moritz samcv, japhb, Zoffix: perl6-all-modules does keep history, so you can dig up old modules from it 05:25
samcv yep i pointed that out moritz
moritz samcv: ok, I didn't backlog carefully enough then 05:26
samcv np
[Tux] This is Rakudo version 2017.04.3-61-g139243971 built on MoarVM version 2017.04-44-gf0db8822 07:03
csv-ip5xs 3.166
test 13.659
test-t 5.092 - 5.143
csv-parser 13.207
Zoffix someone who has reddit access should toss both rakudobrew and panda off the sidebar on www.reddit.com/r/perl6/ 09:57
timotimo good point
any idea for a second entry under the ## Install heading? 09:58
perhaps nxadm's packages, and the docker images?
moritz +1 to the packages 09:59
Geth rakudo/nom: 9f5c8e94c6 | MasterDuke17++ | src/Perl6/Actions.nqp
Fix typed optional Array and Hash parameters

When handling typed optional Array and Hash parameters that don't have a value passed we now nqp::create() the correct empty parameterized type.
Fixes RT #118555		 10:01
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=118555
rakudo/nom: 6231ecb014 | MasterDuke17++ | src/Perl6/Actions.nqp
De-dupe code shared between Array and Hash cases
 10:02
rakudo/nom: a968064c77 | (Zoffix Znet)++ (committed using GitHub Web editor) | src/Perl6/Actions.nqp
Merge pull request #1072 from MasterDuke17/fix_typed_optional_array_and_hash_parameters

Fix typed optional Array and Hash parameters
timotimo Zoffix: see if you like the changes i made to the sidebar 10:08
Zoffix timotimo++ looks good 10:10
timotimo Zoffix++ for noticing 10:16
it's easy to forget about things like that 10:17
Zoffix samcv++ # grant 10:44
Zoffix gulps reading the rejection comments on RPerl grant 10:56
My own record on interaction isn't exactly stellar...
robertle on rakudo.org/how-to-get-rakudo/ the top-right box for 32bit windows says "2016.01", shoudl that not be 2017? 11:08
timotimo nope, there's no rakudo star newer than 2016.01 for the no-jit version 11:09
i.e. for x86
robertle timotimo: ok, just looked odd 11:10
timotimo yeah, hopefully we'll find someone who can build stuff on win32 for the next rakudo star
Zoffix: the part 2 io grant work upgrade information post is still the pinned one, should i put part 3 up instead? 11:15
Zoffix timotimo: yeah, thanks
timotimo took me a few minutes to find the announcement options again 11:18
Geth nqp/master: 9 commits pushed by (Pawel Murias)++ 11:57
Zoffix wonders why atom highlights `TEMP` specially 12:27
m: TEMP 42 12:28
camelia WARNINGS for <tmp>:
Useless use of constant integer 42 in sink context (line 1)
Zoffix huh
m: dd TEMP 42
camelia Non ast passed to WANTED: NQPMu
Weird node in analyze: NQPMu
WARNINGS for <tmp>:
Useless use of constant integer 42 in sink context (line 1)
===SORRY!===
Unknown QAST node type NQPMu
Zoffix m: TOMP 42
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
TOMP used at line 1
timotimo i didn't know about TEMP being a thing 12:29
Zoffix I see it mentioned in design.perl6.org/S06.html#Temporization
timotimo i don't think this whole TEMP method business has made it into anything 12:30
Zoffix m: sub foo { state $x++; TEMP {{ $x = 42}}; $x }; foo.say; foo.say; { say temp foo; foo.say; }; foo.say 12:31
camelia 1
2
Can only use 'temp' on a container
in block <unit> at <tmp> line 1
Zoffix I guess "largely unimplemente" is the status of it all
timotimo well, temp in lowercase exists at leas 12:32
least*
Zoffix Aye
m: use MONKEY; augment class Pair { method Pair { self } }; 'foo' ~~ / $<foo>=.**2 /; say $/.hashĀ».Str.push: :github.Pair 12:41
camelia {foo => fo, github => True}
Zoffix What you think of adding .Pair to Pair?
Geth nqp: 526275c3f1 | (Pawel Murias)++ | src/vm/js/nqp-runtime/core.js
[js] Fix nqp::encode with unsigned buffers
 12:42
nqp: b82f6096d5 | (Pawel Murias)++ | src/vm/js/Operations.nqp
[js] Make nqp::chr supports arguments that are bigger than 2**16
nqp: a5adc67705 | (Pawel Murias)++ | t/nqp/059-nqpop.t
Test nqp::chr with an argument that bigger than 2**16
jnthn Zoffix: Things having themselves as a coercer is fine
Zoffix Cool. I'll add it then. 12:44
jnthn I wonder if we can sensibly .Pair anything else... :)
timotimo single-element hashes potentially, but ... doesn't seem useful 12:46
[Coke] m: "HI".Pair.perl.say 12:47
camelia No such method 'Pair' for invocant of type 'Str'
in block <unit> at <tmp> line 1
Zoffix m: <a b>.Pair.say 12:48
camelia No such method 'Pair' for invocant of type 'List'
in block <unit> at <tmp> line 1
Zoffix eh, no good
timotimo m: <a b>.pairup.perl.say 12:49
camelia (:a("b"),).Seq
Geth rakudo/nom: bd9e57309a | (Zoffix Znet)++ | src/core/Pair.pm
Add Pair.Pair method

  - Returns self
  - Provides another way to pass a Pair as a positional arg,
   instead of a named one.
  irclog.perlgeek.de/perl6-dev/2017-...i_14495138		 12:56
timotimo that's an interesting use case, yeah 12:57
i didn't even notice what you did up there
Geth roast: 019b0b5150 | (Zoffix Znet)++ | S02-types/pair.t
Test Pair.Pair

Rakudo impl: github.com/rakudo/rakudo/commit/bd9e57309a
Zoffix buggable: author zoffixznet 13:14
buggable Zoffix, Found 37 dists for zoffixznet. See temp.perl6.party/buggable/40690276...98883.html
Zoffix \o/
Now I can finally know what modules I wrote :P
buggable: author perl6 13:15
buggable Zoffix, Found 10 dists for perl6. See temp.perl6.party/buggable/18626134...98935.html
Zoffix buggable: author perl6-community-modules
buggable Zoffix, Found 7 dists for perl6-community-modules. See temp.perl6.party/buggable/01452236...98941.html
Zoffix pmurias: no, my Int $foo = NaN; should not work. The test should be removed 13:17
pmurias: and here's the ticket with all the discussions: rt.perl.org/Ticket/Display.html?id=61602 13:18
.tell moritz not sure if you've seen my comment on the Issue, but my suggestion to use CSS to break long lines on logs was wrong. That CSS property does jack all due to auto-resizing table the stuff is in. So lines no longer are broken right now: github.com/moritz/ilbot/issues/51#...-296676513 13:21
yoleaux Zoffix: I'll pass your message to moritz.
moritz that's unfortunate 13:22
yoleaux 13:21Z <Zoffix> moritz: not sure if you've seen my comment on the Issue, but my suggestion to use CSS to break long lines on logs was wrong. That CSS property does jack all due to auto-resizing table the stuff is in. So lines no longer are broken right now: github.com/moritz/ilbot/issues/51#...-296676513
Zoffix Man the channels are really busy. Even just scanningly scrolling through #perl6's log takes too much time :/
Guess it's good. Though I silently grumble when dev questions are discussed in #perl6 instead of #perl6-dev :} 13:23
pmurias Zoffix: I'll ask my roast question here then ;) 13:24
Zoffix Thanks :)
pmurias '#OK not indicate octal' are just regular comments? 13:36
Zoffix pmurias: I think that's "OK to receive warning about 01 not being an octal in Perl 6" 13:38
m: 01 13:39
camelia Potential difficulties:
Leading 0 does not indicate octal in Perl 6. Please use 0o1 if you mean that.
at <tmp>:1
------> 03017ā5<EOL>
WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Zoffix m: use Test; is "{\01;42}", "42"
camelia Potential difficulties:
To refer to a positional match capture, just use $0 (numbering starts at 0).
Parenthesize as \(...) if you intended a capture of a single numeric value.
at <tmp>:1
------> 3use Test; is "{\7ā0501;42}",ā€¦
Zoffix yup
match capture, eh....
That's buggish. It thinks \0 is part of string and not code 13:40
or something
Ah
m: \0
camelia Potential difficulties:
To refer to a positional match capture, just use $0 (numbering starts at 0).
Parenthesize as \(...) if you intended a capture of a single numeric value.
at <tmp>:1
------> 3\7ā050
Zoffix right, nevermind
Hm, I guess those tests will throw when Int.Capture will be made to throw 13:41
waaaaat "# The code of the closure takes a reference to the number 1, discards it" 13:42
We're talking about these BTW: github.com/perl6/roast/blob/master....t#L13-L15
m: use MONKEY; augment class Int { method Capture { die "no cookie" } }; dd \ 1 13:46
camelia 1
Zoffix huh... ok then :/
Ah, that's just an unspace 13:50
though dd \1 also doesn't die and gives \(1). Is it 'cause it treats stuff after it as a list? 13:52
m: dd \"foo", "bar" 13:53
camelia \("foo")
"bar"
Zoffix hm, OK then
pmurias Zoffix: I was just asking if #OK is some special magic to silence warning or just a comment for humans 14:12
Zoffix isn't aware of magic comments in roast, so would assume the latter 14:13
[Coke] hurm. it's magical somewhere, trying to remember if there was a p6 context. 15:16
moritz fudge maybe? 15:17
Zoffix Note to self: $foo && $foo = 42 has wrong precedence :( 15:28
jnthn m: my $foo = 1; $foo && $foo = 42; say $foo 15:29
camelia 42
jnthn m: my $foo = 0; $foo && $foo = 42; say $foo 15:30
camelia 42
jnthn :)
m: my $foo = 0; $foo and $foo = 42; say $foo
camelia 0
jnthn m: my $foo = 1; $foo and $foo = 42; say $foo
camelia 42
jnthn `and` has the correct one for that, though :)
Zoffix yeah :) 15:32
[Coke] win 2 15:37
Geth roast: 71a6423a99 | (Zoffix Znet)++ | S16-io/home.t
[io grant] Add $*HOME tests
 15:41
rakudo/nom: 7412184fcd | (Zoffix Znet)++ | src/core/Process.pm
[io grant] Make $*HOME default to Nil, not Any

RT#130456: rt.perl.org/Ticket/Display.html?id=130456		 15:42
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=130456
Geth rakudo/nom: d22303d019 | (Zoffix Znet)++ | t/spectest.data
Run S16-io/home.t test file
roast: 2d9194fff9 | (Zoffix Znet)++ | S16-io/home.t
Test we can assign to $*HOME in all the permutations
 15:45
Zoffix $*TMPDIR is set to stuff that's checked to be `.rwx & .d` and if nothing is found, it uses ".".IO, while $*HOME is set to Nil 15:57
(got distracted...) And isn't checked that the values obtained from env vars even exist 15:59
I kinda want to make $*HOME be set in the same way as $*TMPDIR: try HOME (+ HOMEDRIVE+HOMEDIR on Windows) check it's .rwx & .d and if didn't find anything, default to ".".IO 16:00
Because otherwise we're basically telling users that any time they reach for $*HOME they need to check it's not Nil AND that they can do whatever things they want to do in it 16:01
Arguments against: people might want to store sensitive info in $*HOME dir and it ending up being some random path the app got run from on some system that unset the HOME is kinda sucky 16:08
Zoffix takes the path of least resistance and leaves it be as is 16:10
Geth rakudo/nom: 3c7cd9333a | (Zoffix Znet)++ | 3 files
Simplify method call

We have .rwx method; no need to use .r .w .x instead		 16:16
Zoffix And for the perms checkage: usecase running a script with privs to read but not write $*HOME; thge .rwx & .d check would fail and result in $*HOME being Nil, so also good to leave as is 16:19
ZofBot: right??
ZofBot Zoffix, m: Failure.new.?slurp-rest
Zoffix m: IO::Path.e 16:25
camelia IO::Path is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in method FALLBACK at src/RESTRICTED.setting line 31
in block <unit> at <tmp> line 1
Zoffix m: Int.absd
camelia No such method 'absd' for invocant of type 'Int'
in block <unit> at <tmp> line 1
Zoffix m: Int.abs
camelia Invocant requires an instance of type Int, but a type object was passed. Did you forget a .new?
in block <unit> at <tmp> line 1
Zoffix
.oO( wonder if it's possible to include the method name called in that last error to give some hint for debug location, when line number is missing :/ )
 16:26
(like when exception is caught and it's .message is printed)
lizmat perhaps someone should say "'file'.IO.words.Bag' to www.reddit.com/r/perl/comments/67v...unt_words/ 16:41
Zoffix will do one-better :) 16:45
As soon as Text::CSV's test suite finishes -_-
"type "perldoc split" into your URL box for more info about the split command." that's pretty neat actually. We should come up with some term that would learn to search results on docs site via search engines 16:46
lizmat: done :) 16:50
Geth roast: 785fea051b | MasterDuke17++ | S06-signature/optional.t
Add tests for RT #118555

Optional typed Array and Hash parameters should act like regular optional Array and Hash paramters (just with a type constraint) and not die.		 16:58
synopsebot6 Link: rt.perl.org/rt3/Public/Bug/Display...?id=118555
MasterDuke_ what can one do with, or what info can be obtained from, a lexpad? 17:01
lizmat Zoffix++ # providing food for next week's P6W :-) 17:02
Zoffix \o/
lizmat Zoffix: fwiw, I'bve just had a flapper in t/spec/S32-io/indir.t 17:04
test 74
can't reproduce
Zoffix OK, I'll take a look later on
Geth rakudo/nom: b233281602 | (Elizabeth Mattijsen)++ | src/core/set_operators.pm
Make Iterable (^) Iterable about 3.5x faster

  - with a 100 element Array with itself (worst case)
  - give Iterable (^) Iterable its own candidate written using nqp ops		 17:06
jdv79_ huh? 17:07
lizmat jdv79_?? 17:08
git Fu question: I want to reapply commit 1534652765, what do I do ? 17:10
jdv79_ idk. i saw someone said something to me but lost it
cherry-pick maybe? 17:11
lizmat yeah, looks like 17:12
Geth rakudo/nom: 3b247096e9 | (Elizabeth Mattijsen)++ | src/core/set_operators.pm
Make (-) a multi

Start of optimization work on (-)		 17:13
travis-ci Rakudo build passed. Zoffix Znet 'Run S16-io/home.t test file' 17:32
travis-ci.org/rakudo/rakudo/builds/226447653 github.com/rakudo/rakudo/compare/7...2303d01920
MasterDuke_ hey, travis is back 17:55
geekosaur hah 18:06
samcv MasterDuke_, it came back yesterday, to non freenode places at least 18:24
or 2 days ago when i added an irc field
so was suprised it actually worked, since it was broken here. hopefully it's fixed on freenode now. wonder if freenode blocked it for joining channels and spamming messnages all day long! 18:25
MasterDuke_ jnthn: in bind_one_param in BOOTSTRAP.nqp, would it be possible to find out the method bind_one_param is binding for from its $lexpad argument? 18:27
jnthn I think there's a nqp::ctxcode or so 18:28
Which you can use nqp::getcodeobj on
MasterDuke_ cool, i'll experiment with them 18:29
ugexe should `try { }<\n>|| something-else()` short-circuit without having to unspace like `try { }\<\n>|| something-else()`? gist.github.com/ugexe/47c6369bead8...be78939bd2 18:37
jnthn Yes. The syntax rule is that a } at the end of a line (or only whitespce after it on the line) terminates the statement 18:38
So the next line will be a new statement
And the || will be interpreted as a prefix 18:39
m: try { }; || something-else()
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
something-else used at line 1
jnthn As there
ugexe I see... I was thinking like I do with map/grep, and now that you mention it those instances are using ({ }), not { } 18:42
`No such method 'e' for invocant of type 'Distribution::Resource'` when doing %?RESOURCES<...>.e - wonder if there is a better way to delegate all that IO stuff in Distribution::Resources so nothing is missed 18:48
Geth rakudo/nom: aa23a91fb2 | (Elizabeth Mattijsen)++ | src/core/set_operators.pm
Optimize a bit for empty (Set|Bag|Mix)Hashes
 18:58
lizmat afk&
(off to see some more Guardians Of The Galaxy) 19:01
Zoffix jdv79_: yeah, it was me. CPAN stuff was brought up yesterday, so I asked how it was going. 19:34
Hmm. I just got my own office (yey), but it has a weird desk that's a bit too narrow for my 3-computer setup :| Typing with my elbows hanging out (ney) 19:35
ZofBot: can't ever win!
ZofBot Zoffix, c: HEAD say 42
MasterDuke_ m: Int.abs 19:40
camelia Invocant requires an instance of type Int, but a type object was passed. Did you forget a .new?
in block <unit> at <tmp> line 1
MasterDuke_ Zoffix: locally i have that saying `Invocant requires an instance of type Int when method 'abs' is called on it, but a type object was passed. Did you forget a .new?`
that wording is awkward, but the functionality is there 19:42
Zoffix Cool
MasterDuke_ anybody have suggestions for a better worded message?
Zoffix "metho X must be called on an instance of XXX, but a type object was passed"? 19:43
timotimo should we also be giving info about the type the method is declared on and the type object that was actually passed? 19:46
geekosaur or Invocant for $method requires...
timotimo because what if the type object you're passing is derived from the one you were supposed to pass 19:47
but that'd not be an actual problem
MasterDuke_ 'Invocant' can also be 'Parameter $varname'
so the wording need to take that into account 19:48
or those two options changed also 19:49
Geth nqp: 2bfafb8721 | (Pawel Murias)++ | 2 files
Reindent code
 21:11
nqp: 78b233ef5b | (Pawel Murias)++ | 3 files
[js] Work around for unicode issue

MoarVM uses NFG while the js backend doesn't do it. The js backend is cross compiling so that causes a problem with our hardcoded as string literals lists of characters. Work around it using nqp::chr(0x....).
pmurias samcv: what's the status of your unicode grant? 21:12
samcv approved :)
yesterday
more than willing to help you with anything you need now/during the grant as well. since i had hoped the script i'm making could also do stuff for other backends as well 21:13
pmurias samcv: thanks, I'll look into the UCD module tommorow 21:16
Geth roast: 38128f4668 | (Pawel Murias)++ | S02-types/nan.t
Remove incorrect fudged test
 22:07
roast: 0b043cd23e | (Pawel Murias)++ | S02-lexical-conventions/unicode-whitespace.t
[js] Remove no longer needed fudges

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!