tbrowder moritz: thnx for the tip 00:58
wow, that is a REAL cleanup tip! 01:01
Geth rakudo/zoffix-post-release: dc800d8933 | (Zoffix Znet)++ | src/core/IO/CatHandle.pm
Fix IO::CatHandle.read switching handles too soon

Fixes #1313: github.com/rakudo/rakudo/issues/1313
The logic is:
   read some stuff
   do we have enough stuff?
... (15 more lines)		 03:05
roast/zoffix-post-release: 6a027d36c5 | (Zoffix Znet)++ | S17-promise/nonblocking-await.t
Fix typo in test description
 04:00
rakudo/zoffix-post-release: c51f1796e6 | (Zoffix Znet)++ | 2 files
Fix issues with &await given conted Iterables

Fixes R#1323: github.com/rakudo/rakudo/issues/1323
  - For 6.c: eagerize return of the map, otherwise nesting of iterables
   prevents its sinkage, never causing it to reify (and thus await)
  - For 6.d: approach similar to 6.c version blew up some tests that
... (5 more lines)		 04:50
synopsebot R#1323 [open]: github.com/rakudo/rakudo/issues/1323 Issues with &await with nested containerazed Iterables
roast/zoffix-post-release: 9a9ec4989f | (Zoffix Znet)++ | 2 files
Test issues with &await

Closes github.com/rakudo/rakudo/issues/1323 Rakudo fix: github.com/rakudo/rakudo/commit/c51f1796e6
Include test for both 6.c and 6.d languages, since we use two separate sets of &awaits for them and each has different failure modes in the bug.		 04:52
Zoffix .tell jnthn Made a change to 6.d &await that's kinda gross. If you wanted to review: github.com/rakudo/rakudo/commit/c51f1796e6
yoleaux Zoffix: I'll pass your message to jnthn.
AlexDaniel fwiw I'll be online for the whole day, but sometimes sporadically :) 05:37
[Tux] Rakudo version 2017.11-133-g334b87226 - MoarVM version 2017.11-48-g0b9f8090a
csv-ip5xs1.077 - 1.082
csv-ip5xs-2013.125 - 13.128
csv-parser12.078 - 12.318
csv-test-xs-200.430 - 0.481
test11.652 - 11.780
test-t2.925 - 2.975
test-t --race1.178 - 1.248
test-t-2053.065 - 53.936
test-t-20 --race18.137 - 18.518
 12:19
lizmat Files=1233, Tests=75894, 338 wallclock secs (14.98 usr 5.21 sys + 2310.69 cusr 234.55 csys = 2565.43 CPU) 12:20
Geth nqp/master: 4 commits pushed by pmurias++ 13:40
lizmat Trying to determine if something is a value type or not, is still difficult 13:51
I was thinking of adding a 'class ObjAtValue is ObjAt { }' to src/core/ObjAt.pm 13:52
and then all classes that have a value type WHICH bless it with ObjAtValue rather than ObjAt
then checking whether something is a value type would be as simple as nqp::istype(ObjAtValue) 13:53
how is that for a plan? jnthn timotimo moritz TimToady ZofBot AlexDaniel` ?
moritz lizmat: sounds good to me, but I'm not aware of all the ramnifications 13:54
lizmat well, any tests for ObjAt would still work, so I think it would be ok 13:55
travis-ci NQP build failed. pmurias 'Expand test for nqp::setinvokespec' 13:57
travis-ci.org/perl6/nqp/builds/319176105 github.com/perl6/nqp/compare/567f6...800ed74fab
Zoffix m: say -«%(:70b) 14:20
evalable6 {b => -70}
timotimo we don't allow :-70b i imagine?
Zoffix No
This is a test in roast and it looks bogus to me. Why is - applied to the value? Pair isn't Iterable 14:21
timotimo you have a % in between
it's working on a hash, not on a single pair 14:22
that's how i'd explain it
Zoffix Same difference
m: say -«:70b
evalable6 {b => -70}
timotimo that's more interesting, then 14:23
Zoffix m: dd %(:4b, :60c)».&{(dd $_ )} 14:24
evalable6 Int % = 60
Int % = 4
Hash % = {:b(Any), :c(Any)}
Zoffix m: dd %(:4b, :60c)».value 14:25
evalable6 (exit code 1) No such method 'value' for invocant of type 'Int'. Did you mean 'values'?
in block <unit> at /tmp/FK1YwiJQ3q line 1
Zoffix m: dd (:4b, :60c)».value
evalable6 (4, 60)
Zoffix Looking at some deepmap tests, looks like the current behaviour is a winner in this case. Otherwise, this: 14:29
m: dd %(:4b, :60c)».flip
evalable6 Hash % = {:b("4"), :c("06")}
Zoffix Would have to return a List of .flipped Pairs which seems kinda useless
And gather { @a».take } will have to remain caveated about breaking Pairs of Hashes 14:30
This was the patch for gather { @a».take } FWIW gist.github.com/zoffixznet/25015bc...d20b47b314 14:33
Geth nqp: ca5c3bd28f | pmurias++ | t/nqp/100-dispatcher.t
[jvm] Skip some failing tests
Zoffix BUT, the routines don't descend into Pairs when they're part of a List :/ 14:38
This is really warty
travis-ci NQP build failed. pmurias '[jvm] Skip some failing tests' 14:48
travis-ci.org/perl6/nqp/builds/319200198 github.com/perl6/nqp/compare/69800...5c3bd28f9c
lizmat Zoffix: agree on wartiness :-) 14:50
TimToady lizmat: I don't see any problem with ObjAtValue, except that objects aren't at values :) 17:17
but I haven't thought of a better name... 17:19
ObjAtPlatonicIdealSpace is a bit unwieldy... 17:21
could pun something like ObjAtom I suppose 17:22
moritz ValueObjIdentity 17:23
or ValueObjectIdentity -- I guess it doesn't need to be huffmannized 17:24
TimToady It'sTheSameOneEverywhere 17:25
samcv i got t/spec/S17-procasync/stress.t test 22 flapping. 18:06
can i ignore it for the MoarVM release or no? 18:07
lizmat checks 18:08
timotimo ValueObjAt 18:09
lizmat like that 18:14
samcv: can't get it to flap :-( 18:17
samcv it did it when i was running stresstest
but not running it again by itself 18:18
lizmat runs a full stresstest
Geth roast: cfaa5d1801 | (Zoffix Znet)++ (committed using GitHub Web editor) | S17-procasync/stress.t
Use larger wait period in doesn't-hang test
 18:26
Zoffix samcv: ^ that'll likely stop it flopping 18:27
And if it still flops, there's likely an issue
lizmat fwiw, stresstest did not fail for that test
Zoffix I ran multiple stresstests this morning and no failures, tho I was using zoffix-post-release branch 18:28
samcv well running it again it is working now (though updated roast version as well) 18:42
so probably okay
Zoffix \o/ 18:47
samcv moarvm has been tagged and released 18:49
lizmat whee!
Zoffix samcv++_ 18:50
[Coke] samcv++
samcv tar.gz pushed to moarvm.org repo now 18:54
bartolin oh, the test pmurias++ added earlier today fails really badly on the JVM: github.com/perl6/nqp/commit/506cea5bcc 19:19
it seems to happen during parsing, so I don't see a way to fudge the test
I'd prefer to remove the tests temporarily and re-add it after the release 19:20
nqp: sub bar() { my $foo := 100; nqp::takedispatcher('$foo') }
anyway, that dies with NegativeArraySizeException 19:21
lizmat bartolin: please do :-) 19:23
Geth nqp: 24ebb039c9 | usev6++ | t/nqp/100-dispatcher.t
Temporarily remove three newly added tests

They explode badly on the JVM. I'll revert this commit after the release.		 19:26
ilmari isn't the .backend.name check in that removed test just the wrong way around? 19:27
bartolin yes, that too. (but fixing that didn't help) 19:28
lizmat wonders whether AlexDaniel` has been nudged now that MoarVM has been tagged and uploaded ? 19:31
twitter.com/zoffix/status/943520971822108672 # Zoffix: I was thinking the same thing :-) 19:42
travis-ci NQP build passed. usev6 'Temporarily remove three newly added tests
travis-ci.org/perl6/nqp/builds/319336538 github.com/perl6/nqp/compare/ca5c3...ebb039c93b
AlexDaniel` Ah OK. Will be back home in 30 mins 20:09
samcv++ 20:10
lizmat m: say "👍🏽".uninames; say "👍🏽".uniname # samcv: is that a bug or not? 20:53
evalable6 (THUMBS UP SIGN EMOJI MODIFIER FITZPATRICK TYPE-4)
THUMBS UP SIGN
samcv uhm what is the bug? also hexchat isn't rendering the icons though i saw there were some in the notification that popped up on my screen 20:54
lizmat isn't it supposed to be a single grapheme ?
samcv the first you did uninames not uniname though
lizmat yup
but I sorta expected the second one to produce the same string, because it is one grapheme ? 20:55
m: say "👍🏽".elems
evalable6 1
lizmat m: say "👍🏽".chars
evalable6 1
lizmat m: say "THUMBS UP SIGN, EMOJI MODIFIER FITZPATRICK TYPE-4".parse-names 20:57
evalable6 👍🏽
lizmat m: say "THUMBS UP SIGN, EMOJI MODIFIER FITZPATRICK TYPE-4".parse-names.chars
evalable6 1
lizmat I'm not sure this is a bug or just a misunderstanding on my end :-) 20:58
samcv lizmat: ah. 20:59
AlexDaniel doesn't it need a ZWJ?
samcv yeah it only returns the first codepoint name
number of graphemes doesn't matter
lizmat okidoki 21:00
Geth rakudo: 1273d064d2 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
Changelog and announcement

Deliberately not logged: 2658fcdf 6f357af0 03e099d7 47ebc4ab 76fcea1e c967cb6c 79390147 f768623d e5b660ee c0f99a39 2f955d55 0a2e5172 1fb51895 d98ccea2 370073fc 8159e383 ae6177ca f82433bf e5879c64 5eb299fb 1fbb9c80 a37a4e52 6a78b9ed 345ea806 790ae04c c8891791 e13d5b6e 20b6b848 7a0eba64 4003b93f a9ec5f92 80dfb30c b285aaf3 deb88aa5 4a32089f f78fd7c1 f6c2e70b f70e20bd 1dbf5f58 f1be29c0 a5635446 e57c771c 83eac0e4 21555412 c398eb7e 11613475 46eae469 664967d0 53a94fe5		 21:06
AlexDaniel releasable6: status 21:14
releasable6 AlexDaniel, Next release will happen when it's ready. No blockers. 130 out of 134 commits logged
AlexDaniel, Details: gist.github.com/6222c3f0151ab5070b...b1edceac93
Geth rakudo/zoffix-post-release: 7df750a108 | (Zoffix Znet)++ | 2 files
Revert "Fix issues with &await given conted Iterables"

This reverts commit c51f1796e627e93b34e1b06b46c6dccb82549e04.
Per: irclog.perlgeek.de/moarvm/2017-12-20#i_15604339
  +1 on not bending over backwards to DWIM with deconts for the user		 22:02
roast/zoffix-post-release: 37d56be46f | (Zoffix Znet)++ | 2 files
Revert "Test issues with &await"

This reverts commit 9a9ec4989f2f99b2f47d8465c50128424161ef61.
This isn't the direction we want to go in. Per: github.com/rakudo/rakudo/commit/7df750a108		 22:03
jnthn While I go along with "torture the implementor on behalf of the users", `await` walking a nested data structure feels a bit too magical 22:04
Especially when we've sometimes speculated about whether we should get a non-Awaitable just pass through `await` 22:05
Rather than be an error
Geth rakudo: 27961edb5f | (Aleks-Daniel Jakimenko-Aleksejev)++ | docs/release_guide.pod
Actual date; Claim next release
 22:06
jnthn Sorry I've been about so little over the last week, was sick.
I've not backlogged at all. 22:07
But I did read the weekly :)
Zoffix . 22:08
yoleaux 21:34Z <jnthn> Zoffix: I don't think await should be doing any flattening or descent beyond what *@awaitables does when slurping
Zoffix jnthn: yeah, that's why I pinged you. It felt a bit weird :) Looking over that commit, I'm going to put back the `eager` in 6.c (I think 6.d doesn't need it). So that await (p, (p, p)) would await it 22:09
jnthn Zoffix: Yeah, that part seemed sensible 22:11
Zoffix m: use v6.d.PREVIEW; my atomicint $x; sub p { start { sleep .3; $x⚛++ } }; await (((p(), (p(), (p(),))), (p(), p(), p())), (p(), p(), p())); say [$x, $x == 9] 22:12
evalable6 (exit code 1) An operation first awaited:
in block <unit> at /tmp/kAZNtZWgVZ line 1

Died with the exception:
Can only specify Awaitable objects to await (got a List)
in block <unit> at /tmp/kAZNtZWgVZ line 1
Zoffix What about this one? It could fit into the slurpy and get flattened there, but it goes through Iterable candidate
m: my atomicint $x; sub p { start { sleep .3; $x⚛++ } }; await (((p(), (p(), (p(),))), (p(), p(), p())), (p(), p(), p())); say [$x, $x == 9] 22:13
evalable6 [0 False]
Zoffix The eager patch in 6.c makes this work.
I'll add a .flat call in 6.d 22:15
jnthn Arguably, we could remove the Iterable candidate in 6.d
Zoffix Ah, OK, even better :)
jnthn And just leave the *@foo
Less code
Zoffix Yeah
jnthn And given that you're willing to pay a context switch overhead for `await` if you're using it...the opt of the other one is a drop in the ocean 22:16
Zoffix If you feel like making more decisions, there was a ping to you on S#370 about whether nodemap should be made private 22:20
synopsebot S#370 [open]: github.com/perl6/roast/issues/370 `nodemap` is virtually untested
jnthn My recollection in this area is that some years back (pre-6.c), TimToady looked at the various meta-op guts impl functions we had and concluded that they should be public and given public names. `produce` (the function form of the [\...] metaop) appeared at that time as a result of this 22:24
iirc, nodemap/hypermap/deepmap came around this time
And so were indeed intended as public
Zoffix OK cool. I'll add a comment on the Issue 22:25
AlexDaniel well… 2017-12-21 is kinda a bit in the future… 22:50
but we'll take it I think :)
Geth nqp: 68df86b901 | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/MOAR_REVISION
[release] Bump MoarVM revision to 2017.12
 23:01
nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2......2017.12
b1e3ed44e5 | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
rakudo: 0700cdbf0d | (Aleks-Daniel Jakimenko-Aleksejev)++ | tools/build/NQP_REVISION
[release] Bump NQP revision to 2017.12
 23:02
rakudo: version bump brought these changes: github.com/perl6/nqp/compare/2017.......2017.12
c84ed2942d | (Aleks-Daniel Jakimenko-Aleksejev)++ | VERSION
Zoffix AlexDaniel++ 23:06
AlexDaniel and there we have it, two years of releases since 2015.12 23:07
Zoffix \o/ 23:08
AlexDaniel 🙌 indeed
squashable6: next 23:09
squashable6 AlexDaniel, Next SQUASHathon in 15 days and ≈10 hours (2018-01-06 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
lizmat AlexDaniel++ 23:11
Geth rakudo/zoffix-post-release: ee8ac60720 | (Zoffix Znet)++ | 2 files
Ensure &await can handle flattenable Iterables

Fixes github.com/rakudo/rakudo/issues/1323
We don't want to do deconts for the user, as previous commits tried to fix the issue with, but we still want[^1] to accept whatever would be accepted by *@ slurpy: ... (10 more lines)		 23:37
rakudo/zoffix-post-release: 1a4df4e100 | (Zoffix Znet)++ | 3 files
Ensure &await can handle flattenable Iterables

Fixes github.com/rakudo/rakudo/issues/1323
We don't want to do deconts for the user, as previous commits tried to fix the issue with, but we still want[^1] to accept whatever would be accepted by *@ slurpy: ... (10 more lines)		 23:38
roast/zoffix-post-release: 8398c47121 | (Zoffix Znet)++ | 2 files
Add tests for &await with slurpable Iterables

Rakudo fix: github.com/rakudo/rakudo/commit/1a4df4e100 Issue: github.com/rakudo/rakudo/issues/1323		 23:39
rakudo/master: 5 commits pushed by (Zoffix Znet)++ 23:50
roast/master: 6 commits pushed by (Zoffix Znet)++ 23:51
rakudo: 08539c43d2 | (Zoffix Znet)++ | 2 files
Revert "Revert "Make Array.gist limit output to 100 els""

This reverts commit 334b87226d5950f85025a99797960753b014ceb3.		 23:52
roast: 5a0beb5afb | (Zoffix Znet)++ | S02-types/array.t
Revert "Revert "Test Array.gist limits output to 100 els""

This reverts commit 1b814c63368ba018fe25b92b92e4972c4165b59c.
lizmat m: use Telemetry; snapper; Nil for ^10000000 23:54
evalable6 Telemetry Report of Process #21867 (2017-12-20T23:54:27Z)
Number of Snapshots: 7
Initial/F…
lizmat, Full output: gist.github.com/22f93ca987c14b866a...53d866842c
lizmat m: use snapper; Nil for ^10000000
evalable6 Telemetry Report of Process #21919 (2017-12-20T23:54:45Z)
Number of Snapshots: 6
Initial/F…
lizmat, Full output: gist.github.com/a3d5b7e73164bd33f3...1475e6fbb0
lizmat weird: I get very different results locally 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!