AlexDaniel does anybody know why this counts up? 02:00
m: say (True ff 0) xx 5
camelia (1 2 3 4 5)
timotimo - QAST::Op(bind) <sunk> 02:09
- QAST::Var(local FLIPFLOP_STATE__1_orig :decl(var))
- QAST::Op(call &prefix:<++>)
- QAST::Var(lexical !FLIPFLOP_STATE__1)
i need something that can connect the vertical lines for me ... 02:11
i'm not sure the state was ever supposed to be returned 02:13
timotimo bedtime 02:14
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Streamline Array.AT-POS for better inlineability 05:22
travis-ci.org/rakudo/rakudo/builds/311928336 github.com/rakudo/rakudo/compare/f...29a227c8b7
Geth roast: 93ee959c08 | usev6++ | S05-mass/properties-general.t
[jvm] Skip dying Unicode test
 06:18
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Streamline List.AT-POS a bit 07:23
travis-ci.org/rakudo/rakudo/builds/312001553 github.com/rakudo/rakudo/compare/a...2f53250ecb
[Tux] After a reboot and minor system update 08:19
Rakudo version 2017.11-71-gae02bc299 - MoarVM version 2017.11-28-gba1fb9542
csv-ip5xs1.085 - 1.260
csv-ip5xs-2012.775 - 12.949
csv-parser12.155 - 12.182
csv-test-xs-200.419 - 0.466
test11.591 - 11.596
test-t2.900 - 2.907
test-t --race1.178 - 1.288
test-t-2051.419 - 52.196
test-t-20 --race18.150 - 18.235
DrForr o/ 08:40
lizmat Files=1229, Tests=75904, 318 wallclock secs (14.83 usr 5.38 sys + 2195.55 cusr 215.98 csys = 2431.74 CPU) 09:29
yoleaux 07:13Z <piojo> lizmat: Would you say it's a bad idea to use Pair as a member in a Set (or check $pair āˆˆ @array)? Because it seems awkward that Pair.WHICH is sometimes static (Pair.new(1,2)) but sometimes dynamic (Pair.new($a,2)). The ambiguity makes causes hard debugging since $foo !== EVAL($foo.perl), but EVAL($foo.perl) === EVAL($foo.perl)
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make Array.ASSIGN-POS about 1.7x faster 10:16
travis-ci.org/rakudo/rakudo/builds/312027117 github.com/rakudo/rakudo/compare/9...afc97e30f1
tbrowder could someone please review PR #1287 (with roast PR #361 for tests)? it affects proper handling of pod tables: when cell data has a visual column separator it must be escaped to avoid being recognized as such. 10:27
this PR will allow the docs (language/regexes) table showing unicode properties combined with set operators to display properly 10:32
.tell samcv PR #1287 would help the unicode docs 10:33
yoleaux tbrowder: I'll pass your message to samcv.
samcv i can review 10:34
yoleaux 10:33Z <tbrowder> samcv: PR #1287 would help the unicode docs
tbrowder thnx!
samcv link?
tbrowder github.com/rakudo/rakudo/pull/1287 10:36
note that you will also need the roast PR #361 you want to see and use the added tests 10:37
github.com/perl6/roast/pull/361 10:38
s/you/if you/
wander \o 10:42
tbrowder AlexDaniel: i closed PR #1291 because it involved too many fixes. i will break it up into at least two PRs: one to address the caption attribute data and one to address the quote issues. 10:43
AlexDaniel GH#1291
synopsebot GH#1291 [closed]: github.com/rakudo/rakudo/pull/1291 Fix table caption RT #126740
AlexDaniel sure 10:44
tbrowder i have submitted a CLA to the perl foundation and am bidding for commit privs after my next two PRs 10:45
AlexDaniel :) 10:46
all POD stuff needs more eyes definitely
tbrowder: by the way, have you seen this? irclog.perlgeek.de/perl6/2017-12-06#i_15537013 10:47
tbrowder yes, i have
AlexDaniel c: bbeb40d55 say 42 10:50
committable6 AlexDaniel, Ā¦bbeb40d: Ā«Cannot find this revision (did you mean ā€œa85b654ā€?)Ā»
DrForr I've got a partially-refactored Pod::To::HTML repo that I'm repurposing to do Blogspot-ready HTML and hoping to be able to refactor it to where the Blogspot changes are just either a role you introduce or a subclass you wrap around the existing HTML.
AlexDaniel hmmm *able is still unable to track branchesā€¦ argh 10:51
DrForr The first "big" change was getting rid of the exported methods that were being used for testing. It also meant that everything had to be just a function.
(looking at backlog here that AlexDaniel was talking about.) 11:01
samcv tbrowder: reviewed! :-) 11:03
about time for bed
++<AlexDaniel> all POD stuff needs more eyes definitely 11:04
tbrowder++
tbrowder .tell samcv thanks! 11:05
yoleaux tbrowder: I'll pass your message to samcv.
wander tbrowder++ 11:11
tbrowder .tell samcv i made yr suggested chgs 12:03
yoleaux tbrowder: I'll pass your message to samcv.
lizmat m: dd (a => my $ = 42) # jnthn: if we .perl this without the container 12:17
camelia :a(42)
lizmat m: dd (a => my $ = 42).WHICH # shouldn't .WHICH follow the same logic ?
camelia ObjAt.new("Pair|57247968")
jnthn m: dd flat 0, (a => my $ = (1,2,3)).value, 4 12:22
camelia (0, $(1, 2, 3), 4).Seq
jnthn m: dd flat 0, (a => my $ = (1,2,3)).perl.EVAL.value, 4
camelia (0, $(1, 2, 3), 4).Seq
jnthn m: dd flat 0, (a => (1,2,3)).perl.EVAL.value, 4
camelia (0, 1, 2, 3, 4).Seq
jnthn huh
m: say (a => my $ = (1,2,3)).perl
camelia :a($(1, 2, 3))
jnthn m: say (a => my $ = 42).perl 12:23
camelia :a(42)
jnthn Oh, interesting. It only shows it when it makes a difference. Hmm :)
lunch, bbiab
tbrowder thou: can you list the pod bugs you need fixed? 12:25
.ask thou can you list the pod bugs you need to have fixed? 12:26
yoleaux tbrowder: I'll pass your message to thou.
pmurias if we have a test that tests for incorrect behavior how should it be fixed? 13:29
we have this in the tests: is "\t \t" ~~ /<:space>+/, ' ', 'space chars'; # RT #130483 13:30
synopsebot RT#130483 [new]: rt.perl.org/Ticket/Display.html?id=130483 [UNI] Regex Unicode properties check string values before checking bool properties
pmurias the problem is: <:space> actually matches \t
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make R:It:ReifiedArray 15% to 30% faster 14:54
travis-ci.org/rakudo/rakudo/builds/312049106 github.com/rakudo/rakudo/compare/c...74de9ba6a1
TimToady m: say "\t" ~~ /<:space>/ 17:00
camelia Nil
TimToady m: say "\t" ~~ /<space>/ 17:01
camelia ļ½¢ ļ½£
space => ļ½¢ ļ½£
tbrowder ref rakudo PR #1287 (and companion roast PR #361): samcv has approved it and i would appreciate someone else reviewing and merging it. 17:19
lizmat m: my @a = ^10; my $b = List.new(|@a); dd $b; $b[0] = 42; dd $b # that's unexpected to me 19:17
camelia List $b = $(0, 1, 2, 3, 4, 5, 6, 7, 8, 9)
List $b = $(42, 1, 2, 3, 4, 5, 6, 7, 8, 9)
lizmat am I wrong in expecting decontainerization ?
timotimo why wouldn't changing $b change $b? 19:28
did you mean to output @a after the change? 19:29
lizmat well, if I turn an Array into a List, I sorta expected it to lose its containers
timotimo oh
now i get what you mean :) 19:30
lizmat m: my @a = ^10; my $b = @a.List; dd $b; $b[0] = 42 # expected this
camelia List $b = $(0, 1, 2, 3, 4, 5, 6, 7, 8, 9)
Cannot modify an immutable List ((0 1 2 3 4 5 6 7 8 9))
in block <unit> at <tmp> line 1
lizmat so, List.new(|@a) and @a.List are subtly different 19:31
Geth rakudo: 00632edb6f | (Elizabeth Mattijsen)++ | src/core/multidim_slice.pm
Fix for GH #1297

The indices were hard checked for Int, it's a bit smarter now		 20:49
synopsebot RAKUDO#1297 [open]: github.com/rakudo/rakudo/issues/1297 Indices for shaped arrays need manually coerced to numbers
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Streamline Array.iterator.push-until-lazy a bit 21:02
travis-ci.org/rakudo/rakudo/builds/312096572 github.com/rakudo/rakudo/compare/a...02bc2994f0
Zoffix R#1260 is related to List.new above though the fix in it won't fix the slipping stuff. 21:11
synopsebot R#1260 [open]: github.com/rakudo/rakudo/issues/1260 List.new creates unwanted containers
Zoffix Dunno if it should outright strip containers. Should prolly be similar to (...). If I put a Proxy up in it, I don't want it vanishing, no? 21:12
lizmat A Proxy *is* a container 21:16
so it *will* be removed
m: use nqp; say nqp::iscont(Proxy.new( FETCH => {}, STORE => {} )) 21:17
camelia 1
Zoffix I know. I'm saying that might not be what the user wants. It won't be removed in (....) 21:20
And if they want decont, they can do it manually. Array.List to me says convert it to immutable list; List.new to me says "make a List out of this stuff" 21:21
m: my $x := Proxy.new: :STORE(-> $, $ {;}), :FETCH{;}; my @l := ($x,); @l[0] = 42 21:22
camelia ( no output )
Zoffix
lizmat m: use nqp; say nqp::decont(Proxy.new( FETCH => { say "fetching"; 42 }, STORE => {} ))
camelia fetching
42
Zoffix It fetched the value and removes the proxy; can't store stuff in it no mote 21:23
*more
Zoffix &
jnthn Having List.new consistent with infix:<,> feels sensible overall to me
lizmat m: use nqp; say nqp::iscont(List.new(1,2,3)[0]) # so this would be inconsistent, right? 21:30
camelia 1
lizmat m: say use nqp; List.new(1,2,3)[0] = 42
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routines:
nqp used at line 1
use used at line 1
lizmat m: say List.new(1,2,3)[0] = 42
camelia 42
lizmat jnthn: you would consider ^^^ wrong, right ? 21:31
jnthn Yes.
Very.
lizmat ok, will look at it bow
*now
jnthn lizmat++
Uff, not many days until my advent post is due
lizmat jnthn: looks like the error is in handling of the sig: method new(**@things) 21:32
that appears to be creating containers for 1,2,3
jnthn Yeah, it would
**@things is raw ?
lizmat no, just **@things 21:33
jnthn I was suggesting the fix ;)
lizmat ah
testing
$ 6 'say List.new(1,2,3)[0] = 42' 21:35
Cannot modify an immutable List ((1 2 3))
jnthn :) 21:37
bbiab
Geth rakudo: d80df073ad | (Elizabeth Mattijsen)++ | src/core/List.pm
Make List.new not create containers

List.new should not create containers, it did through a faulty sig.
  jnthn++ for suggesting the fix.		 21:45
timotimo that's sig, broh 21:46
jnthn :) 22:29
gfldex one of my benchmark got 15% faster in the last 2 weeks 22:35
MasterDuke i think jnthn, lizmat, timotimo, and Zoffix have all commited optimizations in that timespan 22:38
timotimo i only did the tinyest thing
lizmat if it's in the past 2 weeks, it's probably jnthn's inlining prevention removing work 22:39
:w
oops :-)
why don't we have a raffle for that :-)
timotimo i'd maybe call it "inlining enablement work"?
gfldex .oO( inline moarification ) 22:40

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!