Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
Geth rakudo/release-2018.09: f9a3f96d4c | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Log all changes (+ announcement)

Deliberately not logged: a949a161 c1e62f08 607ac154 fa4d3963 1937fe29 39db408a 6a6b0afb df2783ca 8593a0e0 82755848 d08a4145 5af3c501 656b0730 9b297074 741ae6f4 bd8e50b4 be07f300 53e82c9a 85da7b20 d5960313 b1a07f09 0a45ffbc 3b61d4f2 fa4fccb1 76c1f22a 1f44d169 970fafc5 0f05d27c 174e728c 0a416dad 1c3b7248 fc23b197 2f85fa37 5b34f696 ca1124c8 da0a81e3 18a6968b 0a2d8321 213e3282 54bd7ce8		 01:18
rakudo/release-2018.09: 9d772edba4 | (Aleks-Daniel Jakimenko-Aleksejev)++ | docs/release_guide.pod
Actual date, claim next release
AlexDaniel the rest after a nap :) 01:20
[Tux] Rakudo version 2018.08-129-gb22edce40 - MoarVM version 2018.09
csv-ip5xs0.908 - 0.924
csv-ip5xs-207.108 - 7.434
csv-parser24.208 - 25.185
csv-test-xs-200.429 - 0.430
test8.642 - 8.822
test-t2.053 - 2.105
test-t --race0.872 - 0.947
test-t-2035.163 - 35.582
test-t-20 --race11.801 - 12.299
 09:07
Geth rakudo: e54885e5e6 | (Elizabeth Mattijsen)++ | 7 files
Add skip-one method to appropriate PredictiveIterators
 10:14
lizmat hopes that timotimo will look at PR's of JSON::Fast 10:35
(the latest could make JSON::Fast more than 2x as fast)
AlexDaniel that is awesome 11:03
lizmat++ 11:11
timotimo uploaded liz' changes to cpan 11:19
AlexDaniel timotimo: what about github.com/rakudo/rakudo/issues/2311 11:20
timotimo i'd say give it a go 11:22
AlexDaniel tony-o: ↑ ? 11:23
Geth roast: 5738c8a5e2 | (Zoffix Znet)++ | S02-types/int-uint.t
[v6.d REVIEW] Precisify byte.int-bounds test

Orig: github.com/perl6/roast/commit/0331647a6		 11:49
roast: b2dd2bb39e | (Zoffix Znet)++ | 2 files
[v6.d REVIEW] Consistify formatting
 11:51
roast: 8928b850a0 | (Zoffix Znet)++ | 2 files
[v6.d REVIEW] Test actual results in .pick(Rat) tests

Orig: github.com/perl6/roast/commit/e4302e5d7		 11:52
dogbert17 o/ is 2018.09 out yet? 11:59
lizmat dogbert17: in a branch, waiting for AlexDaniel to do the final things when he wakes up 12:01
afaik :-)
AlexDaniel Working on it now, yes! :)
Geth rakudo/release-2018.09: 6dee0ef370 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Markdownify a bit more
 12:09
AlexDaniel 29666 commits in rakudo :) just a little bit more and we'll be at 30k 12:12
tbrowder__ hi, folks, got another question about nqp land inside rakudo, working on pod... 12:13
Geth roast: 09f7684bef | (Zoffix Znet)++ | S02-types/baggy.t
[v6.d REVIEW] Don't spec actual values of stringified .WHICH

We have cases where we changed them to hashes of values and I believe the original test was by 2016-Zoffix who didn't know much.
Orig: github.com/perl6/roast/commit/14af8024d
tbrowder__ when processing a pod chunk during testing i see a contents of pod instance N looking like the contents of a previous pod instance of N-2. it may have something to do with block type but i’ve not seen this before. my immediate concern is how “make” in Actions.nqp affects the compile% 12:18
compile_time_constant objects
s/constant/value 12:20
tbrowder__ for instance, does, say, $<pod_contents> get overwritten by next use or does it hang around in an array or hash on $/ as the cursor moves on? 12:23
i’m expecting every $/ passed to an actions method to be in its own scope but maybe that’s not quite true (although lots of .dumps seem to confirm that) 12:25
lizmat tbrowder__: sorry, I don't have an answer 12:27
tbrowder__ ok, thnx 12:29
dogbert17 AlexDaniel, lizmat: cool, then jnthn should be able to merge the postrelease-opts branch which is quite exciting 12:30
tbrowder__ i have to go for some hours but if anyone can shed some light i will be grateful!
Geth roast: 1de82a8c7a | (Zoffix Znet)++ | S17-supply/basic.t
[v6.d REVIEW] Improve callback test

Test both callbacks got called just once.
Orig: github.com/perl6/roast/commit/927b026c3		 12:32
Zoffix AlexDaniel: also, in 20 issues we'll have 2,000 rakudo issues 12:34
pmurias Zoffix: sometimes the sheer amount of bugs being found in rakudo makes me wonder if Perl 6 being so large is really a good idea :/ 12:39
AlexDaniel Zoffix: how do you count? Reportable disagrees: gist.github.com/Whateverable/2c1d0...83c1090dce
dogbert17: you don't need to wait for the release to merge it 12:40
dogbert17: rakudo and nqp are released from branches, so whatever happens on master after the branch is created no longer matters
that was true for the last few releases actually
note, however, that this is not the current practice in MoarVM 12:41
so you do need to wait for moarvm release, but that already happened
nine pmurias: Perl 6 contains a lot that you would find in libaries elsewhere. I see no indication that those libaries contain significantly fewer bugs. Also our goals are set quite high. How often do you see LTA error messages reported as bugs in libaries that provide things we have in core? 12:42
Zoffix buggable: tags
buggable Zoffix, Total: 1434; 6.D: 1; 9999: 7; @LARRY: 33; BOGUSTEST: 1; BOOTSTRAP: 4; BUG: 479; BUILD: 13; CONC: 36; DOCS: 1; EXOTICTEST: 7; FLAP: 1; GLR: 3; IO: 16; JVM: 44; LHF: 5; LTA: 160; MATH: 4; META: 1; MOAR: 2; MOLD: 91; NATIVECALL: 22; NYI: 61; OO: 13; OPTIMIZER: 5; OSX: 2; PARSER: 3; PERF: 26; PERFTEST: 1; POD: 17; PRECOMP: 19; REGEX: 44;
Zoffix, REGRESSION: 31; REPL: 5; RFC: 60; RT: 1; SECURITY: 2; SEGV: 29; SEVERE: 10; SINK: 1; SITE: 1; SPESH: 1; STAR: 7; TESTCOMMITTED: 12; TESTNEEDED: 33; TODO: 11; UNI: 25; UNTAGGED: 299; WEIRD: 4; WINDOWS: 7; See fail.rakudo.party/ for details
Zoffix AlexDaniel: I went with ^ that total + rakudo issues
pmurias: all the bugs are because people don't write tests 12:43
AlexDaniel Zoffix: I can feel your pain :) 12:46
Zoffix Well, it's not just my pain. We're all wasting valueable resources chasing bugs that didn't have to exist and trying to do research to document features that should've been documented in the first place. 12:47
Like R#2303 12:48
lizmat Zoffix: I think one of the issues is that writing tests codifies a Perl 6 version
synopsebot R#2303 [open]: github.com/rakudo/rakudo/issues/2303 [@LARRY] Are `NumericEnumeration` / `StringyEnumeration` implementation detail?
lizmat and I most definitely have held of writing tests because of that
*back 12:49
Zoffix lizmat: I don't see why it's the issue; or at least, I don't see why it's acceptable to implement a feature the users will likely use, but not acceptable to also spec the behaviour. It's like: what's the plan there? That the feature will change in the future or someone else will just write a partial-test because they're not fully aware of how the feature was designed or what? 12:50
lizmat: your QuantHash ops test are wonderful.
That will be part of the spec.
Zoffix & to get laundry
lizmat well, what happened is that it was unsure if the feature I implemented was going to stay 12:51
AlexDaniel Zoffix++ for pointing out important issues
lizmat and then it stayed, and I didn't write tests
:-(
I guess I should make issues for those things in the future, so test writing doesn't fall through the cracks 12:52
AlexDaniel yes, good point
the number of testneeded tickets does not seem to be growing at a crazy rate, which is good 12:53
and the ones that keep piling up are generally related to stability and other things that are hard to test 12:55
Zoffix lizmat: IMO if there's enough sureness to implement a feature (as opposed to, say, maintain a discussion about it), then it's more likely it'll stay than be removed. It might change slightly, but the tests speccing it can also change until the spec is actually released. 13:00
lizmat otoh, not having tests makes it easier to remove things later, like the awaiterator recently :-) 13:01
afk for a bit&
dogbert17 AlexDaniel: perhaps lizmat or timotimo can merge unless jnthn wants to do it himself 13:10
Geth nqp/release-2018.09: 1e83c929cb | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[release] Bump MoarVM revision to 2018.09
 13:31
nqp/release-2018.09: 8ce9eb88f1 | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
[release] Bump VERSION to 2018.09
rakudo/release-2018.09: fb5c387b37 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[release] Bump NQP revision to 2018.09
rakudo/release-2018.09: d9bdc9cafd | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
[release] Bump VERSION to 2018.09
Geth nqp: 1e83c929cb | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[release] Bump MoarVM revision to 2018.09
 13:37
nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2......2018.09
8ce9eb88f1 | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
rakudo/master: 6 commits pushed by (Aleks-Daniel Jakimenko-Aleksejev)++
¦ rakudo/master: version bump brought these changes: github.com/perl6/nqp/compare/2018.......2018.09
AlexDaniel undersightable6: status 13:38
undersightable6 AlexDaniel, OK! Working on it…
AlexDaniel, 1 error, 5 warnings: gist.github.com/c37ee80604f5edad7a...7747b19cc2 13:40
AlexDaniel stupid bot is not quick enough to realize that there was a release :)
stupid bot is not quick enough to realize that there was a release :) 13:45
undersightable6: status
undersightable6 AlexDaniel, OK! Working on it…
AlexDaniel, 0 errors, 5 warnings: gist.github.com/ceab7ecebf26be89bf...61d02f01ac 13:48
lizmat AlexDaniel++ 15:50
Geth roast: 580a249dc3 | (Zoffix Znet)++ | S17-supply/wait.t
[v6.d REVIEW] Lessen wait time for speedier test

Orig: github.com/perl6/roast/commit/8fa281425		 16:26
roast: ed4298f57f | (Zoffix Znet)++ | S32-io/mkdir_rmdir.t
[v6.d REVIEW] Simplify iffy .mkdir test

The original test doesn't work (the test file was never run, because it's not part of the t/spectest.data)
Test only for the SEGV for the related ticket. The rest of the tested behaviour is currently poorly specced.		 16:33
roast: 6a02b8d185 | (Zoffix Znet)++ | S32-io/mkdir_rmdir.t
[v6.d REVIEW] Precisify mkdir/rmdir throwage tests

Orig: github.com/perl6/roast/commit/8a351395c
rakudo: d3ba449fa8 | (Zoffix Znet)++ | t/spectest.data
Add S32-io/mkdir_rmdir.t to list of test files to run
timotimo lizmat: do you have a measurement for the array-instead-of-hash stuff? i'd expect some of the cost comes from having to stringify integers for every access 16:51
lizmat: actually, i'm measuring a tiny slowdown from your last PR 16:54
dogbert17 Zoffix: perhaps around here github.com/MoarVM/MoarVM/blob/mast...ops.c#L379 17:32
lizmat timotimo: but where are the ints stringified ?? the whole point was that they would't have to be 17:58
timotimo i meant in the hash version 17:59
can you verify my findings? that using arrays instead of hashes there makes parsing slower?
lizmat do you have a sample JSON file ?
or are you using the projects.json ? 18:00
timotimo i'm seeing about 0:03.41elapsed in your PR and 0:03.36elapsed without it
yes, just a sec
hack.p6c.org/~timo/AllCardsWhitespace.json.gz 18:01
lizmat they're all really in the 1.23-1.25 range for me, well within noise, really 18:06
I guess you need a JSON with a lot of hex digits to notice
lizmat but if you don't feel comfy with the PR, then just close it :-) 18:07
Geth nqp/truffle: 1055bd9ee2 | (Paweł Murias)++ | 2 files
[truffle] Remove unused nqp::runtrufflebytecode op
 18:35
nqp/truffle: c412888ebe | (Paweł Murias)++ | 19 files
[truffle] Implement $*foo dynamic variables
timotimo i now see that my moarvm currently has --optimize=0 18:41
lizmat timotimo: pretty sure mine doesn't :-) 18:45
Geth roast: 15478ac108 | (Zoffix Znet)++ | S16-io/readchars.t
[v6.d REVIEW] Clean up readchars tests

  - Use standard tempfile routines
  - Don't assume a particular location of test runner
  - Remove Perl5-isms
Orig: github.com/perl6/roast/commit/d006543e1		 19:31
roast: 392de5303f | (Zoffix Znet)++ | S02-types/set.t
[v6.d REVIEW] Relax set throwage tests

  - Clarify `set` won't throw if parens are there
  - Don't spect exact error message text
Orig: github.com/perl6/roast/commit/e4dec916d		 19:40
roast: 6c3183d4b9 | (Zoffix Znet)++ | S03-operators/identity.t
[v6.d REVIEW] Use better test routine for =:= || =:= tests

Orig: github.com/perl6/roast/commit/5e12710c9		 19:42
roast: 723b3811e0 | (Zoffix Znet)++ | S02-types/baghash.t
[v6.d REVIEW] s/is/is-deeply/ for bag bug cover tests

Orig: github.com/perl6/roast/commit/bf26816bc		 19:48
Geth roast: 8c696d3416 | (Zoffix Znet)++ | S32-list/rotor.t
[v6.d REVIEW] Precisify .rotor return type tests
 19:53
roast: 84974bf372 | (Zoffix Znet)++ | S12-subset/subtypes.t
[v6.d REVIEW] Actually test UInt MD worked

Orig: github.com/perl6/roast/commit/84e530956		 19:54
6.d-prep: e31fae89cf | (Zoffix Znet)++ (committed using GitHub Web editor) | TODO/README.md
Update Zoffix's review commit-reached
 20:37
Geth rakudo/normalized-ZDRs: d2f9b6fc21 | (Zoffix Znet)++ | src/core/Rational.pm6
Remove trailing whitespace
 20:46
roast/normalized-ZDRs: 3a34299eb6 | (Zoffix Znet)++ | S32-num/rat.t
Change expectation of non-normalization of ZDRs

BDFL's blessing: colabti.org/irclogger/irclogger_log...09-21#l248
What follows is a rationalization for the change, although it can be equally construed as mental gymnastics to justify a direct specification violation.
... (11 more lines)		 21:05
roast/6.c-errata: a2630ef074 | (Zoffix Znet)++ | S32-num/rat.t
Change expectation of non-normalization of ZDRs

BDFL's blessing: colabti.org/irclogger/irclogger_log...09-21#l248
What follows is a rationalization for the change, although it can be equally construed as mental gymnastics to justify a direct specification violation.
... (11 more lines)		 21:14
rakudo/normalized-ZDRs: 671c41156a | (Zoffix Znet)++ | src/core/Rational.pm6
Normalize ZDRs on Rational creation
 21:16
Zoffix "Unhandled exception: Lexical index out of bounds: 8 > 0" :/ 21:37
with this patch: gist.github.com/zoffixznet/08d6488...3d21fe28a8 21:38
Zoffix oh, nm. I think it was simply ZScript trying to run in the middle of my system perl6 being rebuilt. 21:39
Geth roast/normalized-ZDRs: 95c19bc4b7 | (Zoffix Znet)++ | 2 files
Don't expect .numerator to be set in DivByZero exceptions.

BDFL's blessing: colabti.org/irclogger/irclogger_log...09-21#l248
We no longer carry that information. The 420 test was also wrong from the start as it used `isa-ok` instead of just `is`.
... (11 more lines)		 23:13
roast/normalized-ZDRs: c4c673c7c9 | (Zoffix Znet)++ | 2 files
Mod propspec for normalization of ZDRs

These tests are not part of 6.c language.
roast/6.c-errata: 54c8a27d10 | (Zoffix Znet)++ | 2 files
Don't expect .numerator to be set in DivByZero exceptions.

BDFL's blessing: colabti.org/irclogger/irclogger_log...09-21#l248
We no longer carry that information. The 420 test was also wrong from the start as it used `isa-ok` instead of just `is`.
... (11 more lines)		 23:16
rakudo/normalized-ZDRs: 75ec2248ce | (Zoffix Znet)++ | src/core/Rational.pm6
Add missing decont in Rational.new for ZDRs
 23:30
rakudo/normalized-ZDRs: a639fa9cdd | (Zoffix Znet)++ | src/core/Rational.pm6
Proactively die on Rational.Str for ZDRs

Don't wait until we explode later in the body		 23:47
rakudo/normalized-ZDRs: 70ea883016 | (Zoffix Znet)++ | src/core/Rat.pm6
Teach DIVIDE_NUMBERS helper to normalize ZDRs

Powered by Cro and the Raku® Programming Language.

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

Thank you!