Geth rakudo: 02d2e2cafd | (Zoffix Znet)++ | src/core/Version.pm
Make Version.new 82x faster

  - Measured with git-describe string as input
  - This opt is mostly done to get rid of regexes, so we could use
   this method earlier in the setting to initialize some objects.
02:07
rakudo: ae7d89d245 | (Zoffix Znet)++ | 2 files
Remove Version.new-from-git-describe

  - This should've been a Rakudo::Internal method instead
  - This is no longer needed, as Version.new was made to work
   without using regexes
02:09
rakudo: 23a8aa87b0 | (Zoffix Znet)++ | src/core/core_epilogue.pm
Remove un-needed Version.new's

There are no concurrency issues in this code. During a race, the cache may come to be initialized twice, but that's fine.
Removing these from this part of the codebase, to reduce the spaghetti factor.
02:10
rakudo: 0bf735aaf7 | (Zoffix Znet)++ | tools/build/gen-version.pl
Differentiate failed version from git describe

If we fail to get version from git describe, append a string to the version from VERSION file, so we'd have a way to differentiate this case from the user using one of our releases.
The appended string is choosen to fit into the assumption that our versions are `YYYY.DD.<[.]+xdigit>+`, to be different from release, and to be $release before $broken-git-ver before $first-working-git-ver
02:14
travis-ci Rakudo build passed. Zoffix Znet 'Differentiate failed version from git describe 03:24
travis-ci.org/rakudo/rakudo/builds/337836709 github.com/rakudo/rakudo/compare/2...f735aaf75c
[Tux] Rakudo version 2018.01-121-g0bf735aaf - MoarVM version 2018.01-77-g9a029b408
csv-ip5xs0.802 - 0.814
csv-ip5xs-207.790 - 7.826
csv-parser11.395 - 11.851
csv-test-xs-200.415 - 0.471
test9.077 - 9.150
test-t2.589 - 2.699
test-t --race1.063 - 1.087
test-t-2047.335 - 47.453
test-t-20 --race16.237 - 17.117
07:42
Geth roast: b545dabf4b | (Zoffix Znet)++ | S06-operator-overloading/prefix.t
Unbust highlights in atom

  github.com/perl6/atom-language-perl6/issues/81
10:38
roast/post-inc-dec-opt: 9bcee77ff7 | (Zoffix Znet)++ | S06-operator-overloading/prefix.t
Test post-inc/dec op optimization does not affect custom ops

Closes github.com/rakudo/rakudo/issues/1477 Closes github.com/rakudo/rakudo/issues/1315
10:52
rakudo/post-inc-dec-opt: 779693ae3c | (Zoffix Znet)++ | src/Perl6/Optimizer.nqp
Fix/add native type post-inc/dec to pre- optimization

Fixes github.com/rakudo/rakudo/issues/1315 Part of github.com/rakudo/rakudo/issues/1477
  - Do not perform optimization if the op is not a core op
  - Add optimization for native num types:
   - rewrite the op call to NQP ops directly
10:54
Zoffix . 13:27
yoleaux 13:05Z <jnthn> Zoffix: It's never been spec'd either way so far. I'm curious when the behavior changed, though. Can probably use the bissect bot, but I'm a bit busy with other stuff at the moment.
Zoffix OK. I'll just leave it alone :) AlexDaniel FYI ^ RE the "supply out of react" thing 13:28
buggable: toast
buggable Zoffix, Between 2018.01-121-g0bf735a and 2018.01: 12 (1.23%) modules got burnt; 15 (1.53%) got unsucced; 255 (26.07%) out of 978 modules appear unusable. See toast.perl6.party/ for details.
Zoffix .tell AlexDaniel mid-month toast run: toast.perl6.party/ 12 burnt; 15 unsucced
yoleaux Zoffix: I'll pass your message to AlexDaniel.
Zoffix "When invoking 41 '', provided outer frame 0x1134a3d0 (43 '') does not match expected static frame 0x1134a1e0 (42 '')" well, that I never seen before 13:29
dogbert2_ m: for ^2 { for ^2 { say "{ (state $a)++ }" } } # RT #129114 13:32
camelia 0
0
0
0
synopsebot RT#129114 [new]: rt.perl.org/Ticket/Display.html?id=129114 [BUG] state variable declared inside of a {} interpolation in a string, attaches to wrong scope
jnthn The state var there is inside of a closure, which is cloned per entry to the outer block. So all 0s is the correct answer there 13:33
state can't mind-read which scope you were thinking of :)
timotimo m: for ^2 { for ^2 { say "$( (state $a)++ )" } } 13:34
camelia 0
1
0
1
jnthn m: for ^2 { for ^2 { say "$((state $a)++)" } }
camelia 0
1
0
1
jnthn heh :)
Zoffix m: say do loop { state $x = 0; last if ++$x > 2; $x }
camelia (1 2)
jnthn That looks correct, though a PR that I think touched that recently had a slightly bothersome description 13:35
Zoffix It was R#1467
jnthn In that the state var shouldn't need to move in this case, and doing so would probably hide a different problem
synopsebot R#1467 [closed]: github.com/rakudo/rakudo/pull/1467 Issue 1341: Elevate statevars to enclosing block when wrapped
Zoffix Oh. it does move them 13:36
dogbert2_ so the RT is bogus then?
Zoffix dogbert2_: yeah
dogbert2_ so I guess we can reject it without AlexDaniel going crazy :-) 13:37
jnthn Zoffix: Yeah, didn't have time to dig any further but the movement of it outside of the block it was declared on feels wrong. I think that'd just be hiding the real problem (and probably making a new one) 13:38
Zoffix OK. I opened R#1484 to take a closer look at this later on. 13:41
synopsebot R#1484 [open]: github.com/rakudo/rakudo/issues/1484 `say do loop { state...` fix might be hiding the actual problem
Zoffix uh-oh. I see one toast burnt by my hands 13:42
m: (@ = <a b c>) .= grep :k 13:43
camelia 5===SORRY!5=== Error while compiling <tmp>
You can't adverb .=
at <tmp>:1
------> 3(@ = <a b c>) .= grep :k7⏏5<EOL>
expecting any of:
pair value
Zoffix c: 2018.01 (@ = <a b c>) .= grep :k
committable6 Zoffix, ¦2018.01: «Cannot resolve caller grep(Array: :k); none of these signatures match:␤ ($: Bool:D $t, *%_)␤ ($: Mu $t, *%_)␤ in block <unit> at /tmp/mL64MAdg6V line 1␤␤ «exit code = 1»»
Zoffix c: 2018.01 (@ = <a b c>) .= grep "a", :k
committable6 Zoffix, gist.github.com/e38d29ffed54ef9c70...844cf32a1e
Zoffix c: 2018.01 (@ = <a b c>) .= grep: "a", :k
committable6 Zoffix, ¦2018.01: «»
Zoffix c: 2018.01,HEAD (@ = <a b c>) .= grep: "a", :k 13:44
committable6 Zoffix, ¦2018.01,HEAD(0bf735a): «»
Zoffix c: 2018.01 class F { method z(|c) { dd c } }; ($ = F) .= z :42foo :70bar 13:45
committable6 Zoffix, ¦2018.01: «\(:bar(70), :foo(42))␤»
Zoffix the fake-infix pairs got broken.
There's a whole bunch of burns due to "When invoking 41 '', provided outer frame 0xab45868 (43 '') does not match expected static frame 0xab45678 (42 '')"... which seems to come from Panda::* module lots of modules use 13:50
AlexDaniel Zoffix: thanks 13:51
yoleaux 10:33Z <Zoffix> AlexDaniel: just a note. In spec, we use full URL to Rakudo's Issues when labeling tests. So here github.com/perl6/roast/commit/396a...ead86cfb84 it'd be `# github.com/rakudo/rakudo/issues/1477` . Just labeling with "GH#1477" doesn't indicate what implementation it is (or is it roast Issue?)
synopsebot GH#1477 [open]: github.com/rakudo/rakudo/issues/1477 [optimizer] Postfix → Prefix ++/-- optimization
yoleaux 10:35Z <Zoffix> AlexDaniel: and just to note: no need to modify that commit. I'll do it when I write more tests for R#1477
13:28Z <Zoffix> AlexDaniel: mid-month toast run: toast.perl6.party/ 12 burnt; 15 unsucced
AlexDaniel Zoffix: what about using R#12345
Zoffix the GH#1477 is already fixed in a branch
AlexDaniel R#1477 13:52
ok…
:)
Zoffix AlexDaniel: that makes it a bit clearer what the issue is from, yeah
AlexDaniel but the bot does not understand that syntax
GH#1477
Zoffix I'm still gonna use full links 'cause it's something I can just copy-paste to visit instead of trying to figure out how to get to it 13:53
AlexDaniel ah… it does
Zoffix AlexDaniel: it does. It's just mentioned that issue recently so it won't mention it again for (I think) 10 m
AlexDaniel R#1467
synopsebot R#1467 [closed]: github.com/rakudo/rakudo/pull/1467 Issue 1341: Elevate statevars to enclosing block when wrapped
dogbert2_ m: { berp => "lerp", ( with "herp" { $_ => "derp" } ) } # RT #126945
camelia When invoking 1 '', provided outer frame 0x1908990 (3 '<unit>') does not match expected static frame 0x1908a88 (2 '')
in block <unit> at <tmp> line 1
synopsebot RT#126945 [new]: rt.perl.org/Ticket/Display.html?id=126945 [BUG] outer frame does not match expected static frame
Zoffix bisect: { berp => "lerp", ( with "herp" { $_ => "derp" } ) }
bisectable6 Zoffix, Bisecting by output (old=2015.12 new=0bf735a) because on both starting points the exit code is 1
Zoffix boo 13:54
bisectable6 Zoffix, bisect log: gist.github.com/3148839775eee07e16...ce09ca119e
Zoffix, (2015-12-25) github.com/rakudo/rakudo/commit/07...dc61f84053
Zoffix, The result looks a bit unrealistic, doesn't it? Most probably the output is different on every commit (e.g. 「bisect: say rand」)
AlexDaniel dogbert2_: wait, what's the reason to go crazy this time? :) 13:55
Zoffix hopes for another tip from jnthn for this bug, akin to "blocks need to be in the same scope" :)
I fixed like a whole ton of bugs now with that tip :)_
dogbert2_ I rejected RT #129114 :-)
synopsebot RT#129114 [rejected]: rt.perl.org/Ticket/Display.html?id=129114 [BUG] state variable declared inside of a {} interpolation in a string, attaches to wrong scope
jnthn Zoffix: I'd think that error would also be thanks to mis-nesting
Zoffix \o/ 13:56
dogbert2_ could RT #130383 be something similar? 13:57
synopsebot RT#130383 [new]: rt.perl.org/Ticket/Display.html?id=130383 [BUG] perl6 -n does not set $_ when it's interpolated via a block
Zoffix dogbert2_: but isn't the ticket showing "0 1 2 3" as the output, while the current and correct output is "0 0 0 0"? So looks like there was a bug that is now fixed and needs to be covered. The OP just didn't describe the correct behaviour correctly
mc: for ^2 { for ^2 { say "{ (state $a)++ }" } }
committable6 Zoffix, ¦2015.12: «0␤1␤2␤3␤»
Zoffix m: for ^2 { for ^2 { say "{ (state $a)++ }" } }
camelia 0
0
0
0
AlexDaniel dogbert2_: yes, would've been easier if the current behavior was shown
but 0 0 0 0 looks correct to me too
dogbert2_ things always goes wrong when I reject stuff :( 13:58
jnthn Yes, there was a missing clone before. That caused some horrible threading bugs, where you got the wrong thread's data interpolated. /o\
dogbert2_ was looking for RT's which would convert Zoffix Monster Kill to an Ultra Kill :-) 14:00
Zoffix dogbert2_: RT#1481 would be it 14:01
dogbert2_: R#1481 would be it 14:02
synopsebot R#1481 [open]: github.com/rakudo/rakudo/issues/1481 `where` with regex miscopes QAST::Blocks inside the regex
dogbert2_ AlexDaniel: thx, I can add a test later
AlexDaniel dogbert2_: also, it is pretty much “resolved” and not “rejected” in a sense that the wrong behavior is now gone and is replaced with the right one (not what was asked, but that's ok) :) 14:03
this difference does not exist on github and I'm very grateful for it :) 14:05
Zoffix: any news on github.com/rakudo/rakudo/issues/1481 ? It prevents me from rolling out a bot xD 14:06
Zoffix R#1481
Ah. 14:07
I'll fix it in 0-72hrs
AlexDaniel thank you very much! 14:08
Zoffix What would be the excted result of `{ berp => "lerp", ( with "herp" { $_ => "derp" } ) }` ?
m: dd %( berp => "lerp", ( with "herp" { $_ => "derp" } ) )
camelia Hash % = {:berp("lerp"), :herp("derp")}
Zoffix This?
Or a block
IIRC speculations say "if starts with Pair -> tis a block" 14:09
s/block/hash/;
m: use QAST:from<NQP>; BEGIN $*W.compile_time_evaluate: $, QAST::Op.new: :op<call>, :name<&say>, QAST::Op.new: :op<with>, QAST::IVal.new(:value(1)), QAST::Block.new(QAST::Stmts.new, QAST::Stmts.new: QAST::Op.new: :op<say>, QAST::SVal.new: :value<meow>).annotate_self: 'blocktype', 'immediate' 14:11
camelia <anon>
Zoffix I don't get how `%( berp => "lerp", ( with "herp" { $_ => "derp" } ) )` ends up with a Pair instead of a Block for the `with` part
AlexDaniel Zoffix: looks like a block to me 14:12
Zoffix AlexDaniel: why a block? 14:13
AlexDaniel well, rakudo always disagrees with me on this one…
Zoffix The OP clearly wanted a hash
AlexDaniel should've used %() then :S 14:14
Zoffix From design.perl6.org/S03.html `Inside must be either empty, or a single list starting with a pair or a hash, otherwise you must use hash() or %() instead.`
That looks like a single list starting with pair to me. So I guess that should be a hash
m: $_ = "foo"; dd {:$_}
camelia -> ;; $_? is raw { #`(Block|62590184) ... }
Zoffix But that doesn't follow that rule 14:15
m: $_ = "foo"; dd {:foo, (if 1 { :bar })} 14:16
camelia ===SORRY!===
Cannot reference undeclared local 'pres_topic__1'
Zoffix eheheh
Looks to be in the same family as the `with` thing 14:18
m: $_ = "foo"; dd {:foo, (if 1 -> $x { :bar })}
camelia When invoking 1 '', provided outer frame 0x32e3940 (3 '<unit>') does not match expected static frame 0x32e3a38 (2 '')
in block <unit> at <tmp> line 1
AlexDaniel also, what's the meaning of “or a hash” in that definiton?
m: dd { %(42 => 5) } 14:19
camelia -> ;; $_? is raw { #`(Block|54536944) ... }
AlexDaniel m: dd { {42 => 5} }
camelia -> ;; $_? is raw { #`(Block|54337856) ... }
AlexDaniel m: dd { {42 => 5}, abc => 42 }
camelia -> ;; $_? is raw { #`(Block|72498144) ... }
Zoffix Yeah, I guess a precise definition of when {...} means Hash needs to be defined (and documented) 14:20
"Note that this construct does not re-parse the contents; rather, the contents are always parsed as a statement list (i.e. like a block), and if the later analysis shows that it needs to be interpreted as a hash, the block is executed and coerced to Hash."
That seems to be awfully implementation-specific prose 14:21
From docs.perl6.org/routine/%7B%20%7D
jnthn Zoffix: I believe the synopses had a pretty detailed one
AlexDaniel github.com/perl6/doc/issues/1758 14:22
Zoffix jnthn: so far we one one mention that doesn't say anything about $_, but does say that hash inside means hash, which isn't the current behaviour
jnthn Uff 14:23
Zoffix S03: "Inside must be either empty, or a single list starting with a pair or a hash, otherwise you must use hash() or %() instead."
jnthn There's a spectest file dedicated to this, though
Zoffix cool, I'll look for it
jnthn block_vs_hash or some similar name
Zoffix ./S06-other/anon-hashes-vs-blocks.t
Zoffix looks at S06 14:24
AlexDaniel m: my %h = :42a, :50b; dd { %h } 14:25
camelia Hash % = {:a(42), :b(50)}
AlexDaniel m: my %h = :42a, :50b; dd { %h, 5 => 6 }
camelia Hash % = {"5" => 6, :a(42), :b(50)}
AlexDaniel that's what it means by “hash” I think?
Zoffix Ah 14:26
m: dd pair(1, 2) 14:27
camelia 1 => 2
Zoffix TIL we have &pair
jnthn Funny thing about the prose: that is *not* how it's presently implemented, so far as I recall. But if we *did* implement it like that, it'd probably be better. :)
The one pass parse part is correct. But iirc, we rewrite the AST to call circumfix:<{ }> with the stuff in the block.
Zoffix But the one pass parse part is implementation-specific, isn't it? 14:28
jnthn No
Zoffix :o
jnthn Pretty sure it's laid down that Perl 6 is parsed in a single pass
Zoffix TIL
jnthn Otherwise, you can't execute BEGIN blocks safely :) 14:29
It's another question entirely whether this prose in the docs helps the language user, however :) 14:30
Zoffix Neither the S06 speculation nor S06-other/anon* seem to define whether `{:foo, (if 1 -> $x { :bar })}` is a hash or a block. There's mention in the spec that $_ or $^foo force a block, which `{:foo, (if 1 -> $x { :bar })}` ain't got, so I'm thinking it should be a hash 14:35
jnthn I think hash also. And I think if we implemented it by leaving it as a block and passing the result to circumfix:<{ }> instead of trying to be clever, then we'd probably not have that bug 14:36
At a slight performance cost, maybe. 14:37
But we can inline any small blocks, and for big ones the overhead will probably be noise
Zoffix doesn't get what "passing the result to circumfix:<{ }> instead" mean 14:39
Like, it can't take a Block
m: circumfix:<{ }>({; say 42})
camelia Odd number of elements found where hash initializer expected:
Only saw: -> ;; $_? is raw { #`(Block|69353528) ... }
in block <unit> at <tmp> line 1
jnthn Call the block, pass what the block returns to circumfix:<{ }>
Zoffix m: circumfix:<{ }>({; say 42}()) 14:40
camelia 42
Odd number of elements found where hash initializer expected:
Only saw: Bool::True
in block <unit> at <tmp> line 1
jnthn m: dd circumfix:<{ }>(-> { a => 1, b => 2 }())
camelia Hash % = {:a(1), :b(2)}
jnthn At the moment we try to turn it into not-a-QAST::Block
Zoffix m: $_ = 42; dd circumfix:<{ }>(-> { a => 1, :$_ }()) 14:41
camelia Hash % = {:_(42), :a(1)}
Zoffix But then this would break the current behavour about params (though the speculations don't seem to say anything about them)
I still don't get it, so I'll leave it alone for now :) 14:43
The "When invoking 1 '', provided outer frame 0x202bba8 (3 '') do" burns are mine too 14:49
m: (*.match(/./).self)(42)
camelia When invoking 1 '', provided outer frame 0x1f86ba8 (3 '') does not match expected static frame 0x1f869b8 (2 '')
in block <unit> at <tmp> line 1
Zoffix So I guess that's most of them :)
ZofBot: burning toast is my second job, bruh
ZofBot Zoffix, On the way I got hold of Charlie
Zoffix Ohhh... *instead* of trying to turn into not-a-QAST::Block, call circumfix:<{ }>. K, got it. 14:51
jnthn Zoffix: Yes, call block, pass result to circumfix:<{ }> 14:53
Should be less/simpler code than we have now 14:54
Zoffix Cool
Geth rakudo: 052a0bbef5 | (Zoffix Znet)++ | t/05-messages/moar/01-errors.t
Make segv test more robust

When using system as root, current version actually fails to produce any errors.
Fix by making the test attempt to use a directory as a file instead
17:50
japhb Is there a DWIM method of finding modules in the ecosystem? Case-insensitive, word-stemming or substring, etc.? 18:15
japhb is tiring of guessing how module writers uppercase, pluralize, and decline 18:17
stmuk_ japhb: panda? :) 18:19
[Coke] modules.perl6.org ? 18:20
tadzik ha :)
doesn't zef search do it nicely though?
stmuk_ AFAIK zef search only does exact searches
[Coke] wonders why 'zef search uni' shows 1 result, DateTime::TimeZone 18:22
japhb zef has some odd ideas about searching. The only thing that really seems reliable to me is exact match of at least one chunk of a module name. 18:40
stmuk I'd favour case insensitive sub-string (as panda did). I think I have mentioned it before 18:41
japhb stmuk: That seems to be what modules.perl6.org does (and it also appears to be WAY faster than zef) 18:42
Nice site actually, thanks for the reminder [Coke] 18:43
Is this a Zoffix creation? 18:44
Seems like something Zoffix might do ....
[Coke] he's at least done most of it. 19:02
dogbert17 c: 2017.01 my @arr = (); for ^2 { for ^2 { @arr.push( "{ (state $a)++ }") } }; say @arr; 19:50
committable6 dogbert17, ¦2017.01: «[0 1 2 3]␤»
Geth roast: dogbert17++ created pull request #385:
Add test for RT #129114
19:52
synopsebot RT#129114 [open]: rt.perl.org/Ticket/Display.html?id=129114 [BUG] state variable declared inside of a {} interpolation in a string, attaches to wrong scope
Geth roast: f5f14b4c4d | (Jan-Olof Hendig)++ | S04-declarations/state.t
Add test for RT #129114
20:06
roast: 0552e64c1f | (Zoffix Znet)++ (committed using GitHub Web editor) | S04-declarations/state.t
Merge pull request #385 from dogbert17/test-rt-129114

Add test for RT #129114
synopsebot RT#129114 [open]: rt.perl.org/Ticket/Display.html?id=129114 [BUG] state variable declared inside of a {} interpolation in a string, attaches to wrong scope
AlexDaniel dogbert17++ 20:10
dogbert17 bisect: my Int @a; @a[5] = 42; $_ = 666 for @a 22:07
bisectable6 dogbert17, Bisecting by exit code (old=2015.12 new=052a0bb). Old exit code: 1 22:08
dogbert17, bisect log: gist.github.com/3bb66eaa0c2116d723...a630f0bbe0
dogbert17, (2016-06-01) github.com/rakudo/rakudo/commit/06...cb82579346
dogbert17 bisect: my Str @a[3;3] = <a b c>, <d e f>, <g h i>; say @a 22:16
bisectable6 dogbert17, On both starting points (old=2015.12 new=052a0bb) the exit code is 0 and the output is identical as well
dogbert17, Output on both points: «[[a b c] [d e f] [g h i]]␤»
dogbert17 grumble
AlexDaniel c: 069b789a^,069b789a my Int @a; @a[5] = 42; $_ = 666 for @a 22:21
committable6 AlexDaniel, ¦069b789a^: «Cannot assign to an immutable value␤ in block <unit> at /tmp/n1SLOybKkH line 1␤␤ «exit code = 1»» ¦069b789: «»
AlexDaniel c: 069b789a^,069b789a my Int @a; @a[5] = 42; $_ = 666 for @a; say @a
committable6 AlexDaniel, ¦069b789a^: «Cannot assign to an immutable value␤ in block <unit> at /tmp/oR4KKlPlxe line 1␤␤ «exit code = 1»» ¦069b789: «[666 666 666 666 666 666]␤»
AlexDaniel dogbert17: what's your grumble about? :) 22:22
current behavior seems to be right?
and the bot output is correct?
dogbert17 AlexDaniel: which commit fixed this? RT #126703
synopsebot RT#126703 [new]: rt.perl.org/Ticket/Display.html?id=126703 Typed Shaped Arrays can't be initialized
AlexDaniel c: 069b789a^,069b789a,HEAD my Int @a; @a[5] = 42; $_ = 666 for @a; say @a 22:23
committable6 AlexDaniel, ¦069b789a^: «Cannot assign to an immutable value␤ in block <unit> at /tmp/a_dieWbIFx line 1␤␤ «exit code = 1»» ¦069b789,HEAD(052a0bb): «[666 666 666 666 666 666]␤»
AlexDaniel 6c: my Str @a[3;3] = <a b c>, <d e f>, <g h i>; say @a
committable6 AlexDaniel, ¦6c (27 commits): «[[a b c] [d e f] [g h i]]␤»
dogbert17 069b789 then
AlexDaniel c: all my Str @a[3;3] = <a b c>, <d e f>, <g h i>; say @a 22:24
committable6 AlexDaniel, gist.github.com/42a4bc2f7befa53043...be53cc6e30
AlexDaniel bisect: old=2015.11 my Str @a[3;3] = <a b c>, <d e f>, <g h i>; say @a
bisectable6 AlexDaniel, Bisecting by exit code (old=2015.11 new=052a0bb). Old exit code: 1
AlexDaniel, bisect log: gist.github.com/9d712dd830c09f4777...9c2696ed35 22:25
AlexDaniel, (2015-11-26) github.com/rakudo/rakudo/commit/13...4a50579a1c
dogbert17 I'm just updating a few old RT's
AlexDaniel c: 131367c7daa89d^,131367c7daa89d my Str @a[3;3] = <a b c>, <d e f>, <g h i>; say @a
committable6 AlexDaniel, ¦131367c7daa89d^: «Type check failed in assignment to ; expected Str but got List␤ in block <unit> at /tmp/t6D1g7_ZSy:1␤␤ «exit code = 1»» ¦131367c: «[[a b c] [d e f] [g h i]]␤»
AlexDaniel dogbert17: 131367c7daa89d ?
dogbert17++ 22:26
dogbert17 AlexDaniel: seems to be the one, thx :-)
for some reason I'm unable to tag the issues as tests needed, ah well 22:28
RT #126394 should also have the tests needed tag
synopsebot RT#126394 [open]: rt.perl.org/Ticket/Display.html?id=126394 [BUG] Assigning Failure to typed variable swallows Failure information
AlexDaniel marked 22:35
dogbert17 thx 22:49
Geth rakudo/post-inc-dec-opt: 407fe24229 | (Zoffix Znet)++ | src/Perl6/Optimizer.nqp
Fix wrong primspec test; MasterDuke++
23:10
rakudo/post-inc-dec-opt: 4a5cc2c200 | (Zoffix Znet)++ | src/Perl6/Optimizer.nqp
Fix/add native type post-inc/dec to pre- optimization

Fixes github.com/rakudo/rakudo/issues/1315 Part of github.com/rakudo/rakudo/issues/1477
  - Do not perform optimization if the op is not a core op
  - Add optimization for native num types:
   - rewrite the op call to NQP ops directly
23:29