Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-03-04

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 masak lue: I'm inclined to agree.
00:00 masak grammar inheritance notwithstanding.
00:00 masak extending rather than modifying will always be the "safe" strategy.
00:01 lue The only time you'd really want to change an existing grammar is to add a feature to it, e.g. to hook into a new slang, and macros could accomplish that perhaps.
00:02 masak lue: I can see people maybe getting away with adding a twigil, or something similar.
00:02 masak but perhaps protoregexes will make that feasible.
00:05 * lue still thinks the syntax inside the braces of  slang { }  needs to be something that pulls together grammar, actions, and other possible related things, instead of precisely the same things as go in  grammar { }
00:06 tgt joined #perl6
00:08 lue Still a lot of thinking that needs to be done on slangs, just thought I'd let you know what I see slangs useful for nowadays :) .
00:10 masak lue: thanks.
00:10 masak I agree we're not quite there yet.
00:10 masak as usual in these situations, I think things will only start to happen once the rubber meets the road -- once someone tries something, and gets feedback from users and from seeing it work in practice.
00:11 lue I'm not surprised it's not even well-spec'd, let alone implemented; who'd be insane enough to allow the user to change the language they're writing in? ;)
00:11 masak aka "rough consensus and working code"
00:12 lue I think there's still some work to do to get something that can be implemented in the first place, but certainly most of the ironing out will be in usage, not speculation :)
00:15 masak agree that we must be fairly certain what we want, and we aren't yet :)
00:16 masak sometimes I wish I could associate an actions class with a grammar, slang or no.
00:17 masak it's very rare that I do something other than 1-grammar-1-actions-class.
00:17 Mouq "chr codepoint cannot be negative" :?
00:17 lue masak: what, isn't :actions enough an association for you? :P
00:18 lue Mouq: check for an E<> somewhere near where the backtrace tells you in the offending Pod6 file.
00:18 masak lue: no, because that association has to be done per-parse-call. I'd like to be able to make the association once, from the grammar.
00:19 Mouq lue: Re-ran, came up with same bug
00:19 lue masak: one of my ideas is that slang could perhaps be a general grammar/action associator, with the special language-level bit being stuffing that slang-thing in a $~THING and using that variable somewhere in the Perl6 grammar/actions/etc.
00:20 * Mouq tries https://gist.github.com/Mouq/82ebe2
00:20 masak lue: mhm.
00:20 * Mouq realizes how much that diff will annoy lue
00:20 lue Mouq: https://gist.github.com/Mouq/82ebe2 is a 404 :/
00:20 skids masak: just overload .parse.
00:20 Mouq lue: https://gist.github.com/Mouq/82ebe2b9633c487d3795
00:22 lue I still like my way for how it follows the POD generator's output. :)
00:22 lue Also, my way uses high-number PUA Unicode chars, so there :P
00:23 fridim__ joined #perl6
00:24 Mouq lue: :) your way definitely wins on coolness (and probably actually-working-ness)
00:25 lue Perhaps I can stick the HTML I want in one giant V<> at some point :) [Unless Pod::To::HTML tries to escape the angle brackets and such...]
00:27 * Mouq realizes the problem is probably something as simple as spaces around the <p>
00:27 lue (fwiw, I originally wanted to use the C0 and/or C1 control chars (because they look so disused, at least in the stuff I find myself doing), but none of them have the precisely right meanings ☺ (closest is SOH/STX/ETX) )
00:28 * Mouq notes that his diff did solve the problem though :)
00:28 lue Have yet to try latest everything on syngen, by the way.
00:29 masak skids: ooh
00:29 Mouq (And you don't have to wait for all the HTML to be output to see if the issue was fixed w/ the removal of the Post-gen stage
00:30 Mouq :) )
00:30 lue Mouq: yeah, though I've been considering turning the pipeline into a concurrent sort of thing to take care of that "waiting on" issue :)
00:31 masak r: class Actions { method TOP { say "OH HAI" } }; grammar G { regex TOP { x }; method parse($input) { nextsame($input, :actions(Actions.new)) } }; G.parse("x")
00:31 camelia rakudo-jvm c8ef8d: OUTPUT«(timeout)»
00:31 camelia ..rakudo-parrot c8ef8d: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤  in block  at gen/parrot/CORE.setting:653␤  in method parse at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
00:31 camelia ..rakudo-moar c8ef8d: OUTPUT«Too many positional parameters passed; got 1 but expected 0␤  in block  at src/gen/m-CORE.setting:652␤  in method parse at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
00:31 masak r: class Actions { method TOP { say "OH HAI" } }; grammar G { regex TOP { x }; method parse($input) { nextwith($input, :actions(Actions.new)) } }; G.parse("x")
00:31 camelia rakudo-parrot c8ef8d: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤  in method TOP at /tmp/tmpfile:1␤  in any !reduce at gen/parrot/stage2/QRegex.nqp:752␤  in any !cursor_pass at gen/parrot/stage2/QRegex.nqp:714␤  in regex TOP at /tmp/tmpfile:1␤  in m…»
00:31 camelia ..rakudo-moar c8ef8d: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤  in method TOP at /tmp/tmpfile:1␤  in any !reduce at gen/moar/stage2/QRegex.nqp:752␤  in any !cursor_pass at gen/moar/stage2/QRegex.nqp:714␤  in regex TOP at /tmp/tmpfile:1␤  in method …»
00:31 camelia ..rakudo-jvm c8ef8d: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤  in method TOP at /tmp/tmpfile:1␤  in any !reduce at gen/jvm/stage2/QRegex.nqp:752␤  in any !cursor_pass at gen/jvm/stage2/QRegex.nqp:714␤  in regex TOP at /tmp/tmpfile:1␤  in method par…»
00:31 masak oh, right.
00:31 masak r: class Actions { method TOP($/) { say "OH HAI" } }; grammar G { regex TOP { x }; method parse($input) { nextwith($input, :actions(Actions.new)) } }; G.parse("x")
00:31 camelia rakudo-parrot c8ef8d, rakudo-jvm c8ef8d, rakudo-moar c8ef8d: OUTPUT«OH HAI␤»
00:31 masak \o/
00:31 masak and with that...
00:32 masak ...'night, #perl6
00:32 lue ♞ masak o/
00:32 skids masak: I do agree that that will be common enough that grammar might best be a parameterized role that takes an actions to use as a default.
00:32 skids Oh goodnight.
00:33 Mouq o/ masak
00:33 * Mouq ought to bed soon as well, getting close to the 36 hour mark of being awake
00:33 lue Heh, if only it wasn't complex enough to require ncurses, I'd have conc-syngen display a "leaderboard" (or whatever it's best called) of all the synopses with progress indicators for each one :P
00:35 lue Mouq: yeesh, you should probably go to bed real soon.
00:41 lue Aw, I get a segfault trying to run ./bootstrap.pl on perl6-m :(
00:44 Mouq lue: Are you sure you don't want rebootstrap.pl?
00:44 lue positive. "No need to rebootstrap, running normal bootstrap"
00:45 Mouq :(
00:58 xenoterracide joined #perl6
00:59 lue Looks like the bootstrap script itself is what segfaults.
01:05 lue That explains it. the script is trying to `mkdir` /usr/local/languages/perl6/site , which it is not allowed to do (nor do I want it to do that, I want ~/.perl6/)
01:07 tgt joined #perl6
01:07 lue .tell tadzik to see http://irclog.perlgeek.de/perl6/2014-03-04#i_8379034  (I get a segfault when perl6-m tries mkdir'ing a path it hasn't the permissions for. Should be creating ~/.perl6 instead anyway)
01:07 yoleaux lue: I'll pass your message to tadzik.
01:07 * lue &
01:12 kbaker_ joined #perl6
01:24 xinming joined #perl6
01:37 * Mouq generates a Perl 6 hash of all lowercase HTML5 entities and the codepoint(s) they represent. Now. Where to put this.
01:38 Mouq I can't imagine this is useful anywhere but in Pod parsing
01:39 BenGoldberg Not necessarily -- I can imagine it being used to convert html to text
01:39 Mouq BenGoldberg: But that would use a full list.
01:39 BenGoldberg Oh, well nevermind then ;)
01:39 Mouq :)
02:05 Mouq m: say "\c[8842,65024]"
02:05 camelia rakudo-moar c8ef8d: OUTPUT«⊊︀␤»
02:07 tgt joined #perl6
02:23 xenoterracide joined #perl6
02:29 Sqirrel joined #perl6
02:31 klapperl joined #perl6
02:32 itz_ joined #perl6
02:41 alc joined #perl6
02:41 itz joined #perl6
02:44 __thou joined #perl6
02:48 regreg joined #perl6
02:56 hoverboard joined #perl6
03:04 [Coke] should I have dalek report on mojo6 here? (slow going at the moment)
03:06 japhb__ [Coke]: Definitely
03:08 tgt joined #perl6
03:12 xinming joined #perl6
03:14 japhb__ Did jnthn++ mention when he would be 6ing again?
03:33 xenoterracide joined #perl6
04:09 tgt joined #perl6
04:15 moritz [Coke]: for Mojo6, do you want to implement Mojolicious-style setter methods?
04:21 lue .tell tadzik looks like it's moar's fault, not panda's. Try  mkdir /test  on all three backends.
04:21 yoleaux lue: I'll pass your message to tadzik.
04:23 cognominal joined #perl6
04:23 xenoterracide joined #perl6
04:28 molaf joined #perl6
04:33 dalek nqp: fb32da5 | larry++ | / (2 files):
04:33 dalek nqp: hook up new uniprop opcodes
04:33 dalek nqp: review: https://github.com/perl6/nqp/commit/fb32da5ea0
04:44 hoverboard joined #perl6
04:55 TimToady so, I'm wondering if there's a magical way to bump revisions, or if one just does that by hand
04:59 lue TimToady: I've not heard of some bumping tool, I believe it's all manual.
05:06 TimToady and in a number like
05:07 TimToady 2014.02-8-geb722b4, what is the significance of the 8?
05:07 lue I have no clue, unfortunately :(
05:08 TimToady the previous rev had a 5 there
05:08 TimToady but it's not the day of month
05:08 TimToady maybe it just has to increase
05:10 tgt joined #perl6
05:11 lue TimToady:  perl tools/build/gen-version.pl  in the top dir of rakudo might be helpful.
05:12 TimToady looking and NQP::Configure.pm it's just a sequence number
05:12 TimToady s/and/at/
05:13 __thou joined #perl6
05:20 dalek nqp: a1e5033 | larry++ | tools/build/MOAR_REVISION:
05:20 dalek nqp: bump moar revision
05:20 dalek nqp: review: https://github.com/perl6/nqp/commit/a1e503377e
05:21 dalek rakudo/nom: 62996cb | larry++ | tools/build/NQP_REVISION:
05:21 dalek rakudo/nom: bump nqp revision
05:21 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/62996cb3d2
05:21 * TimToady hopes he done it rihgt
05:23 TimToady hopefully we can now start playing with uniprops in nom, and do things like return the Rat corresponding to the vulgar fractions
05:31 lue TimToady: currently S15 has a uniprops function that returns a hash of all the properties a given character. I'm wondering if replacing that with a uniprop($char, $prop) type thing would be better.
05:32 lue (would have the advantage that the name &uniprops can be for the function that gives you a property for every character, akin to ord/ords and chr/chrs. If the hash is still needed, perhaps &unihash or something)
05:33 dalek rakudo/nom: b4336af | larry++ | tools/build/NQP_REVISION:
05:33 dalek rakudo/nom: oops, need to bump the nqp that bumps moar
05:33 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b4336af9d9
05:43 TimToady there is no point in returning all the properties; any given program would only be interested in a few properties
05:44 TimToady the tendency is to look at the same set of props for a lot of character rather than a lot of props for a single char
05:45 TimToady I think uniprop(Int $codepoint, Str $propname) should probably be the basic function, and it should cache whatever it needs to access that particular property for other chars too
05:46 TimToady I'm thinking that it should return an Int, a Str, or a Rat depending on the nature of the property
05:46 TimToady alternately, we can split into functions that return each type
05:46 japhb__ TimToady: re: http://irclog.perlgeek.de/perl6/2014-03-04#i_8379668 , the format of these is (most recent tag in commit path)-(revs since the tag)-g(7 leading digits of SHA-1 for this commit)
05:46 japhb__ (Optionally with -dirty on the end, if you've asked for it, and your working directory is not fully committed)
05:46 TimToady hmm
05:47 lue TimToady: yeah, I'm realizing now "return all the properties" provides no benefits, and only detriments. I'm wondering if uniprop($chr, *@props) would be better or part of a multi.
05:47 TimToady we can have a multi uniprop that takes a string instead of a codepoint, but the basic interface is integer based
05:47 lue I can go for that.
05:48 TimToady so we should provide access to the integer API for when that's the natural input
05:48 lue TimToady: the hash-based function as specced has this to say about its values: "Values for properties may be C<Bool> for binary ("Yes"/"No") values, a C<Rat>
05:48 lue for numeric values, and C<Str> objects for all other types of values."
05:49 lue .oO(The *one* place where polymorphic (return) types would be cool to have, though not necessary.)
05:49 TimToady well, a lot of them are Int rather than Rat
05:49 lue Yeah, I just went with Rat to reduce the number of possible types :)
05:49 TimToady but adding in Bool is...possible..
05:49 TimToady but we don't know which ones are bools in the current API
05:50 TimToady well, I think it should .narrow :)
05:51 lue TimToady: this is what I used to discern possible return types: http://www.unicode.org/reports/tr44/tr44-12.html#About_Property_Table
05:51 lue C,E,S,M --> Str, N --> Rat (etc.), B --> Bool
05:53 lue (The Enumeration values could theoretically be enums, but that would be pushing it I feel.)
05:54 renormalist joined #perl6
05:54 TimToady except to calculate N, you need to be numerator and denominator, which are coming out as string enums right now, because of how moar compresses enumerations and things like them
05:55 lue could the returns be subjected to val() processing perhaps?
05:55 lue the N values that is
05:55 TimToady well, we'll construct the right type somehow
05:55 TimToady $num / $denom will work even if those are strings :)
05:57 xinming_ joined #perl6
05:57 TimToady r: say ("5" / "1").WHAT
05:57 camelia rakudo-parrot c8ef8d, rakudo-jvm c8ef8d, rakudo-moar c8ef8d: OUTPUT«(Rat)␤»
05:57 TimToady r: say ("5" / "1").narrow.WHAT
05:57 camelia rakudo-parrot c8ef8d, rakudo-jvm c8ef8d, rakudo-moar c8ef8d: OUTPUT«(Int)␤»
06:10 tgt joined #perl6
06:20 dalek nqp: d6a2776 | larry++ | tools/build/MOAR_REVISION:
06:20 dalek nqp: oops, didn't understand format
06:20 dalek nqp: review: https://github.com/perl6/nqp/commit/d6a2776b85
06:23 dalek rakudo/nom: 82ea648 | larry++ | tools/build/NQP_REVISION:
06:23 dalek rakudo/nom: fix nqp revision, hopefull
06:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/82ea648b2d
06:25 TimToady okay, that's configuring better now
06:35 TimToady well, up to a point
06:35 dalek specs: 33dad1e | lue++ | S15-unicode.pod:
06:35 dalek specs: [S15] Make uniprop/uniprops be more useful.
06:35 dalek specs:
06:35 dalek specs: Instead of a hash of ALL THE PROPERTIES, it's a much more useful
06:35 dalek specs: single-property lookup.
06:35 dalek specs: review: https://github.com/perl6/specs/commit/33dad1e20b
06:38 dalek rakudo/nom: 2582919 | larry++ | tools/build/NQP_REVISION:
06:38 dalek rakudo/nom: If at first you don't succeed...
06:38 dalek rakudo/nom:
06:38 dalek rakudo/nom: ...suck until you do succeed.
06:38 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2582919ecd
06:38 hoverboard xkcd hit 1337 today! a momentous occasion!
06:39 lue .oO(Next order of business for S15: figure out what Unicodey and Stringy precisely do. But not tonight, at least not from me.)
06:42 TimToady japhb++ for explaining the version numbers...now if I only understood how to count revs, or where it came up with that '9'
06:42 lue #perl6 o/
06:42 TimToady o/
06:50 japhb__ TimToady: It's the output of `git describe`
06:51 TimToady japhb__++
06:51 TimToady cool, that makes it a lot easier :)
06:51 japhb__ :-)
06:51 TimToady someone should write a bump script...
06:52 [Sno] joined #perl6
07:05 dalek nqp: e79a521 | larry++ | tools/build/MOAR_REVISION:
07:05 dalek nqp: bump moar rev to get mkdir_p fix for lue++
07:05 dalek nqp: review: https://github.com/perl6/nqp/commit/e79a521a6c
07:06 dalek rakudo/nom: fc2a6e6 | larry++ | tools/build/NQP_REVISION:
07:06 dalek rakudo/nom: bump nqp to get mkdir_p fix
07:06 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/fc2a6e632e
07:09 TimToady one of these years I might be semi-competent
07:11 tgt joined #perl6
07:12 TimToady obviously mkdir is not sufficiently tested--the problem with letting Camelia chase all the mean people away is that the meanest poeple write the meanest tests...
07:13 xfix joined #perl6
07:14 TimToady of course, #?rakudo skip "mkdir NYI" doesn't help a whole lot here
07:16 berekuk joined #perl6
07:26 xinming joined #perl6
07:34 kaleem joined #perl6
07:37 dmol joined #perl6
07:41 berekuk joined #perl6
07:42 FROGGS joined #perl6
07:42 darutoko joined #perl6
07:54 virtualsue joined #perl6
08:04 TimToady m: say nqp::getuniname(0x1f4a9)
08:04 camelia rakudo-moar fc2a6e: OUTPUT«PILE OF POO␤»
08:04 TimToady \o/
08:07 tadzik TimToady++ # |o|
08:07 yoleaux 01:07Z <lue> tadzik: to see http://irclog.perlgeek.de/perl6/2014-03-04#i_8379034  (I get a segfault when perl6-m tries mkdir'ing a path it hasn't the permissions for. Should be creating ~/.perl6 instead anyway)
08:07 yoleaux 04:21Z <lue> tadzik: looks like it's moar's fault, not panda's. Try  mkdir /test  on all three backends.
08:07 tadzik oh oh
08:07 FROGGS TimToady++ # \o/
08:08 tadzik indeed, mkdir segfaults :)
08:08 FROGGS tadzik: please bugreport is (if there is no ticket yet), and I'll care about it if nobody beats me to it
08:09 FROGGS .u PILE OF POO
08:09 yoleaux U+1F4A9 PILE OF POO [So] (????)
08:10 tadzik lue: ah, it seems like the way bootstrap.pl figures out if a directory path is feasible is "try mkdir $that"
08:10 zakharyas joined #perl6
08:10 tadzik so, if it didn't segfault it would go to ~/.perl6 anyway :)
08:10 FROGGS TimToady: I'd like to see a CharProperty class (or so) that gistifies to yoleaux output here
08:10 tadzik I seem to recall writing it as a check for writability (is that a word?), not sure why it's a "try" these days
08:10 FROGGS m: say "pups".w
08:10 camelia rakudo-moar fc2a6e: OUTPUT«No such method 'w' for invocant of type 'Str'␤  in block  at /tmp/nsIyaZHjI1:1␤␤»
08:11 FROGGS m: say "pups".IO.w
08:11 camelia rakudo-moar fc2a6e: OUTPUT«False␤»
08:11 tadzik m: say "/bbbaaa".IO.w
08:11 camelia rakudo-moar fc2a6e: OUTPUT«False␤»
08:11 FROGGS this is supposed to segfault too, no?
08:11 tadzik not to my knowledge
08:11 FROGGS ahh, it is not mkdir
08:11 tadzik but why is it always false?
08:11 tadzik maybe that's why panda uses "try mkdir" :)
08:11 FROGGS m: say "/tmp".IO.w
08:11 camelia rakudo-moar fc2a6e: OUTPUT«True␤»
08:11 tadzik >:(
08:12 FROGGS tadzik: an unknown dir is never writeable :o)
08:12 tgt joined #perl6
08:12 TimToady I already fixed the mkdir thing
08:12 tadzik oh, awesome
08:12 tadzik TimToady++
08:12 FROGGS heeh
08:12 FROGGS TimToady++
08:13 tadzik m: mkdir "/aaaa"
08:13 FROGGS glad to have you on our side now :o)
08:13 TimToady it was just trying to format errno using %s
08:13 camelia rakudo-moar fc2a6e: OUTPUT«mkdir is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting:2␤  in sub mkdir at src/RESTRICTED.setting:10␤  in block  at /tmp/Zu95yW9RJP:1␤␤»
08:13 tadzik bleh
08:13 tadzik hah
08:13 FROGGS >.<
08:13 tadzik I can see why that'd fail :)
08:13 FROGGS yeah
08:20 dalek rakudo/nom: 5f45799 | larry++ | src/core/Str.pm:
08:20 dalek rakudo/nom: implement rudimentary uniname, uniprop, and unival
08:20 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5f45799778
08:21 TimToady > ./perl6 -e 'say unival "⅖"'
08:21 TimToady 0.4
08:22 * timotimo didn't realize that was a thing
08:24 moritz $de eq '1' ?? $nu.Int !! $nu / $de;
08:24 moritz missed the opportunity to write  ($nu/$de).narrow # :-)
08:24 moritz m: say (5/1).narrow.WHAT
08:24 camelia rakudo-moar fc2a6e: OUTPUT«(Int)␤»
08:24 moritz though TimToady++'s code is probably more efficient
08:25 TimToady > ./perl6 -e 'say unival "\c[CUNEIFORM NUMERIC SIGN SHAR2 TIMES GAL PLUS MIN]"'
08:25 TimToady 432000
08:26 zakalwe joined #perl6
08:29 TimToady this will make it really easy to write a roman numeral translator :)
08:30 tgt joined #perl6
08:31 TimToady except, of course, for getting the user to actually input roman numerals as unicode thinks of them... :)
08:32 sergot hi o/
08:35 mathw Let me guess, Unicode has special characters for roman numerals that nobody knows about
08:36 TimToady nope, I think someone knows about 'em
08:36 moritz $ uni ROMAN|grep NUMERAL|wc -l
08:36 moritz 40
08:37 pdcawley joined #perl6
08:39 TimToady some of them are certainly unfamiliar though
08:39 TimToady > ./perl6 -e 'say unival("ↈ")'
08:39 TimToady 100000
08:39 TimToady .u ↈ
08:39 yoleaux U+2188 ROMAN NUMERAL ONE HUNDRED THOUSAND [Nl] (ↈ)
08:41 masak morning, #perl6
08:42 TimToady o/
08:42 TimToady it's morning here too :)
08:45 moritz \o masak, *
08:45 masak japhb__: I think you'll see jnthn++ back here sometime tomorrow.
09:05 rindolf joined #perl6
09:28 kbaker__ joined #perl6
09:33 masak r: say unival "5"
09:33 camelia rakudo-parrot fc2a6e, rakudo-jvm fc2a6e, rakudo-moar fc2a6e: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Undeclared routine:â�¤    unival used at line 1â�¤â�¤Â»
09:33 masak meh.
09:36 moritz one commit behind.
09:36 dakkar joined #perl6
09:37 masak *nod*
09:38 moritz a small task for improving camelia: summarize 'rakudo-parrot fc2a6e, rakudo-jvm fc2a6e, rakudo-moar fc2a6e:' as 'rakudo-{parrot,jvm,moar} fc2a6e:'
09:46 masak +1
09:52 moritz timotimo: ^^ you could add that to your next "how to help" section
10:04 spider-mario joined #perl6
10:05 zakharyas joined #perl6
10:13 zakharyas joined #perl6
10:14 darutoko- joined #perl6
10:18 zakharyas joined #perl6
10:19 masak (OT) what's the name of the CPAN module that does correct matching of nested brackets/parentheses?
10:20 FROGGS Regexp::Common::balanced?
10:20 masak thanks.
10:20 FROGGS that is at least google's #1 for your question :o)
10:21 masak that's the one I was thinking of.
10:22 moritz there's also Text::ParseWords
10:22 moritz which exports quotewords
10:22 kivutar joined #perl6
10:23 xinming joined #perl6
10:27 masak ooh
11:12 denisboyun joined #perl6
11:19 sqirrel_ joined #perl6
11:21 tgt joined #perl6
11:22 colomon TimToady++
11:30 colomon Error while compiling op getuniname: No registered operation handler for 'getuniname'
11:30 colomon erm… possibly TimToady==
11:30 colomon TimToady--
11:40 masak could it be an nqp version skew?
11:46 FROGGS colomon: your moarvm is outdated
11:46 FROGGS and you'd see a hint about that if you'd reconfigure nqp
11:46 colomon FROGGS: that was a parrot build
11:47 FROGGS ahh, interesting fact :o)
11:48 FROGGS colomon: these ops were not added for nqp-p...
11:48 FROGGS so, we'd need to fudge this for !moar: https://github.com/rakudo/rakudo/commit/5f457997782e2f0bb3ddfc4e903892d048c5c077
11:49 FROGGS and perhaps other commits as well
11:49 FROGGS I do that now
11:50 dalek rakudo/eleven: 7c12aaf | Mouq++ | src/core/ (2 files):
11:50 dalek rakudo/eleven: Fix regression caused by correct list flattening
11:50 dalek rakudo/eleven:
11:50 dalek rakudo/eleven: $/.pairs couldn't handle Matches with multiple matches under a single
11:50 dalek rakudo/eleven: index, e.g., where it should output 0 => (Match1,Match2) it would output
11:50 FROGGS that is not it yet
11:51 dalek joined #perl6
11:54 dalek rakudo/nom: f2471ab | (Tobias Leich)++ | src/core/Str.pm:
11:54 dalek rakudo/nom: fudge uniname/-prop/val which is implemented on moar only
11:54 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f2471ab1e0
11:54 FROGGS colomon: ^^
11:55 FROGGS I am rebuilding everything atm so this will take a while
11:55 rindolf joined #perl6
11:56 Hamada joined #perl6
12:01 scottp joined #perl6
12:01 bowtie joined #perl6
12:09 TestIRCClient joined #perl6
12:10 Sawaf Hello
12:11 colomon FROGGS++
12:11 FROGGS hi Sawaf
12:13 TestIRCClient joined #perl6
12:14 Sawaf Please I need to ask how can I start learning perl
12:15 scottp joined #perl6
12:15 vendethiel joined #perl6
12:15 tadzik hello
12:16 tadzik you can start with http://perl6.org/documentation/
12:16 FROGGS Sawaf: If you are interested in Perl 6, then perhaps http://perl6.org/getting-started/
12:16 FROGGS Sawaf: if it is more about Perl 5, then http://www.perl.org/learn.html
12:20 vendethiel joined #perl6
12:20 bowtie_ joined #perl6
12:21 TestIRCClient joined #perl6
12:22 Sawaf Thanks FROGGS but can't I find a video tutorials ?
12:23 FROGGS there are a lot of videos on youtube
12:23 FROGGS Sawaf: but still, are you looking for Perl 5 or Perl 6?
12:25 TestIRCClient joined #perl6
12:25 Sawaf I'm looking for Perl 6
12:26 FROGGS then you should search for "Perl 6", "Jonathan Worthington", "Carl Masak", "Rakudo" to mention a few
12:29 TestIRCClient joined #perl6
12:29 Sawaf Thanks FROGGS I'll serch for it :)
12:30 FROGGS you're welcome
12:30 FROGGS do not hesitate to ask or play around here :o)
12:30 FROGGS r: say "hallo" # :o)
12:30 camelia rakudo-parrot fc2a6e, rakudo-jvm fc2a6e, rakudo-moar fc2a6e: OUTPUT«hallo␤»
12:31 FROGGS you can also do that in a private chat with the camelia bot
12:32 Sawaf thanks alot
12:32 bluescreen10 joined #perl6
12:33 Sawaf left #perl6
12:36 bowtie_ joined #perl6
12:37 bluescreen100 joined #perl6
12:41 bluescreen__ joined #perl6
12:43 kaleem joined #perl6
12:44 masak http://hyperboleandahalf.blogspot.se/2010/04/alot-is-better-than-you-at-everything.html # thanks alot!
12:46 TestIRCClient joined #perl6
12:49 FROGGS *g*
12:49 TestIRCClient joined #perl6
13:06 xinming_ joined #perl6
13:11 xinming joined #perl6
13:12 Mouq joined #perl6
13:15 dmol joined #perl6
13:15 timotimo moritz: i have a bit of code that goes into the right direction to do that already, but i never finished it :o
13:22 bowtie joined #perl6
13:29 bowtie joined #perl6
13:43 stevan_ joined #perl6
13:49 Alina-malina joined #perl6
13:53 lue joined #perl6
13:55 [Coke] moritz: if there's a natural sixian way to do it, I'll prefer that way.
13:56 fridim__ joined #perl6
14:00 moritz [Coke]: well, there are rw attributes in Perl 6, but they tend to be a hassle
14:03 [Coke] moritz: at this point, I'm just getting 15m here and there to pull in basic stuff. (rather than thinking big picture). happy to give out commit bits if anyone wants to poke at things they like.
14:10 guru joined #perl6
14:15 xragnar_ joined #perl6
14:25 skids joined #perl6
14:27 jnap joined #perl6
14:28 bluescreen10 joined #perl6
14:29 bowtie_ joined #perl6
14:41 salv0 joined #perl6
14:59 Adriaaan joined #perl6
15:00 Mouq joined #perl6
15:13 xfix joined #perl6
15:24 test joined #perl6
15:24 FROGGS http://www.stephendiehl.com/llvm/ - Implementing a JIT Compiled Language with Haskell and LLVM
15:37 __thou joined #perl6
15:37 hoverboard joined #perl6
15:41 bowtie joined #perl6
15:44 telex joined #perl6
15:48 guru joined #perl6
16:05 dmol joined #perl6
16:08 Adriaaan joined #perl6
16:10 ajr joined #perl6
16:14 * TimToady -- for thinking "What about jvm and parrot?" and then promptly getting distracted
16:14 TimToady er, TimToady--
16:16 colomon TimToady++ # for being human
16:17 TimToady well, for trying anyway :)
16:18 TimToady m: say unival('⅓').nude
16:18 camelia rakudo-moar f2471a: OUTPUT«1 3␤»
16:19 TimToady m: say unival('3').nude;  # curious
16:19 camelia rakudo-moar f2471a: OUTPUT«No such method 'nude' for invocant of type 'Int'␤  in block  at /tmp/hVbqhLpZGw:1␤␤»
16:19 TimToady I guess Rats aren't cool enough :)
16:20 TimToady m: say unival('⅓')
16:20 camelia rakudo-moar f2471a: OUTPUT«0.333333␤»
16:20 TimToady m: say uniprop('⅓', 'NumericValue')
16:20 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:20 TimToady which gives the *appearance* of being more precise, but isn't
16:21 TimToady m: say uniprop('⅓', 'NumericValue').round(15)
16:21 camelia rakudo-moar f2471a: OUTPUT«0␤»
16:21 TimToady m: say uniprop('⅓', 'NumericValue').Num.round(15)
16:21 camelia rakudo-moar f2471a: OUTPUT«0␤»
16:21 TimToady hmm
16:21 pragma- too much gibberish in here
16:21 pragma- left #perl6
16:21 moritz .oO( too much pragma- in here )
16:22 TimToady m: say uniprop('⅓', 'NumericValue').WHAT
16:22 camelia rakudo-moar f2471a: OUTPUT«(Str)␤»
16:22 moritz m: say uniprop('⅓', 'NumericValue').perl
16:22 camelia rakudo-moar f2471a: OUTPUT«"0.33333333"␤»
16:23 TimToady Str I expected, not .Num-ing I didn't
16:23 moritz m: say "0.33333333".Num
16:23 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:24 TimToady maybe it's really a moar string
16:24 TimToady m: say +uniprop('⅓', 'NumericValue')
16:24 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:24 TimToady that works
16:25 TimToady m: say uniprop('⅓', 'NumericValue').Numeric.WHAT
16:25 isacloud_ joined #perl6
16:25 camelia rakudo-moar f2471a: OUTPUT«(Rat)␤»
16:25 TimToady m: say uniprop('⅓', 'NumericValue').Num.WHAT
16:25 camelia rakudo-moar f2471a: OUTPUT«(Num)␤»
16:25 TimToady m: say uniprop('⅓', 'NumericValue').Numeric
16:25 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:25 TimToady m: say uniprop('⅓', 'NumericValue').Num
16:25 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:25 TimToady say what?
16:26 TimToady oh, duh
16:26 TimToady .round was wrong
16:26 TimToady m: say uniprop('⅓', 'NumericValue').Num.round(0.1 ** 15)
16:26 camelia rakudo-moar f2471a: OUTPUT«0.33333333␤»
16:26 treehug88 joined #perl6
16:26 TimToady m: say unival('⅓').Num.round(0.1 ** 15)
16:26 camelia rakudo-moar f2471a: OUTPUT«0.333333333333333␤»
16:26 TimToady m: say unival('⅓').Num.round(0.1 ** 20)
16:26 camelia rakudo-moar f2471a: OUTPUT«-0.0922337203685478␤»
16:26 TimToady ^^^ there's a bug
16:27 TimToady m: say unival('⅓').round(0.1 ** 20)
16:27 camelia rakudo-moar f2471a: OUTPUT«-0.0922337203685478␤»
16:27 TimToady same bug for Rat, apparently
16:29 masak m: say (1/3).round(0.1 ** 20)
16:29 camelia rakudo-moar f2471a: OUTPUT«-0.0922337203685478␤»
16:31 xinming_ joined #perl6
16:34 FROGGS joined #perl6
16:36 colomon :\
16:36 colomon r: say (1/3).round(0.1 ** 20)
16:36 camelia rakudo-parrot f2471a: OUTPUT«0.333333333333333␤»
16:36 camelia ..rakudo-moar f2471a: OUTPUT«-0.0922337203685478␤»
16:36 camelia ..rakudo-jvm f2471a: OUTPUT«0.3333333333333333␤»
16:37 TimToady I note the others are capped at 15
16:37 TimToady r: say (0.1 ** 20).WAHT
16:37 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«No such method 'WAHT' for invocant of type 'Num'␤  in block  at /tmp/tmpfile:1␤␤»
16:37 TimToady r: say (0.1 ** 20).WHAT
16:37 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«(Num)␤»
16:37 TimToady hmm
16:37 TimToady that I didn't expect
16:38 TimToady r: say (10 ** -20).WHAT
16:38 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«(Num)␤»
16:38 TimToady n: say (0.1 ** 20).WHAT
16:38 camelia niecza v24-109-g48a8de3: OUTPUT«Num()␤»
16:38 TimToady n: say (1 / 10 ** 20).WHAT
16:38 camelia niecza v24-109-g48a8de3: OUTPUT«Num()␤»
16:39 TimToady oh, wait
16:39 colomon r: say (3333333333333333333/100000000000000000000).perl
16:39 camelia rakudo-jvm f2471a: OUTPUT«0.03333333333333333e0␤»
16:39 camelia ..rakudo-parrot f2471a, rakudo-moar f2471a: OUTPUT«0.0333333333333333e0␤»
16:39 colomon r: say (3333333333333333333/100000000000000000000).WHAT
16:39 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«(Num)␤»
16:39 TimToady say 2 ** 64
16:39 TimToady r: say 2 ** 64
16:39 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«18446744073709551616␤»
16:39 TimToady r: say (2 ** 64).chars
16:39 kaleem joined #perl6
16:39 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«20␤»
16:40 TimToady r: say (0.1 ** 19).WHAT
16:40 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«(Rat)␤»
16:40 TimToady that's the issue
16:40 molaf_ joined #perl6
16:40 TimToady well, one of 'em
16:42 TimToady m: (^0x110000).pick.&uniname
16:42 camelia rakudo-moar f2471a: ( no output )
16:42 TimToady m: (^0x110000).pick.&uniname.say
16:42 camelia rakudo-moar f2471a: OUTPUT«<unassigned>␤»
16:42 TimToady well, duh
16:43 TimToady m: (^0x30000).pick.&uniname.say
16:43 camelia rakudo-moar f2471a: OUTPUT«<unassigned>␤»
16:43 hoverboard joined #perl6
16:43 TimToady m: (^0x30000).pick.&uniname(10)».say
16:43 camelia rakudo-moar f2471a: OUTPUT«*␤»
16:43 TimToady bleh
16:43 TimToady m: (^0x30000).pick(10).&uniname».say
16:43 camelia rakudo-moar f2471a: OUTPUT«*␤»
16:43 TimToady m: (^0x30000).pick(10)».&uniname».say
16:43 camelia rakudo-moar f2471a: OUTPUT«<Private Use>␤<CJK Ideograph Extension B>␤<CJK Ideograph Extension B>␤HEAVY RIGHT-POINTING ANGLE BRACKET ORNAMENT␤<unassigned>␤<CJK Ideograph Extension A>␤<unassigned>␤<unassigned>␤<CJK Ideograph Extension B>␤GREEK DASIA AND PERISPOMENI␤»
16:44 TimToady there we go, and I'll quit spamming for at least 30 seconds, lest people take me out and shoot me for designing Camelia :)
16:46 * TimToady apologizes (slightly) for the manic phase into which his pneumonia drugs have (apparently) propelled him... :)
16:47 skids .oO(uninames would make good anime power callouts)
16:48 bowtie_ joined #perl6
16:48 Mouq joined #perl6
16:50 skids .oO(I guess those are called "kiai")
16:51 TimToady 気合 scream; yell; fighting spirit
16:52 skids All I know is I wouldn't want to be hit in the head by a HEAVY RIGHT-POINTING ANGLE BRACKET ORNAMENT.
16:53 skids (or a PILE OF POO)
16:56 bowtie joined #perl6
17:01 rindolf TimToady: I'm sorry to hear you had pneumonia.
17:07 btyler joined #perl6
17:12 dwarring joined #perl6
17:26 [Coke] http://news.perlfoundation.org/2014/03/outreach-program-for-women-201.html - anyone have some intern ideas for six?
17:28 masak TimToady: I look at that and see "steam proper".
17:28 rindolf joined #perl6
17:31 TimToady well, yeah, in japanese the first has been metaphoricalized from "vapor" to "spirit"; see "spirits of ammonia" and such in English, and relationship of breath to spirit or soul in hebrew or greek
17:31 TimToady and the second one is more generally "fits", which in chinese tends to mean "is considered fitting" :)
17:32 TimToady but you can think of it as "proper spirit yell" or some such
17:32 TimToady something fitting for the occasion
17:33 hoverboard joined #perl6
17:34 [Sno] joined #perl6
17:35 [Coke] spirit yell sounds like what we used to do before a show - every cast/crew hold hands, start with a low note, raising pitch and volume, ending in a scream.
17:45 mcglk_ joined #perl6
17:48 bowtie joined #perl6
17:51 __sri joined #perl6
18:04 bowtie joined #perl6
18:13 guru joined #perl6
18:22 tgt joined #perl6
18:24 Adriaaan joined #perl6
18:33 shinobicl joined #perl6
18:35 * FROGGS is eager to see an update to perl6_pass_rates.csv
18:37 Mouq joined #perl6
18:39 shinobicl left #perl6
18:40 timotimo is that because of parrot snapping back to the values it should have or did moar get something cool while i wasn't looking?
18:42 [Coke] FROGGS: thanks for the reminder; the one shot run went ok, now trying it via cron.
18:42 FROGGS I dunno about moar, that is kinda the point
18:42 FROGGS \o/
18:42 FROGGS [Coke]++
18:42 FROGGS timotimo: and yeah, I hope that parrot is clean atm
18:42 timotimo heh.
18:43 rurban joined #perl6
18:43 timotimo i'm kind of annoyed that the spectest pass rate is so excellent, but there are bugs that are not showing up there that cause it to be kind of unusable for module installation
18:44 FROGGS t/spec/S10-packages/basic.rakudo.moar                       (Wstat: 256 Tests: 0 Failed: 0)
18:44 FROGGS Non-zero exit status: 1
18:44 FROGGS Parse errors: No plan found in TAP output
18:44 FROGGS perhaps it is a related issue?
18:44 FROGGS I was about to look at it
18:45 FROGGS ohh dear, it does work when I invoke it directly :/
18:45 timotimo sounds like a good candidate >_>
18:45 [Coke] ok, it's running.
18:46 FROGGS cool!
18:46 [Coke] LHF: find a test that says "nom regression", open a rakudobug for it, then update the test to refer to the RT #.
18:46 [Coke] LHF: /home/coke/sandbox/perl6-roast-data/Pugs.hs/t/spec/fudge: No such test file 't/spec/S05-metasyntax/prior.t'
18:50 FROGGS 66 hits for "nom regression" if I did it right
18:50 FROGGS (in roast)
18:51 pecastro joined #perl6
18:57 FROGGS btw, I bet that quite a high percentage of the failing tests in moar are fudged for jvm
18:57 FROGGS like test eight of t/spec/S06-macros/quasi-blocks.t
18:57 FROGGS fails on moar, fudged for jvm
18:58 timotimo so moar would have an even higher percentage? ;)
18:58 FROGGS yes
18:58 berekuk joined #perl6
18:58 FROGGS /home/froggs/dev/rakudo/t/spec/S04-phasers/enter-leave.t:193:#?rakudo.jvm skip 'unwind'
18:58 FROGGS /home/froggs/dev/rakudo/t/spec/S04-phasers/pre-post.t:161:#?rakudo.jvm skip "POST and exceptions"
18:59 FROGGS the same here, this fails on moar
18:59 berekuk joined #perl6
18:59 pmichaud good afternoon, #perl6
18:59 FROGGS so it is not really backend specific
18:59 FROGGS hi pmichaud
18:59 timotimo pmichaud o/
19:03 FROGGS p: my $c = 0; macro donner { quasi { ++$c } }; say donner; say donner; say $c
19:03 camelia rakudo-parrot f2471a: OUTPUT«1␤2␤0␤»
19:03 FROGGS is that expected?
19:08 TimToady all of the backends have been failing those
19:08 FROGGS r: my $c = 0; macro donner { quasi { ++$c } }; say donner; say donner; say $c
19:08 [Coke] percentage is based on passing tests, so failing vs. skipped doesn't matter.
19:08 camelia rakudo-parrot f2471a: OUTPUT«1␤2␤0␤»
19:08 camelia ..rakudo-jvm f2471a: OUTPUT«Cannot modify an immutable value␤  in sub prefix:<++> at gen/jvm/CORE.setting:1782␤  in sub prefix:<++> at gen/jvm/CORE.setting:1780␤  in any  at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
19:08 camelia ..rakudo-moar f2471a: OUTPUT«No such method 'STORE' for invocant of type 'Mu'␤  in sub prefix:<++> at src/gen/m-CORE.setting:1782␤  in sub prefix:<++> at src/gen/m-CORE.setting:1780␤  in any  at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
19:09 FROGGS [Coke]: but it matters when we abort a test file on moar, where jvm skips a single test and continues
19:09 [Coke] ah, yes, abort is worse than simple failure.
19:09 FROGGS but I'd like to see these tests fixed instead of skipped of course :o)
19:09 [Coke] this new testing box seems much slower.
19:09 FROGGS :/
19:09 [Coke] Still on niecza.
19:10 [Coke] pugs is completely hosed now (though I can probably get it back to the 20% rate it enjoyed on host06)
19:10 TimToady we'd probably have better luck writing a Haskell backend for rakudo these days :)
19:11 FROGGS but I hope we don't have to do that ó.ò
19:11 FROGGS three and one in the queue is enough I'd say
19:12 [Coke] I wouldn't call JS in the queue.
19:12 FROGGS why not?
19:12 [Coke] it has no champion at the moment, yes?
19:12 [Coke] if no one is committed to working on it, it's dead.
19:12 FROGGS ohh, he is going to work on it I think
19:13 FROGGS otherwise he should hand it over to us
19:14 denis_boyun joined #perl6
19:16 masak pmichaud! \o/
19:19 itz2 joined #perl6
19:24 masak FROGGS: no, that macro/quasi/1,2,0 thing is a bug, as far as I'm concerned.
19:24 masak FROGGS: it's related to http://strangelyconsistent.org/blog/lexpads-and-why-roles-need-fixups
19:24 [Coke] anyone looking into the java eval server failures that force the daily run to be done without?
19:24 FROGGS k
19:25 FROGGS masak: I guess it is not an easy fix?
19:25 FROGGS [Coke]: I'm not
19:25 masak FROGGS: I might be wrong, but it *used* to work.
19:25 masak FROGGS: don't know what broke it, or when.
19:26 FROGGS masak: years ago I suppose?
19:26 FROGGS star: my $c = 0; macro donner { quasi { ++$c } }; say donner; say donner; say $c
19:26 camelia star 2013-09: OUTPUT«1␤2␤0␤»
19:28 masak FROGGS: a year ago, I'd say.
19:28 FROGGS k
19:28 masak but I may be wrong on this.
19:28 FROGGS that is a starting point :o)
19:30 timotimo so, if i had a gui system, or something that uses callbacks a lot
19:30 timotimo python would allow you to write things like "callback=self.the_shiny_callback"
19:31 segomos masak: that method foo { method bar { } } thing is throwing me off
19:31 timotimo in perl6 we'd have to write something more like callback => { self.the_shiny_callback; }, don't we?
19:31 FROGGS timotimo: can't you pass it around using an & sigil?
19:31 masak FROGGS: yes, I'm wrong. there's an open ticket for this: https://rt.perl.org/Public/Bug/Display.html?id=120928 -- linked from that post.
19:32 masak segomos: why?
19:32 FROGGS masak: thanks!
19:32 masak segomos: (genuinely curious)
19:33 timotimo FROGGS: not methods
19:33 timotimo the trick with python is that when you write self.the_shiny_callback, you get a "bound method" back
19:33 timotimo which is kind of like a closure in disguise
19:34 segomos masak: it just looks like doing C.bar should be a syntax error, i know that it isn't, i'm just drawn to it
19:34 timotimo to be exact, it's a closure that closes over exactly "self" and delegates any arguments you give to it to the method on that method invoked on that closed-over self
19:34 virtualsue joined #perl6
19:34 timotimo i was about to complain, that p6's way is more wordy, but it's actually much clearer what's going on
19:36 masak segomos: you can declare a method inside a class. that is all. :)
19:37 masak timotimo: Python's way of doing it is definitely a local minimum. it has "user simplicity" going for it. Perl 6's way is more "honest".
19:40 timotimo i think the increased amount of "honesty" makes perl 6 a kind of nicer language to be taught to complete beginners
19:40 timotimo but i'd have to try that to verify it.
19:41 FROGGS masak: that is an awesome post!
19:42 FROGGS but I think "outselves" is a typo :o)
19:43 treehug88 joined #perl6
19:44 Sqirrel joined #perl6
19:44 timotimo maybe masak has just outselved himself once again? :)
19:47 arnsholt joined #perl6
19:47 FROGGS masak: I think "this kind of fixup" as you say in the post already is in place for macros
19:47 FROGGS at least I see something like that in the actions and world
19:49 hoverboard joined #perl6
19:53 masak yeah.
19:53 masak FROGGS: thanks for typo spotting -- fixing.
19:56 * moritz upgrades feather{1,2,3} and host07 to get rid of the gnutls bug
20:01 vendethiel wat wat wat
20:01 vendethiel wait
20:01 vendethiel r: class C { method a { say "a"; method b { say "b"; } } }; C.new.b
20:02 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«b␤»
20:02 vendethiel wtf
20:02 masak vendethiel: :)
20:02 vendethiel That's *really, really, really* surprising IMHO
20:02 vendethiel I actually discovered that from (I think) masak's post ? (lexpads and why roles need fixups)
20:02 masak we were surprised too, first time we discovered it.
20:02 FROGGS a method declaration just looks up the out class, and installs itself in there
20:03 masak but really, it falls out of how classes and method declarations work.
20:03 FROGGS outer*
20:03 vendethiel well, I really don't think it should even allow that. What sense does it make ?
20:03 moritz r: class A { if True { method a { 'a' } } else { method b { 'b' } } }; say A.a; say A.b
20:03 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«a␤b␤»
20:04 FROGGS moritz: nice try :o)
20:04 masak vendethiel: that's generally not the question Perl people ask... :)
20:04 masak moritz: BEGIN time vs runtime.
20:04 jnthn Turns out it's a lot more effort to forbid it than allow it :)
20:04 moritz masak: I know :-)
20:04 vendethiel masak: I ask it though ;). can't you just refuse to parse `method`s inside a non-{class,role}?
20:04 FROGGS yay! I will have a rakudo 2013.01 in a few minutes :o)
20:04 masak vendethiel: you can't really TIMTOWTDI if you run around shouting "but what's the *use*!?"
20:05 vendethiel masak: no, far from that. But really, at first, I thought the code exampled messed up its  { and }
20:05 masak vendethiel: you write a module that refuses to parse it. everyone wins.
20:05 moritz vendethiel: there was a time when we wouldn't allow product of two Duration objects
20:05 vendethiel masak: would people be okay about that ?
20:05 moritz vendethiel: because we couldn't think of a use case
20:06 vendethiel moritz: (not sure what Duration objects are)
20:06 moritz vendethiel: time spans
20:06 masak typos fixed -- http://strangelyconsistent.org/blog/lexpads-and-why-roles-need-fixups brought to you by Rakudo-on-Moar!
20:06 moritz vendethiel: until masak++ wanted to calculate standard deviations of Durations
20:06 FROGGS yay! :o)
20:06 masak vendethiel: we would. all is fair if you predeclare.
20:07 masak vendethiel: my point is, we don't go around forbidding things just because we can't see the use. we're not Python devs.
20:07 vendethiel masak: if you predeclare that methods can't be nested ?
20:07 masak vendethiel: as in 'use Modules::Strict;'
20:07 FROGGS masak: haha, I tell that one my PHP peeps :o)
20:07 pecastro joined #perl6
20:07 * vendethiel has to do PHP for school and is really sad
20:07 vendethiel (I need a JS-to-PHP to that I can compile anything to JS then back to PHP !)
20:08 vendethiel masak: "strict mode" before 6.0.0 seems like a bad idea though ?
20:09 masak vendethiel: not at all. we're experimenting *now*.
20:10 masak vendethiel: though I'm sorry to report that language-modifying modules are Not Quite There yet :/
20:10 vendethiel masak: that's cool, i'm Not Quite Perler either.
20:10 vendethiel not yet, though.
20:10 moritz vendethiel++
20:11 FROGGS masak: 2013.01 is also borken, I am going further back in time now
20:11 vendethiel well, `method { method }` doesn't seem like a bug you should stumble upon unless you want to experiment
20:11 vendethiel (or unless you messed up your }, I guess)
20:13 dalek nqp: f30b65f | larry++ | / (2 files):
20:13 dalek nqp: add bool and match uniprop opcodes
20:13 dalek nqp: review: https://github.com/perl6/nqp/commit/f30b65fed3
20:14 masak FROGGS++
20:14 dalek nqp: d3849dc | larry++ | tools/build/MOAR_REVISION:
20:14 dalek nqp: bump moar rev
20:14 dalek nqp: review: https://github.com/perl6/nqp/commit/d3849dc182
20:15 bluescreen100 joined #perl6
20:16 dalek rakudo/nom: 02329b9 | larry++ | / (2 files):
20:16 dalek rakudo/nom: update to use new opcodes and bump NQP rev
20:16 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/02329b92c5
20:17 * TimToady hopes he did it all right this time
20:17 TimToady there should be an emoticon with crossed fingers
20:18 TimToady I don't think the japanese have that gesture, so we don't have the emoji.  :)
20:18 bluescreen100 joined #perl6
20:18 FROGGS TimToady: it looks good I'd say :o)
20:19 [Coke] daily run up to S03 in parrot
20:19 TimToady .n ????
20:19 TimToady well now
20:20 TimToady m: say '????'.&uniname
20:20 camelia rakudo-moar f2471a: OUTPUT«SMILING FACE WITH OPEN MOUTH AND COLD SWEAT␤»
20:20 TimToady almost don't need a unicode bot anymore :)
20:20 FROGGS yeah :o)
20:21 masak \o/
20:21 [Coke] m: say '�'.uniname
20:21 camelia rakudo-moar f2471a: OUTPUT«No such method 'uniname' for invocant of type 'Str'␤  in block  at /tmp/UCktbOVktq:1␤␤»
20:21 moritz off-topic question: is there a way to configure less to act like cat if the input is small enough to fit into one screen?
20:21 FROGGS [Coke]: it is a sub
20:21 timotimo i like that.
20:21 [Coke] I see that now ,yes. :)
20:21 moritz there's less -F, but that just exits, it doesn't cat the file
20:22 masak moritz: Git behaves like that, but I don't know how it does it.
20:22 masak moritz: ...maybe ask on #git ? :)
20:22 geekosaur for less there's an option
20:22 pippo joined #perl6
20:22 geekosaur (of course, there's only a billion of them)
20:23 [Coke] -F looks like it's supposed to work...
20:23 timotimo isn't -E right?
20:23 timotimo hm. well, almost.
20:23 jercos -E will die when the bottom is hit, -F looks more appropriate
20:23 moritz oh, I meant -E, not -F
20:24 timotimo oh?
20:24 timotimo ah!
20:24 timotimo it exits and clears the screen
20:24 timotimo that's annoying
20:24 moritz neither -E nor -F produce any output for me if the input is short
20:24 geekosaur odd
20:25 Mouq joined #perl6
20:25 [Coke] mmm. I've seen this working, I think there's another option to go along with -F...
20:25 jercos $ echo foo | less -F
20:25 jercos foo
20:25 * jercos shrugs
20:25 moritz oh
20:25 moritz less -XF
20:26 geekosaur might see if -L makes it work for you
20:26 geekosaur oh, yes, the deinit clears screen thing
20:26 moritz ok, thanks everybody, timotimo++ jercos++ geekosaur++ [Coke]++ masak++
20:26 timotimo yeah, -X
20:26 timotimo i just stumbled upon that
20:26 geekosaur which is more likely not clears screen but switches to/from alternate screen, which is something I often disable in the terminal
20:27 geekosaur because it's obnoxious
20:27 timotimo aye, the alternate screen is an interesting concept
20:28 timotimo it has the whole DWIM + WAT thing going on
20:29 FROGGS DO WAT I MEAN
20:29 moritz FROGGS++
20:29 timotimo :D
20:29 moritz oh, now I understand why nothing remains in the terminal after less has terminated
20:30 lue Ah, is the alternate screen the thing where I exit out of vi or whatever and I get to see all the previous terminal output there?
20:30 geekosaur yep
20:31 geekosaur it has its place, sadly there is no way to specify which programs use it
20:31 lue It's weird how so many things I found searching for "alternate screen" tell you how to get rid of it. I get annoyed when I have to use the bare linux console and it doesn't do that.
20:32 geekosaur depending on the terminal description (termcap/terminfo stuff) in use, it may activate on all programs that use cursor addressing or programs that use insert/delete commands
20:36 [Coke] bottom of today's moar run:
20:36 [Coke] https://gist.github.com/coke/9355089
20:37 [Coke] only 20 aborted tests in rakudo.moar
20:39 FROGGS so we are down to 57 fails (from 71), nice!!
20:39 timotimo aborted?
20:39 timotimo as in return value != 0?
20:39 FROGGS like in S06-macros/unquoting.t aborted 10 test(s
20:39 [Coke] TimToady: S32-str/tclc.t 3 - tclc and German sharp s
20:40 [Coke] that test is suspect and I think we should fix it or kill it.
20:40 [Coke] timotimo: aborted = teh test file crashed and we didn't try to run the remaining tests you declared.
20:40 timotimo ah
20:40 [Coke] usually a syntax that isn't supported or an unhandled exception.
20:40 timotimo why didn't you try to run the tests i declared? :(
20:41 FROGGS :o(
20:41 * FROGGS cries too
20:41 lue r: say "ßß".tclc
20:41 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«ßß␤»
20:41 lue Doesn't even do anything; it's supposed to come out as ẞß of course.
20:42 timotimo at least the insanity of "ss" is gone
20:42 timotimo ... or is it?
20:42 [Coke] the test is expecting 'Ssß'
20:43 [Coke] which might work if it was mid-word
20:43 * lue wishes Unicode would make everything easier by setting the uc mapping of ß to ẞ
20:43 timotimo aye
20:43 FROGGS +1
20:44 lue Can't recall if that's against the stability policy though; worst case, they have to make new codepoints for ß and ẞ
20:44 moritz and then have two ß characters that have different case mappings? NOOOOO
20:45 moritz n: say 'ß'.uc, 'ß'.tc, 'ß'.tclc
20:45 camelia niecza v24-109-g48a8de3: OUTPUT«ßSsSs␤»
20:45 moritz p6: say 'ß'.uc, 'ß'.tc, 'ß'.tclc
20:45 camelia rakudo-jvm f2471a: OUTPUT«SSSSß␤»
20:45 camelia ..rakudo-parrot f2471a, rakudo-moar f2471a: OUTPUT«ßßß␤»
20:45 camelia ..niecza v24-109-g48a8de3: OUTPUT«ßSsSs␤»
20:45 timotimo SSsssssssssssss
20:45 timotimo NO LOL
20:45 FROGGS all are rong
20:46 lue timotimo: for the record, the tc mapping of ß *is* Ss
20:46 moritz 4 implementations, 3 answers, all wrong
20:46 moritz lue: but 'ß'.uc is supposed to be 'SS'
20:46 FROGGS either SS or ẞ
20:46 Vlavv joined #perl6
20:46 lue http://www.unicode.org/policies/stability_policy.html#Identity  <-- I wouldn't say changing ß.uc to ẞ counts as changing the identity, does it?
20:47 FROGGS lue: corret
20:47 FROGGS lue: if you would change the meaning of ß, but that is not the plan
20:49 xenoterracide joined #perl6
20:49 * lue feels ß->ss and ẞ->SS should be left to the NFK* forms.
20:50 lue ooc, just how many Germans are here right now? There's me, FROGGS, timotimo, and moritz, right? :)
20:50 * TimToady is part German...
20:50 timotimo oh, i didn't know you were german
20:50 * geekosaur is also part German... bery part
20:50 TimToady well, depends on how you count, of course
20:50 geekosaur *very part, bad typun
20:50 [Coke] I have a german born contractor here who is supposedly fluent, I could drag him in.
20:51 FROGGS I bet quite a few from U.S. and A are part german :o)
20:51 lue timotimo: Well, half-German, but still :)
20:51 timotimo fair enough :)
20:51 lue (don't expect me to start speaking fluent German here though; need a lot of practice (and vocabulary) first)
20:51 timotimo hoelzro has an umlaut in his name, you may count that as german, too :P
20:51 * moritz submits ß case-folding rakudobug
20:51 * skids isn't even germane most of the time, forget German.
20:51 TimToady Ich bin nicht ein Berliner.
20:52 FROGGS lue: Sqirrel is also german :o)
20:52 timotimo but germans can't say squirrel!
20:52 FROGGS hehe
20:53 moritz eichhörnchen!
20:53 lue TimToady: Ich bin nicht ein Berliner; ich bin ein Sachse!
20:54 guru joined #perl6
20:54 FROGGS [Sno] is german too
20:54 timotimo ah yes, [Sno] :)
20:54 * timotimo waves
20:55 FROGGS and the british people are from saxonia anyway :P
20:55 FROGGS so BinGOs++ is also german, period.
20:56 __thou joined #perl6
20:57 * [Sno] waves back and goes to sleep
20:57 FROGGS that reminds me of the ancestry research I did years ago... that was nice
20:57 FROGGS sleep well [Sno]
20:57 grondilu joined #perl6
20:57 * grondilu noticed the 99.83%
20:58 grondilu (on https://github.com/coke/perl6-roast-data/blob/master/perl6_pass_rates.csv that is)
20:58 BinGOs FROGGS: eh?
20:58 lue Hrm. The "Case Pair Stability" section at the bottom took away plenty of my hope.    http://www.unicode.org/policies/stability_policy.html#Case_Pair
20:58 FROGGS yes, TimToady++ fixed a lot of unicode issues
20:58 TimToady quick, add a bunch of tests for unicode props :D
20:58 FROGGS BinGOs: not that important :o)
20:58 vendethiel rakudo jvm 100% oO ?
20:58 BinGOs FROGGS: I am celt, not anglo-saxon.
20:59 FROGGS ohh
20:59 FROGGS my bad
20:59 BinGOs from Welshland
20:59 grondilu vendethiel: the percentage are relative to the maximum.  Since this maximum is obtained for jvm, jvm gets 100%
21:00 vendethiel grondilu: 5 failures, 610 todo, 1k5 skips
21:00 vendethiel alright :p
21:00 FROGGS ummm, now I want to be on that said Island :/
21:00 grondilu the number that count is the "pass" I guess
21:00 itz2 "Taffy Was A Welshman, Taffy Was a Thief "
21:01 lue Seems like we ain't getting ß.uc -> ẞ unless we add at least one new, otherwise-redundant character :/ .
21:01 FROGGS hehe
21:01 grondilu r: printf "%.2f%%", 28931/ 28979
21:01 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.00%»
21:01 pippo o/ #perl6
21:01 moritz \o pippo
21:01 FROGGS when BinGOs was a thief he would probably steal electric power to have even more smokers :o)
21:01 geekosaur did you want a *100 in there?
21:02 timotimo probably
21:02 FROGGS o\ pippo
21:02 geekosaur r: printf "%.2f%", 28931 * 100 / 28979
21:02 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«'.2f%' is not valid in sprintf format sequence '%.2f%'␤␤»
21:02 pippo anybody could help me? Moar from latest git does no more compile. :-(
21:02 geekosaur oh futz
21:02 geekosaur r: printf "%.2f%%", 28931 * 100 / 28979
21:02 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«99.83%»
21:02 geekosaur no cookies for you, xchat...
21:02 pippo Here is the error log: http://nopaste.info/96f407d25c.html
21:02 [Coke] pippo: what error are you getting?
21:02 [Coke] danke
21:02 treehug88 joined #perl6
21:03 vendethiel and
21:03 [Coke] pippo: did you update and reconfigure, or update and just make?
21:03 moritz looks like moar compiled just fine
21:03 moritz that's NQP failling
21:03 vendethiel r: printf "%.2f%%", 28979 * 100 / 28979+5+610+1505
21:03 * moritz tries
21:03 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«2220.00%»
21:03 moritz vendethiel: precedence
21:03 vendethiel r: printf "%.2f%%", 28979 * 100 / (28979+5+610+1505)
21:03 vendethiel yeah
21:03 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«93.18%»
21:04 pippo I have update and recompiled. Then erased everything then recompiled. No way!
21:04 vendethiel so rakudo.jvm passes 93.18% of perl6 specs
21:04 * moritz also tries
21:04 moritz nobody knows the actual number of tests in roast.
21:05 moritz pippo: did you also clean up the install directory?
21:05 timotimo vendethiel: that's more than i was expecting
21:05 timotimo vendethiel: but a lot of spec tests are not yet written ;)
21:06 pippo moritz: yes. It was a subdir of rakudo. So it went erased when I nuked everything.
21:06 grondilu r: printf "%.2f", .09
21:06 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.09»
21:06 [Coke] my german guy says: "I always thought that ẞ was case insensitive"
21:06 grondilu r: printf "%.2f", 9 / 10
21:06 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.90»
21:06 timotimo vendethiel: and pieces of the specs are not yet final anyway; things like S11 and lue's shiny new unicode specs for example
21:06 masak [Coke]: it definitely should have been!
21:06 timotimo and the list and iterators overhaul
21:06 [Coke] daily run now building r-j
21:07 FROGGS [Coke]: no, but for a long time there was no uppercase version of ß
21:07 lue I wonder if the Identity Stability policy or the Case Pair policy would win in an argument for ß.uc -> ẞ
21:07 grondilu r: printf "%.2f", 28931/ 28979
21:07 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.00»
21:08 grondilu r: printf "%.2f", 2891/ 2897
21:08 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.00»
21:08 grondilu wth
21:08 lue (also, there's the issue of losing linguistic info by converting to "SS")
21:08 grondilu r: printf "%.2f", 289/ 289
21:08 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«1.00»
21:08 grondilu r: printf "%.2f", 288/ 289
21:08 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.00»
21:08 FROGGS masak: about the macros... bisecting did not help, I went back 2011.12 and either it still printed 1,2,0 or did not build
21:08 grondilu r: printf "%.2f", 27/ 29
21:08 FROGGS masak: so I am going to debug HEAD
21:08 camelia rakudo-parrot f2471a, rakudo-jvm f2471a, rakudo-moar f2471a: OUTPUT«0.93»
21:10 masak FROGGS: ok.
21:10 masak FROGGS: then it probably never worked.
21:10 masak FROGGS: thanks for the detective work. FROGGS++
21:11 FROGGS I hope I can gain some knowledge about that though
21:11 [Coke] https://gist.github.com/coke/9355722 - parrot log today
21:12 segomos masak: it's the method { method { } } thing that looks funny to me :-)
21:12 segomos nested methods
21:12 timotimo oh, parrot is almost clean!
21:12 timotimo very good
21:13 timotimo S26 is documentation, no?
21:13 timotimo interesting
21:13 lue timotimo: it is for docs, yes
21:14 timotimo probably just a little difference between current implementation and current tests
21:16 masak segomos: sure, I understand that.
21:16 masak segomos: my point is that *distaste* is not a strong enough reason to forbid something.
21:16 virtualsue joined #perl6
21:16 timotimo hm, i think i'm not getting the reference
21:17 timotimo is that about the "macros need fixups" blog post?
21:17 segomos masak: of course
21:17 segomos it's not even a distaste, just a fixation
21:18 masak segomos: it happens ;)
21:18 * lue wonders where/how he files a property change to the Unicode consortium.
21:19 [Coke] masak: https://twitter.com/DannyTanner/status/331284347338493953 # pretty sure that's not the original, but I'm lazy
21:19 FROGGS lue: http://www.unicode.org/pending/proposals.html
21:21 lue FROGGS: that's for new characters, AFAIK, not for proposing changes to characters (could've mis-skimmed though)
21:21 masak [Coke]: weakly autopunnish, yes.
21:22 masak [Coke]: though I can't verify the actual claim. it seems to be related to the verb "to sling".
21:24 FROGGS lue: this? http://www.unicode.org/pending/properties.html
21:26 lue Hey, this looks like the avenue I need to go down: http://www.unicode.org/review/pri251/
21:42 xragnar joined #perl6
21:46 lue Can't install URI (and thus Pod::To::HTML) through Moar:  t/01.t ................... 2/47 Cannot invoke null object
21:46 FROGGS lue: known
21:47 FROGGS RT #121298
21:47 synopsebot Link: https://rt.perl.org/rt3//Public/Bug/Display.html?id=121298
21:53 lue good article for fellow ẞ lovers :)  http://opentype.info/blog/2011/01/24/capital-sharp-s/
21:58 kbaker__ joined #perl6
21:59 * masak .oO( I don't write capital S often, but when I do, I write it sharp )
22:01 masak lue: interesting article. I skimmed it. will read it careful...lier later.
22:01 stevan_ joined #perl6
22:02 lue masak: most interesting thing I learned is that the ß.uc → SS thing was only ever meant to be an interim solution, because typsetters apparently couldn't agree on what the capital would look like :)
22:02 stevan_ joined #perl6
22:06 masak lue: :)
22:06 bluescreen10 joined #perl6
22:07 FROGGS lue: nice article!
22:07 masak well, it's been good talking to y'all.
22:07 masak 'night, #perl6
22:07 FROGGS gnight masak
22:08 lue ♞ mäsak o/
22:14 TimToady for a second there, I thought someone had snuck a Prolog engine into Perl 6:
22:14 TimToady m: say '⅓'.&uniprop
22:14 camelia rakudo-moar 02329b: OUTPUT«No␤»
22:15 TimToady m: say '⅓'.&unimatch('No')
22:15 camelia rakudo-moar 02329b: OUTPUT«True␤»
22:15 TimToady m: say '⅓'.&unimatch('Nd')
22:15 camelia rakudo-moar 02329b: OUTPUT«False␤»
22:16 FROGGS No is numeric other?
22:17 FROGGS but anyway, getting "No" as a response of a compiler to &say something is kinda funny :o)
22:17 FROGGS from a*
22:20 [Coke] m: say unipoo('a');
22:20 camelia rakudo-moar 02329b: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/eRJLLrgr02â�¤Undeclared routine:â�¤    unipoo used at line 1. Did you mean '&uniprop', '&unibool'?â�¤â�¤Â»
22:21 lue I'd suggest that 'uni' can be shortened to 'u' in these sub names, but ambiguity issues come up with uname ... :)
22:25 rurban_ lue: well most typesetters use the lowercase  ß for the uppercase version.
22:25 lue m: say '⅓'.&uniprop('Numeric_Value')
22:25 camelia rakudo-moar 02329b: OUTPUT«0.33333333␤»
22:26 lue rurban_: and that's incredibly sick and wrong :)
22:26 lue [article from same guy as the last one: http://opentype.info/blog/2013/11/18/capital-sharp-s-design-review/]
22:26 * lue afk
22:43 TimToady r: say .chr, :16(.&uniprop("BidiMirroringGlyph")//next).chr when .&unimatch("BidiMirrored") for 0..0xffff;
22:44 camelia rakudo-jvm 02329b: OUTPUT«(timeout)»
22:44 camelia ..rakudo-parrot 02329b: OUTPUT«unimatch NYI on parrot backend␤  in sub unimatch at gen/parrot/CORE.setting:6760␤  in sub unimatch at gen/parrot/CORE.setting:6760␤  in block  at /tmp/tmpfile:1␤␤»
22:44 camelia ..rakudo-moar 02329b: OUTPUT«(timeout)()␤)(␤<>␤><␤[]␤][␤{}␤}{␤«»␤»«␤༺༻␤༻༺␤༼༽␤༽༼␤᚛᚜␤᚜᚛␤‹›␤›‹␤⁅⁆␤⁆⁅␤⁽⁾␤⁾⁽␤₍₎␤₎₍␤⅀)␤∁)␤∂)␤∃)␤∄)␤∈∋␤∉∌␤∊∍␤∋…»
22:45 TimToady hmm, some of them seem to think ) is the correct answer
22:46 timotimo oooh, i like the spelling .&foo for foo($_, ...)
22:46 TimToady m: say '∃'.uniprop("BidiMirroringGlyph")
22:46 camelia rakudo-moar 02329b: OUTPUT«No such method 'uniprop' for invocant of type 'Str'␤  in block  at /tmp/HUVW977xlW:1␤␤»
22:47 TimToady m: say '∃'.&uniprop("BidiMirroringGlyph")
22:47 camelia rakudo-moar 02329b: OUTPUT«0029␤»
22:47 TimToady sucks if you leave out the & though
22:47 TimToady hmm, I wonder why it things right paren is the mirror there...
22:48 TimToady anyway, soon we oughta be able to generate the pairs for our parser automatically
22:48 timotimo hehe.
22:48 TimToady m: say "\c[COMET]".&uniprop("BidiMirroringGlyph")
22:48 camelia rakudo-moar 02329b: OUTPUT«0029␤»
22:48 TimToady aww
22:50 tgt joined #perl6
22:51 timotimo ... wha? it's not snowman?
22:51 * timotimo files a bug to the unicode consortium
22:53 lue timotimo: file your request here, so far as I can tell :)  http://unicode.org/reporting.html
22:53 TimToady m: say "༼".&uniname
22:53 camelia rakudo-moar 02329b: OUTPUT«TIBETAN MARK ANG KHANG GYON␤»
22:54 timotimo TimToady: do you want to write the support to match against the name of unicode chars inside a regex with an embedded regex now? :)
22:54 lue TimToady: I assume unimatch only works on boolean properties?
22:54 timotimo seems likely
22:54 TimToady not sure
22:56 TimToady currently it looks up the table using the same property value, you'd have to separate them
22:56 btyler well, doing my part to spread the culture...used "golf it down" in the context of a mozilla project bug report conversation and ended up sharing that phrase with a few confused devs :)
22:57 timotimo ... that is not a common thing?!
22:57 btyler maybe it is generally, and just this batch of folks hadn't encountered it
22:58 TimToady m: my $table = nqp::unipropcode("BidiMirroringGlyph"), my $val = nqp::unipvalcode("0029"); say nqp:matchuniprop("(".ord, $table, $val)
22:58 camelia rakudo-moar 02329b: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/fd79dMI2l8â�¤Undeclared routine:â�¤    nqp:matchuniprop("(".ord, $table, $val) used at line 1â�¤â�¤Â»
22:58 TimToady m: my $table = nqp::unipropcode("BidiMirroringGlyph"), my $val = nqp::unipvalcode("0029"); say nqp::matchuniprop("(".ord, $table, $val)
22:58 camelia rakudo-moar 02329b: OUTPUT«WARNINGS:␤Useless use of "," in expression "my $table = nqp::unipropcode(\"BidiMirroringGlyph\"), my $val = nqp::unipvalcode(\"0029\")" in sink context (line 1)␤===SORRY!===␤Error while compiling op unipvalcode: Arg count 1 doesn't equal required ope…»
22:59 TimToady m: my $table = nqp::unipropcode("BidiMirroringGlyph"); my $val = nqp::unipvalcode("0029"); say nqp::matchuniprop("(".ord, $table, $val)
22:59 camelia rakudo-moar 02329b: OUTPUT«===SORRY!===␤Error while compiling op unipvalcode: Arg count 1 doesn't equal required operand count 3 for op 'unipvalcode'␤»
22:59 TimToady m: my $table = nqp::unipropcode("BidiMirroringGlyph"); my $val = nqp::unipvalcode($table,"0029"); say nqp::matchuniprop("(".ord, $table, $val)
22:59 camelia rakudo-moar 02329b: OUTPUT«1␤»
22:59 TimToady it seems to say you can
22:59 TimToady m: my $table = nqp::unipropcode("BidiMirroringGlyph"); my $val = nqp::unipvalcode($table,"0029"); say nqp::matchuniprop("[".ord, $table, $val)
22:59 camelia rakudo-moar 02329b: OUTPUT«0␤»
23:00 lue r: say '('.&uniprop('BidiMirroring')
23:00 camelia rakudo-jvm 02329b: OUTPUT«uniprop NYI on jvm backend␤  in sub uniprop at gen/jvm/CORE.setting:6753␤  in sub uniprop at gen/jvm/CORE.setting:6753␤  in block  at /tmp/tmpfile:1␤␤»
23:00 camelia ..rakudo-parrot 02329b: OUTPUT«uniprop NYI on parrot backend␤  in sub uniprop at gen/parrot/CORE.setting:6757␤  in sub uniprop at gen/parrot/CORE.setting:6757␤  in block  at /tmp/tmpfile:1␤␤»
23:00 camelia ..rakudo-moar 02329b: OUTPUT«0␤»
23:00 TimToady yes, I just used it to match in the mirroring glyph table, but we don't expose that with unimatch() currently
23:00 lue m: say '('.&unibool('BidiMirroring')
23:00 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:00 timotimo heh, oops.
23:00 lue doesn't look like unibool works atm
23:01 Mouq joined #perl6
23:01 timotimo well, that's not a property that'd have a boolean value, eh?
23:01 * lue isn't too sure about a separate function for boolean values anyway
23:01 lue timotimo: oh, I thought it was.
23:02 FROGGS m: say "༼".&uniname # I am a bit surprised that this works
23:02 camelia rakudo-moar 02329b: OUTPUT«TIBETAN MARK ANG KHANG GYON␤»
23:02 timotimo FROGGS: what's surprising about that?
23:02 FROGGS ahh wait
23:02 FROGGS ༼ makes the & look differently
23:03 FROGGS (and the second quote)
23:03 FROGGS ༼&
23:03 TimToady you can only use unibool on 1-bit wide props
23:03 TimToady m: say '('.&unibool('BidiMirrored')
23:03 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:03 TimToady hmm
23:03 lue TimToady: so it's a case of LTA error then :)
23:03 TimToady yeah...
23:04 TimToady m: say 'A'.&unibool('BidiMirrored')
23:04 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:04 TimToady m: say 'A'.&unibool('Common')
23:04 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:04 lue m: say '('.&unibool('Bidi_Mirrored')
23:04 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:05 lue TimToady: any particular reason you made a separate unibool apart from uniprop?
23:05 TimToady m: say 'A'.&unibool('Any')
23:05 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:06 TimToady well, uniprop is going to prefer the str output, which may be an enum
23:06 TimToady just seemed like a good thing to let it know exactly what you want on a low level
23:07 cognominal joined #perl6
23:08 lue TimToady: just asking because string and integer are combined in uniprop. (Instead of every form being separated)
23:09 TimToady we can expose those however we like
23:11 TimToady m: my $table = nqp::unipropcode("Any"); say nqp::getuniprop_bool("[".ord, $table)
23:11 camelia rakudo-moar 02329b: OUTPUT«(signal SEGV)»
23:11 TimToady well, looks like a low-level bug, but it's not obvious offhand
23:12 TimToady oh, yes it is :)
23:13 Mouq o/
23:13 Mouq TimToady++
23:14 Mouq Should the following work:
23:14 Mouq p6: say 112344 ~~ /(.)$0/
23:14 Mouq p6: say 112344 ~~ /(.)($0)/ #this one
23:14 timotimo o/ Mouq
23:15 camelia rakudo-parrot 02329b, rakudo-jvm 02329b, rakudo-moar 02329b, niecza v24-109-g48a8de3: OUTPUT«「11」␤ 0 => 「1」␤␤»
23:15 camelia niecza v24-109-g48a8de3: OUTPUT«Use of uninitialized value in string context␤  at /home/p6eval/niecza/lib/CORE.setting line 1389 (warn @ 5) ␤  at /home/p6eval/niecza/lib/CORE.setting line 269 (Mu.Str @ 15) ␤  at <unknown> line 0 (ExitRunloop @ 0) ␤  at /tmp/tmpfile line 0 (A…»
23:15 camelia ..rakudo-parrot 02329b, rakudo-jvm 02329b, rakudo-moar 02329b: OUTPUT«Nil␤»
23:15 Mouq m: say 112344 ~~ /(.){}($0)/
23:15 camelia rakudo-moar 02329b: OUTPUT«Nil␤»
23:24 lue m: say 'z'.&unival
23:24 camelia rakudo-moar 02329b: OUTPUT«NaN␤»
23:27 rjbs Postmodernism is dead.  Perl is postmodern.  Ergo...? -- http://www.prospectmagazine.co.uk/magazine/postmodernism-is-dead-va-exhibition-age-of-authenticism
23:32 Probably_Tadzik joined #perl6
23:32 Probably_Tadzik hey hey
23:32 Probably_Tadzik my feather account seems to be... inactive? Or the password has been changed?
23:33 Probably_Tadzik I can verify myself on github/email/whatever-necessary
23:36 timotimo i just found ⌀⊙⊖⊕ on my keyboard \o/
23:38 rurban joined #perl6
23:40 Mouq timotimo: It was a happy day indeed when I downloaded a keyboard mapper :)
23:41 timotimo well, i've got little stickers glued to my keys
23:41 timotimo i just never looked at all of the symbols on them
23:41 timotimo (for they are many)
23:43 Mouq timotimo: Ahh :)
23:43 timotimo neo-layout.org ← truly a beauty
23:44 lue timotimo: too bad US keyboards don't have the extra keys for German layouts :(
23:44 timotimo huh?
23:47 lue https://en.wikipedia.org/wiki/File:German-T2-Keyboard-Prototype-May-2012.jpg  vs  https://en.wikipedia.org/wiki/File:ANSI_Keyboard_Layout_Diagram_with_Form_Factor.svg
23:47 skids joined #perl6
23:48 pippo my perl6-m broke after git commit fc2a6e632efa06c8c6dc1fd1c5564d2e4ac17c03
23:48 timotimo oh, the terrible one-line-return-key
23:48 timotimo oh? the left key is extra wide, though?
23:49 kaare_ joined #perl6
23:49 lue timotimo: on my specific keyboard, the |\ key is next to a shorter backspace, and I have a larger return as a result. But otherwise, that ANSI is my keyboard.
23:49 lue Yes, extra-wide shift key on the left, only 10 normal keys on the bottom row.
23:49 timotimo keyboards are weird, yo.
23:49 timotimo anyway, i'm off to bed
23:49 timotimo gnite!
23:49 lue Also one less key on the middle row (next to caps lock)
23:49 lue ♞ timotimo o/
23:50 timotimo .o( i have not yet found that key on my keyboard )
23:50 lue s/next to/the row with/
23:51 tgt joined #perl6
23:51 BenGoldberg joined #perl6
23:51 Mouq night timotimo!
23:52 lue So I couldn't use a German keyboard layout if I wanted to; I'm missing two keys! :)
23:52 * Mouq wants to get a http://www.typematrix.com/
23:53 pippo Sigh. Good night #perl6
23:53 pippo left #perl6
23:53 xinming joined #perl6
23:56 lue Mouq: I wish these keyboard were still made: http://www.quadibloc.com/comp/images/scadet.gif
23:59 Mouq night pippo! I'm sorry, i missed that your perl6-m broke

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

Perl 6 | Reference Documentation | Rakudo