Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2017-10-14

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

All times shown according to UTC.

Time Nick Message
00:33 lizmat joined #perl6-dev
00:54 ugexe use local CURIs as a faux metacpan microservice: https://gist.github.com/ugexe/10a1bf67f8849323bc855408e3577b2a
00:57 ugexe idea is to eventually act as both a web front-end to zef/cur as well as easily exposing it as a microservice to act as a darkpan
01:57 ilbot3 joined #perl6-dev
01:57 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
02:03 MasterDuke joined #perl6-dev
02:18 AlexDaniel- “35768 snippets tested, 10524 are known to be unstable”
02:18 AlexDaniel- wait what
02:18 AlexDaniel- no way I'm going through 10524 snippets
02:20 AlexDaniel- :S
02:36 AlexDaniel- so can we finally open rakudo/issues? I don't think I can take it anymore
02:40 AlexDaniel- .tell Zoffix Have you seen RT #132280 ? It looks like something you might enjoy working on (judging by the recent-ish optimizer ticket you resolved)
02:40 yoleaux AlexDaniel-: I'll pass your message to Zoffix.
02:40 synopsebot RT#132280 [new]: https://rt.perl.org/Ticket/Display.html?id=132280 [REGRESSION][PERF] Cool methods that take and pass a Capture are uber slower (.contains: 'meow' for ^10000_00)
02:44 AlexDaniel- .tell lizmat fwiw please feel freer to reject my latest regression tickets
02:44 yoleaux AlexDaniel-: I'll pass your message to lizmat.
02:53 AlexDaniel- this is probably one of the funniest snippets I've seen
02:53 AlexDaniel- e: %*ENV<RAKUDO_NO_DEPRECATIONS>:delete; DEPRECATED("the garbage bin")
02:53 evalable6 AlexDaniel-, rakudo-moar 012c80f33: OUTPUT: «Saw 1 occurrence of deprecated code.␤=====================================================…»
02:53 evalable6 AlexDaniel-, Full output: https://gist.github.com/98bf0ca4675e32ef9883bf239a1f4544
02:53 AlexDaniel- “Please use the garbage bin instead.”
03:13 evalable6 joined #perl6-dev
03:14 AlexDaniel- .tell MasterDuke FWIW RT #132291
03:14 yoleaux AlexDaniel-: I'll pass your message to MasterDuke.
03:14 synopsebot RT#132291 [new]: https://rt.perl.org/Ticket/Display.html?id=132291 [LTA] Too few positionals passed to '' (for [:a] X [:b] -> ($i, $j) { })
03:26 AlexDaniel- “No regressions in stresstest from this, which is a promising sign.” https://github.com/rakudo/rakudo/commit/4a8038c2956e863bc661a2a00e8371eb98002608
03:27 AlexDaniel- how is it called when bisectable points at some commit? “unpromising sign”? :)
03:33 evalable6 joined #perl6-dev
03:44 AlexDaniel- .tell fwiw RT #132292. IMO it's blocker-worthy, but not marking it as such just yet. What do you think?
03:44 yoleaux AlexDaniel-: I'll pass your message to fwiw.
03:44 synopsebot RT#132292 [new]: https://rt.perl.org/Ticket/Display.html?id=132292 [REGRESSION] Recursively .emit-ing from the tap of the same supply bails out
03:44 AlexDaniel- /o\
03:44 AlexDaniel- .tell jnthn fwiw RT #132292. IMO it's blocker-worthy, but not marking it as such just yet. What do you think?
03:44 yoleaux AlexDaniel-: I'll pass your message to jnthn.
04:36 AlexDaniel- .tell lizmat does this look alright? Let me know. c: 21b9a720c7^,21b9a720c7 dd Set.Mix
04:36 yoleaux AlexDaniel-: I'll pass your message to lizmat.
04:45 AlexDaniel- ouch
04:46 AlexDaniel- ah, nvm
05:34 lizmat joined #perl6-dev
06:08 AlexDaniel- .tell Zoffix fwiw RT #132293
06:08 yoleaux AlexDaniel-: I'll pass your message to Zoffix.
06:08 synopsebot RT#132293 [new]: https://rt.perl.org/Ticket/Display.html?id=132293 [REGRESSION] It is no longer possible to create native ints with .new (int32.new(42))
06:15 AlexDaniel- m: 0xD4DB.uniname.say
06:15 camelia rakudo-moar 012c80f33: OUTPUT: «Hangul Syllable PWILH␤»
06:16 AlexDaniel- .tell samcv hey, is it correct that uninames of Hangul syllables have no <…> around them? c: 2017.08,HEAD 0xD4DB.uniname.say
06:16 yoleaux AlexDaniel-: I'll pass your message to samcv.
06:16 samcv AlexDaniel-, yes
06:16 yoleaux 06:16Z <AlexDaniel-> samcv: hey, is it correct that uninames of Hangul syllables have no <…> around them? c: 2017.08,HEAD 0xD4DB.uniname.say
06:16 AlexDaniel- OK, thanks
06:16 samcv the previous name was a placeholder name, and they're generated at database generation time
06:17 samcv since it's algorithmic the name data
06:18 samcv technically we could do that, but there aren't THAT many hangul characters. so i just added it easily. though if we wanted to save a tiny bit of space we could do the algorithmic aproach. but then i was concerned about then having to add extra logic to distinguish that codepoint range and would have been much more work
06:19 AlexDaniel- samcv: but I was thiking, don't we use <> for all autogenerated stuff that's in a way unofficial?
06:20 samcv no it's official
06:20 AlexDaniel- ah, okay
06:20 samcv the < > are for ones that have shared names
06:20 AlexDaniel- hmmmmm
06:20 samcv as in more than one cp with the same designation
06:20 AlexDaniel- m: say 0x10FFFF.uniname
06:20 camelia rakudo-moar 012c80f33: OUTPUT: «<noncharacter-10FFFF>␤»
06:20 AlexDaniel- but that's unique now, isn't it?
06:20 samcv well yes
06:21 samcv that's how unicode designates it with those numbers
06:21 samcv but i meant before the hex digits-
06:21 samcv the < > designates it doesn't have an assigned name
06:21 samcv or at least we are still using the placeholder names in UnicodeData.txt. so like for hangul, even though it shows <Hangul Syllabl> there. they have always had names. just not listed in that file
06:22 AlexDaniel- hmm
06:22 samcv but the names are determined by essentially the same algorithm that's used to decompose hangul's
06:22 samcv so we follow that spec for hangul
06:23 samcv does that make some sense?
06:23 AlexDaniel- yes, thank you very much
06:23 samcv ok :)
06:23 AlexDaniel- samcv: any plans for other grant applications? :)
06:24 AlexDaniel- I'm just so happy with your work :)
06:24 samcv i need to decide what it will be though!
06:24 samcv i am very happy with my work too :)
06:24 AlexDaniel- like, I'm going through tons of snippets that have different output, and not all of them regressions of course, most are actually bug fixes
06:24 AlexDaniel- and unicode related stuff comes up *very* often
06:25 AlexDaniel- something about looking at what we had previously and what we have now makes me think that another grant is a very good idea :)
06:31 AlexDaniel- uhhh
06:31 AlexDaniel- c: 79b8ab9d3f^,79b8ab9d3f say $*KERNEL.signal: SIGUSR1
06:31 committable6 AlexDaniel-, ¦79b8ab9d3f^: «10» ¦79b8ab9: «30»
06:31 llfourn joined #perl6-dev
06:31 AlexDaniel- c: 79b8ab9d3f^,79b8ab9d3f say +$*KERNEL.signal(SIGUSR1)
06:31 committable6 AlexDaniel-, ¦79b8ab9d3f^: «10» ¦79b8ab9: «30»
06:35 AlexDaniel- .tell Zoffix maybe you can also take a look at https://rt.perl.org/Ticket/Display.html?id=132012#txn-1499238
06:35 yoleaux AlexDaniel-: I'll pass your message to Zoffix.
06:36 robertle joined #perl6-dev
06:38 AlexDaniel- w… what
06:39 AlexDaniel- c: 2015.12 my int @x[-2**63]; say @x.shape
06:39 committable6 AlexDaniel-, ¦2015.12: «(-9223372036854775808)»
06:39 AlexDaniel- c: 2016.06 my int @x[-2**63]; say @x.shape
06:39 committable6 AlexDaniel-, ¦2016.06: «Illegal dimension in shape: -9223372036854775808. All dimensions must be integers bigger than 0␤  in block <unit> at /tmp/I1tG8GtUgl line 1␤ «exit code = 1»»
06:39 AlexDaniel- so it was fixed? but
06:39 AlexDaniel- c: HEAD my int @x[-2**63]; say @x.shape
06:39 committable6 AlexDaniel-, ¦HEAD(012c80f): «(-9223372036854775808)»
06:39 AlexDaniel- (屮ಠ益ಠ)屮
06:41 AlexDaniel- arrrrrrrrggggh
06:43 AlexDaniel- .tell lizmat fwiw RT #131966
06:43 yoleaux AlexDaniel-: I'll pass your message to lizmat.
06:43 synopsebot RT#131966 [new]: https://rt.perl.org/Ticket/Display.html?id=131966 [REGRESSION][9999] Shaped native arrays with negative sizes (my int @a[-9999999999999999])
06:52 AlexDaniel- c: c3b044df169^,c3b044df169 my $a = rx/aaaaaab/; 'a' x 999999 ~ 'b' ~~ /$a/; say now - INIT now
06:52 AlexDaniel- errr, what do I do with this
06:52 committable6 AlexDaniel-, ¦c3b044df169^: «3.29355215» ¦c3b044d: «4.84434404»
06:53 wander joined #perl6-dev
06:53 AlexDaniel- MasterDuke is maybe interested in INTERPOLATE stuff
06:53 AlexDaniel- hmm
06:56 AlexDaniel- .tell MasterDuke hey, maybe you can take a look at RT #132294
06:56 yoleaux AlexDaniel-: I'll pass your message to MasterDuke.
06:56 synopsebot RT#132294 [new]: https://rt.perl.org/Ticket/Display.html?id=132294 [REGRESSION][PERF] INTERPOLATE-related change made matching of large strings slower
06:59 AlexDaniel- mch: say (await (^10000).map({ start { $*THREAD.id } })).unique
06:59 committable6 AlexDaniel-, ¦2015.12: «(3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18)» ¦HEAD(012c80f): «(4)»
07:01 AlexDaniel- is it supposed to be this way?
07:02 AlexDaniel- it is!
07:02 AlexDaniel- mch: say (await (^10).map({ start { sleep 0.3; $*THREAD.id } })).unique
07:02 committable6 AlexDaniel-, ¦2015.12: «(3 4 5 6 7 8 9 10 11 12)» ¦HEAD(012c80f): «(4 6 7 8 9 10 11 12 13 14)»
07:02 AlexDaniel- ok nvm
07:10 [TuxCM] joined #perl6-dev
07:54 AlexDaniel- .tell Zoffix maybe also RT #132295 although it is low priority stuff
07:54 yoleaux AlexDaniel-: I'll pass your message to Zoffix.
07:54 synopsebot RT#132295 [new]: https://rt.perl.org/Ticket/Display.html?id=132295 [REGRESSION][LTA] :2(…) radix syntax no longer suggests using .base (:2(42))
08:07 AlexDaniel- c: 2017.06,HEAD my $v = "\c[ZWJ]" x 100; say $v.chars
08:07 committable6 AlexDaniel-, ¦2017.06: «100» ¦HEAD(012c80f): «1»
08:07 AlexDaniel- samcv: what about this? ↑
08:07 AlexDaniel- is it correct that you can chain ZWJ with itself?
09:03 Geth_ ¦ rakudo/nom: 7 commits pushed by (Daniel Green)++, lizmat++
09:03 Geth_ ¦ rakudo/nom: 1761540e96 | Break apart INTERPOLATE into obvious multis
09:03 Geth_ ¦ rakudo/nom: 0a68a18f16 | Break out an Iterable multi for INTERPOLATE
09:03 Geth_ ¦ rakudo/nom: d73d500bf4 | Consolidate $i and $m flags to INTERPOLATE
09:03 Geth_ ¦ rakudo/nom: 1775259a32 | Break out Associatve and Regex into their own...
09:03 Geth_ ¦ rakudo/nom: e8003c873e | Remove unused conditional and variable in some...
09:03 Geth_ ¦ rakudo/nom: 4d3ccd8349 | Comment about how the #?if !moar branch in the...
09:03 Geth_ ¦ rakudo/nom: d56151743c | Merge pull request #1192 from MasterDuke17/convert_INTERPOLATE_into_a_multi
09:03 Geth_ ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/compare/012c80f336...d56151743c
09:31 timotimo ugexe: that microservice looks cool!
10:00 Zoffix .
10:00 yoleaux 02:40Z <AlexDaniel-> Zoffix: Have you seen RT #132280 ? It looks like something you might enjoy working on (judging by the recent-ish optimizer ticket you resolved)
10:00 yoleaux 06:08Z <AlexDaniel-> Zoffix: fwiw RT #132293
10:00 yoleaux 06:35Z <AlexDaniel-> Zoffix: maybe you can also take a look at https://rt.perl.org/Ticket/Display.html?id=132012#txn-1499238
10:00 yoleaux 07:54Z <AlexDaniel-> Zoffix: maybe also RT #132295 although it is low priority stuff
10:00 synopsebot RT#132280 [new]: https://rt.perl.org/Ticket/Display.html?id=132280 [REGRESSION][PERF] Cool methods that take and pass a Capture are uber slower (.contains: 'meow' for ^10000_00)
10:00 synopsebot RT#132293 [new]: https://rt.perl.org/Ticket/Display.html?id=132293 [REGRESSION] It is no longer possible to create native ints with .new (int32.new(42))
10:00 synopsebot RT#132295 [new]: https://rt.perl.org/Ticket/Display.html?id=132295 [REGRESSION][LTA] :2(…) radix syntax no longer suggests using .base (:2(42))
10:04 Zoffix c: 2017.06 multi x (int) { say "native" }; multi x (Int) { say "here" }; x  int32.new(42)
10:04 committable6 Zoffix, ¦2017.06: «here»
10:04 Zoffix AlexDaniel: looks like it wasn't possible before either
10:04 Zoffix c: 2017.04 multi x (int) { say "native" }; multi x (Int) { say "here" }; x  int32.new(42)
10:04 committable6 Zoffix, ¦2017.04: «here»
10:14 jnthn m: say int32.new(42).WHAT
10:14 yoleaux 03:44Z <AlexDaniel-> jnthn: fwiw RT #132292. IMO it's blocker-worthy, but not marking it as such just yet. What do you think?
10:14 camelia rakudo-moar d56151743: OUTPUT: «No such method 'BUILDALLPLAN' for invocant of type 'Perl6::Metamodel::NativeHOW'␤  in block <unit> at <tmp> line 1␤␤»
10:14 synopsebot RT#132292 [new]: https://rt.perl.org/Ticket/Display.html?id=132292 [REGRESSION] Recursively .emit-ing from the tap of the same supply bails out
10:15 jnthn AlexDaniel: Actually I'd count that as a fix.
10:16 jnthn You're not meant to be allowed to be in the closure passed to tap more than once at a time
10:17 jnthn This could accidentally happen
10:17 jnthn Due to the implmentation detail of us using a recursive mutex
10:20 Zoffix m: use nqp; say nqp::box_i(42, int32.WHAT)
10:20 camelia rakudo-moar d56151743: OUTPUT: «This representation (P6int) cannot unbox to other types (for type int32)␤  in block <unit> at <tmp> line 1␤␤»
10:21 jnthn int32 is not meant to be something you can box
10:21 jnthn It's a native type
10:21 Zoffix Yeah. Was just trying to figure out where the error's coming from
10:21 jnthn Calling .new on it is very dubious also
10:21 jnthn Though since type objects are objects, you can do it
10:22 Zoffix jnthn: so should int32.new(42) return a native int32 42?
10:23 jnthn No
10:23 jnthn IMO it should blow up
10:23 jnthn Though likely with a better error than today
10:23 jnthn "Cannot make an instance of a native type" would do it
10:23 Zoffix ok
10:25 jnthn As to how the error happens: the type objects of native types claim to be subtypes of, say, Int
10:25 jnthn This is only true in a type checking sense, not in a representation sense
10:25 jnthn It's why int32 comes out narrower than Int when doing the multi sort though
10:25 [TuxCM] This is Rakudo version 2017.09-404-gd56151743 built on MoarVM version 2017.09.1-594-gb9d3f6da
10:25 [TuxCM] csv-ip5xs        1.175 -  1.184
10:25 [TuxCM] test            13.997 - 14.336
10:25 [TuxCM] test-t           3.071 -  3.071
10:25 [TuxCM] csv-parser       0.854 -  0.863
10:26 Zoffix :o so close to sub-3
10:26 [TuxCM] how's that for a small range :)
10:26 timotimo might want to deprecate method new in native ints
10:26 timotimo so that existing code will just take a performance hit at first
10:26 timotimo rather than straight-up dying
10:26 timotimo then kill it with fire in a later release
10:27 Zoffix timotimo: but it never worked
10:27 timotimo it gave you an Int
10:27 Zoffix ok
10:27 timotimo it probably didn't truncate to the right size though?
10:28 Zoffix timotimo: it was the same as Int.new
10:28 timotimo right
10:28 Zoffix m: use MONKEY; augment class Int { proto method new2(|) {*}; multi method new2(int \SELF: $) { say "here" }; multi method new2(Int \SELF: $) { say "there" } }; int32.new2(42)
10:28 camelia rakudo-moar d56151743: OUTPUT: «there␤»
10:28 Zoffix Don't even know how to make the error better anyway :}
10:30 Zoffix m: use MONKEY; augment class Int { proto method new2(|) {*}; multi method new2(int32 \SELF: $) { say "here" }; multi method new2(Int \SELF: $) { SELF.^name.say; say "there" } }; int32.new2(42)
10:30 camelia rakudo-moar d56151743: OUTPUT: «int32␤there␤»
10:30 Zoffix k, tht works
10:33 Zoffix "works".... trying to figure out if it's a native by name will make it so slow, it'd be no point in using Int.new: native int (if people are even using it)
10:34 Zoffix releasable6: status
10:34 releasable6 Zoffix, Next release in 7 days and ≈8 hours. 1 blocker. Changelog for this release was not started yet
10:34 releasable6 Zoffix, Details: https://gist.github.com/8a5dc4a8820d6d5b16d3c57c56ce871f
10:34 Zoffix Gonna work on this later.
10:34 timotimo you can figure out if self.HOW is NativeHOW
10:35 Zoffix Ah, thanks.
10:39 jdv79 Zoffix: do you sleep?
10:40 Zoffix What do I need to look up to learn how to make stuff like we have for dalek, where it shows up as a service, like `sudo service dalek restart` ?
10:40 Zoffix jdv79: robots don't sleep
10:40 Zoffix I wanna make a geth service
10:40 Zoffix on hack
10:42 timotimo yeah, that's "system file" stuff from systemd
10:43 jnthn Zoffix: fwiw, probably the best way to do it is to make NativeHOW do MethodContainer and accounting for that in its find_method/publish_method_cache. Then we can just write "cannot make an instance of this" methods in the type object
10:44 Zoffix ok
10:45 jnthn A moderately easy way to earn the MOP Hacker badge :-)
10:46 Zoffix \o/
10:46 jnthn (If you didn't already :))
11:18 Geth_ ¦ geth: 5bb36864d3 | (Zoffix Znet)++ | lib/Geth/Plugin/GitHub.pm6
11:18 Geth_ ¦ geth: Handle missing chans gracefuler
11:18 Geth_ ¦ geth:
11:18 Geth_ ¦ geth: Any.split still got a proto for split to the "safe" call fails
11:18 Geth_ ¦ geth: there and still tries to call Any.split triggering no-multi
11:18 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/5bb36864d3
11:19 Geth_ ¦ geth: c362e66b4b | (Zoffix Znet)++ | lib/Geth/GitHub/Hooks.pm6
11:19 Geth_ ¦ geth: Update to handle latest HTTP::Server::Tiny
11:19 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/c362e66b4b
11:19 Geth_ ¦ geth: 5536d31b03 | (Zoffix Znet)++ | util-dev/emulate-github-event.p6
11:19 Geth_ ¦ geth: Add a dev util
11:19 Geth_ ¦ geth:
11:19 Geth_ ¦ geth: For simulating a GitHub push event hook arriving to our server
11:19 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/5536d31b03
11:21 Geth_ ¦ geth: bb70c2cde0 | (Zoffix Znet)++ | META6.json
11:21 Geth_ ¦ geth: Remove bogus `test-depends`
11:21 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/bb70c2cde0
11:25 Geth_ ¦ geth: 5615eae1f1 | (Zoffix Znet)++ | META6.json
11:25 Geth_ ¦ geth: Ensure prereqs match our use
11:25 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/5615eae1f1
11:25 Geth_ ¦ geth: 8ca11367aa | (Zoffix Znet)++ | META6.json
11:25 Geth_ ¦ geth: Use semver
11:25 Geth_ ¦ geth: review: https://github.com/perl6/geth/commit/8ca11367aa
11:37 Geth__ joined #perl6-dev
11:49 Geth__ ¦ geth: 8ca11367aa | (Zoffix Znet)++ | META6.json
11:49 Geth__ ¦ geth: Use semver
11:49 Geth__ ¦ geth: review: https://github.com/perl6/geth/commit/8ca11367aa
11:55 Geth__ ¦ specs: tbrowder++ created pull request #123: changes for Texas => ASCII
11:55 Geth__ ¦ specs: review: https://github.com/perl6/specs/pull/123
11:55 MasterDuke joined #perl6-dev
11:55 Geth__ ¦ specs: 1c979ab5a5 | (Tom Browder)++ | 2 files
11:55 Geth__ ¦ specs: changes for Texas => ASCII
11:55 Geth__ ¦ specs: review: https://github.com/perl6/specs/commit/1c979ab5a5
11:55 Geth__ ¦ specs: f52230c29b | (Zoffix Znet)++ (committed using GitHub Web editor) | 2 files
11:55 Geth__ ¦ specs: Merge pull request #123 from tbrowder/texas
11:55 Geth__ ¦ specs:
11:55 Geth__ ¦ specs: changes for Texas => ASCII
11:55 Geth__ ¦ specs: review: https://github.com/perl6/specs/commit/f52230c29b
11:56 Geth__ ¦ roast: 0bdfb1b157 | usev6++ | packages/Test/Util.pm
11:56 Geth__ ¦ roast: Use longer timeout for doesn't-hang on JVM
11:56 Geth__ ¦ roast: review: https://github.com/perl6/roast/commit/0bdfb1b157
11:58 Geth__ ¦ geth: 13becdb8c0 | (Zoffix Znet)++ | config.json
11:58 Geth__ ¦ geth: Hang in #perl6-toolchain too
11:58 Geth__ ¦ geth: review: https://github.com/perl6/geth/commit/13becdb8c0
12:02 MasterDuke c: c3b044df169^,c3b044df169,HEAD my $a = rx/aaaaaab/; 'a' x 999999 ~ 'b' ~~ /$a/; say now - INIT now
12:02 yoleaux 03:14Z <AlexDaniel-> MasterDuke: FWIW RT #132291
12:02 yoleaux 06:56Z <AlexDaniel-> MasterDuke: hey, maybe you can take a look at RT #132294
12:02 synopsebot RT#132291 [new]: https://rt.perl.org/Ticket/Display.html?id=132291 [LTA] Too few positionals passed to '' (for [:a] X [:b] -> ($i, $j) { })
12:02 synopsebot RT#132294 [new]: https://rt.perl.org/Ticket/Display.html?id=132294 [REGRESSION][PERF] INTERPOLATE-related change made matching of large strings slower
12:03 committable6 MasterDuke, ¦c3b044df169^: «3.42546156» ¦c3b044d: «5.072197» ¦HEAD(d561517): ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
12:04 MasterDuke hm, that's unfortunate, not what my PR was supposed to do
12:06 Geth joined #perl6-dev
12:06 Geth ¦ geth: 13becdb8c0 | (Zoffix Znet)++ | config.json
12:06 Geth ¦ geth: Hang in #perl6-toolchain too
12:06 Geth ¦ geth: review: https://github.com/perl6/geth/commit/13becdb8c0
12:09 Zoffix .tell nine I moved geth to hack. Something was wrong with the geth box. GitHub couldn't connect and I couldn't telnet to geth port's either (said connection refused). I tried restarting geth several times and connecting using localhost to geth's listener worked, but not from outside network. So I stopped the geth service on geth box (but did not remove it entirely) and set up a geth service on hack and
12:09 yoleaux Zoffix: I'll pass your message to nine.
12:09 Zoffix moved webhooks of all of our repo's to hack's geth.
12:09 Zoffix .tell nine moved webhooks of all of our repo's to hack's geth.
12:09 yoleaux Zoffix: I'll pass your message to nine.
12:10 Zoffix Can someone with keys to rakudo/star change the geth web hook to geth.niner change the URL to http://hack.p6c.org:8888/?chan=#perl6-dev  (or give me access and I'll set it up myself)
12:12 Zoffix PSA: Geth now lives on hack, in /home/geth/geth  There's a service setup, so if a restart of it is needed, you'd just do `sudo service geth restart`
12:13 jnthn Zoffix: Given you access
12:14 Zoffix \o/
12:19 BenGoldberg joined #perl6-dev
12:33 MasterDuke c: c3b044df169^,c3b044df169,HEAD my $a = "aaaaaab"; 'a' x 999999 ~ 'b' ~~ /$a/; say now - INIT now
12:34 committable6 MasterDuke, ¦c3b044df169^: «5.4635297» ¦c3b044d: «7.4579997» ¦HEAD(d561517): «4.9643206»
12:34 MasterDuke right, that's the case my recent PR made faster
12:41 MasterDuke goodness, the slow rx version spends 3s in the proto for Match's Bool and then another 1.5s in the multi it chose
12:42 MasterDuke 2999980 calls to both
12:44 Zoffix I'm having a hard time understanding where Perl6::Metamodel::MethodContainer fits into the picture of disabling int32.new... If I make NativeHOW do it, seems all the methods need to overridden. The method_table and submethod_table already exist in NetiveHOW without that role (they return an empty nqp::hash).
12:46 timotimo Zoffix: then those can probably be thrown out
12:47 Zoffix They fix a bug, I think: https://github.com/rakudo/rakudo/commit/9a2f7fcac723a43655a7aa89df1a40bc8dbe31e0
12:48 Geth ¦ rakudo: MasterDuke17++ created pull request #1193: Whitespace fix, convert tabs to spaces
12:48 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1193
12:49 Geth ¦ rakudo/nom: a0279e561a | MasterDuke17++ (committed using GitHub Web editor) | src/core/Match.pm
12:49 Geth ¦ rakudo/nom: Whitespace fix, convert tabs to spaces
12:49 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a0279e561a
12:49 Geth ¦ rakudo/nom: 5092d6e60e | (Zoffix Znet)++ (committed using GitHub Web editor) | src/core/Match.pm
12:49 Geth ¦ rakudo/nom: Merge pull request #1193 from MasterDuke17/patch-1
12:49 Geth ¦ rakudo/nom:
12:49 Geth ¦ rakudo/nom: Whitespace fix, convert tabs to spaces
12:49 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/5092d6e60e
12:54 Geth ¦ rakudo: MasterDuke17++ created pull request #1194: Rest of tabs -> spaces in Match.pm
12:54 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1194
12:55 Geth ¦ rakudo/nom: 0262a997aa | MasterDuke17++ (committed using GitHub Web editor) | src/core/Match.pm
12:55 Geth ¦ rakudo/nom: Rest of tabs -> spaces in Match.pm
12:55 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0262a997aa
12:55 Geth ¦ rakudo/nom: 0d217357aa | (Zoffix Znet)++ (committed using GitHub Web editor) | src/core/Match.pm
12:55 Geth ¦ rakudo/nom: Merge pull request #1194 from MasterDuke17/patch-1
12:55 Geth ¦ rakudo/nom:
12:55 Geth ¦ rakudo/nom: Rest of tabs -> spaces in Match.pm
12:55 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0d217357aa
13:01 Zoffix greppable6, \b(int32|int|int64|int16|int8)\.new
13:01 greppable6 Zoffix, Found nothing!
13:01 Zoffix timotimo: do we really need a deprecation for it?
13:01 Zoffix It's be a deprecation for a feature so rarely used that it's not used in the ecosystem and one that never worked right
13:02 Zoffix c: 2017.06 say int8.new: 999999999999999999999999999999999999999999999999999999999999999999999999999999999999
13:02 committable6 Zoffix, ¦2017.06: «999999999999999999999999999999999999999999999999999999999999999999999999999999999999»
13:03 Zoffix Oh damn: https://github.com/nkh/P6-Data-Dump-Tree/blob/a0374d5355253419a633151faa7e277172b570fc/t/48_type_NativeCall.t#L45
13:05 Zoffix nqp: nqp::getstaticcode(sub ( | ) { nqp::die('Cannot instantiate native type') })
13:05 camelia nqp-moarvm: OUTPUT: «Unable to parse expression in routine_def; couldn't find final ')'  at line 2, near "| ) { nqp:"␤   at gen/moar/stage2/NQPHLL.nqp:707  (/home/camelia/rakudo-m-inst-1/share/nqp/lib/NQPHLL.moarvm:panic)␤ from gen/moar/stage2/NQPHLL.nqp:714  (/home/camelia/raku…»
13:05 Zoffix any way to spec "take any args" in nqp?
13:06 Zoffix More of them: https://github.com/wentam/P6-Graffiks/blob/ec8912519ad52e7083c021c79297466b9a37d428/lib/Graffiks/Camera.pm6#L86
13:06 Zoffix So no use of ints but nums quite a few :(
13:07 moritz nqp: sub f(*@a) { nqp::say(@a) }; f(1, 2)
13:07 camelia nqp-moarvm: OUTPUT: «cannot stringify this␤   at <tmp>:1  (<ephemeral file>:f)␤ from <tmp>:1  (<ephemeral file>:<mainline>)␤ from gen/moar/stage2/NQPHLL.nqp:1542  (/home/camelia/rakudo-m-inst-1/share/nqp/lib/NQPHLL.moarvm:eval)␤ from gen/moar/stage2/NQPHLL.nqp:1779  (/home/c…»
13:07 moritz nqp: sub f(*@a) { nqp::say(nqp::join(', ', @a)) }; f(1, 2)
13:07 camelia nqp-moarvm: OUTPUT: «This representation (P6int) cannot unbox to a native string (for type BOOTInt)␤   at <tmp>:1  (<ephemeral file>:f)␤ from <tmp>:1  (<ephemeral file>:<mainline>)␤ from gen/moar/stage2/NQPHLL.nqp:1542  (/home/camelia/rakudo-m-inst-1/share/nqp/lib/NQPHLL.moarv…»
13:07 moritz nqp: sub f(*@a) { nqp::say(nqp::join(', ', @a)) }; f("a", "b")
13:07 camelia nqp-moarvm: OUTPUT: «a, b␤»
13:08 Zoffix nqp: sub f(*@a) { nqp::say(nqp::join(', ', @a)) }; f("a", "b", :foo(42))
13:08 camelia nqp-moarvm: OUTPUT: «Unexpected named argument 'foo' passed␤   at <tmp>:1  (<ephemeral file>:f)␤ from <tmp>:1  (<ephemeral file>:<mainline>)␤ from gen/moar/stage2/NQPHLL.nqp:1542  (/home/camelia/rakudo-m-inst-1/share/nqp/lib/NQPHLL.moarvm:eval)␤ from gen/moar/stage2/NQPHLL.n…»
13:08 Zoffix nqp: sub f(*@a, *%_) { nqp::say(nqp::join(', ', @a)) }; f("a", "b", :foo(42))
13:08 camelia nqp-moarvm: OUTPUT: «a, b␤»
13:08 Zoffix ok thanks
13:13 Zoffix Well, I done this and it works: https://gist.github.com/zoffixznet/19825d7c42479982d27eeab33f0c3b01
13:14 Zoffix make method_table return an nqp::hash that got a blowy method new in it (in the past it was just an empty nqp::hash)
13:19 Zoffix .in 1d https://irclog.perlgeek.de/perl6-dev/2017-10-14#i_15301581
13:19 yoleaux Zoffix: I'll remind you on 15 Oct 2017 13:19Z
13:21 Zoffix I've no idea how to do deprecation from there
13:21 Zoffix from NativeHow
13:22 Zoffix What does getstaticcode do? Like what's a "static code"?
13:24 moritz when you have a closure, you have a pointer to a thing that holds the captured lexpad and the "actual" code
13:24 moritz that's what "static code" - code reference that's not a closure
13:25 Zoffix Ah
13:31 Zoffix m: use nqp; say nqp::getlexdyn('&DEPRECATED')
13:31 camelia rakudo-moar 0d217357a: OUTPUT: «(signal SEGV)»
13:31 Zoffix :)
13:33 Zoffix Filed as https://rt.perl.org/Ticket/Display.html?id=132300
13:39 MasterDuke what causes lots of time spent in a proto? in my experiments with INTERPOLATE, it was because the find_best_dispatchee wasn't able to cache the right method. but in the profile i'm looking at now, find_best_dispatchee doesn't even appear
13:44 MasterDuke https://github.com/rakudo/rakudo/blob/nom/src/core/Match.pm#L772-L774 the proto on line 772, and it always calls the multi on line 774
13:50 buggable New CPAN upload: Config-Parser-toml-1.0.1.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/Config-Parser-toml-1.0.1.tar.gz
13:50 buggable New CPAN upload: Config-Parser-yaml-1.0.1.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/Config-Parser-yaml-1.0.1.tar.gz
14:14 Zoffix Weird. `DEPRECATED "foo"; DEPRECATED "bar"` just shows 1 deprecation :/
14:29 Zoffix Is it possible to add some sort of annotations to a method? Roughly thinking about how Cool methods that pass captures could be fixed up to be faster and I thought maybe we could stick some annotation to the relevant methods in Cool and then the optimizer could just rewrite the calls to Cool.meow(<some args>) to Cool.Stringy.meow(<some args>)
14:43 Geth ¦ rakudo/nom: 9d9c7f9c3b | (Zoffix Znet)++ | src/Perl6/Metamodel/NativeHOW.nqp
14:43 Geth ¦ rakudo/nom: Deprecate .new on native types
14:43 Geth ¦ rakudo/nom:
14:43 Geth ¦ rakudo/nom: - For now, just maintain old behaviour of returning a boxed version
14:43 Geth ¦ rakudo/nom: - After 2017.12, toss the deprecation and make the methods die
14:43 Geth ¦ rakudo/nom: - Fixes RT#132293: https://rt.perl.org/Ticket/Display.html?id=132293
14:43 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9d9c7f9c3b
14:43 synopsebot RT#132293 [open]: https://rt.perl.org/Ticket/Display.html?id=132293 [REGRESSION] It is no longer possible to create native ints with .new (int32.new(42))
14:44 Geth ¦ rakudo/nom: cc6c05586b | (Zoffix Znet)++ | t/05-messages/01-errors.t
14:44 Geth ¦ rakudo/nom: Test behaviour of .new on native types
14:44 Geth ¦ rakudo/nom:
14:44 Geth ¦ rakudo/nom: RT#132293: https://rt.perl.org/Ticket/Display.html?id=132293
14:44 Geth ¦ rakudo/nom: Rakudo fix: https://github.com/rakudo/rakudo/commit/9d9c7f9c3b
14:44 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cc6c05586b
14:49 Zoffix .in 70d take care of tossing deprecation in https://github.com/rakudo/rakudo/commit/9d9c7f9c3b and https://github.com/rakudo/rakudo/commit/cc6c05586b , bruh
14:49 yoleaux Zoffix: I'll remind you on 23 Dec 2017 14:49Z
15:18 Geth ¦ roast: 4043a63b60 | (Zoffix Znet)++ | packages/Test/Util.pm
15:18 Geth ¦ roast: Revert "Use longer timeout for doesn't-hang on JVM"
15:18 Geth ¦ roast:
15:18 Geth ¦ roast: This reverts commit 0bdfb1b157b5286480d8355cf62ce1a76fd445ea.
15:18 Geth ¦ roast:
15:18 Geth ¦ roast: Will rewrite a better version in next commit
15:18 Geth ¦ roast: review: https://github.com/perl6/roast/commit/4043a63b60
15:18 Geth ¦ roast: bdf8904da6 | (Zoffix Znet)++ | 2 files
15:18 Geth ¦ roast: Rescale ROAST_TIMING_SCALE for JVM
15:18 Geth ¦ roast:
15:18 Geth ¦ roast: 'cause it takes longer to fire up procs
15:18 Geth ¦ roast: review: https://github.com/perl6/roast/commit/bdf8904da6
15:19 Zoffix (this version will scale doesn't-hangs with custom wait times too; not just the default times)
15:22 Zoffix I think the most annoying thing about "ANNOYING" ticket tags is the tag, not the bug the ticket describes
15:23 Zoffix Like wtf is RT#132012 an "annoying" ticket?
15:23 synopsebot RT#132012 [new]: https://rt.perl.org/Ticket/Display.html?id=132012 [ANNOYING] Numeric values of signals are wrong (say +SIGUSR1)
15:23 Zoffix .tell AlexDaniel [ANNOYING] ticket tag is annoying
15:23 yoleaux Zoffix: I'll pass your message to AlexDaniel.
15:49 Zoffix "Show message only if value has method .base"    "if $.num.perl.^can('base')"  why yes, .perl it first :P
15:49 Zoffix ZofBot: good job!
15:49 ZofBot Zoffix, weigh my friend's affection with mine own
16:03 BenGoldberg joined #perl6-dev
16:04 dogbert17 joined #perl6-dev
16:07 dogbert17 ZOFFLOP t/spec/S02-types/baghash.t
16:08 dogbert17 'Cannot unbox 64 bit wide bigint into native integer'
16:12 Zoffix dogbert17: is it actually a flop?
16:12 Zoffix As in, are you sure it fails only some times and not all the time?
16:13 Zoffix Also, do you have latest and greatest rakudo?
16:13 Zoffix And if yes, what's the full output
16:14 Geth ¦ rakudo/nom: 441fc9b32b | (Zoffix Znet)++ | t/05-messages/01-errors.t
16:14 Geth ¦ rakudo/nom: Do not add more stuff into 01-errors.t
16:14 Geth ¦ rakudo/nom:
16:14 Geth ¦ rakudo/nom: It takes forever to run and there's no reason we can't split it
16:14 Geth ¦ rakudo/nom: up into multiple files.
16:14 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/441fc9b32b
16:28 Geth ¦ rakudo/nom: de2b9ff720 | (Zoffix Znet)++ | 2 files
16:28 Geth ¦ rakudo/nom: Remove unwanted .perl call
16:28 Geth ¦ rakudo/nom:
16:28 Geth ¦ rakudo/nom: Fixes RT#132295: https://rt.perl.org/Ticket/Display.html?id=132295
16:28 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/de2b9ff720
16:28 synopsebot RT#132295 [new]: https://rt.perl.org/Ticket/Display.html?id=132295 [REGRESSION][LTA] :2(…) radix syntax no longer suggests using .base (:2(42))
16:38 ugexe m: package Foo { our &foo := -> $a { say $a }; }; Foo::foo(1); # should this work? if not, does it differ from (the working) `our sub ...`?
16:38 camelia rakudo-moar 441fc9b32: OUTPUT: «Cannot invoke this object (REPR: Uninstantiable; Callable)␤  in block <unit> at <tmp> line 1␤␤»
16:40 Zoffix Seems to work if you don't use binding:
16:40 Zoffix m: package Foo { our &foo = -> $a { say $a }; }; Foo::foo(1);
16:40 camelia rakudo-moar 441fc9b32: OUTPUT: «1␤»
16:40 Zoffix No idea if it should work with it as well
16:41 Geth ¦ rakudo: MasterDuke17++ created pull request #1195: Optimize Regex multi for INTERPOLATE
16:41 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1195
16:41 ugexe cool
16:47 timotimo i thought "zofflop" is just "flops on zoffix' system", did dogbert root zoffix' box? :)
16:48 Zoffix no, zofflop is "flop" with a prefix that makes it easier to search for in the log
16:48 Zoffix This looks bogus: https://github.com/rakudo/rakudo/commit/ec79843830588c0ef941e79ea4bab5bce05122f9
16:51 Zoffix m:  my $x = "abc"; $x ~~ tr|abc|{$/.uc}|; say $x
16:51 camelia rakudo-moar 441fc9b32: OUTPUT: «{$/␤»
16:51 Zoffix Just gonna toss all the $/ fumbling in Str.trans and see if any tests explode
16:51 Zoffix (it's also not done for Cool.trans, which is how I noticed it)
17:01 travis-ci joined #perl6-dev
17:01 travis-ci Rakudo build failed. Zoffix Znet 'Do not add more stuff into 01-errors.t
17:01 travis-ci https://travis-ci.org/rakudo/rakudo/builds/287962876 https://github.com/rakudo/rakudo/compare/cc6c05586b62...441fc9b32b70
17:01 travis-ci left #perl6-dev
17:01 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
17:03 Zoffix why do travis reports now have these slanted lines across them :?
17:03 Zoffix One Linux job segfaulted during make Segmentation fault (core dumped)
17:06 Zoffix Tossing everything makes this test fail:
17:06 Zoffix m: say '"foo  &   bar"'.trans(/(' '+)/ => {' ' ~ ('&nbsp;' x ($0.chars - 1))}, /(\W)/ => {"&#{ord($0)};"})
17:06 camelia rakudo-moar de2b9ff72: OUTPUT: «&#34;foo &nbsp;&#38; &nbsp;&nbsp;bar&#34;␤»
17:07 Zoffix m: say '"foo  &   bar"'.trans(/(' '+)/ => {' ' ~ ('&nbsp;' x ($0.chars - 1))}, /(\W)/ => {"&#{ord($0)};"}); say $/
17:07 camelia rakudo-moar de2b9ff72: OUTPUT: «&#34;foo &nbsp;&#38; &nbsp;&nbsp;bar&#34;␤Nil␤»
17:07 Zoffix But then, I still don't get why it's grabbing caller's $/
17:07 Zoffix Gonna just file it for now
17:15 Zoffix Filed as https://rt.perl.org/Ticket/Display.html?id=132302
17:20 Geth ¦ roast: c4f9ac0bbd | usev6++ | 8 files
17:20 Geth ¦ roast: [jvm] Skip hanging tests
17:20 Geth ¦ roast: review: https://github.com/perl6/roast/commit/c4f9ac0bbd
17:21 releasable6 joined #perl6-dev
17:22 Geth ¦ rakudo/nom: c1e5cfcfa5 | usev6++ | t/spectest.data
17:22 Geth ¦ rakudo/nom: [jvm] Don't run files when all tests are hanging
17:22 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c1e5cfcfa5
17:37 Zoffix I don't get how this works at all. It's missing a decont: https://github.com/rakudo/rakudo/blob/nom/src/core/Str.pm#L150
17:37 Zoffix m: use nqp; class Foo { method x ($suffix) { say nqp::getattr($suffix,Str,'$!value') } }.x: "x"
17:37 camelia rakudo-moar c1e5cfcfa: OUTPUT: «P6opaque: no such attribute '$!value' in type Str when trying to get a value␤  in method x at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
17:37 Zoffix m: my $x = "x"; say "x".ends-with: $x
17:37 camelia rakudo-moar c1e5cfcfa: OUTPUT: «True␤»
17:37 Zoffix yet it doesn't explode. Huh?
17:39 MasterDuke and nothing changes if you add the decont?
17:39 Zoffix Changes how?
17:40 Zoffix It's now behaving as if the decont was already there
17:40 MasterDuke .ends-with still produces the same results?
17:41 Zoffix Haven't tried. I'd expect it to. Don't really wanna pay the compilation price
17:41 dogbert17 Zoffix: it's not a flapper, it fails every time, I suspect it's 32 bit related, it's the very last test which causes the problem
17:41 dogbert17 t/spec/S02-types/baghash.rakudo.moar (Wstat: 65280 Tests: 292 Failed: 0)
17:41 dogbert17 Non-zero exit status: 255
17:41 dogbert17 Parse errors: Bad plan.  You planned 293 tests but ran 292.
17:41 Zoffix m: use nqp; class Foo { method x (Str:D $suffix) { say nqp::p6bool(nqp::getattr($suffix,Str,'$!value')) } }.x: "x"
17:41 camelia rakudo-moar c1e5cfcfa: OUTPUT: «True␤»
17:41 Zoffix m: use nqp; class Foo { method x (Str $suffix) { say nqp::p6bool(nqp::getattr($suffix,Str,'$!value')) } }.x: "x"
17:41 camelia rakudo-moar c1e5cfcfa: OUTPUT: «True␤»
17:41 Zoffix m: use nqp; class Foo { method x (Str:D $suffix) { say nqp::p6bool(nqp::getattr($suffix,Str,'$!value')) } }.x: "x"
17:41 camelia rakudo-moar c1e5cfcfa: OUTPUT: «True␤»
17:42 Zoffix k, looks like Str type constraint does some sort of decont voodoo
17:42 dogbert17 m: use Test; is-deeply ("foo" => 10000000000000000000).BagHash.grab(1), ("foo",),"can .grab() a BagHash key with weight larger than 64 bits"
17:42 camelia rakudo-moar c1e5cfcfa: OUTPUT: «ok 1 - can .grab() a BagHash key with weight larger than 64 bits␤»
17:42 Zoffix oh wait
17:42 Zoffix Yeah
17:42 Zoffix m: use nqp; class Foo { method x ($suffix) { say nqp::p6bool(nqp::getattr($suffix,Str,'$!value')) } }.x: "x"
17:42 camelia rakudo-moar c1e5cfcfa: OUTPUT: «P6opaque: no such attribute '$!value' in type Str when trying to get a value␤  in method x at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
17:42 dogbert17 this fails for me with
17:42 dogbert17 Cannot unbox 64 bit wide bigint into native integer
17:42 dogbert17 in sub _is_deeply at /home/dogbert/repos/rakudo/install/share/perl6/sources/C712FE6969F786C9380D643DF17E85D06868219E (Test) line 652
17:42 dogbert17 in sub is-deeply at /home/dogbert/repos/rakudo/install/share/perl6/sources/C712FE6969F786C9380D643DF17E85D06868219E (Test) line 598
17:42 Zoffix dogbert17: well, is your rakudo a HEAD? It was fixed recently
17:43 dogbert17 2017.09-410-gcc6c05586
17:43 Zoffix m: say $PERL.compiler.version
17:43 camelia rakudo-moar c1e5cfcfa: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>␤Variable '$PERL' is not declared. Did you mean 'Perl'?␤at <tmp>:1␤------> 3say 7⏏5$PERL.compiler.version␤»
17:43 Zoffix m: say $*PERL.compiler.version
17:43 camelia rakudo-moar c1e5cfcfa: OUTPUT: «v2017.09.413.gc.1.e.5.cfcfa␤»
17:43 Zoffix dogbert17: and that's on 32-bit box?
17:43 dogbert17 yes
17:43 * dogbert17 pull latest Rakudo
17:43 Zoffix 410 would have had the fix already
17:44 Zoffix dogbert17: you could reopen the ticket and say it still fails on 32-bit: https://rt.perl.org/Ticket/Display.html?id=132279
17:44 * dogbert17 parsing ...
17:45 Zoffix and include --ll-exception of perl6 -e '("foo" => 10000000000000000000).BagHash.grab(1)'
17:49 dogbert17 false alarm, it works with Rakudo head. Sorry for the noise
17:50 Zoffix oh, *phew*
17:54 Zoffix What do the cool kids use to measure something too close to noise? Like which one's faster:   method x($x) { $x.uc } or method x(\x) { x.uc }
17:54 MasterDuke callgrind will give you the actual instruction count
17:54 Zoffix I get from 4.6x faster to 2x slower with Benchy -_-
17:55 Zoffix sudo aptitude search callgrind... nothing found
17:55 MasterDuke it's part of valgrind
17:55 MasterDuke valgrind --tool=callgrind <...>
17:55 Zoffix ok
17:56 MasterDuke and then kcachgrind is a nice viewer for the results if you want more than just the instruction count
17:56 Zoffix doesn't seem to print anything for me. Creates an empty file :/
17:57 dogbert17 how did you run it?
17:58 Zoffix dogbert17: valgrind --tool=callgrind ./perl6 foo.p6
17:58 MasterDuke oh, you might need to call the moar executable directly, not the perl6 shell script
17:58 Zoffix Oh
17:58 Zoffix sounds hard
17:58 * Zoffix puts it on the back burner for now
17:59 MasterDuke eh, i just cat perl6 and copy it onto the command line
17:59 dogbert17 make a copy of perl6-valgrind-m and stick in the '--tool=callgrind' there
17:59 MasterDuke and i also have some bash aliases that do that for me
17:59 Zoffix ok
18:00 yoleaux Zoffix: see if nqp::eqaddr deconting stuff needs to be done anywhere in core
18:00 Zoffix MVM_VALGRIND_OPTS=--tool=callgrind ./perl6-valgrind-m foo.p6  works too
18:00 Zoffix yoleaux: meh
18:00 MasterDuke heh, good timing past Zoffix
18:04 dogbert17 Zoffix: I guess you know that the generated files are, in essence, unreadable. KCachegrind, like MasterDuke wrote or callgrind_annotate if you want to stay in the shell
18:05 Zoffix I opened cachegrind, but I don't know what to look for. The numbers fluctuate on each run of the script
18:07 Zoffix There are 19 fewer `optimize_decont` calls and that's all I can interpret. I guess you need to know what to look for
18:07 dogbert17 I belive that the recommended cmdline should be something like 'MVM_SPESH_BLOCKING=1 valgrind --tool=callgrind ./perl6 foo.p6' the fluctuations might go down
18:07 yoleaux Zoffix: fix (my @a is default(42) = <a b c>)[1]:delete; @a.Slip.List.say, bruh
18:08 Zoffix yoleaux: dude, leave me be :(
18:08 dogbert17 correction 'MVM_SPESH_BLOCKING=1 MVM_VALGRIND_OPTS=--tool=callgrind ./perl6-valgrind-m foo.p6'
18:09 Zoffix I need a better reminder bot
18:09 AlexDaniel- joined #perl6-dev
18:11 Zoffix I was gonna write all the tests today... Still haven't started :/
18:11 Geth ¦ rakudo/nom: 04b171bd12 | (Daniel Green)++ | src/core/Match.pm
18:11 Geth ¦ rakudo/nom: Optimize Regex multi for INTERPOLATE
18:11 Geth ¦ rakudo/nom:
18:11 Geth ¦ rakudo/nom: Move some code into a conditional and don't allocate some variables if
18:11 Geth ¦ rakudo/nom: not needed.
18:11 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/04b171bd12
18:11 Geth ¦ rakudo/nom: 0bd2b27df5 | (Aleks-Daniel Jakimenko-Aleksejev)++ (committed using GitHub Web editor) | src/core/Match.pm
18:11 Geth ¦ rakudo/nom: Merge pull request #1195 from MasterDuke17/optimize_Regex_multi_for_INTERPOLATE
18:11 Geth ¦ rakudo/nom:
18:11 Geth ¦ rakudo/nom: Optimize Regex multi for INTERPOLATE
18:11 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0bd2b27df5
18:13 Zoffix dogbert17: still fluctuating enough to wash the difference
18:13 dogbert17 :(
18:13 Zoffix Maybe there ain't any. Just gonna leave it./
18:20 AlexDaniel- .
18:20 AlexDaniel .
18:20 yoleaux 15:23Z <Zoffix> AlexDaniel: [ANNOYING] ticket tag is annoying
18:21 AlexDaniel- Zoffix: RT #132012 is annoying for the same reason described in https://rt.perl.org/Ticket/Display.html?id=131003#txn-1496218
18:21 synopsebot RT#132012 [open]: https://rt.perl.org/Ticket/Display.html?id=132012 [ANNOYING] Numeric values of signals are wrong (say +SIGUSR1)
18:21 AlexDaniel- some signals have correct values and you don't notice it right away, and there's practically no workaround
18:21 AlexDaniel- ok you can hardcode the values for a particular system, but uhhh…
18:22 AlexDaniel- that's an annoying workaround :)
18:22 Zoffix Seems a pointless tag and it certainly makes me want to fix those less.
18:23 Zoffix "
18:23 Zoffix "» is no longer shuffled (<a b c d>».say)" why is this annoying?
18:23 Zoffix This tag is nonsense
18:24 AlexDaniel- because we keep explaining to people “hey actually it doesn't work and it just happens that these are returned in the same order right now”
18:25 timotimo Zoffix: if you want less noisy results from callgrind, set the spesh blocking env var
18:25 AlexDaniel- the tag differentiates painful issues with other bugs that are mostly unnoticeable (or, when noticed, are easily workaroundable)
18:26 MasterDuke i think you could replace annoying with lta
18:26 AlexDaniel- it's other way round
18:26 AlexDaniel- LTA = it's ok but could be better
18:27 Zoffix timotimo: I did. Still too noisy
18:27 AlexDaniel- ANNOYING = something that keeps getting in your way and makes rakudo look like garbage
18:27 timotimo damn
18:28 Zoffix AlexDaniel-: the tag differentiates issues I'll personally gonna be avoiding. It's like a bug reporter who's shouting at how pissed off they are about a bug. You just phase them out and go fix something nicer.
18:28 AlexDaniel- Zoffix: I'm sorry that it makes you feel that way
18:29 AlexDaniel- Zoffix: maybe if you suggest some other name for that purpose, I can retag those
18:32 Zoffix AlexDaniel-: anything that doesn't require you to explain why we're using such a tag. So far you had to explain "ANNOYING" to at least two people.
18:32 AlexDaniel- it's ok for a new tag
18:32 AlexDaniel- I had to explain what a REGRESSION is a couple of times also…
18:33 AlexDaniel- and it's ok for people to ask why some particular issue is annoying (maybe they don't see the actual issue, it's ok to ask for clarification)
18:36 MasterDuke m: class A { method x($x) { $x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:36 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤5.4048479␤»
18:36 MasterDuke m: class A { method x(\x) { x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:36 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤3.85693285␤»
18:37 MasterDuke Zoffix: ^^^ doesn't look like noise? or was that not what you were actually testing?
18:38 Ben_Goldberg joined #perl6-dev
18:39 Zoffix MasterDuke: you just ran it twice though.
18:39 MasterDuke timotimo: do you know anything about protos?
18:39 Zoffix m: class A { method x(\x) { x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:39 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤3.40128488␤»
18:39 Zoffix m: class A { method x(\x) { x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:39 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤3.7919970␤»
18:39 Zoffix m: class A { method x($x) { $x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:39 MasterDuke i ran it a bunch of times locally
18:39 Zoffix m: class A { method x($x) { $x.uc } }; my $a; $a = A.x("aBc") for ^10_000_000; say $a; say now - INIT now
18:39 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤5.3459838␤»
18:39 camelia rakudo-moar 0bd2b27df: OUTPUT: «ABC␤5.359951␤»
18:39 Zoffix MasterDuke: ah. I ran them both in one program, so I guess that's what wiped the difference in Benchy
18:40 Zoffix ok
18:40 MasterDuke huh, wouldn't have expected that to matter
18:41 Zoffix well, GC is involved no?
18:42 MasterDuke right, that could effect things
18:42 Zoffix Weird, this script consistently shows 2017.09-173-g4d0ead2 to be 1-15ms seconds faster than HEAD https://gist.github.com/zoffixznet/988071e6fb158536a35a75ef0bc88db7
18:43 Zoffix Running as   ./perl6 bar.p6 -e '42.starts-with: "x" for ^100_00; say now - INIT now'
18:43 Zoffix bar.p6 being the script and ./perl6 being head
18:44 MasterDuke benchable6: 4d0ead2,HEAD 42.starts-with: "x" for ^100_000
18:44 benchable6 MasterDuke, starting to benchmark the 2 given commits
18:45 benchable6 MasterDuke, benchmarked the given commits, now zooming in on performance differences
18:45 benchable6 MasterDuke, ¦4d0ead2: «1.0593» ¦HEAD: «1.1259»
18:46 MasterDuke benchable6: 4d0ead2,HEAD 42.starts-with: "x" for ^500_000
18:46 benchable6 MasterDuke, starting to benchmark the 2 given commits
18:47 benchable6 MasterDuke, benchmarked the given commits, now zooming in on performance differences
18:47 benchable6 MasterDuke, ¦4d0ead2: «4.7527» ¦HEAD: «4.8870»
18:58 Zoffix ZOFFLOP: t/spec/S17-procasync/kill.t
18:59 Geth ¦ rakudo/nom: 1275fd4ab7 | (Zoffix Znet)++ | 2 files
18:59 Geth ¦ rakudo/nom: Make Cool.[starts|ends]-with 12x faster
18:59 Geth ¦ rakudo/nom:
18:59 Geth ¦ rakudo/nom: Part of fixing RT#132280:
18:59 Geth ¦ rakudo/nom: https://rt.perl.org/Ticket/Display.html?id=132280
18:59 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1275fd4ab7
18:59 synopsebot RT#132280 [new]: https://rt.perl.org/Ticket/Display.html?id=132280 [REGRESSION][PERF] Cool methods that take and pass a Capture are uber slower (.contains: 'meow' for ^10000_00)
18:59 Zoffix Gonna do rest of 'em later today.
19:00 MasterDuke nice
19:01 Zoffix Not really, 'cause many of these is just gaining back speed lost earlier in 2017. But we might break the 3s test-t barrier, 'cause I think I recall it getting slower when these methods were fixed to be captures, but I didn't know why it'd get slow
19:14 travis-ci joined #perl6-dev
19:14 travis-ci Rakudo build passed. usev6 '[jvm] Don't run files when all tests are hanging'
19:14 travis-ci https://travis-ci.org/rakudo/rakudo/builds/287979127 https://github.com/rakudo/rakudo/compare/de2b9ff7209e...c1e5cfcfa595
19:14 travis-ci left #perl6-dev
19:16 lizmat joined #perl6-dev
19:41 travis-ci joined #perl6-dev
19:41 travis-ci Rakudo build errored. Zoffix Znet 'Merge pull request #1194 from MasterDuke17/patch-1
19:41 travis-ci https://travis-ci.org/rakudo/rakudo/builds/287916515 https://github.com/rakudo/rakudo/compare/5092d6e60e45...0d217357aa80
19:41 travis-ci left #perl6-dev
19:41 buggable [travis build above] ✓ All failures are due to: missing build log (1 failure).
19:43 timotimo MasterDuke: not much; what's up?
19:54 timotimo haha. "not much" is also the common answer to "what's up?"
19:55 timotimo i was going to ask how to best turn "while $data-which-is-a-Channel.receive -> $blop { }" into something that uses await to free up the thread in question ... then i remembered we have "react whenever", which ought to be better
20:02 timotimo oh wow, that's quite the performance hit
20:05 timotimo 1_500_000 messages with a while try receive gives me 8.95s user, 6.15s elapsed; using react whenever gives me 10s user, 5.88s user, but for only 50_000 messages
20:05 timotimo 30x difference ;(
20:06 Zoffix :o
20:09 Zoffix AlexDaniel-: whatdyumean reject RT#131515? If that's the right commit, then it should be optimized. The assumtion in the commit is we can be expensive because it won't impact anything, but if it is the right commit, then the assumption is wrong
20:09 synopsebot RT#131515 [open]: https://rt.perl.org/Ticket/Display.html?id=131515 [BISECTME][PERF] Performance loss in some commits
20:11 AlexDaniel- Zoffix: I was thinking that it only impacts the spectest and normal users will never notice it otherwise… now that you said it, I realize it may impact module tests and anything else that tests exceptions
20:11 AlexDaniel- Zoffix: so you're right
20:11 timotimo *sigh*, i'm glad someone is going to make profiler work with multithreade programs ...
20:12 Zoffix AlexDaniel-: ah, never mind, it's in .message which has since been optimized away to being called only when actually needed, so I'll close the ticket
20:12 Zoffix AlexDaniel-: well, bisect it and if it is the commit, then reject it
20:12 AlexDaniel- Zoffix: oh! OK
20:13 AlexDaniel- I'll do it right now if it's that simple
20:13 timotimo adding --profile to this program gives me "inconsistent bind result" m(
20:13 robertle joined #perl6-dev
20:14 Zoffix c: 2500e50,2500e50^,HEAD for ^1000 { try "foo".meow }
20:14 committable6 Zoffix, ¦2500e50: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»» ¦2500e50^,HEAD(1275fd4): «»
20:14 AlexDaniel- right
20:14 Zoffix c: 2500e50,2500e50^,HEAD for ^100 { try "foo".meow }; say "done"
20:14 committable6 Zoffix, ¦2500e50: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»» ¦2500e50^,HEAD(1275fd4): «done»
20:15 Zoffix c: 2500e50,2500e50^,HEAD for ^10 { try "foo".meow }; say "done"
20:15 committable6 Zoffix, ¦2500e50,2500e50^,HEAD(1275fd4): «done»
20:15 Zoffix c: 2500e50,2500e50^,HEAD for ^10 { try "foo".meow }; say now - INIT now
20:15 committable6 Zoffix, ¦2500e50: «3.46179999» ¦2500e50^: «0.00295601» ¦HEAD(1275fd4): «0.0024989»
20:15 * Zoffix rejects
20:15 AlexDaniel- OK
20:15 AlexDaniel- or maybe resolve instead?
20:16 AlexDaniel- it *is* resolved after all
20:16 Zoffix right
20:16 Zoffix Done.
20:16 AlexDaniel- yay
20:29 Geth ¦ rakudo/nom: a7b0449375 | usev6++ | src/vm/jvm/runtime/org/perl6/rakudo/RakudoContainerSpec.java
20:29 Geth ¦ rakudo/nom: [jvm] Move handling of Nil back to 'store'
20:29 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a7b0449375
20:31 [TuxCM] joined #perl6-dev
20:34 Zoffix ZOFVM: Files=1278, Tests=152667, 151 wallclock secs (21.59 usr  3.47 sys + 3259.22 cusr 180.73 csys = 3465.01 CPU)
20:36 Geth ¦ rakudo/nom: 530e129461 | (Zoffix Znet)++ | 2 files
20:36 Geth ¦ rakudo/nom: Make Cool.contains 11.5x faster
20:36 Geth ¦ rakudo/nom:
20:36 Geth ¦ rakudo/nom: Part of fixing RT#132280:
20:36 Geth ¦ rakudo/nom: https://rt.perl.org/Ticket/Display.html?id=132280
20:36 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/530e129461
20:36 synopsebot RT#132280 [new]: https://rt.perl.org/Ticket/Display.html?id=132280 [REGRESSION][PERF] Cool methods that take and pass a Capture are uber slower (.contains: 'meow' for ^10000_00)
20:44 Geth ¦ roast: 611988f01f | usev6++ | S02-types/mix.t
20:44 Geth ¦ roast: [jvm] Skip test that blows up whole test file
20:44 Geth ¦ roast: review: https://github.com/perl6/roast/commit/611988f01f
20:50 TimToady joined #perl6-dev
20:51 robertle joined #perl6-dev
20:54 Zoffix weird.. Patched Cool.split and `42.split: "x" for ^1000_00` now got 2.7x faster, but `42.split: /\d/ for ^1000_0` saw no change whatsoever :S
20:55 Zoffix Oh, prolly 'cause it's so damn slow, the dispatch is drowned
20:56 timotimo i'd assume so
20:56 Geth ¦ rakudo/nom: 785d2f251d | (Zoffix Znet)++ | 2 files
20:56 Geth ¦ rakudo/nom: Make Cool.split 0x-2.7x faster
20:56 Geth ¦ rakudo/nom:
20:56 Geth ¦ rakudo/nom: Split with Str matcher is 2.7x faster, but looks like Regex
20:56 Geth ¦ rakudo/nom: matcher is just so slow that dispatch improvement is drowned so there's
20:56 Geth ¦ rakudo/nom: no noticalbe improvement there.
20:56 Geth ¦ rakudo/nom:
20:56 Geth ¦ rakudo/nom: Part of fixing RT#132280:
20:56 Geth ¦ rakudo/nom: https://rt.perl.org/Ticket/Display.html?id=132280
20:56 synopsebot RT#132280 [new]: https://rt.perl.org/Ticket/Display.html?id=132280 [REGRESSION][PERF] Cool methods that take and pass a Capture are uber slower (.contains: 'meow' for ^10000_00)
20:56 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/785d2f251d
21:14 Geth ¦ rakudo/nom: 43e7b893b1 | (Elizabeth Mattijsen)++ | src/core/Rakudo/Iterator.pm
21:14 Geth ¦ rakudo/nom: Fix for RT #132288
21:14 Geth ¦ rakudo/nom:
21:14 Geth ¦ rakudo/nom: Make sure we check in HLL before committing to the natives.
21:14 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/43e7b893b1
21:14 synopsebot RT#132288 [new]: https://rt.perl.org/Ticket/Display.html?id=132288 [REGRESSION][9999] .tail with large negative arguments (say (4,5,6).tail(-9999999999999999999))
21:14 Geth ¦ rakudo/nom: d79ac9704d | (Zoffix Znet)++ | 2 files
21:14 Geth ¦ rakudo/nom: Revert fixes to Cool methods with captures
21:14 Geth ¦ rakudo/nom:
21:14 Geth ¦ rakudo/nom: Do not sweet the problem under the rug and instead fix
21:14 Geth ¦ rakudo/nom: slippy method caching
21:14 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d79ac9704d
21:18 Zoffix *sweep
21:43 MasterDuke joined #perl6-dev
21:43 wander joined #perl6-dev
22:04 Geth ¦ rakudo/nom: b6982e6892 | (Elizabeth Mattijsen)++ | src/core/traits.pm
22:04 Geth ¦ rakudo/nom: Streamline Routine traits
22:04 Geth ¦ rakudo/nom:
22:04 Geth ¦ rakudo/nom: - give name to role for better gisting
22:04 Geth ¦ rakudo/nom: - make sure we only mixin if a trueish value was specified
22:04 Geth ¦ rakudo/nom: - traits affected:
22:04 Geth ¦ rakudo/nom:   - is hidden-from -backtrace
22:04 Geth ¦ rakudo/nom:   - is hidden-from-USAGE
22:04 Geth ¦ rakudo/nom: <…commit message has 7 more lines…>
22:04 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b6982e6892
22:11 AlexDaniel- samcv: hello? :)
22:12 AlexDaniel- samcv: nevermind!
22:12 samcv hi!
22:15 AlexDaniel- samcv: actually I have two questions
22:15 samcv go ahead
22:15 AlexDaniel- I feel like I'm bothering you too much though
22:15 AlexDaniel- anyway
22:15 samcv i am at your disposal
22:15 samcv not at all
22:15 samcv i enjoy sharing my knowledge
22:15 AlexDaniel- c: releases "a".uniprop('Numeric').say
22:15 AlexDaniel- what is “Numeric” property? What does it stand for?
22:15 committable6 AlexDaniel-, ¦2015.12,2016.01.1,2016.02,2016.03,2016.04,2016.05,2016.06,2016.07.1,2016.08.1,2016.09,2017.09,HEAD(b6982e6): «Lower» ¦2016.10,2016.11,2016.12,2017.02,2017.03,2017.04.3,2017.05,2017.06,2017.07,2017.08: «ALetter» ¦2017.01: «None»
22:16 AlexDaniel- it's not the same as Numeric_Type
22:16 AlexDaniel- I guess?
22:17 samcv ah
22:17 AlexDaniel- and second question: is there some indexed documentation for unicode properties so that I can add a link to every row here: https://gist.github.com/Whateverable/4f40fa483aeed22d355a948cd62d663d
22:18 AlexDaniel- (and also so that I don't have to ask the first question ever again :) )
22:18 samcv numeric just means it has a numeric type
22:19 samcv so numeric_type != "None"
22:19 AlexDaniel- ok, but what is “Lower”?
22:19 Geth ¦ rakudo/nom: bb1df2cb0c | (Elizabeth Mattijsen)++ | 3 files
22:19 Geth ¦ rakudo/nom: Streamline "is pure" trait handling
22:19 Geth ¦ rakudo/nom:
22:19 Geth ¦ rakudo/nom: - rename role's method to "is-pure", to be consistent with other traits
22:19 Geth ¦ rakudo/nom: - don't bother calling the method: the existence is the flag
22:19 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/bb1df2cb0c
22:20 samcv well Numeric is not really an official property. and i don't know when that happened but it should be Numeric_Type != None
22:20 AlexDaniel- want me to submit it as a ticket?
22:20 samcv hold on let me check something
22:21 samcv well it seems to be doing MVM_UNICODE_PROPERTY_SENTENCE_BREAK when you ask for Numeric
22:21 samcv so yeah make a ticket i guess
22:23 samcv if numeric should be a property that is up to perl 6 i guess
22:23 samcv ah i know what's going on
22:23 samcv Sentence Break has a Numeric property value
22:23 samcv i'm guessing
22:24 samcv so anyway it's just an artifect of us looking up property values and not just property names
22:26 AlexDaniel- buggable: tags
22:26 buggable AlexDaniel-, Total: 1631; 6.D: 2; 9999: 11; @LARRY: 27; ANNOYING: 8; BOOTSTRAP: 4; BUG: 592; BUILD: 12; CONC: 45; DOCS: 1; EXOTICTEST: 3; FLAP: 1; GLR: 3; IO: 20; JVM: 48; LHF: 7; LTA: 176; MATH: 5; META: 2; MOAR: 1; MOLD: 233; NATIVECALL: 21; NYI: 57; OO: 13; OPTIMIZER: 8; OSX: 2; PARSER: 5; PERF: 26; POD: 19; PRECOMP: 15; REGEX: 46; REGRESSION:
22:26 buggable AlexDaniel-, 39; REPL: 6; RFC: 61; RT: 2; SECURITY: 2; SEGV: 28; SINK: 1; SITE: 1; SPESH: 1; STAR: 7; TESTCOMMITTED: 12; TESTNEEDED: 38; TODO: 13; UNI: 26; UNTAGGED: 282; WEIRD: 2; WINDOWS: 4; See http://fail.rakudo.party/ for details
22:29 AlexDaniel- samcv: RT #132305
22:30 synopsebot RT#132305 [new]: https://rt.perl.org/Ticket/Display.html?id=132305 [UNI] In some cases ‘Numeric’ uniprop returns “Lower” but it should be “None” instead (say ‘a’.uniprop(‘Numeric’))
22:53 AlexDaniel- c: 2016.06,HEAD my @a = ^50; my $a = now; for ^100_000 { my $b = Junction.new("all",@a) }; say now - $a; $a  = now; for ^100_000 { Junction.new("all",@a) }; say now - $a
22:53 committable6 AlexDaniel-, ¦2016.06: «0.18258788␤0.160125» ¦HEAD(bb1df2c): «0.0761381␤0.4081705»
22:54 AlexDaniel- c: 2016.06,HEAD my @a = ^50; my $a = now; for ^100_000 { Junction.new("all",@a) }; say now - $a
22:54 committable6 AlexDaniel-, ¦2016.06: «0.2100153» ¦HEAD(bb1df2c): «0.37509849»
22:54 AlexDaniel- c: 2016.06,HEAD my @a = ^50; my $a = now; for ^1000_000 { Junction.new("all",@a) }; say now - $a
22:54 committable6 AlexDaniel-, ¦2016.06: «1.53712789» ¦HEAD(bb1df2c): «3.37815925»
22:55 AlexDaniel- bisect: old=2016.06 my @a = ^50; my $a = now; for ^1000_000 { Junction.new("all",@a) }; say (now - $a) > 2
22:55 bisectable6 AlexDaniel-, Bisecting by output (old=2016.06 new=bb1df2c) because on both starting points the exit code is 0
22:56 bisectable6 AlexDaniel-, bisect log: https://gist.github.com/eb832cc5450405ae9634b0094a0535cb
22:56 bisectable6 AlexDaniel-, (2016-08-11) https://github.com/rakudo/rakudo/commit/977797fa401856e5310155f13469b7e6ff5f620a
22:57 AlexDaniel- c: 977797fa40^,977797fa40 my @a = ^50; my $a = now; for ^1000_000 { Junction.new("all",@a) }; say now - $a
22:57 committable6 AlexDaniel-, ¦977797fa40^: «1.67519507» ¦977797f: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
22:57 * AlexDaniel- moves to #whateverable
23:08 AlexDaniel- I'm really not sure what to make out of it
23:09 AlexDaniel- c: 2016.06,HEAD my @a = ^50; my $a = now; for ^1000_000 { @a.all }; say now - $a
23:09 committable6 AlexDaniel-, ¦2016.06: «1.7508351» ¦HEAD(bb1df2c): «4.41190691»
23:09 AlexDaniel- c: 2016.06,HEAD my @a = ^50; my $a = now; for ^1000_000 { $ = @a.all }; say now - $a
23:09 committable6 AlexDaniel-, ¦2016.06: «1.7970691» ¦HEAD(bb1df2c): «0.8887317»
23:09 AlexDaniel- so sinking a junction is now slower, but not sinking it is faster
23:09 AlexDaniel- why would somebody sink a junction, right?
23:10 * AlexDaniel- shrugs
23:11 timotimo use the coverage functionality on all regressable snippets :3
23:20 buggable New CPAN upload: App-Bob-0.2.2.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/App-Bob-0.2.2.tar.gz
23:20 buggable New CPAN upload: Hash-Merge-0.1.1.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/Hash-Merge-0.1.1.tar.gz
23:20 buggable New CPAN upload: IRC-Client-Plugin-Github-0.1.4.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/IRC-Client-Plugin-Github-0.1.4.tar.gz
23:20 buggable New CPAN upload: MPD-Client-0.1.3.tar.gz by TYIL https://cpan.metacpan.org/authors/id/T/TY/TYIL/Perl6/MPD-Client-0.1.3.tar.gz
23:23 MasterDuke joined #perl6-dev
23:26 MasterDuke timotimo: was just hoping you know something about why this `my $a = rx/aaaaaab/; 'a' x 999999 ~ 'b' ~~ /$a/` spends a couple seconds in Match.pm's Bool proto
23:28 MasterDuke after my recent commit that created a Regex multi for INTERPOLATE
23:33 TimToady joined #perl6-dev
23:54 evalable6 joined #perl6-dev

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