Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2015-08-31

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:26 laouji joined #perl6
00:40 laouji joined #perl6
00:44 llfourn joined #perl6
00:49 laouji joined #perl6
00:52 japhb o/
01:13 BenGoldberg joined #perl6
01:20 vendethiel joined #perl6
01:36 ab5tract_ joined #perl6
01:37 zacts joined #perl6
01:38 ggoebel2 joined #perl6
01:40 Fleurety joined #perl6
01:41 cgfbee joined #perl6
02:13 thou ugexe: FYI: this works like I want it to. https://gist.github.com/softmoth/1da0f8e722e98c2363bf
02:13 noganex_ joined #perl6
02:14 thou ugexe: regarding tap/act, I still contend that you've got a race there. See commented-out .act() calls in this code ^; the note() calls for stderr and stdout will overwrite each other.
02:14 thou I have to use another channel to properly merge the lines. I believe your process code has the same issue.
02:16 thou erm, sorry, nevermind! I do think it's still a potential problem, but this one is due to my running multiple threads a the same time
02:17 thou I'd need to do some random sleeping in the act() closures to test my hunch
02:17 thou jobs
02:45 llfourn joined #perl6
03:01 ugexe if i want the lines properly merged, i access stdout and stderr directly. i would not use $merge
03:01 ugexe but thats assuming im using one persons definition of proper
03:02 kaare_ joined #perl6
03:02 llfourn joined #perl6
03:05 ugexe your problem is because you are calling say/note whatever inside your tap/act
03:07 ugexe if multiple threads are printing out $*OUT, you will get what appears to be incorrectly merged lines
03:11 ugexe if you are really trying to merge the output to the terminal, that is a different thing. a blog post that shows the general idea http://ugexe.com/create-a-perl6-terminal-progress-bar/
03:13 peteretep joined #perl6
03:13 peteretep homebrew wants to know if I want to install rakudo-star with pcre
03:13 ugexe but you have to make choices. if you receive a line of text without a line break at the end, and then another process fires its own tap to print, what do you do?
03:13 peteretep Is there a quick overview of the implications of that choice?
03:15 ugexe zef::cli::statusbar puts a lock inside $*OUT's print method
03:30 rurban joined #perl6
03:44 laouji joined #perl6
03:46 _itz joined #perl6
03:46 prevost joined #perl6
03:51 _itz joined #perl6
03:52 thou joined #perl6
03:53 llfourn joined #perl6
03:55 TEttinger joined #perl6
04:02 _itz_ joined #perl6
05:00 TimToady checking out breakfast at the Abba, then off to the Alhambra
05:01 TimToady we have one free ticket if anyone happens to be here and awake enough to leave 8:00 or so
05:19 ShimmerFairy TimToady: how many more stops until you're home (or if not that, your next destination)? A quick look for Alhambra places at least two in Spain and California, so it's not like I could guess :)
05:36 emilbarton joined #perl6
05:36 emilbarton hi, why does $str.eval returns "method not found for invocant of class string"?
05:39 ShimmerFairy it's spelled EVAL in Perl 6 now. so it'd be EVAL($str) or $str.EVAL :)
05:39 emilbarton ok thanks
05:52 cognominal joined #perl6
05:55 TimToady ShimmerFairy: "The" Alhambra is in Granada, Spain
05:55 ShimmerFairy ah, I somehow missed the "the" there :)
06:05 ShimmerFairy By the way, I think I'll be ready to push my val() and allomorphic types as a branch tonight :) . The only main issue is that my implementation of IntStr dies on bigints, and I'm not sure how to fix that.
06:06 ShimmerFairy (The only thing I can think of is a fallback Int attribute, and then reimplementing all of Int's methods in IntStr, but ick)
06:09 lizmat joined #perl6
06:09 ShimmerFairy m: say "".Numeric  # this is surprising, and feels wrong to me
06:09 camelia rakudo-moar 5ba44f: OUTPUT«0␤»
06:09 GLRelia rakudo-moar 36ea47: OUTPUT«0␤»
06:12 nine ShimmerFairy: that's probably Perl 5 heritage
06:13 ShimmerFairy nine: replacing .Numeric with a val() call in my code caught that in a test, and I'm quite OK with saying "but I don't see a number in that string"  :)
06:13 ShimmerFairy (replacing the implementation of Str.Numeric, that is)
06:15 dalek rakudo/glr: cb9b821 | FROGGS++ | t/01-sanity/19-say.t:
06:15 dalek rakudo/glr: remove test of say() that does not work with .gist, laben++
06:15 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/cb9b821b08
06:15 dalek rakudo/glr: cf3b24c | timotimo++ | t/01-sanity/19-say.t:
06:15 dalek rakudo/glr: Revert "remove test of say() that does not work with .gist, laben++"
06:15 dalek rakudo/glr:
06:15 dalek rakudo/glr: Was supposed to be on the glr branch
06:15 dalek rakudo/glr:
06:15 dalek rakudo/glr: This reverts commit cb9b821b08095c6d4c1157085e7fc2084f2dd72d.
06:15 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/cf3b24c6c8
06:15 dalek rakudo/glr: 5ba44fc | hoelzro++ | tools/build/NQP_REVISION:
06:15 dalek rakudo/glr: Bump NQP_REVISION
06:15 dalek rakudo/glr:
06:15 dalek rakudo/glr: For close_stdin fix in MoarVM
06:15 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/5ba44fc3dd
06:15 dalek rakudo/glr: 78b8f29 | (Stefan Seifert)++ | tools/build/NQP_REVISION:
06:15 dalek rakudo/glr: Merge branch 'nom' into glr
06:15 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/78b8f293d1
06:15 ShimmerFairy nine: I could see it might've been some sort of guarantee for  ?$string === ?$string.Numeric.Stringy  , back when "0" was False :)
06:16 nine Darn...really have to leave for the office now
06:28 FROGGS joined #perl6
06:32 [Sno] joined #perl6
06:37 [TuxCM] joined #perl6
06:49 aborazmeh joined #perl6
06:49 aborazmeh joined #perl6
06:50 xinming joined #perl6
06:52 rurban joined #perl6
06:58 ChoHag joined #perl6
07:05 masak morning, #perl6
07:07 ShimmerFairy o/
07:07 itz joined #perl6
07:10 laben joined #perl6
07:13 laben hello #perl6
07:13 abraxxa joined #perl6
07:19 ShimmerFairy o/ laben
07:27 xfix joined #perl6
07:31 Ven joined #perl6
07:35 RabidGravy joined #perl6
07:38 darutoko joined #perl6
07:42 Ven joined #perl6
07:47 Ven joined #perl6
07:56 laben as i feared, Template::Mustache doesn't work because the grammar or the actions fail to create the correct result. now, where do i sign up for a crash course in grammars and actions?
07:56 rurban joined #perl6
07:57 nine laben: http://doc.perl6.org/language/grammars is quite good
07:58 laben nine: do you know if Grammars and/or Actions internally use some sort of push that would need to be adjusted for glr?
07:58 DrForr laben: I've got a pretty extensive grammar/actions combo in the ANTLR module in the ecosystem.
08:05 FROGGS laben: do you get error messages you could gist?
08:05 yoleaux 30 Aug 2015 21:00Z <laben> FROGGS: the LAST phaser doesnt seem to fire when the last execution is finished with next, see "for 1 {LAST {say "last"};next}"
08:05 FROGGS ohh, the LAST phaser thing might be some very easily fixable...
08:05 laben just a quick question, the TOP action is called last, after every other action has been called, right?
08:06 rarara joined #perl6
08:08 nine laben: my Apache-To-Nginx seems to parse just fine and that's a largish grammar. From what I can see, there's just a couple of GLR fixes necessary in the converter code itself.
08:09 nine FROGGS: do you by chance still have my short micro USB cable?
08:11 JimmyZ isn't the TOP first?
08:12 laben somehow an action doesnt get called, which means the correct ast is never made. let's see why...
08:15 espadrine joined #perl6
08:16 [Tux] joined #perl6
08:18 FROGGS nine: damn, it was yours?
08:18 FROGGS ó.ò
08:18 FROGGS I convinced tadzik to take it -.-
08:19 nine This...may even be an advantage. You aren't coming to YAPC are you?
08:19 FROGGS JimmyZ: the TOP token is called first, so the TOP action method is called last
08:19 FROGGS nine: I'm not, correct
08:19 colomon joined #perl6
08:19 nine tadzik may be, so I may even have a chance of getting it back ;)
08:19 FROGGS :S
08:19 FROGGS nine: Sorry
08:19 nine No worry.
08:19 tadzik nine: I'm not, sorry :(
08:20 tadzik nine: but I'm still in Zurich :)
08:20 tadzik if you want to catch me
08:20 nine I'm already back in Linz
08:20 tadzik ah :/
08:21 nine I should have charged my phone in Switzerland at least once. Then I'd have noticed the missing cable.
08:21 nine Curse you large battery!
08:21 rindolf joined #perl6
08:21 colomon joined #perl6
08:22 tadzik :)
08:23 DrForr DAMN YOU FOR BEING SO EFFICIENT
08:24 nine Maybe I should just give in and register a Google account to suck my battery empty in a day ;)
08:24 laben gosh, the bottom action doesn't get called even once on glr, that's why it does not create any ast. at the same time, why the heck is nom calling the same action 2 times, with the same argument, when there's evidently only one occurrence?
08:25 nine laben: sounds like we overcompensated?
08:26 lizmat_ joined #perl6
08:28 domidumont joined #perl6
08:30 laben seems like it's called twice because it's used in an assertion. but the assertion totally doesn't get called in glr. what's going on?
08:42 laben need to go now, have a good day #perl6
08:51 lizmat_ good *, #perl6
08:51 lizmat messages?
08:51 yoleaux 30 Aug 2015 10:39Z <nine> lizmat: I figured out what's wrong with zip-latest.t! It's just that every way we tried to fix the issue hit an underlying bug. Array.clone re-uses the same $!reified buffer for the clone and List.from-iterator creates a new $!reified but uses the same scalar containers for the values. So they are still linked.
08:52 lizmat .tell nine Good to know what caused the issue...  :-)
08:52 yoleaux lizmat: I'll pass your message to nine.
08:52 nine lizmat: It's also fixed by now :)
08:52 yoleaux 08:52Z <lizmat> nine: Good to know what caused the issue...  :-)
08:52 lizmat cool!
08:53 nine There's now only 2 GLR related spec test failures: Sequences do not yet throw the correct exceptions on invalid input and one of the advent tests is not properly GLRed yet.
08:54 telex joined #perl6
08:54 lizmat I saw 5: 1 of them is a known issue on OS X with IO::Socket::INET
08:54 lizmat the other two I'm not sure about
08:55 nine t/spec/S02-literals/numeric.t fails also on nom
08:55 nine ShimmerFairy++ fixed numeric.t to actually test what it's claiming to test and that's why it now shows failures. They should just be todo'd
08:56 cdc .tell [Coke] about http://irclog.perlgeek.de/perl6/2015-08-29#i_11138242, I use for each arguments of a Bash command-line: sub escape-arg(Str $arg is copy) { if $arg ~~ s:g/([ ' ' | \' | \\ | <:Control> ])/{ $0.ord.fmt('\\u%04x') }/ { [~] <$'>, $arg, <'> } else { $arg } }
08:56 yoleaux cdc: I'll pass your message to [Coke].
08:56 nine many-processes-no-close-stdin.t was also a nom issue that has since been fixed by hoelzro++ (NQP bump was missing)
08:58 araujo joined #perl6
08:58 araujo joined #perl6
09:00 ShimmerFairy sorry for not TODOing those tests, by the way; I think my idea was "oh, I messed around in that part of the grammar recently, I could fix that soon!" and then didn't :P
09:03 espadrine_ joined #perl6
09:13 ShimmerFairy m: say :10<-42>  # this is how I guessed you negate adverbial numbers, but apparently I'm wrong :)  (though to be fair the spec says nothing about a negative number of this form)
09:13 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/jP8rFHCpOz␤Malformed radix number␤at /tmp/jP8rFHCpOz:1␤------> 3say :10<7⏏5-42>  # this is how I guessed you negate␤    expecting any of:␤        number in radix notation␤»
09:13 GLRelia rakudo-moar 78b8f2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/zkxkvTeU3r␤Malformed radix number␤at /tmp/zkxkvTeU3r:1␤------> 3say :10<7⏏5-42>  # this is how I guessed you negate␤    expecting any of:␤        number in radix notation␤»
09:16 RabidGravy If I need to reference a type 'B' in a role 'F' that may be consumed in term by 'F'   is doing something like "self ~~ ::('B')" in the role valid way of not having to load the class (and get blown out for circular dependency)?
09:17 ShimmerFairy do you want ::?CLASS, per chance?
09:17 RabidGravy no I don't
09:18 RabidGravy the role may be consumed by a variety of things, it needs to check if something is one of them
09:18 RabidGravy (actually self was a bad example there as the actual code doesn't do that)
09:20 ShimmerFairy you can always stub the class(es) in question, and hopefully be able to use them well enough in the role.
09:23 RabidGravy the actual code is something more like:
09:23 RabidGravy m: role F { method b(Mu:D $f) { say $f ~~ ::("B"); } }; class B does F { }; B.new.b(B.new)
09:23 camelia rakudo-moar 5ba44f: OUTPUT«True␤»
09:23 GLRelia rakudo-moar 78b8f2: OUTPUT«True␤»
09:24 RabidGravy where obviously the circumstances of the creation of the second B are somewhat more complicated
09:26 Ven joined #perl6
09:27 RabidGravy I'm going to go with it and see what happens
09:38 mprelude joined #perl6
09:45 nine ShimmerFairy: could you please todo them now? I'm not sure about the protocol there. For example, do we require RT tickets for todos?
09:48 ShimmerFairy nah, I'll get to it. Just a case of focusing on other stuff, but it'd only take a second. :)
09:48 nine That's even better :)
09:49 nine Btw. sorry if I've been a bit grumpy about this. Past few weeks have been exhausting.
09:51 ShimmerFairy That's alright :)
09:56 dalek roast: 6b38ee7 | ShimmerFairy++ | S02-literals/numeric.t:
09:56 dalek roast: TODO a couple tests.
09:56 dalek roast:
09:56 dalek roast: nine++ for catching that they needed it.
09:56 dalek roast: review: https://github.com/perl6/roast/commit/6b38ee747d
10:04 dalek roast/glr: 8fe513c | ShimmerFairy++ | S02-literals/allomorphic.t:
10:04 dalek roast/glr: Updating allomorphic tests.
10:04 dalek roast/glr:
10:04 dalek roast/glr: In particular coercing the allomorphic types for 'is' tests, since
10:04 dalek roast/glr: allomorphic types probably shouldn't implement === for an allomorphic
10:04 dalek roast/glr: type and one of its component types.
10:04 dalek roast/glr: review: https://github.com/perl6/roast/commit/8fe513c789
10:04 dalek roast/glr: d47db92 | ShimmerFairy++ | S02-literals/listquote.t:
10:04 dalek roast/glr: Fix accidental pure string expectation
10:04 dalek roast/glr:
10:04 dalek roast/glr: Since the test in question isn't meant to test allomorphic types, I
10:04 dalek roast/glr: changed the tested key/value pair to not make an allomorphic type.
10:04 dalek roast/glr: review: https://github.com/perl6/roast/commit/d47db92d59
10:04 dalek roast/glr: 6b38ee7 | ShimmerFairy++ | S02-literals/numeric.t:
10:04 dalek roast/glr: TODO a couple tests.
10:04 dalek roast/glr:
10:04 dalek roast/glr: nine++ for catching that they needed it.
10:04 dalek roast/glr: review: https://github.com/perl6/roast/commit/6b38ee747d
10:04 dalek roast/glr: ba3393d | (Stefan Seifert)++ | S02-literals/ (3 files):
10:04 dalek roast/glr: Merge branch 'master' into glr
10:04 dalek roast/glr: review: https://github.com/perl6/roast/commit/ba3393d285
10:07 Ven joined #perl6
10:21 rurban joined #perl6
10:27 Loren_ joined #perl6
10:45 espadrine_ joined #perl6
11:05 nine m: (1, 2, 6 ... *)[5]
11:05 camelia rakudo-moar 5ba44f: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤  in block <unit> at /tmp/imuAJihaAN:1␤␤Actually thrown at:␤  in block <unit> at /tmp/imuAJihaAN:1␤␤»
11:05 GLRelia rakudo-moar 78b8f2: ( no output )
11:10 ShimmerFairy m: say (1, 2, 6 ... *)[5]
11:10 camelia rakudo-moar 5ba44f: OUTPUT«Unable to deduce arithmetic or geometric sequence from 1,2,6 (or did you really mean '..'?)␤  in block <unit> at /tmp/ZD7R2u6Sq4:1␤␤Actually thrown at:␤  in block <unit> at /tmp/ZD7R2u6Sq4:1␤␤»
11:10 GLRelia rakudo-moar 78b8f2: OUTPUT«Nil␤»
11:13 ggoebel joined #perl6
11:15 telex joined #perl6
11:18 dalek Inline-Perl6/glr: e508f48 | (Stefan Seifert)++ | t/0 (2 files):
11:18 dalek Inline-Perl6/glr: Adopt GLR semantics in test suite
11:18 dalek Inline-Perl6/glr: review: https://github.com/niner/Inline-Perl6/commit/e508f48af7
11:20 ChoHag Is a capture consisting of all the arguments passed to a function available without explicitely capturing it?
11:21 timotimo no, for performance reasons
11:21 timotimo i think
11:21 timotimo hm, though perhas with the nqp ops
11:22 timotimo m: sub test($a, $b, $c?) { say nqp::usecapture(); }
11:22 camelia rakudo-moar 5ba44f: ( no output )
11:22 GLRelia rakudo-moar 78b8f2: OUTPUT«===============================================================================␤The use of nqp::operations has been deprecated for non-CORE code.  Please␤change your code to not use these non-portable functions.  If you really want␤to keep using nqp:…»
11:22 timotimo m: use nqp; sub test($a, $b, $c?) { say nqp::usecapture(); }; test(1, 2);
11:22 camelia rakudo-moar 5ba44f: OUTPUT«Cannot find method 'gist': no method cache and no .^find_method␤  in sub test at /tmp/7OQ5pmkqAg:1␤  in block <unit> at /tmp/7OQ5pmkqAg:1␤␤»
11:22 GLRelia rakudo-moar 78b8f2: OUTPUT«Cannot find method 'gist': no method cache and no .^find_method␤  in sub test at /tmp/XkS6IdVbWZ:1␤  in block <unit> at /tmp/XkS6IdVbWZ:1␤␤»
11:22 timotimo that gives you a very low-level capture
11:23 timotimo you'd inspect it with nqp::captureposarg and whatever lets you grab the named arguments
11:23 ChoHag Is there any other way to call another function with the same arguments as passed in to this one?
11:24 virtualsue joined #perl6
11:25 timotimo m: sub foo($a, $b) { say "$a $b" }; sub bar($a, $b) { &foo.callsame }; bar(1, 2)
11:25 camelia rakudo-moar 5ba44f: OUTPUT«Method 'callsame' not found for invocant of class 'Sub'␤  in sub bar at /tmp/m7g9cyiqGC:1␤  in block <unit> at /tmp/m7g9cyiqGC:1␤␤»
11:25 GLRelia rakudo-moar 78b8f2: OUTPUT«Method 'callsame' not found for invocant of class 'Sub'␤  in sub bar at /tmp/SGvANG8O3S:1␤  in block <unit> at /tmp/SGvANG8O3S:1␤␤»
11:25 timotimo hm
11:35 virtualsue joined #perl6
11:40 RabidGravy great, a bug report for a twenty year old module against perl 5.14.4 on cygwin
11:40 RabidGravy bless their little pointy heads
11:43 Ven joined #perl6
11:47 hoelzro o/ #perl6
11:48 nine Good morning, hoelzro!
11:48 hoelzro afternoon, nine!
11:48 Ven \o, hoelzro , nine , and #perl6
11:49 hoelzro ahoy Ven!
11:53 BooK just looked at jnthn's talk in yapc asia. pretty cool
11:53 BooK but the code examples do not work for me (just did a `rakudobrew build moar`)
11:54 BooK is there a place where I can peek at the complete code?
11:54 zakharyas joined #perl6
11:54 nine Doesn't his talk contain the features that are only implemented on the glr branch?
11:59 _itz_ I think race is only in glr
12:00 timotimo aye
12:00 BooK ah ok
12:02 BooK switching branches then
12:02 BooK meanwhile, doing my first PR to perl6/doc
12:03 nine \o/
12:03 hoelzro BooK++
12:03 Ven BooK++ :)
12:03 dalek doc: c29b488 | (Philippe Bruhat (BooK))++ | lib/Type/Supply.pod:
12:03 dalek doc: Fix syntax error
12:03 dalek doc: review: https://github.com/perl6/doc/commit/c29b4884f5
12:03 dalek doc: 64d55fb | FROGGS++ | lib/Type/Supply.pod:
12:03 dalek doc: Merge pull request #120 from book/patch-1
12:03 dalek doc:
12:03 dalek doc: Fix syntax error
12:03 dalek doc: review: https://github.com/perl6/doc/commit/64d55fb7d1
12:03 FROGGS BooK++
12:05 RabidGravy yay! BooK!
12:06 virtualsue joined #perl6
12:10 BooK I fondly remember my first patch to libwww, it was marginally bigger than this one (swapping two characters) :-)
12:10 timotimo pretty cool, BooK :)
12:11 timotimo welcome to the doc! :)
12:12 BooK mmm, there are others of the same kind in that page
12:12 BooK gotta catch them all!
12:12 Ven who documents the tests, who tests the documents
12:13 moritz BooK: I've invited you to the perl6 org on github
12:13 moritz BooK: if you accept, you can push to the doc repo (and a few others) directly
12:13 nine And I was just gonna say that BooK needs a commit bit :)
12:14 moritz nine: you're too sloooow :-)
12:15 BooK thanks
12:15 BooK I'll drop out of the github interface then :-)
12:18 RabidGravy BooK, as a complete aside I was looking for a picture of you and Eric in those little cars in Braga the other day.  That was ten years ago :-O
12:21 BooK oh I found that picture again recently
12:21 BooK I think it was taken by DrHyde
12:21 cognominal BooK, time to invinte Ven for P3
12:22 Ven mmh?
12:22 rurban joined #perl6
12:22 BooK Ven: http://patch.pm/p3/ # liz & wendy included
12:22 cognominal maybe he can share a ride with liz and wendy back from Granada
12:23 BooK interesting: to filter a chunk of code through perl6. I need to do !perl6 -
12:23 BooK otherwise I seem to see the output of the REPL
12:23 cognominal that is the picture with mongueurs on tricycle?
12:23 Ven I'm afraid I'm at school on the 12th
12:24 BooK it's a Saturday
12:24 Ven yes. I'm at school that day
12:25 BooK claim you're from a religion that doesn't allow you to work on inconvenient days
12:25 FROGGS Ven: but that's like... weekend :o)
12:25 cognominal and prescribes -O fun, that would be the Perl 6 religion.
12:25 Ven FROGGS: yes. my school often makes me work saturdays and sundays (so basically, 13 days in a row, from 9am to 8pm. I have to sacrifice a weekend every 3 weekends for school)
12:26 BooK I think the religion for Monday is Garfiedinity, but that might actually be just an instance of Lazyness (which prescribes working on Any day)
12:27 BooK if you have to make ritual sacrifices, it is a religion
12:27 Ven I had an incredibly hard time to convince my school letting me go to SPW was a good idea, took me more than a week
12:30 dalek perl6-examples: 45bcffa | (Steve Mynott)++ | categories/cookbook/02numbers/02-05-roman-numerals.pl:
12:30 dalek perl6-examples: import Roman numerals from Rosetta Code
12:30 dalek perl6-examples: review: https://github.com/perl6/perl6-examples/commit/45bcffa57b
12:30 BooK my daughters look forward liz and wendy's arrival: http://klonk.bruhat.net/wendy+flore.jpg
12:31 FROGGS Ven: 13 days in a row is horrible!
12:31 Ven BooK: your daughter seems to be pretty tall already
12:31 Ven FROGGS: it gets long and tiring, yes :(
12:31 * FROGGS .oO( 'oribleaux! )
12:31 FROGGS :/
12:33 cognominal is it 42?
12:34 Ven no
12:34 Ven not important anyways
12:34 BooK RabidGravy: not sure you can find the original on the interwebs, but my notes say it was taken by DrHyde
12:34 BooK http://articles.mongueurs.net/comptes-rendus/fpw-2006/closer.png # this version is online, and you're on it too!
12:34 cognominal so you haet yve not found the answer to the big question :_
12:35 RabidGravy BooK :) I'd forgotten about that
12:36 RabidGravy poor Greggy
12:36 cognominal Guillame still got hair on this photo.
12:37 cognominal * Guillaume
12:37 Ven ha! that matz one is great
12:38 cognominal btw, it seems that Matz was in person in yapc asia.
12:38 Ven really? that's nice to hear
12:38 andreoss joined #perl6
12:40 BooK didn't
12:40 andreoss $ time perl6 prob060.pl
12:40 andreoss real    91m35.536s
12:40 andreoss $ time perl prob060.pl5
12:40 andreoss real    0m4.815s
12:40 andreoss
12:40 timotimo oh my
12:40 RabidGravy oops
12:40 timotimo andreoss: any chance you could try that on rakudo/glr? (and do the necessary modifications to the code, of course)
12:40 virtualsue joined #perl6
12:40 RabidGravy on the bright side there is room for improvement
12:41 timotimo yeah, but someone has to go and do that :P
12:41 cognominal ven, second guest speaker  http://yapcasia.org/2015/  scroll down
12:41 andreoss m: say 91 * 60 / 4.8
12:41 camelia rakudo-moar 5ba44f: OUTPUT«1137.5␤»
12:41 GLRelia rakudo-moar 78b8f2: OUTPUT«1137.5␤»
12:41 * BooK wonders if it would be possible to mark code examples and expected output in a special way, so that some test could check the examples work with the current perl6?
12:41 Ven cognominal: indeed!
12:42 BooK pugs had something like this, and executable spec
12:42 andreoss timotimo: ok
12:42 timotimo thank you very much!
12:43 andreoss can i install glr with rakudobrew?
12:43 _itz_ BooK: that would be very useful
12:43 BooK andreoss: I did a checkout glr in ~/.rakudobrew/moar-nom
12:44 BooK _itz_: is the doc in perl5 pod or perl6 pod?
12:44 ShimmerFairy mostly POD 5, for now
12:45 timotimo with rakudobrew you can use "triple"
12:45 timotimo rakudobrew triple glr master master
12:45 timotimo or something like that
12:46 BooK ".'.IO.watch.act(                     { say "$^file changed" } );
12:46 BooK that looks weird
12:46 BooK should I remove the ".'. at the beginning?
12:47 Ven BooK: the . at the beginning means "$_."
12:47 _itz_ m: say ".".IO.WHAT
12:47 camelia rakudo-moar 5ba44f: OUTPUT«(IO::Path)␤»
12:47 GLRelia rakudo-moar 78b8f2: OUTPUT«(IO::Path)␤»
12:48 Ven m: $_ = 'abc'; say .uc
12:48 camelia rakudo-moar 5ba44f: OUTPUT«ABC␤»
12:48 GLRelia rakudo-moar 78b8f2: OUTPUT«ABC␤»
12:48 BooK right, so one of the quotes is incorrect
12:48 BooK m: say ".'.IO.WHAT
12:48 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/LTGGuk9CCD␤Unable to parse expression in double quotes; couldn't find final '"' ␤at /tmp/LTGGuk9CCD:1␤------> 3say ".'.IO.WHAT7⏏5<EOL>␤    expecting any of:␤        argument list␤        doubl…»
12:48 GLRelia rakudo-moar 78b8f2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/pe6pzycVId␤Unable to parse expression in double quotes; couldn't find final '"' ␤at /tmp/pe6pzycVId:1␤------> 3say ".'.IO.WHAT7⏏5<EOL>␤    expecting any of:␤        argument list␤        doubl…»
12:49 cognominal one typo here and there is a good pedagogical device to force the reader to think.
12:50 BooK I have no doubt the doc is perfect and the mistakes are there for the only purpose of my education :-)
12:50 dalek doc: c45f662 | (Philippe Bruhat (BooK))++ | lib/Type/Supply.pod:
12:50 dalek doc: fix more syntax errors
12:50 dalek doc: review: https://github.com/perl6/doc/commit/c45f6621e8
12:51 BooK maybe it's time to read CONTRIBUTING.md
12:51 CQ joined #perl6
13:02 dalek perl6-examples: e83ca50 | (Steve Mynott)++ | categories/cookbook/08-file-contents/08-0 (3 files):
13:02 dalek perl6-examples: three LHF file contents examples
13:02 dalek perl6-examples: review: https://github.com/perl6/perl6-examples/commit/e83ca502c3
13:03 visitor1 joined #perl6
13:11 andreoss m: say $*VM
13:11 camelia rakudo-moar 5ba44f: OUTPUT«moar (2015.7.108.g.7.e.9.f.29.e)␤»
13:11 GLRelia rakudo-moar 78b8f2: OUTPUT«moar (2015.7.108.g.7.e.9.f.29.e)␤»
13:12 andreoss how do i distinguish glr from nom?
13:12 FROGGS andreoss: look at the nick of the bot
13:13 andreoss m: say $*PERL
13:13 camelia rakudo-moar 5ba44f: OUTPUT«Perl 6 (6.Advent)␤»
13:13 GLRelia rakudo-moar 78b8f2: OUTPUT«Perl 6 (6.Advent)␤»
13:14 FROGGS andreoss: glr itself is just a branch, so it won't show up in a version string
13:14 FROGGS but...
13:14 FROGGS m: say ::('Seq') ~~ Failure ?? "good ol' nom" !! 'GLR'
13:14 camelia rakudo-moar 5ba44f: OUTPUT«good ol' nom␤»
13:14 GLRelia rakudo-moar 78b8f2: OUTPUT«GLR␤»
13:17 andreoss m: my @x := grep Int, 1..10;
13:17 camelia rakudo-moar 5ba44f: ( no output )
13:17 GLRelia rakudo-moar 78b8f2: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Seq'␤  in block <unit> at /tmp/uN6Fhkj1do:1␤␤»
13:19 Ven joined #perl6
13:21 andreoss how do i port my code with := to GLR? I get a bunch of Seq errors
13:22 moritz andreoss: if you need to access @x by index, it's best to add .list to the expression on the right-hand side
13:22 moritz andreoss: if you want to iterate over it, the nicest fix is
13:22 moritz m: my \x := grep Int, 1..10; .say for x;
13:22 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:22 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:25 mprelude joined #perl6
13:29 andreoss >This Seq has already been iterated, and its values consumed
13:30 moritz if you need to iterate it twice or more often, use .list to get a proper List
13:30 muraiki joined #perl6
13:30 andreoss i'm actually iterating on copy
13:35 FROGGS andreoss: are you still using binding by any chance?
13:36 FROGGS m: my @x = grep(Int, 1..10).list; .say for @x;
13:36 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 FROGGS m: my @x := grep(Int, 1..10).list; .say for @x;
13:36 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 FROGGS ahh, hmmm, so that does not trigger the said msg
13:36 FROGGS m: my @x = grep(Int, 1..10); .say for @x;
13:36 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:36 FROGGS hmmm, no idea
13:37 moritz andreoss: if you want help, you need to show us some code that produces the error
13:37 FROGGS m: my @x = grep(Int, 1..10); .say for @x; .say for @x
13:37 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 timotimo m: my $x = grep(Int, 1..10); .say for $x.list; .say for $x.list;
13:37 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 FROGGS m: my @x := grep(Int, 1..10); .say for @x; .say for @x
13:37 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 GLRelia rakudo-moar 78b8f2: OUTPUT«Type check failed in binding; expected 'Positional' but got 'Seq'␤  in block <unit> at /tmp/bG3kvFqvWg:1␤␤»
13:37 FROGGS there we go
13:37 moritz m: my $x = grep(Int, 1..10); .say for $x.List; .say for $x.List
13:37 camelia rakudo-moar 5ba44f: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤»
13:37 GLRelia rakudo-moar 78b8f2: OUTPUT«1␤2␤3␤4␤5␤6␤7␤8␤9␤10␤This Seq has already been iterated, and its values consumed␤  in block <unit> at /tmp/gFHUIlm6Tt:1␤␤»
13:37 FROGGS dont use assignment if you want to have a proper copy
13:37 timotimo oh, .List not .list
13:38 moritz .list caches, .List not
13:38 timotimo 'k
13:38 moritz so, depends on what you want
13:38 timotimo .so depends on what you want? but don't you always get a Bool from .so? :p
13:38 moritz but I haven't seen a good use case for either while still keeping the original Seq around
13:39 FROGGS I guess Seqs are like Junctions... you want to collapse them as early as possible
13:39 moritz or pass them around without storing them in a variable
13:39 timotimo i'd rather say "you don't want to let them loose on the program" than the other thing
13:40 andreoss https://paste.debian.net/309804/
13:40 nine You don't need .list if you assign to an @array anyway
13:41 andreoss feels like i have to unlearn some Perl 6 already
13:42 timotimo a tiny bit, yeah :|
13:42 moritz Variable '@x' is not declared                                                      │········
13:42 moritz at glr-andreoss:60                                                                 │········
13:42 nine I think there's just fewer reasons to use := in the first place, which I see as a good thing.
13:42 timotimo agreed
13:43 andreoss https://paste.debian.net/309806/
13:46 RabidGravy m: say "Foo".WHAT.^name; # is there a better way of doing this if I must get the name of the type as a string?
13:46 camelia rakudo-moar 5ba44f: OUTPUT«Str␤»
13:46 GLRelia rakudo-moar 78b8f2: OUTPUT«Str␤»
13:47 tadzik I don't think so
13:47 nine m: say "Foo".^name;
13:47 camelia rakudo-moar 5ba44f: OUTPUT«Str␤»
13:47 GLRelia rakudo-moar 78b8f2: OUTPUT«Str␤»
13:47 tadzik but why do you?
13:47 tadzik oh :)
13:47 timotimo m: say Str.gist; say Str.Str; say Str.perl
13:47 camelia rakudo-moar 5ba44f: OUTPUT«(Str)␤Use of uninitialized value of type Str in string context  in block <unit> at /tmp/zKBwRBUMzH:1␤␤Str␤»
13:47 GLRelia rakudo-moar 78b8f2: OUTPUT«(Str)␤Use of uninitialized value of type Str in string context  in block <unit> at /tmp/qlxNzApGbY:1␤␤Str␤»
13:47 timotimo m: say Str.perl
13:47 camelia rakudo-moar 5ba44f: OUTPUT«Str␤»
13:47 GLRelia rakudo-moar 78b8f2: OUTPUT«Str␤»
13:47 * moritz likes $obj.^name best
13:48 timotimo m: my $foo = Str; say $foo.^name
13:48 GLRelia rakudo-moar 78b8f2: OUTPUT«Str␤»
13:48 camelia rakudo-moar 5ba44f: OUTPUT«Str␤»
13:48 timotimo ah
13:48 timotimo m: my $foo = Str; say $foo.VAR.^name
13:48 camelia rakudo-moar 5ba44f: OUTPUT«Scalar␤»
13:48 GLRelia rakudo-moar 78b8f2: OUTPUT«Scalar␤»
13:48 timotimo m: my $foo = Str; say $foo.VAR.name
13:48 camelia rakudo-moar 5ba44f: OUTPUT«$foo␤»
13:48 GLRelia rakudo-moar 78b8f2: OUTPUT«$foo␤»
13:48 timotimo so that's how you get that
13:49 timotimo m: sub what's-my-name(\foo) { say foo.name }; my $foo = Str; what's-my-name($foo)
13:49 camelia rakudo-moar 5ba44f: OUTPUT«Method 'name' not found for invocant of class 'Str'␤  in sub what's-my-name at /tmp/K5zwpfvq_2:1␤  in block <unit> at /tmp/K5zwpfvq_2:1␤␤»
13:49 GLRelia rakudo-moar 78b8f2: OUTPUT«Method 'name' not found for invocant of class 'Str'␤  in sub what's-my-name at /tmp/XQHOFjqi3q:1␤  in block <unit> at /tmp/XQHOFjqi3q:1␤␤»
13:49 pink_mist foo.VAR.name surely?
13:49 nine andreoss: I'd just write for grep Prime, 1 .. 10000 -> $prime {
13:49 visitor1 Hi all!
13:50 timotimo yeah, pink_mist, i was just checking :)
13:50 pink_mist right =)
13:50 visitor1 Why this works:
13:50 visitor1 r: class C { has Int $.v is rw = 0; multi method new(Int $v){ self.bless(v=>$v) }; multi method new(@vals){ self.bless(v=>[+](@vals)) } }; my @numlist = (10,20,30); my $c = C.new(@numlist); say $c.perl;
13:50 nine andreoss: oh, sorry, forget that
13:50 GLRelia rakudo-jvm 78b8f2: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
13:50 GLRelia ..rakudo-moar 78b8f2: OUTPUT«C.new(v => 60)␤»
13:50 camelia rakudo-{moar,jvm} 5ba44f: OUTPUT«C.new(v => 60)␤»
13:50 timotimo hi visitor1
13:50 visitor1 and this does not?
13:50 visitor1 r: class C { has Int $.v is rw = 0; multi method new(Int $v){ self.bless(v=>$v) }; multi method new(Int @vals){ self.bless(v=>[+](@vals)) } }; my @numlist = (10,20,30); my $c = C.new(@numlist); say $c.perl;
13:50 khw joined #perl6
13:50 GLRelia rakudo-jvm 78b8f2: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
13:50 GLRelia ..rakudo-moar 78b8f2: OUTPUT«Default constructor for 'C' only takes named arguments␤  in block <unit> at /tmp/tmpfile:1␤␤»
13:50 camelia rakudo-{moar,jvm} 5ba44f: OUTPUT«Default constructor for 'C' only takes named arguments␤  in block <unit> at /tmp/tmpfile:1␤␤»
13:50 andreoss nine: i need @primes as stack, so it wont' check same prime number twice
13:50 andreoss nine: that's why .shift
13:51 visitor1 The difference is using (@vals) or (Int @vals) in an argument list for a new method
13:52 virtualsue_ joined #perl6
13:52 timotimo visitor1: if you have "Int @vals" as a parameter, you have to explicitly pass in an array you've manually declared to be "of Int"
13:52 nine andreoss: so what's the remaining issues with your script?
13:53 visitor1 r: class C { has Int $.v is rw = 0; multi method new(Int $v){ self.bless(v=>$v) }; multi method new(Int @vals){ self.bless(v=>[+](@vals)) } }; my Int @numlist = (10,20,30); my $c = C.new(@numlist); say $c.perl;
13:53 GLRelia rakudo-moar 78b8f2: OUTPUT«C.new(v => 60)␤»
13:53 GLRelia ..rakudo-jvm 78b8f2: OUTPUT«Can't open perl script "/home/steve/rakudo-inst/bin/eval-client.pl": No such file or directory␤»
13:53 camelia rakudo-{moar,jvm} 5ba44f: OUTPUT«C.new(v => 60)␤»
13:53 moritz visitor1: see also my rant about this on p6u: http://www.nntp.perl.org/group/perl.perl6.users/2015/07/msg2155.html
13:53 visitor1 Ahhh ok... thanks timotimo!
13:53 timotimo seems like we want to prevent glrelia to try to run rakudo-jvm
13:53 timotimo well, that or build rakudo-jvm on it, too
13:54 virtualsue joined #perl6
13:54 moritz timotimo: I hope that GLRelia's life time is counted in days, anyway
13:54 timotimo sure
13:54 timotimo hm
13:54 RabidGravy merge imminent?
13:54 timotimo though i'd say it'd be interesting to keep nomelia around after glr becomes the main development branch
13:55 timotimo so that we can easily see differences on the channel if we want to
13:55 timotimo i wouldn't run it for every "m:", though
13:55 timotimo maybe "nom:"
13:55 moritz timotimo: shouldn't be so hard
13:55 timotimo aye
13:56 moritz just like we kept pre-nom rakudo ("Beijing") around for a while
13:56 timotimo ah
13:56 timotimo i don't remember that :)
13:56 moritz timotimo: and if you do a MoarVM release today, I can do a nom-based Rakudo release tonight
13:56 timotimo oh!
13:57 moritz and then we can use that release as the reference thingy for camelia
13:57 timotimo i should get going, then!
13:57 moritz no pressure, but yes, do it now! :-)
13:58 andreoss nine: it fails to work with glr
13:58 timotimo haha, that's "the" reason? :)
13:59 timotimo the remaining issue, i mean?
13:59 andreoss nine: i assume because of multiple copies of Seq
14:00 timotimo and i assume getting a --profile of that 90 minute run thingie would result in a ridiculous amount of data
14:01 timotimo just because we record a bit of data for every single gc run
14:01 timotimo and in 90 minutes, the gc runs *a* *lot*
14:01 timotimo (unless you're only numbercrunching native ints/nums in a tight loop)
14:02 nine andreoss: fails in what way?
14:02 timotimo maybe we want a little bit of intelligence that'll start only recording major collections after a certain number of minor collections
14:04 andreoss nine: >This Seq has already been iterated, and its values consumed
14:04 timotimo i'm really looking forward to getting these two new monitors <3
14:05 nine andreoss: cannot reproduce that here
14:06 * moritz has two glr test failures: t/spec/S03-sequence/misc.t and t/spec/integration/99problems-21-to-30.t
14:06 nine moritz: yes, that's the two remaining ones
14:06 thou joined #perl6
14:07 nine moritz: 99problems-21-to-30.t is a plain missing GLRification one and the sequence test fails because we do not detect invalid input early enough.
14:07 masak wow. two.
14:07 skids joined #perl6
14:08 moritz m: say combination(2, <a b c>)
14:08 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/HcDhsG_w2b␤Undeclared routine:␤    combination used at line 1. Did you mean 'combinations'?␤␤»
14:08 GLRelia rakudo-moar 78b8f2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tacMUsSW36␤Undeclared routine:␤    combination used at line 1. Did you mean 'combinations'?␤␤»
14:09 timotimo +s
14:09 andreoss nine: https://paste.debian.net/309810/ this fails for me
14:09 moritz timotimo: that test file has a custom sub combination
14:09 andreoss may be i have a broken installation or somethimg, sorry if i wasted your time
14:10 virtualsue joined #perl6
14:10 timotimo ah, i see
14:11 nine andreoss: add a .Array in line 45
14:12 nine andreoss: gather returns a Seq, so you created an Array containing Seqs and expected it to be an Array of Arrays
14:13 andreoss nine: thanks
14:14 cognominal joined #perl6
14:16 nine andreoss: FWIW I see quite some opportunity for optimization. If you change your implementation to not use .shift anymore and instead operate only on Lists, it may be quite a bit faster.
14:18 domidumont joined #perl6
14:21 andreoss but List cannot be used as stack
14:23 nine Exactly! Your implementation makes countless copies of the same list. If you use the same static list and use subscripting instead, it may be faster.
14:24 dalek rakudo-star-daily: 023f6a2 | coke++ | log/ (9 files):
14:24 dalek rakudo-star-daily: today (automated commit)
14:24 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/023f6a20cc
14:24 nine m: my @list = grep Prime, 1 .. 100; sub foo(@primes) { say @primes.Str }; foo(@list[1 .. *];
14:24 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/R6cLoEJcGC␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/R6cLoEJcGC:1␤------> 3 { say @primes.Str }; foo(@list[1 .. *];7⏏5<EOL>␤    expecting any of:␤        argu…»
14:24 GLRelia rakudo-moar 78b8f2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/IKdXWb_OM1␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/IKdXWb_OM1:1␤------> 3 { say @primes.Str }; foo(@list[1 .. *];7⏏5<EOL>␤    expecting any of:␤        argu…»
14:24 nine m: my @list = grep Prime, 1 .. 100; sub foo(@primes) { say @primes.Str }; foo(@list[1 .. *]);
14:24 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/iUMEFrBsyZ␤Undeclared name:␤    Prime used at line 1␤␤»
14:24 GLRelia rakudo-moar 78b8f2: OUTPUT«5===SORRY!5=== Error while compiling /tmp/UAHvGFDUaF␤Undeclared name:␤    Prime used at line 1␤␤»
14:25 nine m: my @list = grep { $_.is-prime }, 1 .. 100; sub foo(@primes) { say @primes.Str }; foo(@list[1 .. *]);
14:25 camelia rakudo-moar 5ba44f: OUTPUT«3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97␤»
14:25 GLRelia rakudo-moar 78b8f2: OUTPUT«3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97␤»
14:27 [TuxCM] joined #perl6
14:27 nine Meh...we'll still create a copy of the array in that case.
14:27 timotimo grep-index? :)
14:28 timotimo hm, won't really help, right?
14:28 nine I'd say a plain old get-remarkable(:@primes, :$start-index)
14:30 JimmyZ about parrot backend, I think the better ways is to disable it with one line code in configure.pl insteal of remove the code.
14:31 moritz if we disable it, we can't test it anymore
14:34 zostay m: my $x = 1, 2, 3; $x.WHAT.say
14:34 camelia rakudo-moar 5ba44f: OUTPUT«WARNINGS:␤Useless use of "," in expression "my $x = 1, 2, 3" in sink context (line 1)␤(Int)␤»
14:34 GLRelia rakudo-moar 78b8f2: OUTPUT«(Int)␤»
14:35 zostay m: my $x := 1, 2, 3; $x.WHAT.say
14:35 camelia rakudo-moar 5ba44f: OUTPUT«(Parcel)␤»
14:35 GLRelia rakudo-moar 78b8f2: OUTPUT«(List)␤»
14:35 JimmyZ moritz: the advice is to someone who want remove the code from git repo ;)
14:36 ShimmerFairy Is there a reason there's a 'say' multi that's just for Str just so you don't call .gist on it, even though Str.gist is just the Str itself? :)
14:36 moritz JimmyZ: but it's not a solution for anybody; either it's removed, or it stays. There is no middle ground.
14:36 moritz ShimmerFairy: yes. Performance.
14:36 ShimmerFairy moritz: I figured, though gist(Str:D:) { self } seems so innocuous... :)
14:37 JimmyZ incase someone do Str.gist
14:37 ShimmerFairy (I ask because allomorphic types pick that multi accidentally, and "remove the Str multi" seemed like a potential solution)
14:37 timotimo ShimmerFairy: one of the most expensive things moar can do is invoke a sub, or so it seems :|
14:38 ShimmerFairy interesting, is there any clue as to why?
14:39 ShimmerFairy .oO(The obvious solution is to make sure everything is a method, and thus everything is classes. Where have I heard that before...?)
14:39 JimmyZ java?
14:39 timotimo running a script through perf shows that we spend a big amount of time in increasing the ref count of a frame
14:39 timotimo oh, i meant "invoking a code object"
14:39 timotimo not just subs
14:39 Mhoram joined #perl6
14:40 JimmyZ sometimes moarvm is fast enough, but the  build plan in rakudo like gimme, very slow
14:41 timotimo it's so good we've gotten rid of gimme + reify :)
14:41 ShimmerFairy timotimo: to be fair, I'm sure handling what could in some contrived way be called "just advanced gotos" isn't easy :)
14:41 JimmyZ the build plan is as slow as gimme ...
14:41 timotimo :)
14:51 virtualsue_ joined #perl6
14:51 nine But surely we can inline a gist(Str:D:) { self }?
14:56 domidumont joined #perl6
15:05 timotimo hopefully
15:05 perl6_newbee joined #perl6
15:06 masak perl6_newbee: hola.
15:09 Khisanth joined #perl6
15:12 laubster joined #perl6
15:12 [Coke] bots, wake up, get messages.
15:12 yoleaux 08:56Z <cdc> [Coke]: about http://irclog.perlgeek.de/perl6/2015-08-29#i_11138242, I use for each arguments of a Bash command-line: sub escape-arg(Str $arg is copy) { if $arg ~~ s:g/([ ' ' | \' | \\ | <:Control> ])/{ $0.ord.fmt('\\u%04x') }/ { [~] <$'>, $arg, <'> } else { $arg } }
15:14 telex joined #perl6
15:16 xinming_ joined #perl6
15:20 MilkmanDan joined #perl6
15:26 muraiki hi all. I have a script that creates a number of IO::Notification.watch-path Supplies. what I've noticed is that CPU usage tends to stay at a constant level regardless of how many directories are watched, but that the CPU usage also increases in proportion to the # cpu cores and their speed
15:27 muraiki for instance, on a dual core, core 2 duo machine I have 5% cpu usage. but on a 16 core machine (not sure of the exact cpu type, but obviously quite new) I see 10% cpu usage
15:27 muraiki hrm, well actually these cores on the 16 core machine aren't as fast as I thought. so maybe that's it
15:30 muraiki is there a way I can get statistics on how many threads are being used across all these supplies? (note: I obviously don't really understand how this all works :)
15:30 cognominal joined #perl6
15:34 JimmyZ m: say $*SCHEDULER.loads
15:34 camelia rakudo-moar 5ba44f: OUTPUT«0␤»
15:34 GLRelia rakudo-moar 78b8f2: OUTPUT«0␤»
15:37 JimmyZ m: await do for ^5 { start { sleep 1 }; say $*SCHEDULER.loads }
15:37 camelia rakudo-moar 5ba44f: OUTPUT«0␤1␤2␤3␤4␤»
15:37 GLRelia rakudo-moar 78b8f2: OUTPUT«0␤1␤2␤3␤4␤»
15:37 JimmyZ m: await do for ^5 { start { sleep 1 }; say $*SCHEDULER.loads }; sleep 2; say $*SCHEDULER.loads
15:37 camelia rakudo-moar 5ba44f: OUTPUT«0␤1␤2␤3␤4␤0␤»
15:37 GLRelia rakudo-moar 78b8f2: OUTPUT«0␤1␤2␤3␤4␤0␤»
15:38 JimmyZ muraiki: ^^ maybe this is you want?
15:39 muraiki JimmyZ: thanks, but I think the problem actually just comes down to "things aren't optimized yet." a perl6 -e 'sleep' showed similar cpu usage on these computers, heh
15:39 pflanze masak, was it junctions that you said in your talk you'd only use lexically? If I understand right, e.g. any(@foo) eq $x would be equivalent to @foo.any(* eq $x) assuming that there's an any method on arrays (I don't know if there is)?
15:39 muraiki yeah actually sleeping results in the same amount of cpu usage. :(
15:39 pflanze I just ran into those and was questioning sanity, then I remembered your talk.
15:41 masak pflanze: yes :)
15:41 masak pflanze: I do things like `any(@foo) eq $x` too sometimes
15:41 masak pflanze: the important thing to be aware of is that the result is not a *Bool*, but a *junction of Bools*
15:41 masak pflanze: you can amend this by doing `so any(@foo) eq $x`
15:42 JimmyZ muraiki: I am not sure, since sleep calls the system sleep
15:42 pflanze oi
15:42 masak pflanze: for more information, see https://perl6advent.wordpress.com/2014/12/03/day-3-cap-your-junctions/
15:42 muraiki JimmyZ: yeah, but what i mean is I'd have trouble believing system sleep results in the 5%-30% cpu usage I'm seeing
15:42 pflanze *Is* there a higher-order any method on sequences, too?
15:42 muraiki so I assumed it was overhead from moarvm or something
15:43 pflanze Which would simply return a boolean.
15:44 JimmyZ muraiki: maybe it is the parsing time?
15:45 * pflanze goes trying
15:45 muraiki JimmyZ: the cpu usage initially starts at 100%+ for a few seconds, then dies down to 5-30%. so I figured that parsing is what's going on for those first few seconds
15:45 virtualsue joined #perl6
15:46 * JimmyZ have no iden then :)
15:46 muraiki thanks anyways :)
15:46 muraiki it just stinks because I'm probably going to have to rewrite this in go, as 30% is a dealbreaker T_T
15:46 JimmyZ optimizations is hard
15:46 muraiki I want my supplies :(
15:47 * masak jump-scares muraiki
15:47 masak muraiki: SUPPLIES!
15:47 muraiki ;)
15:49 rurban joined #perl6
15:52 Ven joined #perl6
15:52 pflanze Ok, so there is no higher-order "any" method, or more to the point, the "any" method is the junction-building one. (I still don't get how/when "multi" methods and "method" methods are the same and when not.)
16:01 MadcapJake joined #perl6
16:03 perl6_newbee hi masak!
16:03 perl6_newbee \o/
16:03 perl6_newbee :-)
16:11 zakharyas joined #perl6
16:12 leont joined #perl6
16:15 timotimo muraiki: i can't reproduce such cpu usage for perl6 -e 'sleep 30'
16:15 timotimo it reports the cpu time the job got is about 0%
16:15 xinming joined #perl6
16:16 timotimo 0.07user 0.01system 0:30.08elapsed 0%CPU (0avgtext+0avgdata 61636maxresident)k
16:16 timotimo or what exactly are you doing?
16:17 timotimo wait, i forgot to sleep there, didn't i?
16:17 timotimo oh no, wrong workspace
16:18 dalek perl6-roast-data: 466bfa5 | coke++ | / (10 files):
16:18 dalek perl6-roast-data: today (automated commit)
16:18 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/466bfa5972
16:18 muraiki timotimo: I was just doing a perpetual sleep via `perl6 -e 'sleep'`
16:18 muraiki I've reproduced decent cpu usage on 3 systems, one of which just built perl6 today
16:18 muraiki the other 2 have perl6 from sometime last week
16:19 muraiki er, rakudo
16:19 larion joined #perl6
16:26 JimmyZ muraiki: I found the bug, sleeps costs 5% bug
16:27 muraiki JimmyZ: oh nice!
16:27 JimmyZ well, the  -e 'sleep'
16:27 vendethiel joined #perl6
16:27 JimmyZ sleep(10) works well
16:28 muraiki yeah I just discovered that perl6 -e 'while (1) { sleep(5) }' doesn't really use any cpu while perl6 -e 'sleep' does
16:28 JimmyZ muraiki: try s/5/1e16/
16:28 JimmyZ that's the bug
16:29 nine When should one use fail and when die?
16:30 JimmyZ m: say my uint64 $ = 1e16 * 1e9
16:30 camelia rakudo-moar 5ba44f: OUTPUT«This type cannot unbox to a native integer␤  in block <unit> at /tmp/FkGXkDkr3k:1␤␤»
16:30 GLRelia rakudo-moar 78b8f2: OUTPUT«This type cannot unbox to a native integer␤  in block <unit> at /tmp/fyHQBk_bMV:1␤␤»
16:30 muraiki JimmyZ: thanks so much for figuring it out! in my script I'll use the while/sleep(5) form for now
16:31 JimmyZ m: say my uint64 $ = (1e16 * 1e9).Int
16:31 camelia rakudo-moar 5ba44f: OUTPUT«Cannot unbox 84 bit wide bigint into native integer␤  in block <unit> at /tmp/4pYkdgWzcN:1␤␤»
16:31 GLRelia rakudo-moar 78b8f2: OUTPUT«Cannot unbox 84 bit wide bigint into native integer␤  in block <unit> at /tmp/iReVR8XJrY:1␤␤»
16:31 JimmyZ yeah, the bug
16:31 leont Hmmm, using an enum as MAIN argument doesn't DWIM, bummer
16:32 JimmyZ muraiki: I will make a patch to rakudo
16:32 muraiki JimmyZ: you are awesome
16:34 vendethiel leont: how do you mean? enum A <a b c>; sub MAIN(A) {}?
16:34 leont I was hoping for the getops implementation to do something sensible if the argument is either a number(value) or a string (name)
16:37 gfldex joined #perl6
16:41 raiph joined #perl6
16:44 JimmyZ well , maybe to moarvm
16:46 FROGGS[mobile] joined #perl6
16:48 FROGGS[mobile] o/
16:48 yoleaux 30 Aug 2015 16:55Z <nine> FROGGS[mobile]: do you by chance still have my micro USB cable? I seem to remember giving it to you on Wednesday
16:48 yoleaux 30 Aug 2015 17:37Z <laben> FROGGS[mobile]: the LAST phaser doesnt seem to fire when the last execution is finished with next, see "for ^3 {.say; next if $_ < 3; LAST {say "last"}}"
16:48 dalek rakudo/glr: 8a9d288 | (Stefan Seifert)++ | t/01-sanity/99-test-basic.t:
16:48 dalek rakudo/glr: s/done/done-testing in 01-sanity/99-test-basic.t
16:48 dalek rakudo/glr:
16:48 dalek rakudo/glr: Obviously no-one is running the sanity tests regularily...
16:48 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/8a9d288bba
16:49 araujo joined #perl6
16:49 spider-mario joined #perl6
16:52 pullphinger joined #perl6
16:55 [Coke] still 10 failures left on glr. running again to see what I've missed.
16:58 dalek rakudo/glr: d47a42f | (Stefan Seifert)++ | src/core/operators.pm:
16:58 dalek rakudo/glr: Fix (1,2,6...*)[5] not failing as it should
16:58 dalek rakudo/glr:
16:58 dalek rakudo/glr: Fixes the remaining sequence spec tests.
16:58 dalek rakudo/glr: Sequences are now pretty lazy, so even throwing the deduction errors lazily
16:58 dalek rakudo/glr: on top might just be too lazy.
16:58 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/d47a42f58b
16:58 timotimo cool! :)
16:58 nine [Coke]: ^^^
16:58 [Coke] folks running "spectest" aren't running all the tests, which might explain the diff.
16:59 [Coke] nine: will insure my current run includes that.
17:03 [Coke] (I haven't been running the inline::perl5 tests in my summaries, but I assume nine has. :)
17:04 nine Correct :) Intriguingly, they have been passing for weeks already
17:04 domidumont joined #perl6
17:06 * skids has trouble understanding why (1,2,6...*)[5] would not just iterate up to the 6th element and return it.
17:06 pink_mist iterate how?
17:07 skids Oh, because 1,2,6..* itself is a problem.  I see now.
17:08 TimToady but it should still try, and fail, as it does with [0]
17:11 skids btw I have some 3/4th complete code for (someseq)[...] = someotherseq and to fix @a[...] = some_iterable.
17:11 skids But for some reason MMD seems to not be very happy in the mire of all the postcircumfix:<[ ]> candidates.
17:12 skids I don't know if that's a new GLR thing or whether it's always been iffy in that area.
17:13 TimToady um, assigning to elements of a list seems iffy, since lists are immutable and do not guarantee containers (plus they'd be anonymous there anyway)
17:13 skids Especially constraints on \SELF in those cnadidates seem to be ignored.
17:13 TimToady or was that a sequence of containers?
17:13 skids Sequence of containers.
17:14 skids (and get a cannot-assign error if you run into an immutable in ther)
17:16 skids A lingering question is what the eagerness of list assignment for the seq = iteratable case should be -- none? chained to the left?
17:18 mohij joined #perl6
17:20 FROGGS joined #perl6
17:20 FROGGS uhh, it is getting toady in here...
17:20 FROGGS ahh, better :o)
17:22 lizmat joined #perl6
17:23 nine On a make stresstest we're now failing t/spec/integration/99problems-21-to-30.t and t/spec/integration/advent2012-day04.t
17:23 FROGGS lizmat: btw, my plan is to make the CU.load stuff work (== spectest clean), and then hand it over to you for api refactoring if there is any need and if you feel like
17:24 lizmat FROGGS: sounds like a plan to me...
17:24 FROGGS :o)
17:24 lizmat will be driving for most of tomorrow anyways
17:24 spidermario joined #perl6
17:24 FROGGS though, will have to make sure Inline::Perl5 keeps working of course
17:25 lizmat FROGGS: of course :-)
17:25 Khisanth joined #perl6
17:26 andreoss joined #perl6
17:26 dalek roast/glr: 0cfc360 | (Stefan Seifert)++ | integration/advent2012-day04.t:
17:26 dalek roast/glr: Fix a source of "expected 'Positional' but got 'Int'" in advent2012-day04.t
17:26 dalek roast/glr: review: https://github.com/perl6/roast/commit/0cfc3609a7
17:27 andreoss timotimo: with moar-glr it's about the same
17:27 andreoss 92 min
17:28 cognominal joined #perl6
17:29 andreoss .is-prime is slower than Math::Prime::XS but not 1000 times
17:30 JimmyZ muraiki: https://github.com/MoarVM/MoarVM/commit/68c67166a456fce142aa972867ba7681897c168d
17:31 muraiki JimmyZ: awesome! thanks :)
17:34 timotimo JimmyZ: good catch!
17:35 JimmyZ thank you
17:36 nine m: my @a = 1; say 0, @a Z+ @a, 0;
17:36 camelia rakudo-moar 5ba44f: OUTPUT«1 1␤»
17:36 GLRelia rakudo-moar d47a42: OUTPUT«(1 1)␤»
17:36 pullphinger joined #perl6
17:36 nine m: my @a = 1, 1; say 0, @a Z+ @a, 0;
17:36 camelia rakudo-moar 5ba44f: OUTPUT«1 2 1␤»
17:36 GLRelia rakudo-moar d47a42: OUTPUT«(2 2)␤»
17:39 dalek roast/glr: 91cd7dc | (Stefan Seifert)++ | integration/advent2012-day04.t:
17:39 dalek roast/glr: Fix generating Pascal's triangle in GLR
17:39 dalek roast/glr:
17:39 dalek roast/glr: Fixes integration/advent2012-day04.t
17:39 dalek roast/glr: review: https://github.com/perl6/roast/commit/91cd7dc6e9
17:40 nine So there's really only a single spec test file failing now :)
17:40 [ptc] \o/
17:41 nine And that will not require any changes to rakudo. It's just a basic GLRification of a single test.
17:43 TimToady m: my @a = 1, 1; say 0, |@a Z+ |@a, 0;
17:43 camelia rakudo-moar 5ba44f: OUTPUT«5===SORRY!5===␤Variable '&prefix:<|>' is not declared␤at /tmp/Wi9cs20cEx:1␤------> 3my @a = 1, 1; say 0, 7⏏5|@a Z+ |@a, 0;␤Variable '&prefix:<|>' is not declared␤at /tmp/Wi9cs20cEx:1␤------> 3my @a = 1, 1; say 0, |@a Z+ 7⏏5|@a, 0…»
17:43 GLRelia rakudo-moar d47a42: OUTPUT«(1 2 1)␤»
17:43 nine Oh, so the parens would not have been necessary
17:44 TimToady list infixes are, like, list infixes, and stuff
17:45 TimToady that's why hypers aren't :)
17:46 woolfy joined #perl6
17:47 nine Anyway, why am I fixing spec tests? I should be a) making dinner b) work on my YAPC talk or c) both
17:48 * TimToady & Glo --> Restaurante Páprika
17:50 dalek roast/glr: 34116e7 | (Stefan Seifert)++ | integration/99problems-21-to-30.t:
17:50 dalek roast/glr: GLRify test in integration/99problems-21-to-30.t
17:50 dalek roast/glr:
17:50 dalek roast/glr: This is it! The final spec test!
17:50 dalek roast/glr: review: https://github.com/perl6/roast/commit/34116e7331
17:50 nine It is...done.
17:50 lizmat nine++  !!!!
17:50 nine [Coke]: ^^^
17:50 lizmat congratulations!!!
17:51 woolfy nine++  woohooo!!!
17:51 cdc nine++
17:52 muraiki w00t
17:52 muraiki nine++
17:52 nine Yesterday on the train I wasted half an hour on this test. Today I just discarded that, went in, added a couple of characters and it just worked on the very first try :)
17:53 vendethiel nine: AMAZING :D
17:53 muraiki the joy of programming
17:53 nine And I still don't know what this thing is actually doing *g*
17:54 woolfy Magical programming!
17:54 FROGGS nine++ # \o/
17:55 nine All tests successful.
17:55 nine Files=1093, Tests=121114, 289 wallclock secs (17.66 usr  2.29 sys + 1969.59 cusr 73.32 csys = 2062.86 CPU)
17:55 nine Result: PASS
17:55 telex joined #perl6
17:56 dalek roast/glr: b458004 | TimToady++ | integration/advent2012-day04.t:
17:56 dalek roast/glr: remove superstitious parens
17:56 dalek roast/glr: review: https://github.com/perl6/roast/commit/b4580044e0
17:59 lizmat nine: I was wondering why the $ was needed on the return $[]
17:59 lizmat perhaps that's a sig issue with "sub" return?
18:00 larion joined #perl6
18:02 lizmat Files=1082, Tests=121039, 355 wallclock secs (22.69 usr  4.15 sys + 2341.63 cusr 135.15 csys = 2503.62 CPU)
18:02 lizmat 1 fail: IO::Socket::INET (OS X specific fail)
18:04 nine lizmat: it's needed because group is a recursive sub and its result is used in a map call
18:09 dalek perl6-roast-data: 6879af9 | coke++ | / (2 files):
18:09 dalek perl6-roast-data: another glr run for 'today'
18:09 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/6879af911e
18:11 [Coke] http://shapecatcher.com/ - can't find a unicode glyph? draw it.
18:22 mohij Rakudos build process is targeting perl 5.8.0, right?
18:32 geekosaur 5.10.0 I think
18:32 geekosaur there's been discussion about wanting stuff from 5.10.1 but not wanting to bump the dependency
18:32 FROGGS m: require Flubber:from<Foo>
18:32 camelia rakudo-moar 5ba44f: OUTPUT«Could not find Flubber in any of:␤  file#/home/camelia/.perl6/2015.07.1-182-g5ba44fc/lib␤  inst#/home/camelia/.perl6/2015.07.1-182-g5ba44fc␤  file#/home/camelia/rakudo-inst-1/share/perl6/lib␤  file#/home/camelia/rakudo-inst-1/share/perl6/vendor/lib…»
18:32 GLRelia rakudo-moar d47a42: OUTPUT«Could not find Flubber in any of:␤  file#/home/steve/.perl6/2015.07.1-679-gd47a42f/lib␤  inst#/home/steve/.perl6/2015.07.1-679-gd47a42f␤  file#/home/steve/sandbox/perl6/rakudo/install/share/perl6/lib␤  file#/home/steve/sandbox/perl6/rakudo/install/…»
18:32 FROGGS m: require "Flubber":from<Foo>
18:32 camelia rakudo-moar 5ba44f: OUTPUT«Could not find file 'Flubber' for module Flubber␤  in any load_module at src/gen/m-ModuleLoader.nqp:200␤  in block <unit> at /tmp/bIJCLEJZrx:1␤␤»
18:32 GLRelia rakudo-moar d47a42: OUTPUT«Could not find file 'Flubber' for module Flubber␤  in any load_module at src/gen/m-ModuleLoader.nqp:200␤  in block <unit> at /tmp/K3X3oSaLPd:1␤␤»
18:32 FROGGS bug™
18:34 lizmat :-)
18:38 [ptc] nine++  # final spec test!
18:42 mohij geekosaur: thanks!
18:44 leont Depending on 5.10.0 is a bit silly, it's not very commonly used in the wild, and generally considered a flawed release. Might as well depend on 5.10.1
18:47 lizmat dinner&
18:47 pyrimidi_ leont: IIRC most of the 5.x.0 releases kinda fall into that 'flawed' category (same was said of 5.8.0)
18:47 blackcat_ joined #perl6
18:47 leont The same is true for 5.8.0 actually, for different reasons
18:47 leont pyrimidi_: yeah, before 5.12 at least
18:48 Sgeo_ joined #perl6
18:48 leont Nowadays it's just ".1 has a few bugs fewer" instead of "you've got to be insane to use this in production"
18:48 cschwenz joined #perl6
18:48 woolfy left #perl6
18:49 masak on some days I feel like you've got to be insane to use any software in production.
18:50 nine Or hardware
18:50 huf or people, really.
18:51 masak nothing works
18:51 leont L'enfer, c'est les autres!
18:51 huf masak: so how come a lot of things work?
18:51 huf how do things get better? that's the great mystery
18:51 masak huf: measurement error.
18:51 masak or dumb luck
18:52 pyrimidi_ chaos
18:52 moritz it's delusion to think lot of things work
18:52 moritz speaking of not working
18:53 moritz I can't make a rakudo release because there's no moarvm release
18:53 pyrimidi_ https://www.youtube.com/watch?v=JmzuRXLzqKk
18:53 moritz and I can't make a maorvm release, because I don't understand moarvm well enough to write a sensible changelog file
18:53 moritz so, no rakudo release this month, it seems
18:53 masak moritz: well, you tried
18:54 moritz well, I failed.
18:54 masak at least until someone who knows how makes a moarvm release.
18:55 moritz where I live, the month has only about 3 more hours, of which I don't plan to be awake more than 2 or so
18:55 moritz so... unlikely
18:57 japhb dangitall
18:57 vendethiel masak: it's fine if it's only some days!
18:57 rurban joined #perl6
18:58 nine [ptc]: couldn't you write the MoarVM ChangeLog?
18:59 xinming_ joined #perl6
18:59 japhb moritz: At any given moment, there are an uncountable infinity of things not working, but thanks to the wonders of quantum mechanics, we don't notice them.  On the other hand, this says absolutely nothing about the number of things that *are* working, which anecdotally I would say is a lot.  Where by 'anecdote', I mean "apparent continuing existence of Earth, the Solar System, our galaxy, and in fact (as far as we can tell) the entire visible universe".
19:00 vendethiel japhb: I'm not sure it's quantum mechanics, I think it's rather my terrible eyesight
19:00 japhb vendethiel: Additive effects.
19:00 vendethiel *g*
19:00 japhb Nothing like philosophising on a full stomach and an empty brain.
19:01 [ptc] nine: I could give it a go.  When does it need to be done by?
19:02 nine [ptc]: well moritz's still awake for some 2 hours ;)
19:02 japhb [ptc]: Unless you can convince moritz to stay awake later, I'd say "quite soon"
19:02 nine I'm packing for a conference. Seems like I've only been at one yesterday...
19:02 nine Oh wait, I was!
19:02 [ptc] nine: I'm also packing for a conference ;-P
19:06 [ptc] I'm giving it a go, however hoelzro, timotimo and JimmyZ have done a lot of stuff that I don't know about (or yet understand)
19:07 moritz japhb: there you have it, just anecdotes, no data :-)
19:14 RabidGravy m: class C:ver<v0.0.1> { }; say C.^ver ; # anyone know where I can start looking to fix this?
19:14 camelia rakudo-moar 5ba44f: OUTPUT«Method 'gist' not found for invocant of class 'NQPMu'␤  in block <unit> at /tmp/WuBp2uGt4Y:1␤␤»
19:14 GLRelia rakudo-moar d47a42: OUTPUT«Method 'gist' not found for invocant of class 'NQPMu'␤  in block <unit> at /tmp/KDzGgWyHem:1␤␤»
19:15 yqt joined #perl6
19:15 larion joined #perl6
19:15 FROGGS RabidGravy: perhaps look at Perl6::Grammar and Perl6::Actions and check what the package_declarator does with colonpairs
19:16 FROGGS RabidGravy: (colonpair via longname)
19:16 [ptc] wow, timotimo used "fabbergasted" in a commit message.  I'm impressed!  Haven't seen that used in long time :-)
19:16 * Ulti has a go at getting his modules to run from the GLR branch...
19:17 moritz is that a misspelling of "flabbergasted"?
19:19 rindolf joined #perl6
19:20 mprelude joined #perl6
19:20 zostay just published a blog post regarding P6SGI and the latest draft i finished this morning after having an epiphany last night, which probably should have come sooner
19:20 zostay http://blogs.perl.org/users/sterling_hanenkamp/2015/08/p6sgi-revising-and-reviewing.html
19:20 zostay for anyone interested
19:20 dalek rakudo/glr: 0f8cdfb | moritz++ | src/core/operators.pm:
19:20 dalek rakudo/glr: Simplify usage of die().
19:20 dalek rakudo/glr:
19:20 dalek rakudo/glr: When it was still fail(), it needed to be nested in a function call,
19:20 dalek rakudo/glr: because fail() exits the current function and returns the failure.
19:20 dalek rakudo/glr:
19:20 dalek rakudo/glr: die() needs no such sophistication
19:20 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/0f8cdfbbdc
19:21 moritz m: await start { say 42 }
19:21 camelia rakudo-moar 5ba44f: OUTPUT«42␤»
19:21 GLRelia rakudo-moar d47a42: OUTPUT«42␤»
19:21 moritz zostay: start { } has been deprecated? really?
19:21 moritz why is there no deprecation warning?
19:22 moritz how did I miss this?
19:22 RabidGravy it hasn't
19:22 zostay it says it has in asyncops
19:22 zostay as of 2015.06
19:23 RabidGravy no, thats some other usage
19:23 ugexe proc::async, and thats in some other branch
19:24 zostay okay, then i misread the source... i'll clear that out and adjust the spec then
19:24 RabidGravy m: sub foo {}; start(&foo)
19:24 camelia rakudo-moar 5ba44f: ( no output )
19:24 GLRelia rakudo-moar 0f8cdf: OUTPUT«Saw 1 occurrence of deprecated code.␤================================================================================␤start({ ... }) seen at:␤  /tmp/dD8_FHcXNS, line 1␤Deprecated since v2015.6, will be removed with release v2015.9!␤Please use sta…»
19:25 [ptc] timotimo, JimmyZ, hoelzro, jnthn, FROGGS: just updated the MoarVM ChangeLog and pushed a PR.  If I've missed something or got something wrong, please update.
19:26 RabidGravy so start as a function has been deprecated basically
19:26 zostay that makes me happier
19:28 hoelzro [ptc]++
19:29 nine moritz: ah, I've been wondering what the whole take sub { fail business was about
19:29 mohij joined #perl6
19:30 mj41 joined #perl6
19:33 dalek perl6-roast-data: b8c3976 | coke++ | / (2 files):
19:33 dalek perl6-roast-data: another glr run for 'today'
19:33 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/b8c3976960
19:33 larion joined #perl6
19:33 moritz nine: so, are we waiting for something in particular before merging?
19:34 [Coke] note I still have one failure, but it's probably a flapper.
19:34 nine moritz: I'd actually rather rename glr to master... But really we're just waiting for someone to say, that it's ok.
19:35 moritz nine: I've had the same thought (renmaing to mater)
19:35 [Coke] S17-promise/stress.t aborted 1 test(s)
19:35 rangerprice joined #perl6
19:35 moritz we've been wanting to switch back to a 'master' branch for a while now, but didn't have a good trigger
19:35 nine moritz: I don't think I remember anyone at the SPW being against it.
19:36 [Coke] +1 from me on going to master... but also +1 from me on merging glr back to nom first before we decide on that.
19:36 rangerprice Where i can find a good book reference to create my own interpreter/compiler for a programming language ?
19:36 hoelzro I second [Coke]'s suggestion
19:36 nine [Coke]: S17-promise/stress.t is passing here. Considering that it's testing thready features, I really wouldn't be surprised if it was a flapper.
19:36 rangerprice hi moritz :)
19:37 [Coke] nine: my thoughts exactly. it did fail here twice in a row, and I only run one test at a time.
19:37 Ulti does Positional still exist?
19:37 nine Ulti: yes
19:37 Ulti hmmm
19:37 nine Ulti: though in most cases, you can probably require Iterable instead and be more flexible
19:37 Ulti and is Seq a Positional?
19:38 nine Ulti: no
19:38 nine Seq does Iterable
19:38 nine If you get some "cannot bind to Positional" error, just use assignment instead
19:39 dalek rakudo/cu_load: 7262b98 | FROGGS++ | src/ (7 files):
19:39 dalek rakudo/cu_load: move module loading code to CompUnitRepo
19:39 dalek rakudo/cu_load:
19:39 dalek rakudo/cu_load: This means we are calling CompUnitRepo.load_module for every
19:39 dalek rakudo/cu_load: use/require/need we hit. Later we move the CU loading code to
19:39 dalek rakudo/cu_load: CompUnit.load, and will work with CompUnit objects in CUR
19:39 dalek rakudo/cu_load: instead. This patch fails one test of integration/precompiled.t
19:39 dalek rakudo/cu_load: review: https://github.com/rakudo/rakudo/commit/7262b98d3e
19:40 Ulti yeah Iterable looks like the right change here thanks nine
19:40 mj41 Can anybody try to reproduce glr vs. nom speed regression of "say $_ for 1..2_000_000;" https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L129 ? Seems like rakudo issue (3m25s instead of 0m10s). nqp-m is a bit faster.
19:40 Ulti at least the only thing being relied upon is .elems
19:41 zostay type coercion is possible when a type has a method named for the target type, right?
19:42 moritz mj41: it's probably an optimization that wasn't reinstanted for GLR yet
19:43 nine time perl6 -e 'use nqp; nqp::say($_.Str) for 1..2_000_000;' > /dev/null takes only 2.4 seconds instead
19:43 mj41 moritz: merge blocker? as perl5 -MMoose takes 0m0.356s
19:44 mj41 so pre glr 10s was slow and now it is super slow :-)
19:46 AlexDaniel joined #perl6
19:47 moritz mj41: I don't quite see the connection between a "for"/say benchmark and -MMoose
19:48 _itz_ m:  say Parcel.new.perl
19:48 camelia rakudo-moar 5ba44f: OUTPUT«()␤»
19:48 GLRelia rakudo-moar 0f8cdf: OUTPUT«5===SORRY!5=== Error while compiling /tmp/uZHw8pltfX␤Undeclared name:␤    Parcel used at line 1␤␤»
19:48 mj41 moritz: only startup time penalty :-)
19:48 telex joined #perl6
19:48 AlexDaniel Hi!
19:48 Ulti weird I am getting a SORRY! without any line number
19:48 Ulti "Method 'iterator' must be implemented by Iterable because it is required by a role"
19:48 hoelzro AlexDaniel: greetings!
19:49 _itz_ I did wonder if a clearer deprecation message for Parcel use might help .. the one from instally Test::Mock is particularly odd
19:49 mj41 moritz: https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L23 vs. https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L28
19:49 moritz Ulti: do you have any class that implements the Iterable role?
19:49 moritz Ulti: if yes, it needs updating
19:49 AlexDaniel It seems like a lot of reports were merged into https://rt.perl.org/Public/Bug/Display.html?id=61602 and so I thought, maybe this report should be merged as well https://rt.perl.org/Public/Bug/Display.html?id=125336 ?
19:49 Ulti nope or at least not that I'm aware of
19:50 _itz_ s/instally/installing
19:50 Ulti just a package with a tonne of exported subs no OO at all
19:51 AlexDaniel it looks like it is related. However, I have no idea what should these infinite ranges actually mean, so perhaps it is a separate issue
19:51 nine Iterable used to be a class. s/is Iterable/does Iterable/ can help wonders
19:52 AlexDaniel m: .say for Inf..0
19:52 GLRelia rakudo-moar 0f8cdf: ( no output )
19:52 Ulti I'm going to just assume its a List being passed in I only really had Positional because Parcel would turn up
19:52 AlexDaniel oh wow
19:52 camelia rakudo-moar 5ba44f: OUTPUT«(signal XFSZ)-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-9223372036854775808␤-922337203685…»
19:52 AlexDaniel m: .say for NaN..0
19:52 camelia rakudo-moar 5ba44f: OUTPUT«(signal XFSZ)NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN…»
19:52 GLRelia rakudo-moar 0f8cdf: OUTPUT«(timeout)NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤NaN␤N…»
19:53 zostay is there a brief syntax for performing coercion that doesn't involve a signature or assignment?
19:53 Ulti is there a nice picture of what the types look like post GLR anywhere?
19:53 Ulti so I can investigate without diving into Rakudo source
19:54 baest_ joined #perl6
19:54 AlexDaniel Ulti: http://postcircumfix.com/media/typegraph_array_rakudo.svg?2 ?
19:54 AlexDaniel Ulti: (from http://postcircumfix.com/2015-08-29-untangling-the-perl-6-hashpair-type-hierarchy)
19:54 moritz m: say Int("42").perl; # for zostay
19:54 camelia rakudo-moar 5ba44f: OUTPUT«42␤»
19:54 GLRelia rakudo-moar 0f8cdf: OUTPUT«42␤»
19:55 zostay m: say Promise(Supply.new).perl;
19:55 camelia rakudo-moar 5ba44f: OUTPUT«Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned)␤»
19:55 GLRelia rakudo-moar 0f8cdf: OUTPUT«Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Planned)␤»
19:55 zostay m: say Promise(Promise.new).perl;
19:55 camelia rakudo-moar 5ba44f: OUTPUT«Cannot find method 'Promise'␤  in block <unit> at /tmp/apKIDc5HYd:1␤␤»
19:55 GLRelia rakudo-moar 0f8cdf: OUTPUT«Cannot find method 'Promise'␤  in block <unit> at /tmp/dwC9KYs57w:1␤␤»
19:55 Ulti thanks AlexDaniel though that's kind of what I imagined
19:55 zostay that doesn't always work
19:57 moritz m: say Promise.new.Promise
19:57 camelia rakudo-moar 5ba44f: OUTPUT«Method 'Promise' not found for invocant of class 'Promise'␤  in block <unit> at /tmp/XtqRIWial_:1␤␤»
19:57 GLRelia rakudo-moar 0f8cdf: OUTPUT«Method 'Promise' not found for invocant of class 'Promise'␤  in block <unit> at /tmp/18NsrjiTqg:1␤␤»
19:57 moritz zostay: just like oridinary coercion doesn't always work
19:59 zostay well, Promise(Any) $p as a parameter always works... seems like there should be a syntax that always works without having to say $p ~~ Promise orself $p.=Promise or something similar
19:59 zostay *orself
19:59 zostay *orelse
19:59 timotimo [ptc]: these are not all things i would have put into the changelog
19:59 zostay dang
20:00 timotimo like the parse_jitgraph thing just came out of nowhere, because the tool it refers to wasn't mentioned in the changelog; and it's only for internal use anyway
20:01 timotimo and the "jit graph is now an expression tree" thing refers to brrt's excellent work that - alas - hasn't reached the master branch yet
20:03 timotimo but there's good stuff in there, too :)
20:04 [ptc] timotimo: feel free to update :-)
20:04 [ptc] timotimo: I made lots of guesses as to what I should add while doing some repo archaeology
20:04 * [Coke] enjoys his actual day off before going back to work.
20:04 [ptc] should one be spectesting nom or glr for the current release?
20:05 [ptc] t/04-nativecall/16-rt125408.t still fails for me
20:05 RabidGravy yeah
20:05 RabidGravy grr
20:06 RabidGravy it keeps getting fixed "by accident"
20:06 Ulti is there a new and wonderful way to create sequences?
20:06 RabidGravy that's why I put the test in
20:07 dalek rakudo/glr: 994fb0a | (Stefan Seifert)++ | src/core/io_operators.pm:
20:07 dalek rakudo/glr: Add generic single arg candiate of say back in
20:07 dalek rakudo/glr:
20:07 dalek rakudo/glr: Speeds up say $_ for 1..2_000_000; by about 10x
20:07 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/994fb0aa3f
20:07 nine mj41: ^^^
20:08 nine moritz: ^^^
20:08 RabidGravy :)
20:08 moritz nine++
20:08 RabidGravy wtawtaw
20:08 moritz nine: is the slurpy still called?
20:08 nine moritz: passes all spec tests
20:09 mj41 nine++ ... anybody to try another 10x ? :-)
20:09 nine yep, seems to be doing just fine
20:11 [Coke] current release has to be nom, not glr.
20:11 bronco_creek joined #perl6
20:12 bronco_creek .tell lizmat Saw Yes in concert last night.  For some reason thought of you when they played some Tormato songs. (Sorry to be far off topic.)
20:12 yoleaux bronco_creek: I'll pass your message to lizmat.
20:13 mj41 0m16s  .. much better
20:13 diana_olhovik_ joined #perl6
20:14 RabidGravy m4lj, 94% of that time  is in the "say"
20:16 RabidGravy that is that loop calling an empty function is ~15x faster
20:16 moritz RabidGravy: with or without nine's last patch?
20:18 RabidGravy without
20:18 RabidGravy let me check again
20:18 mj41 nqp is 4x faster https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L38 now
20:19 colomon nine++
20:21 llfourn joined #perl6
20:22 RabidGravy moritz, the last was against nom
20:23 RabidGravy nom is twice as fast for that as glr with nine's patch
20:24 cschwenz joined #perl6
20:24 RabidGravy calling an empty function in that loop is 23x faster on glr with nine's patch
20:24 RabidGravy so it's say still slo(er)
20:24 RabidGravy slow
20:25 nine I think we're still missing the for -> while loop optimization
20:25 mj41 nope https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L38
20:25 mj41 line 53 line 58
20:25 mj41 https://gist.github.com/mj41/6761c021135ae1cb17f9#file-gistfile1-txt-L137
20:27 nine Please note that the version in glr will probably produce nicer output in threaded programs that just say without any locking
20:30 RabidGravy also 'say join "\n", 1..2_000_000' is 4x faster than the loop version (there's a moral in there)
20:31 moritz well yes, "do less"
20:31 RabidGravy aye
20:31 mj41 Ok. Good too know before I go to bed. Tiaveling to Granada.
20:32 mj41 err, Tired traveling to Granada.
20:32 mj41 good night
20:32 RabidGravy have fun
20:35 dalek rakudo/glr: 57b7ebc | (Stefan Seifert)++ | src/core/io_operators.pm:
20:35 dalek rakudo/glr: Add print multi candidates for single arguments
20:35 dalek rakudo/glr:
20:35 dalek rakudo/glr: Have an optimized version of print to gain the same speed advantages
20:35 dalek rakudo/glr: as say. For print we need to have separate candidates for Iterables
20:35 dalek rakudo/glr: and the rest because @array.Str would include spaces we don't want
20:35 dalek rakudo/glr: with print. Unless @array is itemized of course (GLRR).
20:35 dalek rakudo/glr: review: https://github.com/rakudo/rakudo/commit/57b7ebc586
20:38 dalek rakudo-star-daily: 79609fe | coke++ | log/ (9 files):
20:38 dalek rakudo-star-daily: today (automated commit)
20:38 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/79609fe4b1
20:40 [Coke] GLR R?
20:42 [ptc] I'm getting "MVMArray: Can't shift from an empty array" in t/spec/S17-promise/start.rakudo.moar  in the spectests on nom.  Dunno if that's useful to anyone...
20:43 rurban joined #perl6
20:47 lizmat joined #perl6
20:49 nine [Coke]: GLR Revised ;) Means that we don't flatten a single arg if it's itemized
20:52 [Coke] only if it's a release blocker, I guess.
20:53 [Coke] (to ptc)
20:53 _itz_ m: say 'GLR'~ ('R' x 1..Inf)
20:53 camelia rakudo-moar 5ba44f: OUTPUT«GLRR S T U ...␤»
20:53 GLRelia rakudo-moar 57b7eb: OUTPUT«GLR...␤»
20:55 TimToady useless use of x 1 ?
20:56 TimToady m: say 'GLR' ~ ('R' .. Inf)
20:56 camelia rakudo-moar 5ba44f: OUTPUT«GLRR S T U ...␤»
20:56 GLRelia rakudo-moar 57b7eb: OUTPUT«GLR...␤»
20:56 _itz_ oh yes
20:56 TimToady m: say 'GLR' ~ ('R' Xx (1 .. Inf))
20:57 camelia rakudo-moar 5ba44f: OUTPUT«(timeout)»
20:57 GLRelia rakudo-moar 57b7eb: OUTPUT«(timeout)»
20:57 TimToady ~ forces eager, I guess
20:58 [Coke] we've got a bunch of tests marked "submix behavior still under discussion" that are passing in glr. ok to unfudge?
20:59 [Coke] seems like testing that current functionality stays working is good.
21:00 moritz +1
21:00 lizmat [Coke] +1
21:00 yoleaux 20:12Z <bronco_creek> lizmat: Saw Yes in concert last night.  For some reason thought of you when they played some Tormato songs. (Sorry to be far off topic.)
21:00 lizmat bronco_creek: I'm jealous
21:00 [Coke] ah... 15 tests in a row; every other test passes.
21:00 [Coke] (previously all 15 were skipped)
21:00 [Coke] That is slightly more dubious.
21:01 [Coke] (going for it anyway.)
21:01 lizmat bronco_creek: On the other hand, I will be able to see Yes in the Cruise To The Edge in November  :-)
21:01 lizmat and yes, "Tormato" is one of my favourite Yes albums  :-)
21:02 lizmat alas, I will not be able to see the Rickenbacker Master in action  :-(
21:04 bronco_creek Lizmat: Vary sad to loose Chris. Sherwood filled in remarkably well on both bass and vocal IMO.
21:04 lizmat looking forward to that  :)
21:04 lizmat Yes will go on, as Chris wanted it  :-)
21:05 _itz_ . o O ( they should get Trevor Horn back )
21:05 lizmat bronco_creek: did Sherwood do The Fish ?
21:06 bronco_creek No. Only bass solo was on the encore rendition of Starship Trooper.
21:07 * virtualsue watches _itz_ 's troll go unanswered
21:07 bronco_creek Also,  Jon D. is excellent.  Close your eyes and it could be Jon A.  Open them, and the same holds true...
21:11 lizmat can't wait for november: http://cruisetotheedge.com
21:12 bronco_creek _itz_:  ;-)  Yes line-up permutations should be a rosetta code problem.
21:14 _itz_ using the  Pete::Frame::Tree module
21:20 [Coke] 1,031 tickets
21:21 llfourn joined #perl6
21:22 jdv79 i don't think i've ever done a 8h layover.  should have checked for more alternative routes maybe:(
21:24 timotimo so ... what do i do about the test for #125408 staying fixed?
21:24 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125408
21:24 timotimo just release anyway?
21:25 [Coke] jdv79: I had a 9 hour layover scheduled on the way back but managed to rebook .
21:26 dalek roast/glr: 3d283da | TimToady++ | S06-signature/unpack-array.t:
21:26 dalek roast/glr: only needs .List, not .list, doesn't cache
21:26 dalek roast/glr: review: https://github.com/perl6/roast/commit/3d283dafd8
21:28 [ptc] 'night everyone!
21:28 timotimo gnite [ptc]!
21:28 ggoebel joined #perl6
21:30 lizmat gnight [ptc]
21:30 * lizmat also gets some sleep&
21:32 timotimo * [new tag]         2015.08 -> 2015.08
21:32 hoelzro night [ptc], lizmat
21:32 timotimo maybe i should have waited for the spec test to run through, though?
21:36 dalek roast/glr: cdab420 | coke++ | S0 (10 files):
21:36 dalek roast/glr: (auto) unfudge several tests.
21:36 dalek roast/glr:
21:36 dalek roast/glr: ... and close some RTs.
21:36 dalek roast/glr: review: https://github.com/perl6/roast/commit/cdab4203a1
21:36 dalek roast/glr: 97b6b02 | coke++ | S05-capture/match-object.t:
21:36 dalek roast/glr: Comparing types doesn't need .gist
21:36 dalek roast/glr:
21:36 dalek roast/glr: Update comment for RT #124998
21:36 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=124998
21:36 dalek roast/glr: review: https://github.com/perl6/roast/commit/97b6b027e2
21:38 colomon t/spec/S17-procasync/many-processes-no-close-stdin.t        (Wstat: 65280 Tests: 0 Failed: 0)
21:38 timotimo http://t.h8.lv/MoarVM-2015.08.tar.gz
21:38 colomon Non-zero exit status: 255
21:38 colomon Parse errors: Bad plan.  You planned 1 tests but ran 0.
21:38 Ulti well the Stats module is now GLR ready :3
21:38 timotimo .tell jnthn when you're back, could you upload http://t.h8.lv/MoarVM-2015.08.tar.gz to moarvm.org?
21:38 yoleaux timotimo: I'll pass your message to jnthn.
21:40 hoelzro colomon: ='(
21:40 timotimo moritz: we have a moarvm release, at least in theory
21:40 Ulti Bag.kxxv.flat.list to get back to a list you might call bag on is a bit weird though
21:40 hoelzro colomon: is that on GLR? have you pulled recently?
21:40 colomon hoelzro: I pulled earlier this afternoon
21:41 colomon I know I didn’t get cdab420 above
21:41 hoelzro colomon: if you run that test with just ./perl6-m, do you get any output?
21:41 colomon might be a flapper or something
21:41 hoelzro (it will probably throw up a bunch of output about /tmp/test-file missing)
21:41 hoelzro it shouldn't flap
21:46 [Coke] someone opened a twitter account with coke@cpan.org
21:46 japhb .tell bronco_creek Years ago I came across a Santana album in which the liner notes consisted of a complex infographic of the changing band lineup, just so fans could get it straight.
21:46 yoleaux japhb: I'll pass your message to bronco_creek.
21:46 * [Coke] just stole it back. weird.
21:52 mrocks77 joined #perl6
21:53 colomon hoelzro:
21:53 colomon cat: /tmp/test-file: No such file or directory
21:53 colomon too many open files
21:53 colomon in block <unit> at t/spec/S17-procasync/many-processes-no-close-stdin.t:7
21:53 hoelzro =(
21:53 hoelzro colomon: what do moar/nqp-m/perl6 --version all output?
21:53 hoelzro I just bumped the NQP_REVISION file to use the fix this morning
21:54 colomon hoelzro: oh, I’m probably missing that.
21:54 hoelzro if you've pulled but not re-Configure.pl'd, that would explain it
21:54 colomon give me a minute
21:55 masak BooK: the supply/whenever stuff only works in the glr branch, IIUC.
21:55 [Sno] joined #perl6
21:55 masak oh, someone already told you. nvm.
21:57 colomon hoelzro: right, that’s probably it.  In middle of full rebuild at the moment.
21:57 hoelzro whew
22:00 masak 'night, #perl6
22:01 colomon hoelzro++ # works after more complete rebuild
22:01 hoelzro \o/
22:01 hoelzro night masak
22:01 colomon o/ masak
22:06 colomon with complete update, I get a completely clean spectest from GLR  \o/
22:06 * awwaiid hops up and down
22:08 colomon bad news (for the moment) is that Audio::Taglib::Simple won’t install on GLR.  guess I need to switch back to nom for a minute.
22:09 _itz_ there is some ecosystem breakage as would be expected
22:09 bbkr joined #perl6
22:11 skids joined #perl6
22:15 colomon _itz_: of course.  I’m hoping to get a decent measure of how much breakage and where later today or tomorrow.
22:17 _itz_ some of the errors look a little odd
22:18 _itz_ the text of the messages I mean
22:19 [TuxCM] joined #perl6
22:23 TEttinger joined #perl6
22:41 RabidGravy how soon is the glr going to become the mainstream? I've been holding off testing my modules on the grounds I couldn't be arsed to do it multiple times.
22:42 BooK masak: yes, I've been told that and the examples from jnthn talk worked better on it when I tried :-)
22:42 muraiki joined #perl6
22:43 BooK mmm, ok so I was reading my backlog too :-)
22:45 rangerprice joined #perl6
22:45 RabidGravy of course until the rt#125408 is fixed I'm not sure what the point would be
22:45 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125408
22:56 muraiki m: say chrs(0x1D047, 0x1D0F0)
22:56 camelia rakudo-moar 5ba44f: OUTPUT«𝁇𝃰␤»
22:56 GLRelia rakudo-moar 57b7eb: OUTPUT«𝁇𝃰␤»
22:56 muraiki hrm. I don't suppose anyone can recommend a decent intro to unicode? I thought that 0x1D0F0 was a combining character, but it seems like it's not behaving that way
23:07 muraiki darn, maybe it's not a combining character. :(
23:10 geekosaur .u U+1D0F0
23:10 yoleaux U+1D0F0 BYZANTINE MUSICAL SYMBOL KENTIMATA NEO MESO [So] (𝃰)
23:10 muraiki yeah, I think I can't use this the way I hoped to
23:10 muraiki oh well
23:11 aborazmeh joined #perl6
23:11 aborazmeh joined #perl6
23:11 muraiki I guess it was too much to expect unicode to not only have byzantine notation characters but also to work as a layout engine for byzantine chant :)
23:17 japhb Given the machine-native guts of https://github.com/rakudo/rakudo/commit/994fb0aa3f , why doesn't IO::Handle.print have a str multi to avoid the box and unbox of $str in the call to $out.print($str)?
23:19 japhb I would think that might actually chew up a fair amount of GC and performance during heavy output, especially when you consider that multi sub print() has the same problem.
23:21 cognominal joined #perl6
23:22 llfourn joined #perl6
23:23 TEttinger muraiki: I'm curious what kind of project actually uses the really rare parts of unicode
23:24 TEttinger (I guess they were added because they thought they would be used, but Byzantine is starting to describe some parts of implementing unicode it seems, heh)
23:27 muraiki TEttinger: hehe. I actually sing byzantine chant using byzantine notation. I'd like to write a FLOSS editor for it. it's actually the project that led me to switching careers to become a programmer
23:28 TEttinger wow!
23:28 muraiki but now that I'm a professional programmer, I haven't actually done much work on that project. seeing additional byzantine chant code points added in unicode 8 got my hopes up
23:28 muraiki at the very least I now have a way to encode neumes using an international standard
23:28 muraiki instead of "some weird font, character 7"
23:29 muraiki I wrote the initial version in racket, which turned a type of byzantine chant lisp into graphical output
23:29 TEttinger nice
23:29 muraiki there's a picture here https://github.com/muraiki/byzscribe
23:33 TEttinger I mean I have no idea if it's right or not, but that does look like a fair amount of work to input each character. hopefully perl6 can be easier to write this with?
23:33 muraiki yeah, I was hoping to use p6's great unicode support and grammars
23:34 muraiki and now I would be able to store the actual characters with some additional info, instead of long english names
23:35 muraiki I suppose I will have to pick up The Haskell School of Music again... it goes through learning music theory + haskell by developing a way to both encode music and render it. I've never done any kind of real programming that had to deal with laying out complex things graphically
23:48 aborazmeh joined #perl6
23:48 aborazmeh joined #perl6
23:59 thou joined #perl6

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

Perl 6 | Reference Documentation | Rakudo