Geth rakudo/post-release: 43b9c82945 | (Zoffix Znet)++ | 3 files
Make `$ where Type1|Type2|Type…` typecheck ~8x faster

Rake the `where` code block for single "any" Junction with type objects only and rewrite it into multiple nqp::istype's right in the ParamTypeCheck op.		 00:20
roast/post-release: 17b2e0bc54 | (Zoffix Znet)++ | S12-subset/subtypes.t
Cover `$ where Type1|Type2|Type…` typecheck

Covers github.com/rakudo/rakudo/commit/43b9c82945		 00:22
rakudo/post-release: 9 commits pushed by (Elizabeth Mattijsen)++, (Itsuki Toyota)++, (Zoffix Znet)++ 00:23
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Micro-optimize exit_handler, part #4 00:41
travis-ci.org/rakudo/rakudo/builds/328445329 github.com/rakudo/rakudo/compare/e...bc5916c0de
timotimo seeing the old "larry wall speaks at google" video; what used to be "suppose points" in the grammar are now worries? 00:46
Geth nqp: d93f6ed700 | (Zoffix Znet)++ | tools/build/MOAR_REVISION
[MoarVM Bump] Brings 4 commits

MoarVM bump brought: github.com/MoarVM/MoarVM/compare/2...4-g4a0a912 4a0a912 New spesh-bisect.pl tool 7ea0802 Fix off-by-one error in uninline 56c03d9 Merge pull request #780 from Kaiepi/compilers 28eaf97 Try to use egcc when compiling on OpenBSD		 08:17
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...4-g4a0a912
rakudo: bdbfdff257 | (Zoffix Znet)++ | tools/build/NQP_REVISION
[NQP Bump] Brings 8 commits

NQP bump brought: github.com/perl6/nqp/compare/2017....3-gd93f6ed d93f6ed [MoarVM Bump] Brings 4 commits 4e647a7 Implement QAST::Node.annotate_self 88fbd34 Remove trailing whitespace d20574d Document if/unless return condition when no else ... (10 more lines)
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2017....3-gd93f6ed
64803b0af8 | (Zoffix Znet)++ | src/Perl6/World.nqp

Original fix[^1] causes breakage on JVM[^2]. Since I don't fully know whether not installing a block is safe, I reverted that and instead simply mark the inserted chunk as wanted, to avoid the warning.
  [1] github.com/rakudo/rakudo/commit/c6b7012a9c61133
  [2] rt.perl.org/Ticket/Display.html?id...xn-1523921
[Tux] Rakudo version 2017.12-241-g2ac1ceaa1 - MoarVM version 2017.12.1-32-g56c03d99c
csv-ip5xs1.120 - 1.141
csv-ip5xs-2013.331 - 13.566
csv-parser12.621 - 12.852
csv-test-xs-200.447 - 0.477
test11.376 - 12.049
test-t2.789 - 2.922
test-t --race1.163 - 1.180
test-t-2050.953 - 52.105
test-t-20 --race18.832 - 20.065
 08:25
lizmat Files=1232, Tests=76024, 327 wallclock secs (14.78 usr 5.31 sys + 2260.39 cusr 229.89 csys = 2510.37 CPU)
TimToady timotimo: STD had a "suppose" hypothetical construct, but hypotheticals are largely limited to <before> and backtracking in rakudo 08:33
"suppose" was "try parsing with this assumption, but throw everything away when you're done" 08:34
Geth rakudo: cf15435575 | (Elizabeth Mattijsen)++ | 4 files
Remove src/core/UInt64.pm

This appears to be a leftover from at about 7 years ago. It created a subset that was not being used anywhere and a constant that was needed for Rat/FatRat. UInt64 (not to be confused with uint64) is not documented anywhere and not tested. Moved the constant to src/core/Rat and changed it into a real constant.
Kudos to Andrew Shitov for spotting this:
   perl6.online/2018/01/15/26-native-...in-perl-6/		 09:09
Zoffix .tell AlexDaniel I reverted that questionable commit I mentioned earlier. Also did a toaster run: toast.perl6.party/ 11:18
yoleaux Zoffix: I'll pass your message to AlexDaniel.
AlexDaniel .
yoleaux 11:18Z <Zoffix> AlexDaniel: I reverted that questionable commit I mentioned earlier. Also did a toaster run: toast.perl6.party/
AlexDaniel oh, thanks!
drive-by development :) 11:19
AlexDaniel looks at the results 11:24
wow, X11::Xlib::Raw tests pass but it fails to *install*. 11:28
moritz that's why clever people put a "zef install ." into their travis config 11:30
AlexDaniel that wouldn't help at all, last commit Mar 27, 2017 :) 11:35
in this case I mean
bisect: old=2017.12 use lib ‘data/all-modules/github/p6-css/CSS-Grammar-p6/lib’; use CSS::Grammar::Actions; 11:40
bisectable6 AlexDaniel, Bisecting by exit code (old=2017.12 new=cf15435). Old exit code: 0
AlexDaniel, bisect log: gist.github.com/e8c4533a349a0f442a...7d11211de4
AlexDaniel, (2018-01-15) github.com/rakudo/rakudo/commit/ee...75528ac61d
AlexDaniel c: eeb3cc726ada^,eeb3cc726ada use lib ‘data/all-modules/github/p6-css/CSS-Grammar-p6/lib’; use CSS::Grammar::Actions; 11:42
committable6 AlexDaniel, ¦eeb3cc726ada^: «» ¦eeb3cc7: «===SORRY!===␤P6opaque: no such attribute '$!storage' on type CSSValue in a Hash when trying to bind a value␤ «exit code = 1»»
jnthn Any way to get the --ll-exception output? 11:45
That error makes no sense, looking at the code in that commit
AlexDaniel jnthn: gist.github.com/AlexDaniel/da5ec5b...659270c786 11:47
jnthn Odd..I wonder if the module redefines the names Map and or Hash in a lexical scope, though 11:48
AlexDaniel I'll move on with the list of failed modules for now, so here's a ticket: github.com/rakudo/rakudo/issues/1407
jnthn Will toss :setting-only on those find_symbol lookups; they should probably have had them in the first place.
AlexDaniel Zoffix++ for running toaster, I usually do it 2 days before the release which is arguably too late :) 11:50
pmurias is an 'Int' constraint in setting code a performance penalty or a benefit? 11:55
we have erroneusly added int constraints in Date and I'm not sure if should remove them or turn them into Int?
timotimo no impact, we already check for Any everywhere anyway 11:56
pmurias so I should use an Int constraint? 11:58
timotimo if you want an Int, then yes :)
AlexDaniel bisect: old=2017.12 gist.githubusercontent.com/AlexDan...b/06-cmd.t 12:07
bisectable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, Bisecting by exit code (old=2017.12 new=cf15435). Old exit code: 0
AlexDaniel, bisect log: gist.github.com/da2b802b17459298f2...5e5cf2bcbf 12:08
AlexDaniel, (2017-12-30) github.com/rakudo/rakudo/commit/ad...a9b5c2e398
AlexDaniel hmm changing << $cmd >> to ($cmd,) makes it work 12:11
mch: my $cmd = ‘a’; multi sub getopt(@args) {}; getopt(« $cmd ») 12:13
committable6 AlexDaniel, ¦2015.12: «» ¦HEAD(cf15435): «Cannot resolve caller getopt(Str); none of these signatures match:␤ (@args)␤ in block <unit> at /tmp/Z3lZxJOKGn line 1␤␤ «exit code = 1»»
AlexDaniel c: ad684de383bb^,ad684de383bb my $cmd = ‘a’; multi sub getopt(@args) {}; getopt(« $cmd ») 12:14
committable6 AlexDaniel, ¦ad684de383bb^: «» ¦ad684de: «Cannot resolve caller getopt(Str); none of these signatures match:␤ (@args)␤ in block <unit> at /tmp/t2ngYdxMno line 1␤␤ «exit code = 1»»
Geth rakudo: tbrowder++ created pull request #1408:
Remove file src/Perl6/DebugPod.nqp		 12:15
AlexDaniel would be nice to have some comments on github.com/rakudo/rakudo/issues/1409 12:23
jnthn: fwiw there's another ticket that you may be interested in, but there's not much info about it at this point: github.com/rakudo/rakudo/issues/1410 13:39
jnthn I'll wait a bit to see if there's a golf 13:41
Will patch the earlier on this afternoon 13:42
Geth rakudo: efdbfecad0 | (Zoffix Znet)++ | src/Perl6/World.nqp
Only look in the setting when we want core symbols

Phixes github.com/rakudo/rakudo/issues/1407		 14:07
roast: 960e9060ec | (Zoffix Znet)++ | MISC/bug-coverage.t
Test enums with core symbol names don't explode unexpectedly

Closes github.com/rakudo/rakudo/issues/1407
AlexDaniel oh, now I understad why it's “phixes” :D
so that github doesn't auto-close the issue :D
jnthn ah, Zoffix++ beat me to the fix :) 14:08
Geth rakudo: 354a384847 | (Zoffix Znet)++ | src/Perl6/World.nqp
Do not crash on parametarization of DefiniteHOWs

Phixes github.com/rakudo/rakudo/issues/1411
Before DefiniteHOW coercers, parametarization of DefiniteHOWs was silently ignored. Revert to that behaviour until it can be properly implemented. See
  github.com/rakudo/rakudo/issues/1412		 14:22
roast: 5394ea5a9e | (Zoffix Znet)++ | MISC/bug-coverage.t
Test Parametarization of DefiniteHOWs does not cry about coercers

Closes github.com/rakudo/rakudo/issues/1411		 14:24
roast: db9872d665 | (Zoffix Znet)++ (committed using GitHub Web editor) | MISC/bug-coverage.t
Lists cannot be parametarized
 14:32
AlexDaniel I'll be away for the rest of the day. I drafted some rakudo tickets for all the issues from toaster 14:35
releasable6: next 14:36
releasable6 AlexDaniel, Next release in 4 days and ≈4 hours. No blockers. Unknown changelog format
AlexDaniel releasable6: next 14:40
releasable6 AlexDaniel, Next release in 4 days and ≈4 hours. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Unknown changelog format
Geth rakudo: 1b21292e2e | (Tom Browder)++ | 3 files
Remove file src/Perl6/DebugPod.nqp

The file is considered unneeded at the moment.		 15:21
rakudo: 20354470f5 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | 3 files
Merge pull request #1408 from tbrowder/delete-debugpod.nqp

Remove file src/Perl6/DebugPod.nqp
rakudo: 6977680bd9 | (Elizabeth Mattijsen)++ | src/core/Rat.pm
Fixed comment, ashitov++
 15:36
moritz ^w 15:37
Geth roast: d186368c0e | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | S32-io/IO-Socket-Async-UDP.t
Fix typo in message
 17:03
[Coke] star: use IO::String; 17:30
camelia ===SORRY!===
P6M Merging GLOBAL symbols failed: duplicate definition of symbol Handle
[Coke] O_o;
I've asked this before, but should this generate a warning on stderr: 17:31
(well, something like this)
use IO::String; my $*ERR = IO::String.new; my $a; my $a;
you get a potential difficulty without the $*ERR shenanigans, but the compiler doesn't know about $*ERR, only stderr, so the warning still happens. 17:32
(that part of the compiler, anyway) 17:33
(this one is tricky because I think the difficulties are emitted at the end of the compile, but that is still before runtime $*ERR changes; the more reasonable one is when that's in an EVAL, which happened after the runtime $*ERR change. 17:35
Zoffix m: use nqp; my $*ERR := open :w, $*SPEC.devnull; my $e := nqp::newexception(); nqp::setmessage($e, "meow"); warn "meow"; nqp::throw($e) 18:12
camelia open is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in sub open at src/RESTRICTED.setting line 9
in block <unit> at <tmp> line 1
Zoffix c: HEAD use nqp; my $*ERR := open :w, $*SPEC.devnull; my $e := nqp::newexception(); nqp::setmessage($e, "meow"); warn "meow"; nqp::throw($e)
committable6 Zoffix, ¦HEAD(6977680): «meow␤ in block <unit> at /tmp/C688ChiN9Q line 1␤␤ «exit code = 1»»
Zoffix [Coke]: briefly reading the code, looks like "Potential difficulties" throwage happens in the machinery that doesn't care about $*ERR; like ^ nqp::throw above 18:13
m: BEGIN $*LANG.set_pragma("no-worries", 1); my $a; my $a 18:15
camelia ( no output )
Zoffix :P
(will be just `no worries;` after release)
m: BEGIN $*LANG.refine_slang: "MAIN", role { method typed_worry (*@args) { dd [@args, %_] } }; my $a; my $a 18:18
camelia [["X::Redeclaration"], {:symbol("\$a")}]
Zoffix [Coke]: ^ that might be even more suitable for your purpose than trying to capture and then parse STDERR
[Coke] zoffix, thanks. 18:56
Geth roast: ba8db9ab75 | usev6++ | 2 files
[jvm] Unskip some recently fudged tests

  Zoffix++ modified the fix for RT #131305 with
  github.com/rakudo/rakudo/commit/64803b0af8
and the tests are passing again.		 20:29
synopsebot RT#131305 [open]: rt.perl.org/Ticket/Display.html?id=131305 Useless Useless Use Warning when invoking colonpaired routines
Geth roast: d0df3dc8ef | usev6++ | S17-supply/syntax.t
[jvm] Help fudger to get number of tests right
 20:40

Powered by Cro and the Raku® Programming Language.

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

Thank you!