llfourn Zoffix++, I just read your blog post on quibbles, babbles and nibbles etc. Thanks it was really useful. 02:11
Zoffix \o/ 02:31
llfourn I'm not sure if I should be using nibbles to other grammars in my own grammars and how to actually do that. 02:59
lizmat Files=1150, Tests=53364, 211 wallclock secs (12.85 usr 3.81 sys + 1288.98 cusr 123.14 csys = 1428.78 CPU) 06:06
commute& 06:44
[Tux] This is Rakudo version 2016.10-2-gb744de3 built on MoarVM version 2016.10 07:35
csv-ip5xs 3.258
test 15.809
test-t 6.979
csv-parser 17.524
nine Zoffix: ah sorry! That's what I get for making changes while playing Diablo II. Fixed the test. 08:57
timotimo nine: is optimization for Inline::Python going to happen only when you have a nice benchmark for it? :)
dalek ar: 8ea8fb8 | (Steve Mynott)++ | docs/perl6intro.pdf:
import new build of perl6intro pdf
 08:58
nine timotimo: a benchmark will greatly increase the chances of me optimizing ;) 09:00
timotimo i don't have anything yet, just wondering 09:02
nine I guess I'm a gamer at heart. I see a number, I want to improve it ;) 09:03
FROGGS o/ 09:08
RabidGravy m: gist.github.com/jonathanstowe/6926...1363ba6ba3 09:26
camelia rakudo-moar b744de: OUTPUTĀ«IO::Socket::INET is disallowed in restricted settingā¤ in sub restricted at src/RESTRICTED.setting line 1ā¤ in method new at src/RESTRICTED.setting line 32ā¤ in block <unit> at <tmp> line 5ā¤ā¤Ā»
RabidGravy boo
dalek ar: e017ba0 | (Steve Mynott)++ | tools/star/release-guide.pod:
semi automate the module changelog creation for announcement
 09:40
ar: bd8bdab | (Steve Mynott)++ | tools/star/Makefile:
bump rakudo, nqp and moar versions
ar: bba83a7 | (Steve Mynott)++ | tools/star/release-guide.pod:
correct changelog create cmd
ar: a49ba73 | (Steve Mynott)++ | modules/ (9 files):
bump submodule versions
nine Zoffix: at least I finally had a look at Inline::Perl5's travis now. Seems like we need a rakudo from between 2016.06 (still broken) to 2016.09 (works) for Inline::Perl5 to pass its tests. Or actually it looks like a MoarVM issue. 09:54
dalek ar: 7fcd019 | (Steve Mynott)++ | docs/announce/2016.10.md:
draft announce
 10:27
timotimo + * Overflow check has been removed from infix:<**>(int, int) 10:29
^- was this actually in the last rakudo star?
star: 1000 ** 1000 10:30
camelia star-m 2016.04: OUTPUTĀ«WARNINGS for <tmp>:ā¤Useless use of "**" in expression "1000 ** 1000" in sink context (line 1)ā¤Ā»
timotimo committable6: releases 1000 ** 1000
how do i work this bot again ...
committable6: help
committable6 timotimo, Ā¦Ā«2015.10,2015.11Ā»: WARNINGS:ā¤Useless use of "**" in expression "1000 ** 1000" in sink context (line 1)ā¤Ā¦Ā«2015.12,2016.02,2016.03,2016.04,2016.05,2016.06,2016.07.1,2016.08.1,2016.09,HEADĀ»: WARNINGS for /tmp/dxaq0BOTzG:ā¤Useless use of "**" in expression "1000 ** 1000" in sink context (line 1)
timotimo, Like this: committable6: f583f22,HEAD say ā€˜helloā€™; say ā€˜worldā€™
timotimo maybe that doesn't overflow yet %) 10:31
dalek ar: ec1e99a | (Steve Mynott)++ | docs/announce/2016.10.md:
minor typo fixed
ar: 5d08154 | (Steve Mynott)++ | docs/announce/2016.10.md:
md fix
timotimo m: say 10_000 ** 10_000
camelia rakudo-moar b744de: OUTPUTĀ«1000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000ā€¦Ā»
timotimo oooh, i forgot to native ints
m: my int $a = 10_000; say $a ** $a
camelia rakudo-moar b744de: OUTPUTĀ«0ā¤Ā»
timotimo committable6: releases my int $a = 10_000; say $a ** $a
committable6 timotimo, gist.github.com/8aefa4b51267f0e001...519d942e53
timotimo hehe. 10:32
it b0rked for many releases
dalek ar: f07dc87 | (Steve Mynott)++ | / (2 files):
bump more versions
ast: 5e51f94 | (Tom Browder)++ | README (2 files):
Revise README (#171)

  * rename to use markdown, strive for consistency, add TOC and details on contributing		 10:51
ar: 7138bfb | (Steve Mynott)++ | / (3 files):
add perl6-pod-to-bigpage needed as dep for p6doc
 11:21
ar: 3d06917 | (Steve Mynott)++ | docs/announce/2016.10.md:
update announce
nine Aaaand updated openSUSE packages submitted less than 24 hours after the release :) 12:00
FROGGS nine++ 12:02
nine There were just minor issues to fix in my update scripts. I guess I'll do one or two more releases this way before I add the "osc commit" commands to the scripts and put them all into a single command 12:04
Most of the code is actually just collecting the changelogs for putting them into the .spec file ;) 12:05
dalek ar: 4e70ec2 | (Steve Mynott)++ | / (3 files):
install prove6 only, after chat with leont and install last so not as to slow star src build - which needs perl5 anyway
 12:08
ar: 7ca15a5 | (Steve Mynott)++ | docs/announce/2016.10.md:
doc addition of prove6
FROGGS nine: after that we'd just need to automate bug fixing ;o) 12:11
dalek ast: 67c494c | FROGGS++ | S32-list/pick.t:
add .pick test for RT #109586
 12:25
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=109586
psch grr my isp is being really spotty :| 12:28
also, the attributive sig binding bug on jvm is gonna be fun 12:29
FROGGS :S
psch the issue is, from what i can tell, that we directly as the repr to get the assignee 12:30
which we can't, because that only works for boxed attributes
FROGGS ohh, I guess I need to bump the revisions to that the added test passes....
directly what? 12:31
psch ask
+k :)
but that also means that we can't getattr there either, cause don't get the attribute slot for assignment but the value in a new box... 12:32
+that way
+we
FROGGS hehe
psch the interesting thing is that we don't even seem to hit the check for the case that's not working... 12:33
github.com/rakudo/rakudo/blob/nom/....java#L606 12:34
j: class A { has int $!a; method f(:$!a) { } }; A.new.f: :1a
camelia rakudo-jvm 2a1605: OUTPUTĀ«org.perl6.nqp.sixmodel.reprs.P6OpaqueBaseInstance$BadReferenceRuntimeException: Cannot access a native attribute as a reference attributeā¤ in method f at <tmp> line 1ā¤ in block <unit> at <tmp> line 1ā¤ā¤Ā»
psch so that & SIG_ELEM_BIND_ATTRIBUTIVE doesn't even get hit
MasterDuke psch: is this at all related to the BOOTSTRAPATTR stuff? 12:35
psch MasterDuke: no, i wouldn't think so. i thought you're working on that? 12:37
MasterDuke yeah, i haven't really had a chance to look at it since last chatting, but i hope to get some time to play around today
(but /me wouldn't complain *too* loudly if someone else did all the work) 12:39
psch yeah i don't think this is working here... 12:48
my putty connection dropped like 3 times in the last 20 minutes :/
stmuk pl6anet.org/drop/rakudo-star-2016.10-RC0.tar.gz 12:49
dalek p: 02fd5f5 | FROGGS++ | tools/build/MOAR_REVISION:
bump moar for .pick fix
 12:52
kudo/nom: 34e515d | FROGGS++ | tools/build/NQP_REVISION:
bump nqp/moar for .pick fix
 12:53
stmuk stevemynott.blogspot.co.uk/2016/10/...akudo.html 13:12
travis-ci Rakudo build errored. Tobias Leich 'bump nqp/moar for .pick fix' 13:37
travis-ci.org/rakudo/rakudo/builds/168041380 github.com/rakudo/rakudo/compare/b...e515d64025
Zoffix travis-ci.org/rakudo/rakudo/builds/168041380 13:41
buggable [travis build above] āœ“ All failures are due to timeout (1), missing build log (0), or GitHub connectivity (0). All failures are on JVM only.
stmuk Zoffix: I messed up and didn't assign a title to my blog (due to blogger being junk). Although I corrected it it's still not showing with title on perl6.org 13:44
will that correct itself or does a cache need clearing?
Zoffix stmuk, try again 13:45
stmuk ah yes
:D
Zoffix IIRC it's updated every 15 minutes 13:46
nine "There are a number of advantages to doing so, including elimination of some high-order memory allocations, improved security, and better diagnostic output when a stack overflow occurs. There was just one little problem: an additional 1.5Āµs of kernel overhead during process creation, a cost that Linus was unwilling to accept." 13:59
Oh I'm glad. That would have slowed down rakudo startup by almost 0.002 % ;) 14:00
mst nine: what's that from?
Zoffix What's the opposite of "texas"? What do we call our fancy pants unicode ops?
nine lwn.net/Articles/692953/
Just thought it amusing how relative performance numbers are. 14:01
mst yeah
nine Btw. they succeeded in getting all those benefits without the 1.5us slowdown for those who don't fancy reading the article. 14:02
[Coke] Zoffix: non texas is the normal, regular version; doesn't have a special name. (maybe 'unicode') ?
Zoffix OK
mst I'd just say "unicode operators" and "texas operators" 14:03
I think
stmuk "Rhode Island" 14:20
not very catchy I admit
mst stmuk++
Zoffix m: use nqp; say nqp::div_n(0e0, 0e0) 14:21
camelia rakudo-moar 34e515: OUTPUTĀ«NaNā¤Ā»
Zoffix m: use nqp; say nqp::mod_n(0e0, 0e0) 14:22
camelia rakudo-moar 34e515: OUTPUTĀ«0ā¤Ā»
Zoffix FWIW, currently we have native num division by zero and % check for zero division and return a Failure, so we have that "more than one type returned; will it inline" thing going on there that was the reasoning behind removing overflow detection for infix:<*> and infix:<**> for ints 14:24
Hm. My bench shows adding an explicit `returns int` to a native-operating sub makes it about 2% slower, but I recall jnthn saying whatever you do, don't remove that return type stuff or inlining will mess up. Am I doing something wrong with my test? 14:33
m: gist.github.com/zoffixznet/98dd51b...18d9f7c391
camelia rakudo-moar 34e515: OUTPUTĀ«startingā¤Before: 1.003957ā¤ After: 1.01247476ā¤ Diff: -0.84%ā¤Ā»
Zoffix m: gist.github.com/zoffixznet/98dd51b...18d9f7c391
camelia rakudo-moar 34e515: OUTPUTĀ«startingā¤Before: 1.0089375ā¤ After: 0.9993584ā¤ Diff: 0.95%ā¤Ā»
nine Zoffix: he could simple have meant possible future optimizations that need this return type info 14:35
Zoffix OK
nine Are we able to profile rakudo's startup somehow? 14:36
dalek ast: eeb5887 | (Zoffix Znet)++ | S02-types/num.t:
[coverage] Cover nummy prefix:<->, abs, infix:<+>, and infix:<*>
 14:41
kudo/nom: 2b5226c | (Zoffix Znet)++ | src/core/Num.pm:
Add return type constraints for native nummy subs

To allow for current/future inlining/optimizations based on knowledge of return types.		 16:53
Zoffix m: my num $x = -125e0; my num $y = (1/3).Num; dd $x**$y 17:27
camelia rakudo-moar 2b5226: OUTPUTĀ«NaNā¤Ā»
Zoffix Ah, it's the logged one already rt.perl.org/Public/Bug/Display.htm...et-history 17:29
travis-ci Rakudo build passed. Zoffix Znet 'Add return type constraints for native nummy subs 17:39
travis-ci.org/rakudo/rakudo/builds/168076021 github.com/rakudo/rakudo/compare/3...5226ccd35d
Zoffix m: say NaN**0 17:47
camelia rakudo-moar 2b5226: OUTPUTĀ«1ā¤Ā»
Zoffix I thought any operation on a NaN gave NaN
AlexDaniel committable6: releases say NaN**0
committable6 AlexDaniel, Ā¦Ā«2015.10,2015.11,2015.12,2016.02,2016.03,2016.04,2016.05,2016.06,2016.07.1,2016.08.1,2016.09,HEADĀ»: 1
AlexDaniel m: say 42**NaN 17:48
camelia rakudo-moar 2b5226: OUTPUTĀ«NaNā¤Ā»
Zoffix Perl 5 gives a NaN
AlexDaniel m: say 1**NaN
camelia rakudo-moar 2b5226: OUTPUTĀ«1ā¤Ā»
Zoffix :o
AlexDaniel m: say 0**NaN
camelia rakudo-moar 2b5226: OUTPUTĀ«NaNā¤Ā» 17:49
Zoffix Feels like a shortcut logic there somewhere isn't watching for a nan
AlexDaniel SourceBaby: infix:<**>(1, NaN)
Zoffix Nananananananananana bugman
AlexDaniel SourceBaby: help
SourceBaby AlexDaniel, Use s: trigger with args to give to sourcery sub. e.g. s: Int, 'base'. See modules.perl6.org/dist/CoreHackers::Sourcery
Zoffix s: &infix:<**>, \(1, NaN) 17:50
SourceBaby Zoffix, Sauce is at github.com/rakudo/rakudo/blob/2b52...al.pm#L141
AlexDaniel ah
what is Bridge
Zoffix The issue is at nqp level (or below)
m: use nqp; dd nqp::pow_n(NaN, 0e0) 17:51
camelia rakudo-moar 2b5226: OUTPUTĀ«1e0ā¤Ā»
Zoffix j: use nqp; dd nqp::pow_n(NaN, 0e0)
camelia rakudo-jvm 2a1605: OUTPUTĀ«1e0ā¤Ā»
AlexDaniel bisect: say 1**NaN
bisectable6 AlexDaniel, On both starting points (old=2015.12 new=34e515d) the exit code is 0 and the output is identical as well
AlexDaniel, Output on both points: 1
AlexDaniel RT #129894 17:55
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=129894
Zoffix m: NaN**1i
camelia rakudo-moar 2b5226: OUTPUTĀ«WARNINGS for <tmp>:ā¤Useless use of "**" in expression "NaN**1i" in sink context (line 1)ā¤Ā»
Zoffix m: say NaN**1i
camelia rakudo-moar 2b5226: OUTPUTĀ«NaN+NaN\iā¤Ā»
AlexDaniel Zoffix: haha, that's a good one: -1**NaN 17:56
Zoffix m: say (-1)**NaN
camelia rakudo-moar 2b5226: OUTPUTĀ«NaNā¤Ā»
Zoffix Works fine.
AlexDaniel ahhhhhā€¦ parensā€¦
I'll goā€¦ create a doc issue perhapsā€¦ 17:57
Zoffix m: use Test; is-deeply (my num $ = 125e0) ** (my num $ = (1/3).Num), (my num $ = 5e0) 17:59
camelia rakudo-moar 2b5226: OUTPUTĀ«not ok 1 - ā¤ā¤# Failed test at <tmp> line 1ā¤# expected: 5e0ā¤# got: 5e0ā¤Ā»
Zoffix heh
AlexDaniel surprisingly I don't have many RTs with NaNs 18:03
dalek ast: 47aba2d | (Zoffix Znet)++ | S02-types/num.t:
[coverage] Cover nummy infix:<%> and infix:<**>(num, num)

Includes partial fudge for RT#129894: rt.perl.org/Ticket/Display.html?id=129894
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=129894
AlexDaniel chances are there is a bunch of other similar bugs related to NaN 18:04
Zoffix: should I also create a ticket for say NaN**0 ?
Zoffix m: say NaN**0
camelia rakudo-moar 2b5226: OUTPUTĀ«1ā¤Ā»
Zoffix AlexDaniel, it's all part of that, innit
AlexDaniel Zoffix: OK I'll then leave a comment on that ticket 18:05
dalek kudo/nom: 3cebff6 | (Dominique Dumont)++ | docs/running.pod:
Fix pod syntax error in running.pod

Added the missing =back found by podchecker		 18:06
kudo/nom: ae1f2f8 | (Zoffix Znet)++ | docs/running.pod:
Merge pull request #905 from dod38fr/nom

Fix pod syntax error in running.pod
Zoffix I'm a bit suspicious about this // 0 here: github.com/rakudo/rakudo/blob/ae1f...ngy.pm#L26 21:53
What's the usecase for it? Right now it silences errors.
m: my $x = 5 x '10sadsadas'; say $x
camelia rakudo-moar ae1f2f: OUTPUTĀ«ā¤Ā»
Zoffix Hmmmm... github.com/rakudo/rakudo/commit/23...c51bc75a69 21:56
Zoffix changes that to something better 21:58
lucasb_ Zoffix: do you really think that '// 0' is necessary? 22:13
after all, 'a' + 'b' raises an error
I think, so should 'a' x 'b'
Zoffix I know
lucasb_ m: say 'a' xx 3 22:19
camelia rakudo-moar ae1f2f: OUTPUTĀ«(a a a)ā¤Ā»
lucasb_ oops, sorry
m: say 'a' xx 'b'
camelia rakudo-moar ae1f2f: OUTPUTĀ«This type cannot unbox to a native integer: P6opaque, Failureā¤ in block <unit> at <tmp> line 1ā¤ā¤Ā»
lucasb_ ^^ I think this settles the discussion. 'a' x 'b' should throw :)
Zoffix thought `* Zoffix changes that to something better` did it :) 22:20
'cause the commit indicates failure silencing wasn't the reason for the addition 22:21
lucasb_ Zoffix: oh, sorry. I missed that :)
dalek rakudo/nom: 2dd0ddb | (Zoffix Znet)++ | src/core/Stringy.pm: 22:40
rakudo/nom: Fix infix:<x> accidentally silencing failures
rakudo/nom:
rakudo/nom: An earlier fix[^1] to interpret type objects as 0 iterations had an unintended
rakudo/nom: effect that Failures from the coercion of the RHS argument to Int would also be
rakudo/nom: silently treaded as 0.
rakudo/nom:
rakudo/nom: Add a separate multi for type objects to handle that case, leaving
rakudo/nom: coercion Failures able to freely explode.
rakudo/nom:
rakudo/nom: Fixes RT#129899: rt.perl.org/Ticket/Display.html?id=129899
rakudo/nom: [1] github.com/rakudo/rakudo/commit/23...c51bc75a69
ast: df528f7 | (Zoffix Znet)++ | S03-operators/repeat.t:
Test infix:<x> does not silence Failures

RT#129899: rt.perl.org/Ticket/Display.html?id=129899
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=129899
lizmat waves from London 23:04
Zoffix waves from an hour away from London
... but I'm guessing the wrong London :P
geekosaur ontario? :p 23:06
Zoffix Yeah :)
japhb waves from San Jose (the one in California)
lizmat hits the bar& 23:08
japhb hits the space bar 23:09
... but not too hard
geekosaur tries to remember where he is relative to there...
hah, not quite directly south 23:10
but about equidistant from the border 23:11
gfldex naive implementation of ==> : gist.github.com/gfldex/798aa121207...ff1900660f 23:17

Powered by Cro and the RakuĀ® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!