dalek kudo/nom: 4abef22 | (Brian Gernhardt)++ | tools/build/Makefile-Moar.in:
Move --nqp-lib=blib into M_RUN_PERL6

Any option that has to be included every time you use the variable should just be in the variable.		 08:10
kudo/nom: 9855adc | (Brian Gernhardt)++ | Configure.pl:
Add blib argument to create-moar-runner.pl for debug
rakudo/nom: b67c1e4 | niner++ | / (2 files):
rakudo/nom: Merge pull request #936 from Benabik/nom
rakudo/nom:
[TuxCM] This is Rakudo version 2016.11-140-gb67c1e45a built on MoarVM version 2016.11-27-g3171dbbe 09:07
csv-ip5xs 3.199
test 14.075
test-t 6.488
csv-parser 13.978
nine It's funny how often I read my name in the weekly and think "oh, what did I do now?! Ah, of course..." 09:53
dalek p: 443fb95 | (Pawel Murias)++ | src/NQP/Actions.nqp:
Support calling the handlepayload and handle ops using the nqp:: syntax.
 11:00
p: 568fcc2 | (Pawel Murias)++ | src/vm/js/ (4 files):
[js] Make nqp::exception work when it's not directly inside a nqp::handle.
p: 0fe387b | (Pawel Murias)++ | t/nqp/044-try-catch.t:
Test having nqp::exception be used outside of a nqp::handle.
babydrop Yeah, the same, "wait... what? Oh yeah... Now I remember doing that"
dalek p: 2de9805 | (Pawel Murias)++ | t/nqp/044-try-catch.t:
Remove debugging leftover.
 11:56
kudo/nom: 0ee6bc0 | lizmat++ | / (2 files):
Make native array.reverse|rotate about 20x faster

By using the same logic as for 1dimmed shaped arrays, instead of HLLing up and down.		 12:53
babydrop .tell lizmat Not sure if RT#130274 rendered for you OK, but it references stackoverflow.com/questions/4097008...not-in-201 Core thing being Str.match now sets $/ when it didn't before. We made it that way on purpose, right? 14:27
yoleaux2 babydrop: I'll pass your message to lizmat.
dalek kudo/nom: 16fbd71 | (Zoffix Znet)++ | src/ (3 files):
Fix incorrect error for extra clauses with unless/without

  - Add X::Syntax::WithoutElse
  - Add .keyword attribute to X::Syntax::UnlessElse
  - Split unless/without proto rule into two separate ones
   - Lets us avoid a conditional and use of `{}` to refresh
   $/, which is constly to update.
   - Less mess when figuring out which exception to throw
  - Capture the offending keyword so it could be properly displayed
   in the error message
Fixes RT#130279: rt.perl.org/Ticket/Display.html?id=130279		 14:40
ast: d7328b7 | (Zoffix Znet)++ | S04-statements/ (2 files):
Test extra conditionals with unless/without show proper error

RT#130279: rt.perl.org/Ticket/Display.html?id=130279 Fix commit: github.com/rakudo/rakudo/commit/16fbd7147f		 14:41
lizmat babydrop: that was my understanding, yes 14:55
yoleaux2 14:27Z <babydrop> lizmat: Not sure if RT#130274 rendered for you OK, but it references stackoverflow.com/questions/4097008...not-in-201 Core thing being Str.match now sets $/ when it didn't before. We made it that way on purpose, right?
babydrop Cool. I'll respond to the ticket. 14:56
lizmat commute to Amsterdam.pm meeting& 15:07
gfldex .tell lizmat neither do I or I would have proposed something. No worries, nobody is gonna die from a missing docs confusion. 15:35
yoleaux2 gfldex: I'll pass your message to lizmat.
babydrop .ask lizmat would recall whether the $* twigils were meant to be here in this commit? The tests actually result in a Failure, which is masked by the @ coercert: github.com/perl6/roast/commit/2d48...5e68e7R146 16:32
yoleaux2 babydrop: I'll pass your message to lizmat.
babydrop .tell lizmat conversation on that: irclog.perlgeek.de/perl6/2016-12-06#i_13691323 16:33
yoleaux2 babydrop: I'll pass your message to lizmat.
babydrop .tell lizmat so I'm wondering whether we fix the tests by removing the '*' from twigils or were they meant to work with the twigils and the tests need to be fudged...
yoleaux2 babydrop: I'll pass your message to lizmat.
[Coke] will coercion types ever work in variable declarations? (#130253) 16:34
[Coke] adds [citation needed] to the ticket. 16:36
babydrop .tell lizmat nevermind, TimToady answered in irclog.perlgeek.de/perl6/2016-12-06#i_13691502 16:39
yoleaux2 babydrop: I'll pass your message to lizmat.
dalek kudo/nom: ab3a59c | lizmat++ | src/core/List.pm:
Make permutations/combinations die instead of fail
 16:56
babydrop hm, so on that SO someone is curious about the rationalle for the .match change... 17:23
committable6, 2016.07 'foo'.match: /./; say $/
committable6 babydrop, ¦«2016.07»: ｢f｣
babydrop But I see it was setting $/ even before the refactor
psch committable6: 2015.12 'foo'.match: /./; say $/
committable6 psch, ¦«2015.12»: ｢f｣
babydrop committable6, 2016.07 sub ($/) { 'foo'.match: /./; say $/ }($/)
committable6 babydrop, ¦«2016.07»: Nil
babydrop committable6, 2016.07 sub ($/) { 'foo'.match: /./; say $/ }(42)
committable6 babydrop, ¦«2016.07»: 42
babydrop m: sub ($/) { 'foo'.match: /./; say $/ }(42) 17:24
camelia rakudo-moar ab3a59: OUTPUT«Cannot assign to a readonly variable ($/) or a value␤ in sub at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
babydrop Except in ^ that case, we now cry about readonly.
So... what's the rationalle? Just a bug fix since old behaviour is kinda wonky?
psch i strongly remember the readonly error already happening in Action methods that fiddle with $/ 17:25
i.e. for a Grammar
so i'm mostly curious why that old case doesn't complain, honestly
babydrop yeah, wehn you use ~~ / ... / for example, but taht didn't happen with .match
psch ah, okay
babydrop committable6, 2016.07 sub ($/) { 'foo' ~~ /./; say $/ }(42)
committable6 babydrop, ¦«2016.07»: Cannot assign to a readonly variable or a value␤ in sub at /tmp/rAz0Rtxn2P line 1␤ in block <unit> at /tmp/rAz0Rtxn2P line 1␤ «exit code = 1»
babydrop I'll just respond that old behaviour was a bug, since old .match sets $/ and in the particular case of $/ being already in scope it silently failed 17:26
psch committable6, 2016.07 sub ($/ is read-only) { 'foo'.match: /./; say $/ }(42)
committable6 psch, gist.github.com/5d2b871f75e15b401e...d96030fa03
psch oh shush :l 17:27
m: sub f($ is ro) {}
camelia rakudo-moar ab3a59: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Can't use unknown trait 'is ro' in a parameter declaration.␤at <tmp>:1␤ expecting any of:␤ rw␤ readonly␤ copy␤ required␤ raw␤ leading_docs␤ …»
psch oh, no dash, i see
committable6, 2016.07 sub ($/ is readonly) { 'foo'.match: /./; say $/ }(42)
committable6 psch, ¦«2016.07»: 42
psch right, yeah, definitely somewhat buggy because it just clobbers over whatever already is around
babydrop doesn't see the clobbering 17:28
psch err, yes
i'm not writing good anymore apparently
babydrop But it's still buggy, since you may expect it reading from the .match result but it's the old $/ 17:29
psch well, infix:<~~> surely isn't supposed to have magic that assigns to $/
so .match itself must do that
otoh, maybe not 17:30
i'm also not thinking good anymore i think :P
psch &
babydrop :)
IIRC infix:<~~> on /.../ has some sort of magicks in the grammar that makes it do a regex match instead of going through the .ACCEPTS dance
lizmat m: <a b c>.grep( True ).elems.say # another case 17:31
camelia rakudo-moar ab3a59: OUTPUT«1␤»
babydrop :( 17:32
dalek kudo/nom: bc13bb5 | lizmat++ | src/core/Any-iterable-methods.pm:
Make sure .grep doesn't fail but dies
lizmat $ 6 '<a b c>.grep( True ).elems.say' 17:33
Cannot use Bool as Matcher with '.grep'. Did you mean to use $_ inside a block?
babydrop drive-by bug fixes \o/ 17:37
dalek kudo/nom: 2e1b82c | (Zoffix Znet)++ | src/core/Str.pm:
Fix spurious warnings on .trans with regex pair complements

On an unsuccessful match, the $/ ends up being Nil and .chars on a Nil warns.
Found by tojo++ irclog.perlgeek.de/perl6/2016-12-06#i_13692293		 19:42
ast: f01416c | (Zoffix Znet)++ | S05-transliteration/trans.t:
Test .trans with regex pair complement mode does not produce warnings

Bug finding: irclog.perlgeek.de/perl6/2016-12-06#i_13692293 Fix commit: github.com/rakudo/rakudo/commit/2e1b82cfa2		 19:43
p: db76673 | (Pawel Murias)++ | src/vm/js/nqp-runtime/ctx.js:
[js] Remove debugging leftover.
 19:53
[Coke] removing debug output seems like a good sign. :) 20:18
pmurias: how goes? 20:19
babydrop m: class Foo { method x { "42" } }; for ^1000_000 { my $x = Foo.new }; say now - INIT now; 20:44
camelia rakudo-moar 2e1b82: OUTPUT«1.4886558␤»
babydrop m: class Foo { method x { "42" } }; for ^1000_000 { my $x = Foo.new.x }; say now - INIT now;
camelia rakudo-moar 2e1b82: OUTPUT«1.6499938␤»
babydrop m: say 1.64999/1.48865
camelia rakudo-moar 2e1b82: OUTPUT«1.1083801␤»
babydrop What I was gonna say is IMO say and put and print should not cheat with Str types and call appropriate .Str or .gist methods regardless. 20:48
But it's not important enough to warrant a speed decrease at the current time. So *shrug* 20:49
dalek kudo/nom: 4f4737d | lizmat++ | src/core/Shaped (2 files):
Make sure 1+ dimmed Arrays can .sum
 21:01
kudo/nom: 4fd6e94 | lizmat++ | src/core/ (2 files):
Make sure all .sum are nodal
[Coke] brrt: any chance to move your advent post up to the 9th? (or the the 10th or 11th)? We have gaps, and I'd like to try to move the gaps towards the end of the month. 21:02
jdv79 is this year's holiday season less stressful at least around here? 21:11
moritz oh yes
jdv79 very nice
babydrop .ask TimToady perhaps we need a different approach than dying in leu of Failures in things that may return a list? For example this code does not throw and I only scratched the surface: dd [.elems, .pick.elems, .roll.elems, .eager.elems, .Slip.elems, .Array.elems, .List.elems, .Capture.elems, .rotor(42).elems ] given Failure.new 23:55
yoleaux2 babydrop: I'll pass your message to TimToady.

Powered by Cro and the Raku® Programming Language.

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

Thank you!