samcv AlexDaniel: yay github.com/rakudo/rakudo/issues/1770 is fixed now 00:00
AlexDaniel samcv: are there any tests? 00:01
Geth roast: 43f48c2b83 | (Samantha McVey)++ | S15-unicode-information/uniname.t
Fix incorrect CJK ideograph uniname

This is related to github.com/rakudo/rakudo/issues/1770 Removes the brackets and fixes the name which also changes to uppercase.
samcv in nqp there are new tests
AlexDaniel ah I see
samcv i'll add some to roast as well in a bit
AlexDaniel \o/ 00:02
samcv hopefully that won't break anything too badly 00:03
i had to fix rakudo's tests (which get loaded from Unicode's data file in the rakudo/rakudo repo). there was only one test that changed in roast 00:04
Zoffix t/nqp/106-unicodenames.t seems to still fail for me 00:05
samcv should be passing with updated moar
unless you just recompiled it
Zoffix I'm doing a clean build with zscript: gist.github.com/zoffixznet/5192257...0027bacd9f 00:07
samcv you seem to be right. the test number is wrong 00:08
Geth nqp: 45ab6e373b | (Samantha McVey)++ | t/nqp/106-unicodenames.t
Fix incorrect # of tests from previous commit
 00:09
rakudo: a51112ebc4 | (Samantha McVey)++ | tools/build/NQP_REVISION
Bump NQP to correct failing nqp test

Changes: 2018.04-2-g81373e88f..2018.04-3-g45ab6e373 45ab6e373 Fix incorrect # of tests from previous commit		 00:11
Ā¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....g45ab6e373
travis-ci NQP build failed. Samantha McVey 'Bump MoarVM and add tests for the Unicode name changes 00:12
travis-ci.org/perl6/nqp/builds/371325829 github.com/perl6/nqp/compare/bc0b5...373e88f2f6
Zoffix (didn't do anything today with the nqp num precisionā€”too tired or something; plan to work on it tomorrow in 19 hours, unless someone else beats me) 01:59
(and if anyone wants to try to beat me; the wrongness starts off here: github.com/MoarVM/MoarVM/blob/mast...num.c#L295 and there are two problems: the function doesn't choose the closest representable num for the given string such as parsing 3.145e-20 into 3.14149999999999969e-20 instead of the closest representable num 3.14150000000000029e-20 and it also can't handle denormals 02:01
(5e-324 parses as 0) )
(the precision stuff affects at least 32-bit boxes; fom 02:02
(the precision stuff affects at least 32-bit boxes; don't know if 64-bit are also affected and the denormals are affected on 64 for sure)
nqp: say(nqp::sprintf("%.19g", [3.145e-20])) 02:04
camelia 3.1449999999999997e-20
Zoffix (yeah, precision issue affects 64-bit too) 02:05
Really wonder how come the failing test fails only on 32-bit. 02:10
Ah, mistery solved. I missed a 1 in my eval above. 02:17
nqp: say(nqp::sprintf("%.19g", [3.1415e-20]))
camelia 3.1415e-20
Zoffix nqp: say(nqp::sprintf("%.20g", [3.1415e-20]))
camelia 3.1415e-20
Zoffix And on 32-bit it gives 3.1414999999999997e-20 02:18
C gives 3.1415000000000002898e-20 and it rounds to 142 02:19
m: say 3.1415e-20
camelia 3.1415e-20
Zoffix (^ that is OK; we use the Grisu stuff to stringify, which shortens it as long as it can be roundtripped back to original and it would be if we didn't have the pricision issue in parsing) 02:20
Kaiepi make stresstest is a real pain to get to work with rakudobot without it disconnecting 04:35
using schedulers or threads works but i haven't tested whether i can get the command output with those or not 04:36
samcv m: my %hash = <a a b b c c>; say %hash.keys 04:40
camelia (a b c)
samcv my %hash = <a a b b c c>; say %hash.keys
evalable6 (c b a)
[Tux] Rakudo version 2018.04-13-ga51112ebc - MoarVM version 2018.04-18-ge7d79d869
csv-ip5xs0.877 - 0.899
csv-ip5xs-208.864 - 8.909
csv-parser38.282 - 39.119
csv-test-xs-200.444 - 0.464
test8.769 - 9.192
test-t2.465 - 2.469
test-t --race1.010 - 1.040
test-t-2043.957 - 44.839
test-t-20 --race15.619 - 15.874
 06:56
due to a paste error, I was doing some tests where $PERL6LIB was set to 'inst' instead of the full spec 'inst#/full/path/to/install' 06:57
that causes a lot of noise. I'd prefer a warning/error if PERL6LIB is set to something illogical
Geth roast: b2c864690b | (Stefan Seifert)++ | S06-signature/slurpy-and-interpolation.t
Fix flappy test in slurpy-and-interpolation.t

We must not rely on hash keys appearing in a certain order		 07:15
nine That was probably one of the easiest flappers to fix :D 07:17
lizmat hmmm...now "make install" stopped working for me :-( 07:50
Invocant of method 'path-spec' must be an object instance of type 'CompUnit::Repository::Locally', not a type object of type 'CompUnit::Repository::Staging'. Did you forget a '.new'? 07:51
in block <unit> at tools/build/install-core-dist.pl line 17
El_Che wow
lizmat full backtrace: gist.github.com/lizmat/b40058cee47...e3a0c888a3 07:53
Kaiepi PufferBot, all 08:42
PufferBot [openbsd] Running complete Rakudo build and tests (this will take a while)...
Kaiepi it's not kidding btw, don't expect it to finish for 20-30 minutes 08:51
El_Che lizmat: you mean rakudo or a module? 08:57
PufferBot [openbsd] Failed to build Rakudo and run all tests... See the output at fpaste.scsys.co.uk/577283 09:01
Kaiepi anyone know which of these tests are already known about? ignore the IO::Socket::Async failures, that's beecause i don't have ipv6 set up 09:04
BeastieBot, all 09:10
BeastieBot [freebsd] Running complete Rakudo build and tests... (this will take a while)...
[freebsd] Complete build and tests failed. See the output at fpaste.scsys.co.uk/577284 09:12
Kaiepi oh please 09:13
forgot to reinstall the bot after updating 09:14
BeastieBot, all 09:20
BeastieBot [freebsd] Running complete Rakudo build and tests... (this will take a while)...
[freebsd] Complete build and tests failed. See the output at fpaste.scsys.co.uk/577285
Kaiepi odd that doesn't happen on my system 09:21
BeastieBot, test 09:23
BeastieBot [freebsd] Running tests...
Kaiepi (builds if it isn't already) 09:25
guess we won't know if it builds or not until i get more ram for my vps, the kernel just killed it 09:29
Geth rakudo: 7b33651962 | (Jonathan Worthington)++ | src/core/IO/Socket/Async.pm6
Make sure to wrap error string in Exception object

The user-facing parts of the Supply API all make sure of this. If we
  are directly implementing the Tappable interface, however, then it is
on us to make sure that we only ever call the provided `quit` callback with an `Exception` object, not a `Str`. This mistake was introduced in f58ac9991 during an optimization of IO::Socket::Async.
With this, #1775 now reports the error ("connection reset by peer") correctly, rather than complaining about being unable to throw a string.		 10:04
synopsebot RAKUDO#1775 [open]: github.com/rakudo/rakudo/issues/1775 [ASYNC] Crash in .Supply.tap + .print on an IO::Socket::Async
nqp: d0b1b68bd6 | (Elizabeth Mattijsen)++ | tools/build/MOAR_REVISION
Bump Moar
 10:34
Ā¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...9-g1b78b04
AlexDaniel Kaiepi: pretty sure your question was about something else, but the list of floppers is here: gist.github.com/Whateverable/70e60...3bd7e045b2 10:51
travis-ci NQP build failed. Elizabeth Mattijsen 'Bump Moar'
travis-ci.org/perl6/nqp/builds/371491583 github.com/perl6/nqp/compare/45ab6...b1b68bd614
Geth rakudo: 50f768315d | (Elizabeth Mattijsen)++ | src/core/IO/Path.pm6
Revert "Condense three IO::Path.new multis down to one..."

This reverts commit 4120b226b550a485ddcbc045b9d5fd63d853d6af.
For some reason, this borked "make install" for me on MacOS.
AlexDaniel o_o 10:58
Geth rakudo: 29b092c940 | (Elizabeth Mattijsen)++ | tools/build/NQP_REVISION
Bump NQP to get latest NFA fixes
 11:01
Ā¦ rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2018....4-gd0b1b68
Zoffix .tell MasterDuke FYI: github.com/rakudo/rakudo/commit/50f768315d 11:09
yoleaux Zoffix: I'll pass your message to MasterDuke.
Zoffix Weird. Don't really see a difference that multi change commit introduces :S 11:14
lizmat neither :-(
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Revert "Condense three IO::Path.new multis down to one..." 11:30
travis-ci.org/rakudo/rakudo/builds/371497277 github.com/rakudo/rakudo/compare/7...f768315dc9
buggable [travis build above] ā˜  All failures are due to: failed make test (1 failure). Across all jobs, only t/02-rakudo/repl.t test file failed.
lizmat repl.t test passes for me :-( 11:31
travis-ci Rakudo build failed. Elizabeth Mattijsen 'Bump NQP to get latest NFA fixes' 12:07
travis-ci.org/rakudo/rakudo/builds/371500707 github.com/rakudo/rakudo/compare/5...b092c940b7
buggable [travis build above] ā˜  All failures are due to: failed make test (1 failure). Across all jobs, only t/06-telemetry/01-basic.t test file failed. 12:08
El_Che AlexDaniel: I transferred the nxadm/rakudo-pkg-canary to perl6/rakudo-pkg-canary, but now I have no admin access. Can you or an other admin enable the travis and dockerhub integration in settings?
AlexDaniel hmmm I have no admin access there also 12:09
Zoffix ā†‘ ?
github.com/perl6/rakudo-pkg-canary
El_Che thx
chansen_ m: say 0x4E00.uniname; 12:10
camelia CJK UNIFIED IDEOGRAPH-4E00
Zoffix I gave both you admins
chansen_ m: say 0x17000.uniname;
camelia TANGUT IDEOGRAPH-17000
AlexDaniel Zoffix: thank you ā™„
chansen_ samcv++
AlexDaniel \o/
El_Che Zoffix: thx, will enable the integration now
chansen_ m: say 0xAC00.uniname; 12:11
camelia Hangul Syllable GA
chansen_ :(
AlexDaniel chansen_: if that's wrong, please report it
chansen_ Unicode 4.8: 12:12
Only Latin capital letters A to Z (U+0041..U+0056), ASCII digits (U+0030.. U+0039), U+0020 space, and U+002D hyphen-minus occur in character names.
timotimo ok, so should be uppercase
chansen_ m: say 0x4E00.uniname; 12:13
camelia CJK UNIFIED IDEOGRAPH-4E00
chansen_ m: say 0x0.uniname; 12:14
camelia <control-0000>
chansen_ m: say 0xD800.uniname;
camelia <Non Private Use High Surrogate-D800>
chansen_ :(
AlexDaniel: will do 12:15
timotimo does unicode actually give control characters a name, too? or is this something we invented so it doesn't just give The Empty String? 12:19
chansen_ In unicode, only graphical characters are named, for non-graphical characters Unicode specify a code point label 12:21
0xD800.uniname should have returned <private-use-D800> 12:23
It's documented in Unicode Name Property, chapter 4.8
AlexDaniel 6c: say 0xD800.uniname; 12:24
committable6 AlexDaniel, gist.github.com/6e702838b03a905191...01a6fd0c29
chansen_ m: say 0x0378.uniname; 12:44
camelia <reserved-0378>
chansen_ m: say 0x10FFFF.uniname; 12:45
camelia <noncharacter-10FFFF>
AlexDaniel m: say 0x100FFFF.uniname; 12:46
camelia <unassigned>
chansen_ Thats way beyond Unicode, not covered by the spec =) 12:48
AlexDaniel :) 12:49
chansen_ Someone with powers, please reopen <github.com/rakudo/rakudo/issues/1770> 12:50
Kaiepi AlexDaniel, yeah that was what i was looking for, thanks 12:51
AlexDaniel chansen_: reopened, but next time feel free to start a new one 12:52
althogh, ok, samcv said to reopen :)
fine :)
e: for ^0x10FFFF { say .uniname if !.uniname.starts-with(ā€˜<ā€™) and .uniname.match(/<[a..z]>/) } 12:55
evalable6 Hangul Syllable GA
Hangul Syllable GAG
Hangul Syllable GAGG
Hangul Syllable GAGS
Hangul Syā€¦
AlexDaniel, Full output: gist.github.com/ced864235834bfeff9...f1d4710791
AlexDaniel chansen_: is there a tool or something that can return correct name for any codepoint? 12:57
so that we can test perl6 uninames against it 12:58
chansen_ AlexDaniel: <www.unicode.org/Public/10.0.0/ucd/...me.txt> 13:00
Entries with trailing -* should be replaced with -%.4X, ordinal
It doesn't cover code-point-labels but it should be trivial to test 13:03
AlexDaniel ok I think only hangul syllables are not matching 13:04
chansen_ nogaste? 13:09
Rakudo derives it's character names from UnicodeData.txt, these are the ranges specified in UnicodeData.txt (v10.0.0) <gist.github.com/chansen/f8471ea2ff...f4e49f> 13:12
lizmat fwiw, t/06-telemetry/01-basic.t doesn't fail for me 13:14
afk again&
AlexDaniel well yeah, most of these ranges are arguably wrong 13:17
[Coke]_ zoffix privmsg me when you have a sec. 13:25
robertle was there and outcome of that i386 test failure with 2018.04? do we know what needs fixing? 14:58
dogbert2_ m: say $*IN.lines.head(2) 15:02
camelia (Ā»Wann treffen wir drei wieder zusamm?Ā« Ā»Um die siebente Stundā€˜, am BrĆ¼ckendamm.Ā«)
Geth roast: 8e5ac9c294 | (Steve Mynott)++ | 3 files
replace use of README.md (which doesn't exist on 6.c-errata) by $?FILE
 15:03
roast: cdfc8b39db | (Steve Mynott)++ (committed using GitHub Web editor) | integration/advent2011-day24.t
Update advent2011-day24.t
roast: c10315f5b4 | (Steve Mynott)++ (committed using GitHub Web editor) | 3 files
Merge pull request #421 from stmuk/master

replace use of README.md (which doesn't exist on 6.c-errata) by $?FILE
Zoffix robertle: irclog.perlgeek.de/perl6-dev/2018-...i_16096169 15:05
robertle great, thanks! 15:06
tony-o timotimo: i see file::find and shell::command in the cro meta depends: github.com/croservices/cro/blob/ma...6.json#L10 15:58
oh, oops, i see what you mean
you're right, one is the build-depends 15:59
Geth roast/6.c-errata: 6a39961d60 | (Steve Mynott)++ (committed by Aleks-Daniel Jakimenko-Aleksejev) | 3 files
Replace use of README.md by $?FILE

  README.md doesn't exist on 6.c-errata.		 16:47
AlexDaniel stmuk++
lizmat off for some infinitum bellum 17:23
&
AlexDaniel robertle: hey, I'm wondering, is there any way to get rakudo tested using debian infrastructure before the release? 17:33
robertle: for example, we now have El_Che's packages tested with prerelease rakudo: github.com/perl6/rakudo-pkg-canary
not yet functional but it will be soon I believe
robertle AlexDaniel: it's difficult...
AlexDaniel robertle: that at least will give us some info about i386 and maybe in the future also arm 17:34
robertle it's easy for amd64, armhf and other common architectures. harder for mips* ppc* s390x...
AlexDaniel robertle: but big endianā€¦
yeahā€¦
robertle I think I could find a mipsel box, but bigendian mips and otehrs are hard to come by 17:35
AlexDaniel right
robertle: ok but even for common archs, what should be done on our side to get it tested a bit before the release?
robertle: e.g. we now have this, does it help? rakudo.perl6.org/downloads/prerelease/ 17:37
robertle not sure, I guess we could set something CI-like up
but the biggest problem at the moment, from my point of view, is that the mips failures are not reproducible at all
if we had reproducible failures across the board, life would be aesy 17:38
AlexDaniel do we even have any failures after all of the fixes?
or what do you mean?
robertle: also, looking at the current buildd page, does it mean that 2018.04 is good on all archs except i386? 17:40
robertle I get this: buildd.debian.org/status/fetch.php...&raw=0 but I have so far been unable to get a fix on it. 17:41
if I run iot by hand on another mips box, I get ~30 test suite runs before I get a failure
and mips is slow
AlexDaniel well, that's somewhat positive :) 17:42
robertle I'll have another look later, perhaps 2018.04 is easier to get a fix on
AlexDaniel robertle++
samcv j: 0xAC00.uniname.say 17:57
camelia HANGUL SYLLABLES AC00
samcv :( thanks JVM! 17:58
haha
El_Che samcv: your work on Solaris, was it tested on Solaris 11? 18:19
samcv El_Che: yes 18:21
El_Che I saw it on the comments :)
pkg install ...
:)
tried on 10 yesterday, but it didn't build
samcv ah 18:22
El_Che machines were too slow to figure out why
(they were having some hw issues)
samcv i only tested it on 32 bit. but that's good since i fixed both an issue that'd apply to both 32 and 64 bit with libtommath not building
El_Che they == CSW
samcv and then a 32 bit only issue
i should dl solaris 10 then and test
El_Che I think you'd be surprise by the % of the SOlaris 10 share 18:23
samcv El_Che: not at all hah
El_Che 11 was so different that a lot of people just moved to linux from 10
samcv can you open a moarvm issue?
El_Che those who sayed often run 10
samcv it's moarvm that doesn't build right?
El_Che I'll run it again once the hw is ok 18:24
disk was deadslow nfs
samcv i'm dling solaris 10 x86 right now. did you test on sparc?
El_Che on both 18:25
samcv ok
El_Che AlexDaniel, samcv: there seems to be a bug on Rakudo + --gen-moar that I don't hit with rakudo-pkg because I build each component separatedly: bbs.archlinux.org/viewtopic.php?id=227185 19:31
example failure: bbs.archlinux.org/viewtopic.php?id=227185 19:32
oops: travis-ci.org/nxadm/rakudo-pkg-can...6961#L3176
AlexDaniel We couldn't find the repository nxadm/rakudo-pkg-canary 19:33
El_Che it's perl6/rakudo-pkg-canary
AlexDaniel stmuk: ā†‘ you should probably see that too :) 19:34
El_Che AlexDaniel: moved to test a know working release while testing the setup 19:35
AlexDaniel k
El_Che the prerelease requires a non existent moar
2018.05
AlexDaniel lol rly 19:37
El_Che (I moved to building each component apart on rakudo-pkg because Zoffix invented dot releases :P
AlexDaniel El_Che: hey uh, can you clarify where's that?
El_Che: I looked at prerelease nqp MOAR_REVISION and it's 2018.04-25-g717ee970f 19:38
El_Che with that moving around is not easy
once opensuse works, I'll put it on prerelease again and we'll see
AlexDaniel k
Kaiepi i'm upgrading my vps, so beastiebot should be able to run make build and make stresstest properly soon 19:39
El_Che do we live with failed alpine builds for now, or do we do a workaround (build the 3 components separatedly)?
AlexDaniel weekly: El_Che++ was working on a rakudo package canary ( github.com/nxadm/rakudo-pkg-canary ). The idea is to have CI-like thingy which tests that the upcoming release builds on different platforms correctly 19:40
notable6 AlexDaniel, Noted!
AlexDaniel is clearly very good at explaining stuff 19:41
El_Che weekly: it has an older brother: rakudo-pkg-for-doc: a package with Rakudo , zef + all the perl6 dependencies needed to test the perl6 doc. 19:42
notable6 El_Che, Noted!
El_Che jmerelo is trying it out
compared to rakudobrew, the releasable run time is waaay down. To 3m or something like that instead of 15+ 19:43
AlexDaniel ye that's pretty cool 19:44
stmuk AlexDaniel: thanks! 19:47
Kaiepi BeastieBot, all 20:03
BeastieBot [freebsd] Running complete Rakudo build and tests (this will take a while)...
El_Che as a musing. I wonder if the build-rakudo-in-one-command is a good idea. It breaks unix conventions by writing at configure time (needed, otherwise rakudo can not build) and people that compile the code wouldn't mind the extra steps. Maybe a wrapper script will suffice as a one line compile instruction? 20:05
Kaiepi ah perfect upgrading fixed the kernel constantly trying to kill make 20:06
El_Che AlexDaniel: ok, found a workaround for alpine 20:07
it's a shell thing
Configure.pl is a bashist! 20:08
Kaiepi i'm not sure what to think about having a script that builds rakudo all at once
El_Che (alpine comes with sh by default, bash can be installed
)
Kaiepi ditto *BSD
AlexDaniel El_Che: what is using bashism exactly?
El_Che Kaiepi: something for people that really want that. I prefer the 3 step way
AlexDaniel that doesn't sound good at all
Kaiepi there's --make-install to do it in two steps 20:09
El_Che AlexDaniel: need to pinpoint it, but fact is that it works in bash and not in sh
Kaiepi sh doesn't have arrays 20:10
iirc
El_Che it's related to path or var
in sh --prefix=/opt/rakudo-pkg ends in
/opt/rakudo-pkg/opt/rakudo/pkg
moar is installed there 20:11
and the rest of the install can't find it
AlexDaniel: everything fails with the prerelease: travis-ci.org/nxadm/rakudo-pkg-can.../371733453 (while everything builds with 2018.03) 20:16
error: pathspec '2018.05' did not match any file(s) known to git.
travis-ci.org/nxadm/rakudo-pkg-can...3457#L2986
Kaiepi BeastieBot, stresstest 20:17
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
AlexDaniel El_Che: ok, so what is this 2018.05?
El_Che: that's nqp version right?
El_Che AlexDaniel: you tell me, it's from the release tar
AlexDaniel El_Che: so-o-oā€¦ can you use prerelease nqp?
El_Che prerelease
so, we're back to separate builds? I am all for it 20:18
Kaiepi BeastieBot, stresstest
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
AlexDaniel El_Che: well, I guess soā€¦ obviously you can't pull nqp 2018.05 because that doesn't exist, but the nqp prerelease tar *is* 2018.05, or at least it claims to be 20:19
El_Che woudn't we need a prerelease moarvm?
or which one should I use? 20:20
I would say put one there, even if it's a copy of a release
AlexDaniel El_Che: you can just use --gen-moar when building nqp
that soundsā€¦ werid
Kaiepi hm the bot's hanging when it tries to install Inline::Perl5
El_Che I think my suggestion is better, in case samcv wants to test as well?
Kaiepi but doesn't when i do it manually 20:21
El_Che AlexDaniel: you're the main user, so it's up to you
AlexDaniel El_Che: IMO we don't need an extra effort to make moarvm prereleases 20:22
El_Che deal, i'll go with --gen-moar
AlexDaniel El_Che: alternatively I can change releasable sakefile so that it doesn't use ā€œ2018.05ā€ nqp release, and instead points at some commit like 2018.04-666-g123123 20:23
this way we'd only need rakudo tars really
El_Che deal
i'll implement the alpine workaround
AlexDaniel El_Che: wait, so we go with --gen-nqp and --gen-moar at the same time, right? 20:24
El_Che yes
AlexDaniel k
El_Che I had this: perl ./Configure.pl --backend=moar --gen-moar --prefix=$PREFIX 20:25
and nqp got built as well
Kaiepi nqp has some code related to building with moar iirc 20:27
that's probably why
AlexDaniel El_Che: ok just leave it like this I guess 20:29
El_Che: OK building it, it'll take ā‰ˆ1 hour actually 20:31
the sakefile basically follows all steps in the release guide just like if it was making a release, which includes stresstest and stuff
and when something flops, it tries againā€¦
so yeah, patience is a virtue :) 20:32
El_Che AlexDaniel: I am under the impression that --gen-nqp fixes the alping problem 20:34
nope :) 20:43
AlexDaniel oh wow github just reported a potential security vulnerability in one of the deps for nqp js 21:01
that's pretty cool!
not that it matters because I guess no one is using rakudo on js yet, but still, that's lovely
Kaiepi oh yeah i forgot i wanted to take a look at the js compiler 21:03
there's a lot of code in it that's suboptimal 21:05
though it's modtly stuff that you wouldn't expect to be so
El_Che AlexDaniel: it's an option on the setting, I saw recently 21:12
pmurias Kaiepi: you mean in the runtime code or the generated stuff? 21:35
Kaiepi: (I haven't optimized the code generated for full Perl 6 at all)
Kaiepi what do you mean by the generated stuff? 21:36
El_Che testing rakudo on ubuntu 18.04
pmurias Kaiepi: the code that the compiler emits
Kaiepi the generated stuff i'm assuming 21:37
pmurias the code that's emitted needs a ton of improvement 21:38
Kaiepi: optimizing the code generated for NQP seems more important at first as that will speed up the compiling of the setting etc. 21:44
AlexDaniel c: HEAD say ('WAT~' xx 15000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems 21:46
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«15000ā¤Ā»
AlexDaniel c: HEAD say ('WAT~' xx 15000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«15000ā¤Ā»
Kaiepi i'm asking to compare it to node's, not so much to work on it
AlexDaniel c: HEAD say ('WAT~' xx 15000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«15000ā¤Ā»
AlexDaniel come onā€¦
c: HEAD say ('WAT~' xx 15000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«15000ā¤Ā»
AlexDaniel :|
c: HEAD say ('WAT~' xx 150000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«149643ā¤Ā»
AlexDaniel c: HEAD say ('WAT~' xx 150000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«149643ā¤Ā»
AlexDaniel c: HEAD say ('WAT~' xx 150000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems 21:47
Kaiepi i could work on it, but i need more experience working on moar and nqp first
committable6 AlexDaniel, Ā¦HEAD(29b092c): Ā«149533ā¤Ā»
AlexDaniel m: say ('WAT~' xx 150000).grep({ / ^ <[ \w z ]>+ ā€˜~ā€™ $ / }).elems
camelia 150000
AlexDaniel hehe
dogbert17++
dogbert17: I don't think it golfs any shorter than that
dogbert17: and yes, it looks like it needs the system to be busy quite a bit
right now there's a spectest running on whateverable server 21:48
dogbert17: btw where did you find this?
Gah! 21:59
El_Che: ok not so quick reallyā€¦
there's something wrong with Inline::Perl5
AlexDaniel tries it 22:00
oh, actually it's some sort of a problem with File::Which
dogbert17 AlexDaniel: creds go to dmaestro: github.com/MoarVM/MoarVM/issues/775
tried to golf i down a tiny bit 22:01
AlexDaniel this: gist.github.com/AlexDaniel/3389a67...65fd4468ec 22:07
Kaiepi oh i forgot i don't need to kill pufferbot/beastiebot anymore when i'm upgrading rakudo 22:30
El_Che AlexDaniel: repo is ready. Released rakudo build fine on all OSes including Ubuntu 18.04. Once the Docker Hub finishes building the latest images we'll share a few minutes off the build time. 22:34
AlexDaniel El_Che++ 22:35
El_Che mm, rakudo fails on alpine with the prerelease tar (as expected), but it's marked as OK. Weird 22:39
Kaiepi BeastieBot, stresstest 22:41
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
Kaiepi alright this should work with --force-test for Inline::Perl5 (i hope)
or not since Zef's stuck at ===> Searching for: Inline::Perl5 22:45
fuck it i'll install it manually
El_Che ok fixed
AlexDaniel: my work is done. Ping me if you need something changed.
AlexDaniel El_Che: thanks
El_Che AlexDaniel: and certainlu when the dot release is out so I can release the packages 22:46
AlexDaniel sure
El_Che going back to the proletarian #perl6 :)
MasterDuke i could be convinced we should hardcode a check in X::TypeCheck::Binding::Parameter for the `expected Positional[Array] but got Array` case and suggest something specifically for that 23:02
yoleaux 11:09Z <Zoffix> MasterDuke: FYI: github.com/rakudo/rakudo/commit/50f768315d
Kaiepi BeastieBot, stresstest 23:07
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
Kaiepi whoops wrong dir 23:08
BeastieBot, stresstest 23:12
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
Kaiepi BeastieBot, stresstest 23:13
BeastieBot [freebsd] Running Roast's stress test suite (this will take a while)...
Kaiepi i'll need to work out why zef hangs when searching for Inline::Perl5 the way i'm using it 23:14
Geth nqp/nummy-num: a4688c751b | (Zoffix Znet)++ | t/nqp/115-nums.t
Cover denormals and num value drift bugs
 23:31
b2gills t/04-nativecall/01-argless.t `An exception occurred while parameterizing Pointer` 23:41
t/04-nativecall/19-function-pointers.t `This type cannot unbox to a native string: P6opaque, List` 23:46
I'm getting all kinds of errors in t/04-nativecall 23:48

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!