Geth rakudo: 97b2b3f073 | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Make uniprops a proper multi and about 4x as fast

  - abstract the code/propname -> prop/pref logic
  - create a proper iterator for generating values
10:06
lizmat m: dd unimatch(-1,"Latin") # whee ! 10:34
camelia Internal Error: MVM_codepoint_to_row_index call requested a synthetic codepoint that does not exist.
Requested synthetic 1 when only 6 have been created.
in block <unit> at <tmp> line 1
Geth roast: 2b556e0aac | (Elizabeth Mattijsen)++ | S15-unicode-information/unimatch-general.t
Make unimatch test fail for the right reason

unimatch should always be called with at least 2 parameters as a sub, or one parameter as a method.
11:14
rakudo: e1c2d792a7 | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Make unimatch a proper multi

No functional changes: just make it pass on whatever it gets, and locate the backend specific bits at one place.
11:25
rakudo: 84eac183ae | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Forgot to remove an obsolere sub uniprops candidate
rakudo/no-uniprop-foo: 0f79bd89f6 | (Elizabeth Mattijsen)++ | 11 files
Remove uniprop-(int|str|bool)

They appear to have been added at some point, but they are not tested nor are they documented, nor are they needed in the setting nor any core modules.
Their functionality seems to intended as a performance shortcut for the more general "uniprop" and "uniprops" functions. My argument would be that *if* one knows about these, one might as well use the underlying nqp::ops as these appear to be a very thin HLL wrapper for them.
11:47
rakudo: lizmat++ created pull request #4308:
Remove uniprop-(int|str|bool)
lizmat wonders whether an .UP coercion in Raku would allow for an .IO like abstraction of the Unicode database 11:50
which, looking at this more and more, is more like a hash of hash of hashes
and maybe Raku should provide such an interface for it 11:51
Geth rakudo: e6823863bb | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Make uniparse a proper multi
12:40
rakudo: 2562d57461 | (Patrick Böker)++ | 3 files
.p6 -> .raku in compiled release README
12:48
rakudo: 2c7acb74c4 | (Patrick Böker)++ (committed using GitHub Web editor) | 3 files
Merge pull request #4307 from patrickbkr/p6-to-raku-bin-release

  .p6 -> .raku in compiled release README
lizmat I think this concludes my unicodey work for now 13:05
I hope I didn't break too much on non-MoarVM backends
Geth rakudo: da34b40abc | (Elizabeth Mattijsen)++ | src/core.c/Str.pm6
Make Str.substr(-rw) handle subclasses of Str properly
14:02
rakudo: f4bfeea66b | (Elizabeth Mattijsen)++ | src/core.c/Str.pm6
Make Str.chop|chomp handle subclasses of Str properly
dogbert17 lizmat: there might possibly be a problem with the module Test::Async 14:35
nine MasterDuke: got a fix for you! 16:36
lizmat dogbert17: in what context? what commit ? 17:07
nine lizmat: ah, see #moarvm. I thought I was there 17:08
nine lizmat: oh, sorry. Turns out I'm even more confused than I thought. Didn't see that you were talking to dogbert17 17:10
Geth rakudo: 813d4f1aa9 | (Elizabeth Mattijsen)++ | src/core.c/Str.pm6
Make Str.pred|succ handle subclasses of Str properly
18:18
dogbert17 lizmat: yeah, I forgot the important part and went for a walk :) 18:29
one of the modules tests fails with the cryptic message 'not ok 3 - The type Num can do the method 'NFKD'' 18:30
This test was supposed to fail but didn't and since it mentioned NFKD I thought it might have to do with some recent change 18:31
Let me check a bit more ... 18:33
lizmat ++dogbert17 18:34
dogbert17 m: say Num.^can("NFKD") 18:40
camelia ()
dogbert17 lizmat: ^^ 18:41
lizmat yeah, it's become a Cool method
dogbert17 Bug or should the test in the module be changed? 18:42
lizmat $ r 'dd 42e1.NFKD'
Uni.new(0x0034, 0x0032, 0x0030).NFKD
it's now a proper Cool method, so any Cool class (like Num, but more importantly Match) now just work
dogbert17 If vrurg shows up we can tell him 18:43
lizmat so I guess the test in the module is incorrectly assuming it shouldn't work ?
dogbert17 That would seem to be the case yes 18:44
test-flunks;
can-ok Num, "NFKD";
anyway, your change is pretty Cool :) 18:45
lizmat :-) 18:50
sena_kun 19:10
releasable6, status 19:11
releasable6 sena_kun, Next release in ≈9 days and ≈23 hours. 2 blockers. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/2724a4aacfa8a49916...5ffdedfaba
sena_kun what, 9 days
that's scary
ugexe its only scary for all the bugs that got fixed and the release manager 20:04
sena_kun fair enough 20:10
we still have a regression blocker to beat though 20:11
nine sena_kun: but the regression can be fixed by a simple revert, can it not? 20:12
sena_kun nine, alas we released it last release, so I would be really glad to see it getting a proper fix. 20:13
it is not the first time something like this happens, but still I feel like reverting an introduced feature is the last resort.
MasterDuke vrurg: i think i fixed at least one of the changes due to the nqp::time PR  that caused module breakage. could you do another blin run? 20:15
tellable6 MasterDuke, I'll pass your message to vrurg
dogbert17 vrurg: did you see the discussion I had with lizmat above about a failing test in raku-Test-Async? 20:21
it's in a file called t/110-can-ok.t
vrurg dogbert17: No, but I know the problem and it is already fxied.
tellable6 2021-04-07T20:15:35Z #raku-dev <MasterDuke> vrurg: i think i fixed at least one of the changes due to the nqp::time PR  that caused module breakage. could you do another blin run?
dogbert17 you're faster than lightning :) vrurg++ 20:22
vrurg dogbert17: pure coincidence. I wanted to try new module tests with Test::Async and needed to install it first. 20:23
.tell MasterDuke Not sure if will be able to run blin today. But will try to remember. 20:24
tellable6 vrurg, I'll pass your message to MasterDuke
MasterDuke thanks
vrurg dogbert17: anyway, it'll take me a little time to finalize new Test::Async version and I'm going to publish it ASAP on zef ecosystem. 20:25
vrurg is afk for a couple of hours.
[Coke] fez? 20:26
(is zef the original ecosystem?)
ugexe zef is the tool for downloading. fez is the tool for authoring. the service they both use can probably be called either fez or zef ecosystem 20:31
MasterDuke nqp: multi sub a() { say("hi") }; a 21:31
camelia Cannot find method 'reverse' on object of type NQPArray
at gen/moar/stage2/NQP.nqp:207 (/home/camelia/rakudo-m-inst-2/share/nqp/lib/nqp.moarvm:get_legacy_block_list)
from gen/moar/stage2/NQP.nqp:3092 (/home/camelia/rakudo-m-inst-2/share/nqp/lib/nqp.moa…
MasterDuke comes from here github.com/Raku/nqp/blame/master/s...ld.nqp#L59 via github.com/Raku/nqp/blame/master/s...s.nqp#L903 21:34
.ask jnthn any idea if github.com/Raku/nqp/blame/master/s...ld.nqp#L56 happened? the code below dies with `Cannot find method 'reverse' on object of type NQPArray` if you run `multi sub a() { say("hi") };` in nqp 21:37
tellable6 MasterDuke, I'll pass your message to jnthn
Geth nqp: MasterDuke17++ created pull request #711:
Finish getting rid of NQPCursor
21:56