Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2016-12-19

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

All times shown according to UTC.

Time Nick Message
00:33 pyrimidine joined #perl6-dev
00:47 pyrimidine joined #perl6-dev
00:48 dalek rakudo/nom: e08f069 | (Lucas Buchala)++ | src/core/ (3 files):
00:48 dalek rakudo/nom: Fix {List,Seq,Range}.is-lazy method for type objects invocants
00:48 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e08f069fd9
00:48 dalek rakudo/nom: b8f10c7 | (Zoffix Znet)++ | src/core/ (3 files):
00:48 dalek rakudo/nom: Merge pull request #963 from lucasbuchala/is-lazy
00:48 dalek rakudo/nom:
00:48 dalek rakudo/nom: Fix {List,Seq,Range}.is-lazy method for type objects invocants
00:48 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b8f10c75e5
01:39 pyrimidi_ joined #perl6-dev
02:19 awwaiid I see that Inline::Perl5 is now autoloaded when you do :from<Perl5>. Can we expand it to work with Inline::* ?
02:26 pyrimidine joined #perl6-dev
02:32 dalek roast: a6d3e26 | MasterDuke17++ | S19-command-line-options/02-dash-n.t:
02:32 dalek roast: Add some todo'd tests for RT #129093
02:32 dalek roast:
02:32 dalek roast: FIRST/NEXT/LAST phasers are currently NYI for the '-n' (and by extension
02:32 dalek roast: '-p') command line options.
02:32 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=129093
02:32 dalek roast: review: https://github.com/perl6/roast/commit/a6d3e269d4
02:48 ilbot3 joined #perl6-dev
02:48 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today
02:54 kalkin- joined #perl6-dev
03:16 raiph joined #perl6-dev
04:15 dalek joined #perl6-dev
04:23 stmuk_ joined #perl6-dev
05:52 cognominal joined #perl6-dev
06:49 RabidGravy joined #perl6-dev
07:18 [Tux] This is Rakudo version 2016.12-12-gb8f10c75e built on MoarVM version 2016.12
07:18 [Tux] csv-ip5xs        3.147
07:18 [Tux] test            13.844
07:18 [Tux] test-t           6.205
07:18 [Tux] csv-parser      13.902
07:45 pyrimidine joined #perl6-dev
07:58 pyrimidine joined #perl6-dev
08:20 pyrimidine joined #perl6-dev
08:31 pyrimidine joined #perl6-dev
08:39 pyrimidine joined #perl6-dev
08:40 lizmat good *, #perl6-dev
08:40 yoleaux2 18 Dec 2016 22:55Z <Zoffix> lizmat: any ideas for RT#128005? The problem is unvivified item is an `Any` and Any.cache.elems returns 1, because it's a "1-item list" thing. I tried a blunt-force solution but the impact is rather pervasive and gives ~~11% slowdown: https://gist.github.com/zoffixznet/8574ee473e590d709d5e0051914249bd
08:40 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=128005
08:41 [TuxCM] joined #perl6-dev
08:56 [Tux] joined #perl6-dev
09:03 pyrimidine joined #perl6-dev
09:04 lizmat Zoffix: ah, that can of worms  :-(
09:04 lizmat looking into it now...
09:13 pyrimidine joined #perl6-dev
09:23 pyrimidine joined #perl6-dev
09:43 pyrimidine joined #perl6-dev
09:52 jnthn lizmat, Zoffix: checking for Any will likely break the moment that it's a typed array
09:53 jnthn Could use .DEFINITE
09:53 lizmat jnthn: good point
09:54 lizmat was thinking of adding Any:U candidates, actually
09:56 jnthn Yeah, that's equivalent and  very likely faster :)
10:02 lucasb joined #perl6-dev
10:05 pyrimidine joined #perl6-dev
10:13 lizmat jnthn: so when can the .count of a Block become Inf ???
10:14 jnthn Slurpy
10:16 lizmat aha...
10:17 lizmat m: dd (-> *@a { }).count  # jnthn++
10:17 camelia rakudo-moar b8f10c: OUTPUT«Inf␤»
10:35 dalek rakudo/nom: 57e733e | samcv++ | Configure.pl:
10:35 dalek rakudo/nom: Use #!/usr/bin/env perl instead of #! perl
10:35 dalek rakudo/nom:
10:35 dalek rakudo/nom: Unless there is a good reason to keep this, it would be nice
10:35 dalek rakudo/nom: if this was changed for better compatibility.
10:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/57e733e6c7
10:35 dalek rakudo/nom: 9120ace | lizmat++ | Configure.pl:
10:35 dalek rakudo/nom: Merge pull request #964 from samcv/configure.pl
10:35 dalek rakudo/nom:
10:35 dalek rakudo/nom: Use #!/usr/bin/env perl instead of #! perl
10:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9120ace3ff
10:40 lizmat m: use nqp; say nqp::isnanorinf(1)   # was hoping that would just return 0  :-(
10:40 camelia rakudo-moar b8f10c: OUTPUT«This type cannot unbox to a native number: P6opaque, Int␤  in block <unit> at <tmp> line 1␤␤»
10:41 jnthn Nope
10:41 lizmat any specific reason for that ?
10:41 lizmat the name doesn't seem to indicate to me it needs a num
10:42 jnthn nqp:: ops in general boil down to VM ops, so are pretty close to the metal
10:43 lizmat ok
10:43 jnthn If we go and stick a bunch of extra checks in at that level, then we hurt every single code path that doesn't need them
10:43 lizmat gotcha
10:43 jnthn Like, when you're in a multi candidate on the Num class and you know self is a Num
10:49 AlexDaniel joined #perl6-dev
10:49 pyrimidine joined #perl6-dev
10:52 brrt joined #perl6-dev
10:53 dalek nqp: 4f29733 | samcv++ | Configure.pl:
10:53 dalek nqp: Use the #!/usr/bin/env perl shebang for increased compatibility
10:53 dalek nqp: review: https://github.com/perl6/nqp/commit/4f29733479
10:56 lizmat jnthn: is there a chance that a WhateverCode ever gets a .block == inf ?
10:56 lizmat .count == Inf  *sigh*
10:59 jnthn I can't imagine how that could happen
10:59 * lizmat neither
11:00 jnthn I mean, syntax wise...
11:00 * lizmat fetches another opt opportunity
11:00 jnthn You'd probably have to piece one together "by hand"
11:00 timotimo optortunity
11:00 jnthn So yeah, I think we can rely on WhateverCode never having Inf count
11:14 pyrimidine joined #perl6-dev
11:25 Zoffix joined #perl6-dev
11:26 Zoffix m: say do for <a1 a2 a3 a4> { S/a/z/; } # BROKEN
11:26 camelia rakudo-moar 9120ac: OUTPUT«(z4 z4 z4 z4)␤»
11:26 Zoffix m: say do for <a1 a2 a3 a4> -> $_ is copy { s/a./z/; } # WORKS
11:26 camelia rakudo-moar 9120ac: OUTPUT«(「a4」 「a4」 「a4」 「a4」)␤»
11:26 Zoffix um
11:26 Zoffix I mean both broken
11:27 Zoffix m: $_ = "aaabbb"; dd -> $_  is raw { S/b/B/ }(  $/ = $_ ); # BROKEN
11:27 camelia rakudo-moar 9120ac: OUTPUT«Str $/ = "bB"␤»
11:27 Zoffix m: $_ = "aaabbb"; dd -> $_  is raw { S/b/B/ }(  $/ = $_ ); # BROKEN
11:27 camelia rakudo-moar 9120ac: OUTPUT«Str $/ = "bB"␤»
11:27 Zoffix m: $_ = "aaabbb"; dd -> $_  { S/b/B/ }(  $/ = $_ ); # WORKS
11:27 camelia rakudo-moar 9120ac: OUTPUT«Str $/ = "aaaBbb"␤»
11:27 Zoffix m: $_ = "aaabbb"; dd -> $_  is raw { S/b/B/ }(  $_ ); # WORKS
11:27 camelia rakudo-moar 9120ac: OUTPUT«Str $/ = "aaaBbb"␤»
11:27 Zoffix Any ideas what what could be the reason?
11:27 Zoffix Seems like both issues are related and something to do with the way $_ and $/ are handled in blocks or something?
11:27 lizmat this goes wrong somewhere in Any-iterable
11:28 lizmat will look at it after having fixed RT #128005
11:28 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=128005
11:28 lizmat currently in the maze that is src/core/array_slice.pm
11:28 Zoffix lizmat, do you mean my examples?
11:29 lizmat yeah, I think I can fix it *and* make [] a bit faster overall
11:29 Zoffix huh. OK
11:30 * Zoffix doesn't see how Any-iterable is even involved in the second ones
11:30 lizmat Zoffix: ah, then maybe my hunch is wrong or there ar 2 bugs  :-)
11:32 Zoffix I'll wait for attempted fix
11:33 Zoffix Not sure if my comments on the first one made it to the mailing list: https://rt.perl.org/Ticket/Display.html?id=130355#txn-1440031
11:33 Zoffix "The problem is the `S///` and `s///` ops use the $/ and return it. This value becomes an item in the loop's result list, but the next loop iteration updates the $/, so all the previous items in the list get updated to it as well, which is why we see all items in the output set to the last item.
11:33 Zoffix "
11:33 Zoffix And in the map's version the issue only happens when we start fetching stuff that's already in $!reified
11:34 * Zoffix goes to dig for more tickets to fix
11:34 Zoffix left #perl6-dev
11:37 lizmat joined #perl6-dev
11:38 pyrimidine joined #perl6-dev
11:44 cognominal joined #perl6-dev
11:56 lizmat grr...  had 10+ Terminal crashes in the past hour...
11:57 jnthn o.O
11:57 jnthn That sounds...pretty busted
11:57 lizmat yeah...  :-(
11:58 lizmat hope 10.12.2 update will fix...
11:58 pyrimidine joined #perl6-dev
12:00 jnthn .oO( Maybe my next upgrade should not be to an OSX machine... :) )
12:07 lizmat ah, that just goes to show how backward you are:  it's macOS nowadays  :-)
12:07 lizmat so very important difference
12:07 lizmat :-)
12:08 jnthn Next you'll be telling me they went back to using \r as the line sep... :P
12:09 lizmat lol
12:09 * jnthn wanders off to nom a bowl of dal
12:11 * lizmat wanders off for a restart
12:11 pmurias joined #perl6-dev
12:23 pyrimidine joined #perl6-dev
12:26 lizmat joined #perl6-dev
13:09 pyrimidine joined #perl6-dev
13:16 pyrimidine joined #perl6-dev
13:23 pyrimidine joined #perl6-dev
13:29 dalek rakudo/nom: 08ff3db | lizmat++ | src/core/ (3 files):
13:29 dalek rakudo/nom: Fix RT #128005 and make @a[*-1] 2x as fast
13:29 dalek rakudo/nom:
13:29 dalek rakudo/nom: - abstract the position calculation logic into (Block|WhateverCode).pos
13:29 dalek rakudo/nom:   Because we can slighltly more optimize the WhateverCode case, as that
13:29 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=128005
13:29 dalek rakudo/nom:   can never have have Inf for its .count.  And because we abstracted it
13:29 dalek rakudo/nom:   into those classes, it felt less bad to actually not use a method call
13:29 dalek rakudo/nom:   to get the parameter count, but instead go direct there using nqp.
13:29 dalek rakudo/nom: - Simplify the postcircumfix:<[ ]> candidates accordingly.
13:29 dalek rakudo/nom:
13:29 dalek rakudo/nom: Zoffix++ for the nudge.
13:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/08ff3dbfac
13:34 dalek roast: 1c353d7 | lizmat++ | S02-types/array.t:
13:34 dalek roast: Unfudge now passing test
13:34 dalek roast: review: https://github.com/perl6/roast/commit/1c353d7a20
13:39 dalek rakudo/nom: 2776164 | (Zoffix Znet)++ | tools/contributors.pl6:
13:39 dalek rakudo/nom: Be smarter when figuring out date of last release
13:39 dalek rakudo/nom:
13:39 dalek rakudo/nom: Current algo only works when script is run in the month
13:39 dalek rakudo/nom: following the month of last release.
13:39 dalek rakudo/nom:
13:39 dalek rakudo/nom: Improve by using the date of last commit to VERSION file.
13:39 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/27761645c5
13:42 pyrimidine joined #perl6-dev
13:48 dalek rakudo/nom: cd8ab2e | (Zoffix Znet)++ | tools/contributors.pl6:
13:48 dalek rakudo/nom: s/shell()/run()/
13:48 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cd8ab2e23f
13:59 dalek rakudo/nom: b39c0d8 | lizmat++ | src/core/array_slice.pm:
13:59 dalek rakudo/nom: Squeeze another 3% out of @a[*-1]
13:59 dalek rakudo/nom:
13:59 dalek rakudo/nom: Normally I wouldn't bother, but for such an essential capability, it
13:59 dalek rakudo/nom: seems worth it.
13:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b39c0d8b98
14:17 dalek rakudo/nom: ab26b58 | lizmat++ | src/core/ (5 files):
14:17 dalek rakudo/nom: Remove unnecessary caching on .elems
14:17 dalek rakudo/nom:
14:17 dalek rakudo/nom: Only a Seq needs caching of .elems nowadays, and it does so internally.
14:17 dalek rakudo/nom:
14:17 dalek rakudo/nom: Makes @a[*-1] another 10% or so faster and probably has speed advantages
14:17 dalek rakudo/nom: in other situations as well.
14:17 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/ab26b5896b
14:17 * lizmat hopes for a noticeable improvement on test-t
14:18 lizmat [Tux]: ^^^  :-)
14:19 dalek nqp: 1400616 | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js:
14:19 dalek nqp: [js] Fix signature.
14:19 dalek nqp: review: https://github.com/perl6/nqp/commit/14006161fd
14:19 dalek nqp: 22a147a | (Pawel Murias)++ | src/vm/js/ (5 files):
14:19 dalek nqp: [js] Repossession fixes.
14:19 dalek nqp: review: https://github.com/perl6/nqp/commit/22a147a956
14:30 dalek rakudo/nom: 47f5300 | (Zoffix Znet)++ | LICENSE:
14:30 dalek rakudo/nom: Revert copyright year in LICENSE to original
14:30 dalek rakudo/nom:
14:30 dalek rakudo/nom: The original license[^1] lists 2006 as the end year and the first sentence
14:30 dalek rakudo/nom: in the license documents reads "...changing [this document] is not allowed"
14:30 dalek rakudo/nom:
14:30 dalek rakudo/nom: [1] https://opensource.org/licenses/Artistic-2.0
14:30 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/47f5300f41
14:35 perlpilot joined #perl6-dev
14:36 dalek rakudo/nom: bbeec04 | (Zoffix Znet)++ | / (3 files):
14:36 dalek rakudo/nom: ❄️☃ Happy New Year!
14:36 dalek rakudo/nom:
14:36 dalek rakudo/nom: Update copyrights and stuff
14:36 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/bbeec045c1
14:38 pyrimidine joined #perl6-dev
14:49 dalek rakudo/nom: 454dd45 | zoffixznet++ | CREDITS:
14:49 dalek rakudo/nom: Add Zoffix Znet to CREDITS
14:49 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/454dd45175
14:52 dalek rakudo/nom: 60a310b | zoffixznet++ | tools/contributors.pl6:
14:52 dalek rakudo/nom: Remove unused option
14:52 dalek rakudo/nom:
14:52 dalek rakudo/nom: Sneaked in in the previous commit
14:52 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/60a310bc8f
14:55 dalek rakudo/nom: 1b8c89d | (Zoffix Znet)++ | CREDITS:
14:55 dalek rakudo/nom: Remove U: for Zoffix Znet
14:55 dalek rakudo/nom:
14:55 dalek rakudo/nom: I don't like how dalek uses it for some reason....
14:55 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1b8c89d947
15:01 dalek rakudo/nom: e64dd9a | (Zoffix Znet)++ | tools/contributors.pl6:
15:01 dalek rakudo/nom: Load CREDITS files from all repos
15:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e64dd9a411
15:09 dalek rakudo/nom: 8219acf | (Zoffix Znet)++ | tools/contributors.pl6:
15:09 dalek rakudo/nom: Add protection against publishing email addresses
15:09 dalek rakudo/nom:
15:09 dalek rakudo/nom: It's possible (like happened last release) for a contributor to
15:09 dalek rakudo/nom: end up being an email address. Even though the address is public,
15:09 dalek rakudo/nom: some may not wish for us to plaster it all over the place.
15:09 dalek rakudo/nom:
15:09 dalek rakudo/nom: Fix by redacting the hostname of the email address if one shows up.
15:09 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8219acf631
15:11 pyrimidine joined #perl6-dev
15:14 cognominal joined #perl6-dev
15:25 |Tux| This is Rakudo version 2016.12-19-This is Rakudo version 2016.12-19-gab26b5896 built on MoarVM version 2016.12
15:25 |Tux| csv-ip5xs        3.072
15:25 |Tux| test            13.535
15:25 |Tux| test-t           6.118
15:25 |Tux| csv-parser      13.757
15:25 |Tux| gab26b5896 built on MoarVM version 2016.12
15:25 |Tux| csv-ip5xs        3.072
15:25 |Tux| test            13.535
15:25 |Tux| test-t           6.118
15:25 |Tux| csv-parser      13.757
15:25 |Tux| lizmat, ^^
15:30 pyrimidine joined #perl6-dev
16:01 dalek rakudo/nom: 0010f15 | (Zoffix Znet)++ | docs/release_guide_automated.md:
16:01 dalek rakudo/nom: Update Automated Release Guide
16:01 dalek rakudo/nom:
16:01 dalek rakudo/nom: - Remove pieces that were planned to be automated but weren't
16:01 dalek rakudo/nom:     (some stuff like Wikipedia update and Email sending are just
16:01 dalek rakudo/nom:     not worth the code/debugging time)
16:01 dalek rakudo/nom: - Document the actual working process instead of imagined steps
16:01 dalek rakudo/nom: - Document stuff that's still work in progress
16:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0010f15ffa
16:04 dalek rakudo/nom: aa7e20e | (Zoffix Znet)++ | docs/release_guide_automated.md:
16:04 dalek rakudo/nom: Fix broken markdowhn format
16:04 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/aa7e20ea15
16:10 pyrimidine joined #perl6-dev
17:11 TimToady m: say (** + 1).count
17:11 camelia rakudo-moar aa7e20: OUTPUT«Inf␤»
17:11 TimToady m: say (** + 1).(1,2,3)
17:11 camelia rakudo-moar aa7e20: OUTPUT«(2 3 4)␤»
17:12 TimToady m: say (** + 1).WHAT
17:12 camelia rakudo-moar aa7e20: OUTPUT«(Sub)␤»
17:12 TimToady except somehow it's not a WhateverCode
17:13 TimToady maybe it should be a HyperWhateverCode...
17:13 jnthn o.O
17:13 jnthn I didn't even though about that :P
17:13 jnthn When did that go in?!
17:13 TimToady been in the spec for years
17:13 jnthn Sure, I meant into Rakudo :)
17:14 * TimToady didn't do it, but more than a year
17:14 jnthn m: say (**.abs).(-1, 3, -5)
17:14 camelia rakudo-moar aa7e20: OUTPUT«(1 3 5)␤»
17:14 jnthn wow
17:14 jnthn I coulda used that a bunch of times instead of writing map :P
17:15 timotimo %)
17:15 timotimo i have had trouble understanding how hyperwhatever does its thing
17:16 TimToady certainly we've had people come to grief for donkey's years trying to write *** to mean * * *  :)
17:17 [Coke] I don't feel so bad when jnthn is surprised to find things in Rakudo.
17:17 TimToady m: say (** * *)([1,2,3],4)
17:17 camelia rakudo-moar aa7e20: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Multiple HyperWhatevers and Whatevers may not be used together␤at <tmp>:1␤------> say (** * *⏏)([1,2,3],4)␤»
17:18 TimToady aww
17:18 jnthn Patches welcome :P
17:18 jnthn But wait, if it slurps...
17:18 jnthn That's gonna be a problem :)
17:19 TimToady presumably it would slurp an unpack instead
17:20 TimToady m: say (** Z+ **)([1,2,3],[4,5,6])
17:20 camelia rakudo-moar aa7e20: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Multiple HyperWhatevers and Whatevers may not be used together␤at <tmp>:1␤------> say (** Z+ **⏏)([1,2,3],[4,5,6])␤»
17:20 TimToady m: say (* Z+ *)([1,2,3],[4,5,6])
17:20 camelia rakudo-moar aa7e20: OUTPUT«(5 7 9)␤»
17:20 TimToady well, but then why not just that...
17:34 FROGGS joined #perl6-dev
17:36 pyrimidine joined #perl6-dev
17:36 ZoffixW joined #perl6-dev
17:37 ZoffixW Curious, what's the usecase for undefined Numerics to be treated as zero, after issuing a warning?
17:37 ZoffixW Seems like cruft left over from Perl 5.
17:38 ZoffixW That poorly fits into the model with types in it
17:38 ZoffixW m: quietly say (+Rat).^name;
17:38 camelia rakudo-moar aa7e20: OUTPUT«Int␤»
17:38 ZoffixW m: say Rat + i
17:38 camelia rakudo-moar aa7e20: OUTPUT«Cannot look up attributes in a Rat type object␤  in block <unit> at <tmp> line 1␤␤»
17:39 ZoffixW m: Int == 42
17:39 camelia rakudo-moar aa7e20: OUTPUT«WARNINGS for <tmp>:␤Useless use of "==" in expression "Int == 42" in sink context (line 1)␤Invocant requires an instance of type Int, but a type object was passed.  Did you forget a .new?␤  in block <unit> at <tmp> line 1␤␤»
17:39 ZoffixW m: quietly say Int < 2
17:39 camelia rakudo-moar aa7e20: OUTPUT«Invocant requires an instance of type Int, but a type object was passed.  Did you forget a .new?␤  in block <unit> at <tmp> line 1␤␤»
17:39 * ZoffixW is fixing all of that ATM
17:39 ZoffixW And hence the question came up.... why the hell do we even let undefined numerics to be used?
17:44 ZoffixW oh gee....
17:44 * ZoffixW sees a Rat hole at the end of the rabbit hole
17:46 ZoffixW I'm just gonna leave it as is. All the Rat ops would need additions of :U candidates.... that's on top of these changes http://temp.perl6.party/unum.diff.txt
17:46 ZoffixW ... On top of adding .Numeric and .Real candidates to each type, so all :Us don't numify to Int
17:48 ZoffixW m: quietly say Complex + 42
17:48 camelia rakudo-moar aa7e20: OUTPUT«42␤»
17:48 ZoffixW ... On top of adding stuff to Complex too
17:50 ZoffixW In summation... our :U-numerics-are-zeros concept is quite broken, which makes its utility dubious and I wouldn't shed a tear if we dumped it in 6.d and made all things expect :D numerics
17:50 ZoffixW :)
17:51 ZoffixW Tip your waitresses!
17:51 ZoffixW and waiters
17:51 ZoffixW left #perl6-dev
17:54 ZoffixW joined #perl6-dev
17:54 ZoffixW Forgot to mention. The ** + 1 is a Sub, because it converts into a WhateverCode * + 1 that's given to HYPERWHATEVER(&c) that returns  sub (*@_) { map &c, @_ }
17:55 ZoffixW m: HyperWhatever.new.^name.say; # magic
17:55 camelia rakudo-moar aa7e20: OUTPUT«Sub␤»
17:55 ZoffixW m: (new HyperWhatever:).^name.say; # magic
17:55 camelia rakudo-moar aa7e20: OUTPUT«Cannot make a HyperWhatever object using .new␤  in block <unit> at <tmp> line 1␤␤»
17:55 ZoffixW ^_^
17:56 cognominal joined #perl6-dev
17:59 stmuk joined #perl6-dev
18:09 pyrimidine joined #perl6-dev
18:22 * lizmat is back
18:22 lizmat Files=1155, Tests=53960, 197 wallclock secs (11.22 usr  4.60 sys + 1187.29 cusr 125.55 csys = 1328.66 CPU)
18:22 lizmat [Tux]: thanks
18:29 lizmat joined #perl6-dev
18:32 pyrimidine joined #perl6-dev
18:37 lizmat afk again&
18:56 pyrimidine joined #perl6-dev
18:58 gfldex do we have a list of changes to the language since v6.c?
19:00 notviki joined #perl6-dev
19:01 notviki gfldex: I'd say it's not wrong to say that the number of changes to the *language* has been exactly zero
19:01 notviki huggable: 6.d
19:01 huggable notviki, Proposals for 6.d language: https://github.com/perl6/specs/blob/master/v6d.pod
19:02 notviki And that's the list of proposals for 6.d changes
19:02 gfldex i believe to remember that we had some errata
19:03 notviki Ah, that should be this: https://github.com/perl6/roast/compare/6.c...6.c-errata
19:03 notviki Seems most of those are just adjusting tests than about a change in the language
19:09 [Coke] yes, otherwise they couldn't safely go into something claiming to be 6.c
19:14 pyrimidine joined #perl6-dev
19:15 go|dfish joined #perl6-dev
19:31 pyrimidine joined #perl6-dev
19:35 dalek nqp: fb18d29 | (Pawel Murias)++ | src/vm/js/Compiler.nqp:
19:35 dalek nqp: [js] Fix the default value for variables.
19:35 dalek nqp: review: https://github.com/perl6/nqp/commit/fb18d293ab
19:38 dalek rakudo/nom: 3fe5893 | lizmat++ | src/core/ (2 files):
19:38 dalek rakudo/nom: Fix for RT #130363
19:38 dalek rakudo/nom:
19:38 dalek rakudo/nom: This breaks one spectest, which is faulty because it expects
19:38 dalek rakudo/nom: Any ~~ m/%hash/ to fail.  It did so before, but for the wrong
19:38 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=130363
19:38 dalek rakudo/nom: reason.
19:38 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3fe589309d
19:40 huggable joined #perl6-dev
19:41 NeuralAnomaly joined #perl6-dev
19:42 NeuralAnomaly joined #perl6-dev
19:43 huggable joined #perl6-dev
19:44 buggable joined #perl6-dev
19:45 yoleaux2 joined #perl6-dev
19:47 SourceBaby joined #perl6-dev
19:48 Undercover joined #perl6-dev
19:49 pyrimidine joined #perl6-dev
19:50 huggable joined #perl6-dev
19:56 notviki m: my %hashl Any ~~ m/%hash/
19:56 camelia rakudo-moar 3fe589: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> my %hashl⏏ Any ~~ m/%hash/␤    expecting any of:␤        infix␤        infix stopper␤        statement end␤        statement modifier␤        st…»
19:56 notviki m: my %hash; Any ~~ m/%hash/
19:56 camelia rakudo-moar 3fe589: ( no output )
19:56 notviki star: my %hash; Any ~~ m/%hash/
19:56 camelia star-m 2016.10: OUTPUT«No such method 'match' for invocant of type 'Any'␤  in block <unit> at <tmp> line 1␤␤»
19:59 notviki star: multi foo ($ where m:i/foo/) { say "got it" }; foo class { method Str { "foobar" } }.new
19:59 camelia star-m 2016.10: OUTPUT«No such method 'match' for invocant of type '<anon|68186352>'␤  in sub foo at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
19:59 notviki star: multi foo ($ where /foo/) { say "got it" }; foo class { method Str { "foobar" } }.new
19:59 camelia star-m 2016.10: OUTPUT«got it␤»
19:59 notviki m: multi foo ($ where m:i/foo/) { say "got it" }; foo class { method Str { "foobar" } }.new
19:59 camelia rakudo-moar 3fe589: OUTPUT«Cannot resolve caller match(<anon|57744400>: Regex, :i(Int)); none of these signatures match:␤    (Any:U $: | is raw)␤  in sub foo at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
20:01 * notviki wonders if it'd be sane for Any:D candidate to forward to self.Str.match
20:05 notviki I also wonder what's the differnece between /foo/ and m/foo/ that makes one fail there but not the other :/
20:06 pyrimidine joined #perl6-dev
20:08 lizmat nine: shouldn't we merge the lexical module loading branch now ?
20:08 pmurias [Coke]: re fudging, some of it in roast is for bugs that where in rakudo at the time of the 6.c release
20:09 lizmat notviki: /./ codegens differently bypassing Str.match altogether, since it can be sure no adverbs are specified
20:10 notviki Ah. Thanks.
20:10 lizmat m/./ apparently follows a code path that assumes named parameters *are* specified
20:10 lizmat because that's the only way you can e.g. say: m:2nd/./
20:17 pyrimidine joined #perl6-dev
20:17 AlexDaniel joined #perl6-dev
20:20 [Coke] pmurias: sure, some of it. some is 10-year old leftovers.
20:20 [Coke] really needs a clean sweep before 6.d IMO
20:28 pyrimidine joined #perl6-dev
20:29 lucasb m: Any ~~ m/x/
20:29 camelia rakudo-moar 3fe589: ( no output )
20:30 lucasb sorry, I don't feel very confortable with this
20:30 lizmat lucasb: what's wrong ?
20:30 lizmat m: Any ~~ /x/  # worked like forever ?
20:30 camelia rakudo-moar 3fe589: OUTPUT«Use of uninitialized value of type Any in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.␤  in block <unit> at <tmp> line 1␤»
20:30 lizmat ah, you mean no warning ?
20:30 lizmat good point
20:30 lucasb yes, that :)
20:31 lucasb not a warning, but an error
20:31 notviki m: quietly m: Any ~~ /x/; say "not an error"
20:31 camelia rakudo-moar 3fe589: OUTPUT«not an error␤»
20:31 lizmat well, /x/ has always been a warning
20:32 lucasb star-m: Any ~~ m/x/; say 42
20:32 camelia star-m 2016.10: OUTPUT«No such method 'match' for invocant of type 'Any'␤  in block <unit> at <tmp> line 1␤␤»
20:33 lizmat star-m: Any ~~ /x/; say 42
20:33 camelia star-m 2016.10: OUTPUT«Use of uninitialized value of type Any in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.␤  in block <unit> at <tmp> line 1␤42␤»
20:34 lucasb IMO, the same way I expect a error if I do Int+10 or Str~'a', I would also expect with (SomethingUndefined).match
20:35 travis-ci joined #perl6-dev
20:35 travis-ci Rakudo build failed. Zoffix Znet 'Fix broken markdowhn format'
20:35 travis-ci https://travis-ci.org/rakudo/rakudo/builds/185192183 https://github.com/rakudo/rakudo/compare/0010f15ffa25...aa7e20ea15ab
20:35 travis-ci left #perl6-dev
20:35 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
20:35 lucasb oops, no. the examples I said are not errors, just warnings :)
20:36 lucasb more acurate, Int+1 is an error, but Str~'x' is a warning
20:38 lizmat yeah, testing a fix
20:38 notviki travis: t/04-nativecall/08-callbacks.t ........... Failed 5/8 subtests
20:40 notviki m: (my Int $).match: /^/
20:40 camelia rakudo-moar 3fe589: ( no output )
20:40 notviki star: (my Int $).match: /^/
20:40 camelia star-m 2016.10: OUTPUT«Cannot resolve caller match(Int: Regex); none of these signatures match:␤    (Cool:D $: $target, *%adverbs)␤  in block <unit> at <tmp> line 1␤␤»
20:42 dalek rakudo/nom: cc0f836 | lizmat++ | src/core/Any.pm:
20:42 dalek rakudo/nom: Make sure we warn, just like /./ would do, lucasb++
20:42 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cc0f836e4a
20:59 nine lizmat: yes, we should. Just came home from teaching radio procedures at the airfield. Will start now
21:00 lizmat roger, over!
21:01 lucasb lizmat++ thanks
21:03 pyrimidine joined #perl6-dev
21:04 camelia joined #perl6-dev
21:09 pyrimidine joined #perl6-dev
21:29 evalable6 joined #perl6-dev
21:29 statisfiable6 joined #perl6-dev
21:32 nine m: my %var; m/%var/
21:32 camelia rakudo-moar cc0f83: OUTPUT«Use of uninitialized value of type Any in string context.␤Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.␤  in block <unit> at <tmp> line 1␤»
21:33 nine We seem to have a spec test that expects this to die?
21:33 nine Ah, found it in the backlog
21:34 dalek rakudo/nom: 4b529c8 | niner++ | / (5 files):
21:34 dalek rakudo/nom: Make importing globals from loaded modules lexical
21:34 dalek rakudo/nom:
21:34 dalek rakudo/nom: Right now if A depends on C:auth<foo> and B depends on C:auth<bar> you
21:34 dalek rakudo/nom: cannot use A and B in the same program. Fix this by having not only
21:34 nine It is done!
21:34 nine mst: ^^^
21:35 lizmat nine++  :-)
21:35 dalek joined #perl6-dev
21:36 pyrimidine joined #perl6-dev
21:39 mst nine: MUHAHAHAHAHAHAHAHAHAHAHA
21:39 RabidGravy so it would be oppoortune to re-test all my modules against that now then right?
21:39 nine RabidGravy: yes :)
21:40 mst nine: ... ... ... ... *GRINS*
21:40 mst you realise that buys limited hot reloading, right?
21:41 RabidGravy it also means you can use HTTP::Server::Async and HTTP::UserAgent in the same program now ;-)
21:41 mst psch
21:41 mst who'd ever want to do that
21:42 RabidGravy I'd fixed both of them a while back so you could have them *installed* at the same time
21:43 lizmat and another Perl 6 Weekly hits the Net: https://p6weekly.wordpress.com/2016/12/19/2016-51-flowing-towards-christmas/
21:45 [Coke] lizmat++
21:47 jdv79 nice
21:48 lizmat notviki: just realized that the last paragraph on http://rakudo.org/2016/12/17/lexical-module-loading/ is technically correct, but would read better if it would say "Undeclared name: Bar" .....  add a: "use Bar;"  in light of the example code just above it
21:51 travis-ci joined #perl6-dev
21:51 travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make sure we warn, just like /./ would do, lucasb++'
21:51 travis-ci https://travis-ci.org/rakudo/rakudo/builds/185270105 https://github.com/rakudo/rakudo/compare/3fe589309da1...cc0f836e4ad0
21:51 travis-ci left #perl6-dev
21:53 pyrimidine joined #perl6-dev
22:10 dalek roast: 8dc725a | lizmat++ | S05-interpolation/regex-in-variable.t:
22:10 dalek roast: Fudge tests for /%hash/, we don't check for that yet
22:10 dalek roast: review: https://github.com/perl6/roast/commit/8dc725acac
22:13 dalek roast/6.c-errata: 36eaf30 | lizmat++ | S05-interpolation/regex-in-variable.t:
22:13 dalek roast/6.c-errata: Fudge test for /%hash/, we don't check for that yet
22:13 dalek roast/6.c-errata: review: https://github.com/perl6/roast/commit/36eaf301b3
22:13 lizmat and that concludes my hacking / blogging for the day
22:13 lizmat good night, #perl6-dev!
22:15 pyrimidine joined #perl6-dev
22:26 notviki lizmat: fixed thanks.
22:26 jnthn lizmat++ # weekly
22:45 pyrimidine joined #perl6-dev
23:07 stmuk_ joined #perl6-dev
23:10 pyrimidine joined #perl6-dev
23:51 pyrimidine joined #perl6-dev
23:55 pyrimidi_ joined #perl6-dev

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