perhaps we should start bundling a "nqp.pm"?
15:56 moritz pmichaud: I have a local patch for the R* announcement that mentions the future need for 'use nqp;'
15:57 alpha- m: "v6.2015.12.25" lt "v6.2015.12.26"
15:57 camelia rakudo-moar 756a4a: ( no output )
15:57 TimToady use "before"
15:57 TimToady and no quotes
15:57 moritz and say
15:57 TimToady and say :0
15:57 alpha- :D
15:57 moritz m: say v6.2015.12.25 before v6.2015.12.26
15:57 camelia rakudo-moar 756a4a: OUTPUT«True␤»
15:58 alpha- sorry, I write am archaic perl5 user
15:58 TimToady well, we've had a bit longer to get used to it :)
15:59 TimToady m: say v6.2015.12.25 ~~ v6.2015.12
15:59 camelia rakudo-moar 756a4a: OUTPUT«True␤»
15:59 TimToady m: say v6.2015.12.25 ~~ v6.2015.12.*
15:59 camelia rakudo-moar 756a4a: OUTPUT«True␤»
15:59 TimToady m: say v6.2015.12.25 ~~ v6.2015.11+
15:59 camelia rakudo-moar 756a4a: OUTPUT«True␤»
16:00 TimToady m: say v6.2015.12.25 ~~ v6.2015.12.25+
16:00 camelia rakudo-moar 756a4a: OUTPUT«True␤»
16:00 TimToady m: say v6.2015.12.25 ~~ v6.2015.12.26+
16:00 camelia rakudo-moar 756a4a: OUTPUT«False␤»
16:00 TimToady super
16:01 TimToady though, of course, we have no idea yet whether we'll be using versions in that form for specs
16:01 adu joined #perl6
16:06 kaleem joined #perl6
16:07 sirdancealot joined #perl6
16:09 andreoss` m: v6.2015.12.25.WHAT.say
16:09 camelia rakudo-moar 756a4a: OUTPUT«(Version)␤»
16:12 pmichaud moritz++  # use nqp announcement patch
16:15 nine FWIW I'd much rather have version numbers that perl5 can deal with sensibly. "Perl v6.0.0 required--this is only v5.20.1" is so much better than "Can't locate v2005a.pm in @INC" or even "Bareword found where operator expected at -e line 1, near "2005a""
16:16 FROGGS nine: good point
16:16 moritz nine++ # interop thinking
16:17 andreoss` 6.2015 looks ugly
16:17 dalek star: 23f05ef | moritz++ | docs/announce/2015.01.md:
16:17 dalek star: announce that "use nqp;" will be required
16:17 dalek star: review: https://github.com/rakudo/star/commit/23f05ef02a
16:18 nwc10 $ ./perl -e 'use 6.2015'
16:18 nwc10 Perl v6.201.500 required--this is only v5.21.7, stopped at -e line 1.
16:18 nwc10 BEGIN failed--compilation aborted at -e line 1.
16:18 nwc10 it also doesn't seem to do quite what you might expect :-)
16:19 TimToady but if you put the v there, it should not do the floater
16:19 nwc10 true, Perl v6.2015.0 required--this is only v5.21.7, stopped at -e line 1.
16:20 alpha- date is good
16:20 * TimToady agrees that it's ugly though
16:21 adu joined #perl6
16:22 telex joined #perl6
16:22 [Sno] joined #perl6
16:23 dalek rakudo/nom: b401f76 | moritz++ | / (4 files):
16:23 dalek rakudo/nom: add empty nqp.pm6 and install it
16:23 dalek rakudo/nom:
16:23 dalek rakudo/nom: this enables use to write "use nqp;" now, which will be required for
16:23 dalek rakudo/nom: nqp:: ops in future. Compare RT #123728
16:23 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=123728
16:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b401f76d3b
16:24 lizmat moritz: do we really want an external file for that ??
16:24 lizmat why not have it be handled in the grammar ?
16:24 lizmat like MONKEY_TYPING, strict, etc. ?
16:25 moritz lizmat: well, once we actually support it, that's proably easier in the grammar
16:25 moritz lizmat: but as long as it's a no-op, we can do it however we like
16:25 lizmat ok, fair enough: it would be easily removed then
16:26 moritz a simple 'git revert' will help :-)
16:26 lizmat yup
16:28 konsolebox joined #perl6
16:32 dalek perl6-roast-data: 50adb43 | coke++ | / (5 files):
16:32 dalek perl6-roast-data: today (automated commit)
16:32 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/50adb43e7f
16:33 moritz how do I do string literals in QAST::?
16:33 * moritz forgot
16:33 pmichaud yes, my intent is that "nqp.pm6" is a no-op for now.
16:33 moritz is that QAST::VWal these days?
16:34 Mouq joined #perl6
16:34 moritz ah no, QAST::SVal
16:34 prammer joined #perl6
16:35 TimToady I suppose wrt the setting we can just say that YOU_ARE_HERE is opaque to that
16:35 FROGGS moritz: perhaps grep for add_string_constant
16:35 TimToady s/that/use nqp/
16:35 TimToady that is, for now that is really file-scoped information, not lexical
16:36 TimToady where "now" is when we actually first implement it
16:38 Sir_Ragnarok joined #perl6
16:39 * moritz is working on implementing it
16:40 pmichaud Also, I want there to be an "nqp.pm6" so that people will see that it is possible for other implementations to also create an "nqp.pm6" that DTRT for that implementation.
16:41 abraxxa joined #perl6
16:41 pmichaud Even if DTRT is simply "You can't nqp:: here."
16:41 moritz other implementations aren't bound by how we implement 'use nqp'
16:41 pmichaud no, but people don't always see that.
16:42 pmichaud people see "use nqp;" and if it's implemented only in a grammar they instantly think that every implementation has to handle "use nqp;" in the grammar and that's just wrong.
16:42 lizmat https://p6weekly.wordpress.com/?p=​356&amp;shareadraft=54d24bff59fb0   # draft of P6W so far
16:43 lizmat dinner&
16:46 retupmoca lizmat: Can now pass Buf/Blob to nativecall functions (me + FROGGS); IO::Socket::SSL can now wrap existing perl6 socket connections (me + sergot)
16:49 mr-foobar joined #perl6
16:53 moritz are the command line options available from Action.nqp? if yes, how?
16:54 moritz oh, is that the %*COMPILING<%?OPTIONS> thingy?
16:57 gfldex joined #perl6
17:01 FROGGS moritz: aye
17:04 pmichaud lizmat: I'm unlikely to have my blog post ready by 20h00 CET.
17:04 pmichaud I'm drafting it now, but it's not coming together cleanly.
17:09 woolfy1 joined #perl6
17:10 lizmat_ joined #perl6
17:13 skids joined #perl6
17:14 Mouq lizmat_: I finished what I did on LoL's: Autovivification used to fail when assigning to an LoL access. So now `my @a; @a[0,1;0] = <foo bar>;` works and @a will be Array.new(["foo"], ["bar"])
17:15 Mouq Also, lizmat++
17:25 moritz Mouq: bonus points for adding that to the changelog
17:26 Mouq moritz: Double bonus for adding my more comprehensive tests to roast? :P
17:27 moritz Mouq: indeed. Tripple bonus points for doing both
17:28 arnsholt If anyone else wanted to see my intro to NativeCall, it should be available from https://plus.google.com/hangout​s/_/gzov74ijqbgay5kyikxmmn7b7ia
17:30 dalek rakudo/nom: dc1855e | Mouq++ | docs/ChangeLog:
17:30 dalek rakudo/nom: Add my contributions to the ChangeLog
17:30 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/dc1855e218
17:31 Kristien joined #perl6
17:32 Kristien hi
17:32 masak arnsholt++
17:32 masak Kristien: aloha.
17:33 arnsholt (With apologies for random Norwegian before the actual talking begins)
17:35 hoelzro I feel like many of the regulars here would consider the random Norwegian a bonus =)
17:36 arnsholt There is that ^_^
17:37 masak some of my best friends are random Norwegians
17:38 dalek rakudo/nom: 71339a6 | Mouq++ | docs/ChangeLog:
17:38 dalek rakudo/nom: Add another ChangeLog entry
17:38 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/71339a6bff
17:38 moritz some of my friends and relatives are deterministic Norwegians
17:38 moritz Mouq++
17:44 skids .oO("Flyoid" is both insect and distinctly "Duke".  Unfortunately the only insect enemy in Nukem Forever is the Pregnator and... do... not ...want)
18:02 pmurias jnthn: is there a simple way to set up a minimal Perl 6 setting so that I could have a working Str for nqp::say("Hello") without getting the whole Perl 6 metamodel to work?
18:02 pmichaud RFC:  https://gist.github.com/pmi​chaud/7494d7630caf66951282
18:06 mohij joined #perl6
18:07 itz_ OT but https://twitter.com/erowidrecruiter
18:11 pmurias moritz: the Pelr 6 string literals end up a combination of a QAST::Want/QAST::WVal/QAST::SVal
18:14 vendethiel .oO( the deep truth of the assignment and the binding )
18:14 vendethiel s:g/the//
18:16 sjn cosimo: too bad you're not at arnsholt's talk, it's very interesting :)
18:16 sabujp joined #perl6
18:21 sabujp joined #perl6
18:24 lizmat pmichaud: too bad, but quality comes before the deadline in my book  :-)
18:28 raydiak mornin #perl6
18:28 raydiak adu: pong
18:29 adu raydiak: I worked on the AST last night
18:29 lizmat retupmoca Mouq : added your suggestions
18:29 raydiak adu: nice, will go look right now
18:29 adu raydiak: but I'm stuck on a missing rule, which appears as a Nil somewhere, but I'm not sure where it's coming from
18:30 adu other than that, the AST version will parse zmq now
18:31 raydiak awesome...I'll try using it later today
18:31 jnthn evening, #perl6
18:31 yoleaux 09:08Z <pmurias> jnthn: how do I do pipe stuff to a process in nqp?
18:32 jnthn Dunno, how do we do that in Perl 6? :) I guess "find that and figure out what NQP ops it calls"
18:32 jnthn pmurias: I never tried to do a minimal setting, no. I just worked through the files in the build in order.
18:33 jnthn pmurias: Though I think I commented out some things here and there in the Makefile when they weren't immediately needed.
18:34 colomon joined #perl6
18:34 raydiak adu: btw the next thing I'll be looking at after zmq is libpng which I actually have specific uses for atm...was just using zmq as a test cuz I'd played with it in the past and I know it's something people think is cool
18:36 alini joined #perl6
18:36 jnthn pmichaud: It looks good to me.
18:37 nwc10 ^ /msg ?
18:37 pmurias jnthn: the piping stuff FROGGS figured out
18:38 pmurias jnthn: just trying to do stuff in build order might be more sane
18:38 nwc10 oh, I see
18:38 jnthn nwc10: No, it's to what Pm posted here 37 mins ago
18:38 nwc10 yes,just figured that.
18:38 jnthn pmurias: Yeah, there's not a fast win, but in the end you need it all anyway for a viable Rakudo :)
18:38 nwc10 slow, I am
18:39 nwc10 but I already admitted that yesterday or the day before
18:39 pmichaud any other comments on the parrot announcement before I post it?
18:40 pmichaud I can post to my blog, but also could post to perl6-users
18:41 raydiak adu: and libpng has some __extension__ thing in it...when I grep that out, there are some other failures though...things like "typedef unsigned (*in_func) (void *, unsigned char * *);" which is a syntax I'm unfamiliar with but I think it's actually part of zlib which gets included via gcc -E
18:45 nwc10 pmichaud: I can't see any way to improve it, and I think that posting it to perl6-users is a good idea
18:47 _dolmen_ joined #perl6
18:47 [Coke] I am skeptical that we're going to have to rip it out rather than just not add new things, but I trust your judgement.
18:48 fhelmberger joined #perl6
18:52 PerlJam I wonder how long before I learn that "say %hash.keys" isn't going to show me *all* the keys.  (it's a common error I tend to make when debugging *sigh*)
18:52 vendethiel PerlJam: what'd you expect?
18:53 flussence I think the part of that announcement in .lines[33..37] is a good idea all on its own, regardless of the reasons behind it
18:53 PerlJam vendethiel: I keep expecting a list of all of the keys (forgetting, that I'm really getting a gist)
18:53 PerlJam s/,//
18:53 * PerlJam afk (food) &
18:56 pmichaud [Coke]: well, I'm not planning to actively remove all parrot support at this point.  If when doing the GLR it's easy for me to add the code that supports Parrot, I'll do that.
18:57 pmichaud but if I get to a point where "Gee, this is hard" comes into play, I'll just skip it.
18:57 [Coke] yup, that's fine.
18:58 jnthn The natives stuff is looking a bit harder to do in a graceful downgrade way. Maybe I'll come up with something, but again, I just want to get the darn stuff in place, I'm not going to go miles out of may way.
18:58 [Coke] let me know if doing the big three stuff for jvm gets stuck on anything.
18:59 jnthn I think I know roughly how to port the natives bits for JVM so it at least works.
19:00 pmichaud stuff like https://github.com/rakudo/rakudo/​blob/nom/src/core/Numeric.pm#L182  is really annoying
19:00 pmichaud yeah, we can leave it in, but... geez
19:04 colomon joined #perl6
19:05 rurban_ joined #perl6
19:06 [Coke] on the flip side, there's no reference to an issue in RT or parrot as to why that change is there.
19:08 * [Coke] sees if he can rip it out.
19:08 [Coke] yes, I know this is just one example.
19:10 FROGGS pmichaud / [Coke]: there was a problem that an optimization (inlining) caused trouble on parrot
19:11 FROGGS that's why we had to keep both declaration
19:11 FROGGS s
19:11 beastd joined #perl6
19:11 lizmat P6W published: https://p6weekly.wordpress.com/20​15/02/04/2015-05-fosdem-happened/
19:12 pmichaud is it possible to just avoid the optimization on parrot?
19:12 * FROGGS reads
19:12 FROGGS pmichaud: I dunno, I just can guess... it is too long ago since I did that
19:13 jnthn lizmat++ # weekly
19:13 dalek roast: 7aae046 | Mouq++ | S09-subscript/multidim-assignment.t:
19:13 dalek roast: Start tests for LoL-access assignment
19:13 dalek roast: review: https://github.com/perl6/roast/commit/7aae046191
19:13 dalek roast: 99bfd5a | Mouq++ | S04-declarations/my.t:
19:13 dalek roast: Merge branch 'master' of https://github.com/perl6/roast
19:13 dalek roast: review: https://github.com/perl6/roast/commit/99bfd5a7a5
19:13 dalek roast: 31571a4 | Mouq++ | S09-subscript/multidim-assignment.t:
19:13 dalek roast: Test for Hash LoL assignment
19:13 dalek roast: review: https://github.com/perl6/roast/commit/31571a4ae9
19:14 Mouq lizmat++
19:16 FROGGS lizmat++
19:16 FROGGS and Mouq++ for the 'sub MAIN;', I did not even know :o)
19:17 pmichaud iblogged:  http://pmthium.com/2015/02/suspe​nding-rakudo-support-for-parrot/
19:19 pmichaud actually, I'm deleting the post for a moment.
19:20 skids pmichaud++ lizmat++ good for the project to have fresh blog entries up, post FOSDEM.
19:20 nwc10 and neuralizing the channel?
19:20 * lizmat opens her eyes again and does not know where she is
19:20 FROGGS pmichaud++
19:21 FROGGS (for the post itself)
19:21 jnthn .oO( This is an ex-post )
19:21 FROGGS :P
19:21 FROGGS jnthn: you got one or two spare synapses for me?
19:21 jnthn FROGGS: Sorta
19:22 jnthn :)
19:22 FROGGS :o)
19:22 * jnthn did teaching all day, and is fine, but a bit tired :)
19:22 nwc10 near the beer fridge, or on tour?
19:22 FROGGS k, I want to nativecast a VMArray on the jvm to something else, problem is that a VMArray has no pointer to its data that I can use for the cast
19:23 jnthn On a train that set off on time but will probably be late...
19:23 jnthn FROGGS: You probably should be looking at VMArrayInstance
19:23 jnthn FROGGS: VMArray is the REPR
19:23 FROGGS jnthn: it just has a 'public byte[] slots' for example
19:23 FROGGS I am looking at VMArrayInstance_i8 atm
19:23 jnthn Ah
19:24 FROGGS and I don't know what to do with these slots...
19:24 jnthn Well, if start is 0 then youcan just pass that array
19:25 jnthn If start is *not* zero then you might want to put a method on VMArrayInstnace that shuffles the elements so that start is 0, and then you can just pass the array.
19:25 FROGGS hmmm
19:25 jnthn You can likely steal the code from the method used to grow the thing on a push
19:26 jnthn You're passing it to C, so it doesn't matter if there's too many slots off the end.
19:26 jnthn C has no notion of how long a buffer is, for better or worse. ;)
19:26 jnthn Alternatively, maybe there is some kind of array view thingy that can be handled by JNA
19:26 FROGGS I want to nativecast from VMArrayInstance_* to something else, like CStruct or so
19:27 lizmat Q: should you be able to just add a method trait like this: "multi sub trait_mod:<foo>(Method:D $m, &thunk)" ?
19:27 lizmat or do you need additional foo?
19:27 jnthn FROGGS: Oh...heck.
19:27 FROGGS jnthn: you're welcome :o)
19:27 jnthn FROGGS: You'll have to probably look at doing unsfe things to make that work...
19:28 FROGGS but yeah, passing a VMArrayInstance to C is also on my list to port to jvm
19:28 jnthn Rght, I thought that's the one you were doing, so that's what my answer was for.
19:28 FROGGS hmmm
19:28 FROGGS lemme try something
19:29 lizmat jnthn: re the "aka" trait on multi's: $m.multi is *always* false, it appears
19:30 FROGGS err, the code I touched was about to cast *to* a vmarrayinstance... but that is a todo for laters
19:30 timotimo good evening everybody
19:30 FROGGS hi timotimo
19:30 jnthn lizmat: No, the set of available traits is defined in the Perl 6 grammar, and different ones parse differently. So adding more is certainly a slangy thing.
19:30 nwc10 timotimo: heresy!
19:30 lizmat ah, ok
19:30 jnthn m: (multi foo() { }).multi.say
19:30 camelia rakudo-moar 71339a: OUTPUT«True␤»
19:30 lizmat timotimo o/
19:30 jnthn lizmat: Wrong. :)
19:30 nwc10 good UGT, timotimo
19:30 lizmat well, perhaps in that case
19:31 lizmat but not inside the aka trait :-(
19:31 lizmat perhaps the dispatcher isn't there yet in the setting ?
19:31 Ovid_ joined #perl6
19:31 jnthn m: multi trait_mod:<is>(Routine $m, :$oh-yes-it-is!) { say $m.multi }; multi foo() is oh-yes-it-is { }
19:31 camelia rakudo-moar 71339a: OUTPUT«True␤»
19:32 jnthn m: multi trait_mod:<is>(Routine $m, :$oh-yes-it-is!) { say $m.multi }; sub foo() is oh-yes-it-is { }
19:32 camelia rakudo-moar 71339a: OUTPUT«False␤»
19:32 jnthn lizmat: Looks fine enough to me.
19:32 lizmat jnthn: those examples are *not* running in the setting
19:33 jnthn lizmat: I struggle to see that making much difference, tbh. Maybe if you were trying to use "is aka" in the setting, but I hope you're not.
19:34 nwc10 jnthn: [ptc] had some fun earlier getting an "an unhandled exception error when running htmlify.p6 through perl6-debug-m". It doesn't happen with regular perl6-m
19:34 nwc10 jnthn: http://irclog.perlgeek.de/p​erl6/2015-02-04#i_10058771
19:34 nwc10 I wondered if you could quickly/easily spot a way get him unstuck.
19:34 lizmat jnthn:  multi sub trait_mod:<aka>(Method:D $m, &thunk) {
19:34 lizmat +say "aka: $m.multi()";
19:34 lizmat $ 6 'class A { multi method a aka <b> { ... } }'
19:34 lizmat aka: False
19:34 spider-mario joined #perl6
19:35 moritz ingy: quick 'git subrepo' question: how do I delete a subrepo?
19:35 nwc10 "Unhandled exception: ctxlexpad needs an MVMContext"
19:35 nwc10 just seems very strange
19:35 FROGGS m: multi trait_mod:<is>(Routine $m, :$oh-yes-it-is!) { say $m.multi }; class Foo { multi foo() is oh-yes-it-is { } };
19:35 camelia rakudo-moar 71339a: OUTPUT«True␤»
19:35 FROGGS m: multi trait_mod:<is>(Routine $m, :$oh-yes-it-is!) { say $m.multi }; class Foo { method foo() is oh-yes-it-is { } };
19:35 camelia rakudo-moar 71339a: OUTPUT«False␤»
19:35 FROGGS m: multi trait_mod:<is>(Routine $m, :$oh-yes-it-is!) { say $m.multi }; class Foo { multi method foo() is oh-yes-it-is { } };
19:35 camelia rakudo-moar 71339a: OUTPUT«False␤»
19:35 FROGGS ^^
19:35 jnthn lizmat: Uhh...since when wsa "aka" valid syntax like that?
19:35 jnthn *was
19:35 lizmat since I added it  :-)
19:36 * FROGGS .oO( This golf was brought to you by FROGGS™ )
19:36 jnthn I'm reluctant to see us add "is aka" *at all*. There's no way I want it going in as a totally new trait mod without a LOT more consensus.
19:36 pmichaud I feel a need to make a public comment on   10:03 <woolfy> The term sister languages was just used in the past to keep the Perl 5 people happy.
19:36 pmichaud I disagree.
19:36 Ovid_ moritz: isn’t it just “git rm $subrepo”? (but you have to also delete it from your .git dir, It hink)
19:37 pmichaud The term "sister languages" came about to help us explain the difference between Perl 5 and Perl 6 to people unfamiliar with them.
19:37 lizmat jnthn: it;s not "is aka", it's "aka"
19:37 moritz Ovid_: I don't know. Hence the question.
19:37 jnthn I really wish more folks would follow my (correct) strategy of doing stuff *outside* of or CORE.setting *first*, as we've done with NativeCall.
19:37 pmichaud Previously the statement had been "Perl 6 is the successor language to Perl 5", and that's not really accurate.
19:37 Ovid_ @moritz: https://git.wiki.kernel.org/index​.php/GitSubmoduleTutorial#Removal
19:38 jnthn (Not that NativeCall is going into CORE.setting)
19:38 dalek rakudo/nom: 8e6a1a4 | Mouq++ | t/spectest.data:
19:38 dalek rakudo/nom: Add multidim-assignment.t to spectest
19:38 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8e6a1a49f2
19:38 spider-mario joined #perl6
19:38 Ovid_ Sucks that it’s so complicated.
19:38 timotimo jnthn: i was about to ask :)
19:38 moritz Ovid_: not submodule, subrepo
19:38 pmichaud jnthn  +1
19:38 moritz Ovid_: https://github.com/ingydotnet/git-subrepo this one
19:38 Ovid_ Ah.
19:38 jnthn (but it does seem it's going to become a module we include in the Rakudo repo)
19:38 Ovid_ Sorry :)
19:38 moritz Ovid_: no problem
19:38 pmichaud jnthn: on the other hand, how do we reconcile "prototype outside of CORE.setting" with "forgiveness > permission"  ?
19:39 dalek rakudo/nom: 337b4c8 | lizmat++ | src/core/traits.pm:
19:39 dalek rakudo/nom: Removing "aka" trait until more consensus
19:39 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/337b4c8c6e
19:40 Mouq joined #perl6
19:40 lizmat aka also neuralized in blogpost
19:41 timotimo neuralized in blogpost?
19:41 moritz removed from
19:41 timotimo oh
19:41 pmichaud neuralized, as in MiB
19:41 timotimo into the memory hole it goes
19:41 lizmat afk&
19:43 alini joined #perl6
19:43 anaeem1 joined #perl6
19:44 jnthn pmichaud: Good question. I'm not expecting folks to get permission to put stuff in CORE.setting, mind. I just wish to see a bit more consensus, and a good dose more hesitancy. Especially when the change in question is involving a grammar addition as well as a CORE.setting addition.
19:45 pmichaud and when there's also some (unexplored) potential relationship with the already-existing 'handles' trait
19:45 mst jnthn: "don't ask permission, do consider asking for opinions" is often a good heuristic
19:46 pmichaud if there's something that is close to what we want but not exactly, that's often an indication that we need to refactor+unify, rather than simply add a closely-themed new thing that we have to explain the difference for
19:46 woolfy1 I'm going to take a break of a couple of weeks. AFK.
19:46 woolfy1 left #perl6
19:46 FROGGS :/
19:47 jnthn Yes, *I* was rather surprised to discover the current behavior implemented (and in the design documents) is what it is.
19:47 Mouq jnthn: I agree, on the other hand I don't think there's really a good mechanism for "RFC this feature". RT *could* work, but that's not how it's used and there's not good visibility for [RFC] things
19:47 jnthn Mouq: Sure, but it's also very easy to explore things in modules in many cases.
19:47 Mouq jnthn: Very true
19:47 pmichaud We do have perl6-language, I think.
19:48 nwc10 you can also make pull requests, even if you have commit rights
19:48 moritz well, I wouldn't ask perl6-language unless I wanted some hilariously off-topic and lofty discussion
19:49 pmichaud but, more to the point -- as I answered ovid's question in my talk --  the real design pattern ought to be "prototype as a module, then adopt into the language as consensus grows about its design and usage"
19:49 jnthn pmichaud: +1
19:49 Mouq nwc10: Also very true, and that's often used, I think, but PR =/= RFC, and I think mixing the proposal with some specific implimentation isn't so great
19:49 jnthn Also, we're *already* struggling to get CORE.setting size and base memory use down.
19:50 jnthn Both myself and others are doing what we can on systemic issues in that regard.
19:50 jnthn But it needs a little thought as to "what really should be in there" too.
19:52 pmichaud I mean, a big part of the reason for making Perl 6 extensible is so that language additions can be tried and tested _without_ having to put them into the compiler.
19:52 pmichaud *core compiler
19:52 jnthn Yeah. otoh, common things should be convenient.
19:53 dalek rakudo/nom: e804903 | lizmat++ | src/Perl6/ (2 files):
19:53 dalek rakudo/nom: Make core a bit smaller
19:53 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e804903868
19:54 pmichaud I would like to see us come up with a better consensus about the way things get adopted into the language.  Not sure that's ripe yet, though.
19:54 lizmat FWIW, the "aka" trait woud be very useful in the core to reduce the core settings
19:54 pmichaud *the ways
19:55 lizmat there are *many* methods that do exactly the same thing with different names
19:55 pmichaud lizmat: examples?
19:55 lizmat this would allow them to exist only once
19:55 * jnthn would also be curious to see examples, and wonders if said duplication is itself a warning sign.
19:56 mr-foobar joined #perl6
19:57 lizmat pmichaud: https://github.com/rakudo/rakudo​/blob/nom/src/core/Setty.pm#L11 for instance
19:57 pmichaud earlier language design documents allowed for a single Routine body to have multiple signature declarators
19:57 pmichaud I don't think that ever got implemented or adopted
19:57 pmichaud so, my first reaction is, "why is there a 'total' method"?
19:57 * pmichaud looks.
19:58 Mouq pmichaud: Bag and Mix consistency
19:58 pmichaud so, 'total' here is something like a degenerate case
19:59 pmichaud notably, the 'elems' method could be eliminated by using 'handles' on %!elems, I think.
19:59 lizmat List.Int / List.Numeric
20:00 jnthn lizmat: Are there many examples, besides these two? Those methods are both quite tiny, whereas after pondering the various cases an "is aka" trait has to handle means it will not be an especialy small amount of code. Guess maybe there are just enough to reach break-even...
20:00 jnthn Though I do really wonder if the current "handles" behavior on methods is really desirable.
20:00 pmichaud jnthn: there is that, also.  :)
20:01 lizmat various methods in Nil
20:01 jnthn I mean, I was really surprised when I read what it does. :)
20:02 moritz probably some Str, Stringy and gist methods overlap
20:02 jnthn OK, so mebbe it can pay for itself.
20:02 Mouq (frankly, I'm pro "aka", but I'm very glad it spurned this discussion)
20:03 pmichaud I think "aka" and "handles" need thinking
20:03 dalek Inline-Perl6: 4f67da5 | (Stefan Seifert)++ | / (3 files):
20:03 dalek Inline-Perl6: Remove superfluous p6_ prefix from function names
20:03 dalek Inline-Perl6: review: https://github.com/niner/Inl​ine-Perl6/commit/4f67da5234
20:03 lizmat really afk&
20:04 flussence I think I've said this before and had it shot down, but... can't a lot of the classes in the core setting be loaded on demand? Only self-contained OOP-ish classes, I mean, not things like the set ops that alter how code is parsed.
20:04 FROGGS flussence: that's what lazy serialization does for you
20:04 jnthn flussence: Well, that's part of what the lazy deserialization work I was doing was about.
20:04 flussence hah :)
20:04 jnthn And we already delay verifying bytecode until we actuall run it.
20:04 jnthn *actually
20:05 flussence okay, it's smarter than I remember it being...
20:06 moritz flussence: I think it was never shot down, the reaction two years ago was just "we don't know how to do that"
20:07 moritz jnthn: playing the devil's advocate, does Proc::Async need to be in core?
20:07 dalek Inline-Perl6: c0f4066 | (Stefan Seifert)++ | / (2 files):
20:07 dalek Inline-Perl6: Fix destroy() ending the program prematurely
20:07 dalek Inline-Perl6: review: https://github.com/niner/Inl​ine-Perl6/commit/c0f4066a8b
20:07 jnthn moritz: Not particularly.
20:07 moritz jnthn: I mean, it uses nqp:: ops, and needs to, so it should probably be shipped with rakudo, but I don't think it needs to be in src/core/*
20:08 jnthn moritz: Yes, that's an interesting distinction.
20:08 jnthn I guess it wound up there 'cus other process stuff *is* in core
20:08 moritz and on the language level, I'd be fine with it requiring a 'use'
20:08 jnthn And it feels a tad odd for the async variety to be a second class citizern.
20:09 bronco_creek joined #perl6
20:09 * FROGGS votes for moving all math stuff out of the setting so we can do: use Math
20:10 FROGGS (j/k)
20:10 japhb (Not having read much backlog ...)  I certainly would like to think of async and sync as peers, not having async be second-class.  That feels like part of the "Perl 6 does concurrency well" story
20:10 FROGGS aye
20:10 flussence +1 - async is hard enough already
20:10 dalek Inline-Perl6: bd9e834 | (Stefan Seifert)++ | / (3 files):
20:10 dalek Inline-Perl6: Make the interface consistent with Inline::Perl5
20:10 dalek Inline-Perl6:
20:10 dalek Inline-Perl6: Rename call_method to invoke and eval_code to run
20:10 dalek Inline-Perl6: review: https://github.com/niner/Inl​ine-Perl6/commit/bd9e834908
20:11 Sqirrel joined #perl6
20:12 jnthn 'course, I need to hunt the Proc::Async buglets folks continually hit for us to claim we do it well... :(
20:13 lucas__ joined #perl6
20:13 nwc10 use more jnthns;
20:14 bronco_creek I saw that Elon Musk is naming his robotic rocket landing barges in honor of Ian Banks' Culture minds: "Just Read the Instructions" and "Of Course I Still Love You.  I'd love to see "Perl 6 Of-Course-I-Still-Love-You" as a standard and a message to the P5 community, although it might be a little long.
20:15 nwc10 new pypy. bullet list of improvemnts looks nice: http://morepypy.blogspot.co.at/​2015/02/pypy-250-released.html
20:17 jnthn Indeed.
20:17 timotimo i always love reading the pypy release notes
20:20 zakharyas joined #perl6
20:21 [Coke] https://gist.github.com/coke/ad33b575ede7428eed5b - removing one of those parrot ifdefs from Numeric causes no spectest failures.
20:23 moritz [Coke]: then by all means, remove it
20:28 pmichaud [Coke]++
20:28 FROGGS nice
20:29 * moritz hopes that can be generalized to all the proto-related #?if parrot's
20:30 dalek roast: b4b8ab1 | usev6++ | S04-statements/for.t:
20:30 dalek roast: Add test for RT #122095
20:30 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122095
20:30 dalek roast: review: https://github.com/perl6/roast/commit/b4b8ab1cf2
20:32 Mouq joined #perl6
20:33 moritz do we have a a glob() in Rakudo?
20:35 hoelzro moritz: last I checked, I think it was NYI
20:35 hoelzro is glob even spec'd?
20:35 moritz :(
20:35 moritz dunno
20:35 moritz but it's so damn practical, I want it.
20:35 [Tux] use Inline::Perl5 :)
20:35 hoelzro it's mentioned in S32
20:36 hoelzro S32 only mentions it in an example though; it's referred to in IO-OLD.pod
20:37 * flussence just typed "perl6" to try something and was shocked at how fast the repl came up...
20:37 [Tux] lizmat, has "aka" been backed out again?
20:37 lizmat yes, by lack of consensus
20:38 flussence (giving a repl no input is approximately 5 times faster than -e '')
20:38 lizmat I think it's practical, and will help in redicing core sttings
20:38 lizmat *reducing
20:38 lizmat but that is not enough
20:38 moritz flussence: well, -e '' still has to fire up the grammar
20:39 jnthn lizmat: I think it's the grammar addition that really bothered me, rather than it just being "is aka". But I really do feel we want to question the current "handles" on methods.
20:39 lizmat handles has its uses as well
20:39 [Tux] can I get the one-liner that tadzik used?
20:40 moritz [Tux]: used for what?
20:40 [Tux] for aka
20:40 jnthn detrain, bbiab
20:40 flussence moritz: I guess it's cheating, but still pretty nice. :)
20:41 alini joined #perl6
20:41 moritz [Tux]: not aware of that, but I'd use ::?CLASS.^add_method('new_name', ::?CLASS.^find_method('old_name'))
20:41 moritz oh, and s/^/BEGIN /
20:42 lizmat yeah, much more succinct than "aka <new_name>
20:42 lizmat sorry, I should not be at a keyboard right now
20:42 * moritz hugs lizmat
20:42 flussence .oO( /me mumbles something about macros... )
20:43 moritz lizmat: using IO::Path, do you know some simple-ish way to get the last two directory parts from a path? ie from "/a/b/c/d" to "c/d"?
20:43 [Tux] lizmat, just putting the removed line in my script doesn't help. tadzik had a one-liner to show that would
20:45 dalek rakudo-star-daily: a944c54 | coke++ | log/ (10 files):
20:45 dalek rakudo-star-daily: today (automated commit)
20:45 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/a944c544ba
20:46 moritz lizmat: never mind, .relative works in my case \o/
20:46 FROGGS [Tux]: I use that in XML::LibXML:
20:47 flussence r: say IO::Path.new('/tmp/foo/bar/baz').ab​solute.split('/')[*-2..*].join('/') # this is ugly but I can't do better. :(
20:47 camelia rakudo-parrot e80490: OUTPUT«IO::Path is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:2␤  in method new at src/RESTRICTED.setting:32␤  in block <unit> at /tmp/tmpfile:1␤␤»
20:47 camelia ..rakudo-moar e80490: OUTPUT«IO::Path is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:1␤  in method new at src/RESTRICTED.setting:32␤  in block <unit> at /tmp/tmpfile:1␤␤»
20:47 FROGGS m: multi trait_mod:<is>(Routine $r, :$aka!) is export { $r.package.^add_method($aka, $r) }; sub foo is aka<bar { 42 }; say bar
20:47 camelia rakudo-moar e80490: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/JSsWR8Mhtaâ�¤Unable to parse expression in quote words; couldn't find final '>'â�¤        â�¤at /tmp/JSsWR8Mhta:1â�¤------> [32mr) }; sub foo is aka<bar { 42 }; say bar[33mâ��[31m<EOL>[0mâ�¤    â€¦Â»
20:47 FROGGS m: multi trait_mod:<is>(Routine $r, :$aka!) is export { $r.package.^add_method($aka, $r) }; sub foo is aka<bar> { 42 }; say bar
20:47 camelia rakudo-moar e80490: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/yFX9t3zCUf�Method 'add_method' not found for invocant of class 'Perl6::Metamodel::PackageHOW'�at /tmp/yFX9t3zCUf:1�------> �»
20:47 flussence r: say '/tmp/foo/bar/baz'.split('/')[*-2..*].join('/') # bah
20:47 camelia rakudo-{parrot,moar} e80490: OUTPUT«bar/baz␤»
20:47 FROGGS err, for methods, not subs
20:48 FROGGS m: multi trait_mod:<is>(Routine $r, :$aka!) is export { $r.package.^add_method($aka, $r) }; class Foo { method foo is aka<bar> { 42 } }; say Foo.bar # [Tux]
20:48 camelia rakudo-moar e80490: OUTPUT«42␤»
20:48 Ovid_ m: sub a returns Bool { b() }; sub b returns Str {}; say “Hi there!”
20:48 camelia rakudo-moar e80490: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/BLnLp4wLUu�Unsupported use of bare 'say'; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument�at /tmp/BLnLp4wLUu:1�------> [32mBool { b() }; sub b returns Str {…»
20:49 Ovid_ That’s interesting. Works on the command line.
20:49 FROGGS Ovid_: these smart quotes break it here
20:49 [Tux] FROGGS, your version does not take a list as in
20:49 [Tux] method diag_verbose (*@s) is aka < diag-verbose verbose_diag verbose-diag >
20:49 FROGGS [Tux]: true, so you would have to put a for loop into the trait
20:49 dalek doc: bb5cbc1 | paultcochrane++ | lib/P (6 files):
20:49 dalek doc: Add consistent vim coda to module files
20:49 dalek doc: review: https://github.com/perl6/doc/commit/bb5cbc19d9
20:49 Ovid_ m: sub a returns Bool { b() }; sub b returns Str {}; say "Hi"
20:49 camelia rakudo-moar e80490: OUTPUT«Hi␤»
20:50 Ovid_ (Didn’t realize the Colloquy was using smart quotes)
20:50 FROGGS m: sub a returns Bool { b() }; sub b returns Str { "Hi" }; say b
20:50 camelia rakudo-moar e80490: OUTPUT«Hi␤»
20:50 FROGGS m: sub a returns Bool { b() }; sub b returns Str { "Hi" }; say a
20:50 camelia rakudo-moar e80490: OUTPUT«Type check failed for return value; expected 'Bool' but got 'Str'␤  in any return_error at src/vm/moar/Perl6/Ops.nqp:649␤  in sub a at /tmp/pfE6nfhDve:1␤  in block <unit> at /tmp/pfE6nfhDve:1␤␤»
20:50 FROGGS looks fine to me
20:50 Ovid_ How hard would it be for the optimizer to realize that the above code will fail since b() is supposed to return a string?
20:51 Ovid_ It sort of looks like the information is there.
20:51 FROGGS Ovid_: since I never implemented optimizations I cannot tell
20:51 moritz Ovid_: doable without too much tricky hackery, I'd say
20:51 * moritz has implemented some constant folding in the optimizer
20:53 denis_boyun joined #perl6
20:54 alpha- foreach (thing that can return a value) {find expression that creates the return value; if information about type that will be returned is available check if it matches the thing's return type}
20:59 moritz iirc perl 5 doesn't complain at optimization time about code that will die
20:59 moritz because it might not be reachable/reached
20:59 jnthn Ovid_: It should be do-able, but I'm still kinda waiting for the dust to settle on what exactly Failure can sneak through type-check wise before I invest time on it, since it matters deeply from the optimizer's point of view.
21:00 Ovid_ jnthn: fair enough. I was just curious :)
21:01 moritz perl -e 'if (rand() < 0.001) { print 1/0 }' # doesn't complain here, even though the peephole optimizer knows for sure what's going to happen in that branch
21:01 jnthn It would be a neat one to catch :)
21:02 moritz comparison: perl -MO=Deparse -e 'if (rand() < 0.001) { print 2 * 5 }' # constant-folds the '2 * 5' to 10
21:02 jnthn m: say 1/0
21:02 camelia rakudo-moar e80490: OUTPUT«Divide by zero␤  in method Numeric at src/gen/m-CORE.setting:14593␤  in sub infix:<-> at src/gen/m-CORE.setting:5136␤  in method Str at src/gen/m-CORE.setting:12046␤  in method gist at src/gen/m-CORE.setting:4973␤  in sub say at src/gen/m-CORE.se…»
21:02 jnthn I see we leave that one until runtime too :)
21:02 moritz jnthn: aye
21:03 jinxter joined #perl6
21:03 moritz jnthn: that was to make the optimization less intrusive. Maybe that was a mistake.
21:03 moritz ... and should be easy to correct
21:04 _dolmen_ joined #perl6
21:08 jnthn moritz: Don't immediately have a good feel for it.
21:08 jnthn moritz: You're not that likely to write a literal / 0 unless wanting to demonstrate something, I guess.
21:09 jnthn Finding out that some compile-time constant going to zero is going to break things feels useful to know early...
21:15 Ugator joined #perl6
21:18 * jnthn wonders if anyone else is having trouble pulling from GitHub, or just him...
21:19 jnthn Ah, seems it was transient.
21:20 Kristien LLVM is very neat.
21:20 avuserow_ m: my $foo = "foo"; $foo [R~]= "bar"; say $foo  # should this or something similar result in "barfoo"?
21:20 camelia rakudo-moar e80490: OUTPUT«===SORRY!===␤Unknown QAST node type NQPMu␤»
21:21 Mouq joined #perl6
21:21 avuserow_ in other words, is there a "prepend" version of ~= ?
21:21 jnthn avuserow_: I *think* that one should do what you want, and that error is certainly a bug.
21:24 Diederich ~
21:25 avuserow_ (not sure I'd want to make a habit of using such an operator, but it's amazing that Perl 6 has [or will have] this flexibility)
21:27 * itz_ wonders what birdless could be :)
21:32 telex joined #perl6
21:37 * vendethiel never thought of R~
21:37 vendethiel the possibilities!
21:37 vendethiel meta-R assign is going to be very useful indeed
21:41 rindolf joined #perl6
21:47 Mouq my @foo = <foo bar>; @foo [Z~]= <d n>; say @foo
21:47 Mouq m: my @foo = <foo bar>; @foo [Z~]= <d n>; say @foo
21:47 camelia rakudo-moar e80490: OUTPUT«Cannot find method 'postcircumfix:<( )>'␤  in block  at src/gen/m-CORE.setting:20498␤  in block <unit> at /tmp/7Msc1JdAqi:1␤␤»
21:48 * jnthn wonders if the = postfix meta-op thingy is making a bogus transformation when there's a meta-op on the inside.
21:48 Mouq m: my @foo = <foo bar>; @foo Z[~=] <d n>; say @foo
21:48 camelia rakudo-moar e80490: OUTPUT«food barn␤»
21:48 jnthn hah, test data win :P
21:50 PerlJam too bad that won't work with R :)
21:51 PerlJam (well, the original problem that is)
21:52 alini joined #perl6
21:52 brrt joined #perl6
21:54 Juerd "Z[~=]". Whoa.
21:54 skids Nownow just because you can does not mean you should :-)
21:55 Juerd I don't understand why the = is in the [] though
21:55 psch Juerd: precedence resolution
21:55 psch m: my $x = 5; $x R[+=] 4; say $x
21:55 camelia rakudo-moar e80490: OUTPUT«Cannot modify an immutable Int␤  in block  at src/gen/m-CORE.setting:20498␤  in block  at src/gen/m-CORE.setting:20510␤  in block <unit> at /tmp/8eBsl9wEVQ:1␤␤»
21:55 psch m: my $x = 5; 4 R[+=] $x; say $x
21:55 camelia rakudo-moar e80490: OUTPUT«9␤»
21:56 Juerd Er, what's R?
21:56 Juerd Reverse?!
21:56 vendethiel Juerd: reverse meta-operator
21:56 PerlJam aye
21:56 Juerd Funky
21:56 Juerd So Z[] is a thing, and it's not the same [] as in [+] LIST?
21:57 vendethiel Juerd: indeed.
21:57 Juerd And not the same Z as in @foo Z @bar, then?
21:57 vendethiel Juerd: and it's not the same as in <<[+]>>
21:57 Juerd It's beginning to make sense to me
21:57 vendethiel Juerd: @a Z @b is @a Z, @b, if you squint enough
21:57 vendethiel (or @a Z[,] @b)
21:58 Juerd Oh wow
21:58 psch in general, [$op] is kind of magic
21:58 vendethiel or rather really magic
21:58 PerlJam Juerd: that's exactly the reaction we want people to have with Perl 6  :)
21:58 psch in my head at least
21:58 Juerd psch: I knew that, but I didn't know there were several kinds of [op] now. :)
21:58 psch as in, i know the two spots where it does things, and the two things are kind of related, but not really...
21:59 psch m: say 5 [[[+]]] 5
21:59 camelia rakudo-moar e80490: OUTPUT«10␤»
21:59 vendethiel m: my &add = -> { $^a * 2 + $^b }; say [[&add]] ^10
21:59 camelia rakudo-moar e80490: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/YJSjkQS97qâ�¤Placeholder variable '$^a' cannot override existing signatureâ�¤at /tmp/YJSjkQS97q:1â�¤------> [32mmy &add = -> { $^a * 2 + $^b }[33mâ��[31m; say [[&add]] ^10[0mâ�¤    expecting any…»
21:59 Juerd Mind blown.
21:59 vendethiel m: my &add = { $^a * 2 + $^b }; say [[&add]] ^10
21:59 camelia rakudo-moar e80490: OUTPUT«1013␤»
21:59 vendethiel Juerd: ^  :P
22:00 Juerd What is it called?
22:01 tadzik [Tux]: hmm, I don't remember any oneliner :/
22:01 vendethiel Juerd: it's the same meta-reduce. but you can pass a function
22:01 b2gills m: my $a = 5; $a [[+]=] 5; say $a
22:01 camelia rakudo-moar e80490: OUTPUT«10␤»
22:01 andreoss joined #perl6
22:02 Juerd vendethiel: What's the difference between [] and [[]]?
22:02 vendethiel Juerd: sorry, sorry, not trying to confuse you here
22:02 vendethiel Juerd: [[]] is necessary to pass a function as reducer.
22:02 Juerd Oh, I see
22:03 Juerd So [+] is sugar for [[&infix:<+>]]?
22:03 vendethiel m: say [[&infix:<+>]] ^10; # just try it :-)
22:03 camelia rakudo-moar e80490: OUTPUT«45␤»
22:03 psch m: sub add { $^a + $^b }; say 5 [&add] 5
22:03 camelia rakudo-moar e80490: OUTPUT«10␤»
22:03 Juerd No, I meant definitionwise
22:03 vendethiel Juerd: also, []-wrapping comes very useful when there are ambiguities. Like "@a <<[<]>> @b"
22:04 vendethiel psch: oh, didn't know that actually worked. TIL
22:04 psch Juerd: not quite, [&sub] is the infix form of &sub
22:04 FROGGS_ joined #perl6
22:04 psch Juerd: but &infix:<+> is already an infix without the infix "marker"
22:04 vendethiel Juerd: http://design.perl6.org/S03.html#Tur​ning_a_binary_function_into_an_infix
22:04 Juerd Is [something] expected to behave like [[&infix<something>]] in all cases, or does this just happen to be the same in this specific case?
22:04 vendethiel Juerd: and see just before, "nesting of metaoperators"
22:04 Juerd psch: Oh!
22:04 psch (where «infix "marker"» means the longname)
22:04 vendethiel Juerd: sorry :P. I'll let the spec do the speaking now
22:05 vendethiel (and psch++)
22:05 Juerd psch: That helps. I thought [[ itself was one token, but the inner [] are different from the outer [], they just happen to look a lot alike.
22:05 PerlJam m:  sub add { $^a + $^b }; say 5 [[[[[&add]]]]] 5
22:05 camelia rakudo-moar e80490: OUTPUT«10␤»
22:06 PerlJam It's turtles all the way down
22:06 itz_ grrr ecosystem-api.p6c.org is blocked by one of the UK ISP web filters as "malware"
22:09 kjs_ joined #perl6
22:12 dalek zavolaj: b51ff75 | FROGGS++ | t/05-arrays.t:
22:12 dalek zavolaj: unfudge native-sub(Buf) test
22:12 dalek zavolaj: review: https://github.com/jnthn/zavolaj/commit/b51ff757f2
22:13 dalek nqp: 369084b | FROGGS++ | src/vm/jvm/runtime/org/perl6/nqp/ (2 files):
22:13 dalek nqp: [jvm] allow to pass VMArrays (Blob) to native subs
22:13 dalek nqp: review: https://github.com/perl6/nqp/commit/369084b98c
22:14 brrt left #perl6
22:14 FROGGS_ ohh, now that I pushed it I think I'd need to take the .start into account :/
22:15 FROGGS_ I'm going to bump the revision anyway
22:15 dalek rakudo/nom: 613c0bc | FROGGS++ | tools/build/NQP_REVISION:
22:15 dalek rakudo/nom: bump nqp rev for nativecast improvements on jvm
22:15 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/613c0bcbb1
22:16 FROGGS_ jnthn: is there a way to force a case where .start is non-zero?
22:16 jnthn FROGGS_: nqp::unshift something
22:16 FROGGS_ ahh
22:16 FROGGS_ that is what .start is for
22:17 jnthn yes ;)
22:17 FROGGS_ :o)
22:17 Juerd m: say [,] 1, 2, 3  # makes no sense
22:17 camelia rakudo-moar e80490: OUTPUT«1 2 3␤»
22:17 Juerd Possibly anyway. Nice.
22:17 FROGGS_ Juerd: the best thing is that it is not a hack
22:18 FROGGS_ jnthn: I care about .start tomorrow... I am too tired atm
22:19 nwc10 FROGGS_: when do the FROGGSlets wake up?
22:20 FROGGS_ nwc10: well, the tiniest will wake up every now and then, but the mid sized one tends to wake up at half past six
22:20 FROGGS_ which is the right time to get up luckily
22:20 nwc10 :-)
22:20 FROGGS_ :o)
22:20 nwc10 the small alarm clock here is not totally reliable
22:21 nwc10 on the other hand, start time at $work is flexible (and public transport is frequent) so it doesn't matter massively
22:21 FROGGS_ our alarm clock was malfunctioning for a week due to an illness
22:21 nwc10 ours have been ill too
22:21 FROGGS_ nwc10: same here
22:21 nwc10 so have we
22:21 FROGGS_ aye
22:22 nwc10 seems that we also got the grandparents
22:22 FROGGS_ seems the right time to be ill or so
22:22 nwc10 two other people at work were ill. I hope I didn't get them
22:22 jnthn I just taught my 3rd course of the year.
22:22 nwc10 is that a high number or low number?
22:22 jnthn It was, however, the first course of the year when I didn't spend part of it suffering with some illness.
22:22 nwc10 yay!
22:23 jnthn (First one - an icky cold. Second one - either stomach virus or food poisoning.)
22:26 FROGGS_ gnight
22:28 jnthn 'night, FROGGS++
22:31 andreoss joined #perl6
22:44 virtualsue joined #perl6
22:45 colomon joined #perl6
22:46 daxim joined #perl6
22:52 masak pmichaud: re http://irclog.perlgeek.de/p​erl6/2015-02-04#i_10061249 , see https://github.com/perl6/specs/commit/9​950d83ede3d1aa77a96ad83fd7549a69fc9e4a5
22:55 Kristien joined #perl6
23:09 Kristien m: say 1 xx 10 Z+< 1..*
23:09 camelia rakudo-moar e80490: OUTPUT«2 4 8 16 32 64 128 256 512 1024␤»
23:09 Kristien absolutely amazing :D
23:10 avuserow_ m: say 1 xx 10 <<+<<< 1..*
23:10 camelia rakudo-moar e80490: OUTPUT«$(1,)..Inf␤»
23:10 avuserow_ m: say 1 xx 10 >>+<>> 1..*
23:10 camelia rakudo-moar e80490: OUTPUT«$(1,)..Inf␤»
23:10 avuserow_ hm, thought there was a similar way to write that
23:10 timotimo m: say 1 <<+<<< 1..*
23:10 camelia rakudo-moar e80490: OUTPUT«(2,).list.item..Inf␤»
23:10 Kristien m: say map * ** 2, (1, 2, 4 ... 64)
23:10 timotimo m: say 1 <<+<<< (1..*)
23:10 psch m: say 1 <<+<<< ^10
23:10 camelia rakudo-moar e80490: OUTPUT«1 4 16 64 256 1024 4096␤»
23:10 camelia rakudo-moar e80490: OUTPUT«␤»
23:10 camelia rakudo-moar e80490: OUTPUT«1 2 4 8 16 32 64 128 256 512␤»
23:10 timotimo m: say 1 <<+<<< (^10)
23:10 camelia rakudo-moar e80490: OUTPUT«1 2 4 8 16 32 64 128 256 512␤»
23:10 timotimo exactly, psch
23:11 Kristien what does <<=<<< mean?
23:11 Kristien and <<+<<<
23:11 psch Kristien: hyper meta op
23:12 timotimo may want to use «+<« instead to make it stand out more what's the operator being hyper'd
23:12 psch Kristien: <<$op<< means "take the right as total result list length and repeat the left to match" iirc
23:12 Kristien I like how when someone posts code here others posts millions of other ways to do the same thing :)
23:12 psch m: say 1 <<+<< 1,2,3
23:12 camelia rakudo-moar e80490: OUTPUT«223␤»
23:12 psch eh, [,] precedence
23:13 psch m: say 1 <<+<< (1,2,3)
23:13 camelia rakudo-moar e80490: OUTPUT«2 3 4␤»
23:13 masak 'night, #perl6
23:13 psch g'night masak
23:13 psch m: say (1, 2) <<+<< (1,2,3)
23:13 camelia rakudo-moar e80490: OUTPUT«2 4 4␤»
23:13 psch m: say (1,2,3) >>+<< (3,2,1)
23:13 camelia rakudo-moar e80490: OUTPUT«4 4 4␤»
23:14 psch Kristien: S03:Hyper operators
23:14 Kristien m: (1 xx * Z+< 1..*).say
23:14 camelia rakudo-moar e80490: OUTPUT«2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768 65536 131072 262144 524288 1048576 2097152 4194304 8388608 16777216 33554432 67108864 134217728 268435456 536870912 1073741824 2147483648 4294967296 8589934592 17179869184 34359738368 68719476736 1…»
23:14 psch hrm, did i break synopsebot with my PR? :/
23:14 Kristien interesting
23:15 psch http://design.perl6.org/S03.html#Hyper_operators is the link
23:15 Kristien thanks!
23:16 Kristien m: (map * ** 2, (1..*)).say
23:16 camelia rakudo-moar e80490: OUTPUT«1 4 9 16 25 36 49 64 81 100 121 144 169 196 225 256 289 324 361 400 441 484 529 576 625 676 729 784 841 900 961 1024 1089 1156 1225 1296 1369 1444 1521 1600 1681 1764 1849 1936 2025 2116 2209 2304 2401 2500 2601 2704 2809 2916 3025 3136 3249 3364 3481 3600…»
23:17 Kristien now this is interesting
23:17 muraiki_ joined #perl6
23:17 Kristien nah not really :P
23:17 avuserow_ m: (map * ** 2, (1..*)).elems.say
23:17 avuserow_ ... but watch for timeouts :)
23:18 camelia rakudo-moar e80490: OUTPUT«Memory allocation failed; could not allocate 60928 bytes␤»
23:18 psch m: say (** ** 2)(1..*) #
23:18 camelia rakudo-moar e80490: OUTPUT«1 4 9 16 25 36 49 64 81 100 121 144 169 196 225 256 289 324 361 400 441 484 529 576 625 676 729 784 841 900 961 1024 1089 1156 1225 1296 1369 1444 1521 1600 1681 1764 1849 1936 2025 2116 2209 2304 2401 2500 2601 2704 2809 2916 3025 3136 3249 3364 3481 3600…»
23:19 Kristien m: ((1 xx * Z+< 0..*) Z= (map 2 ** *, (0..*)))[^10].say
23:19 camelia rakudo-moar e80490: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/5hXx9FOyrX�Cannot zip with = because list assignment operators are too fiddly�at /tmp/5hXx9FOyrX:1�------> [32m((1 xx * Z+< 0..*) Z=[33m�[31m (map 2 ** *, (0..*)))[^10].say[0m�»
23:19 Kristien m: ((1 xx * Z+< 0..*) Z== (map 2 ** *, (0..*)))[^10].say
23:19 camelia rakudo-moar e80490: OUTPUT«True True True True True True True True True True␤»
23:20 Kristien joined #perl6
23:21 Kristien I keep being beaten by Z= vs Z==
23:21 timotimo m: say 1, 2 ... *
23:21 camelia rakudo-moar e80490: OUTPUT«1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 8…»
23:21 timotimo m: say 1, 2, 4 ... *
23:21 camelia rakudo-moar e80490: OUTPUT«1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768 65536 131072 262144 524288 1048576 2097152 4194304 8388608 16777216 33554432 67108864 134217728 268435456 536870912 1073741824 2147483648 4294967296 8589934592 17179869184 34359738368 68719476736…»
23:25 labster Ovid_++ saw your blog linked on HN, good work.  Minor nitpick: "So why isn't .1 + .2 + .3 zero?"  Uh, because it's .6?
23:31 Kristien m: say (.1 + .2 - .3)
23:31 camelia rakudo-moar e80490: OUTPUT«0␤»
23:31 Kristien m: say (.1 + .2 - .3) == 0
23:31 camelia rakudo-moar e80490: OUTPUT«True␤»
23:31 Kristien m: say (.1 + .2 - .3) == .0
23:31 camelia rakudo-moar e80490: OUTPUT«True␤»
23:32 Kristien wow such precision
23:38 Kristien I think I kinda get it, though I don't see the difference between »+« and Z+
23:38 lsm-desktop joined #perl6
23:39 muraiki_ "But long before I get to Perl 6, I want to explain why I like is COBOL"
23:39 muraiki_ why I like is COBOL?
23:43 Juerd You no like is cobol then?
23:44 muraiki_ heheh
23:44 Kristien m: sub SORRY { }; say 'a' ===SORRY!=== 'b'
23:44 camelia rakudo-moar 613c0b: OUTPUT«True␤»
23:44 Juerd Are you say you is not like cobol!
23:44 Kristien Redirect perl6' stderr to perl6!
23:48 jnthn Kristien: :P
23:48 * jnthn gets some rest :)
23:48 jnthn 'night, #perl6
23:56 jack_rabbit joined #perl6

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo