gfldex m: sub s(*@a){}; say &s.arity 01:46
camelia 0
gfldex I wonder if that should return Whatever. 01:47
or ∞ 01:49
Zoffix gfldex: just use .count if you want the maximum number 02:24
m: dd [.arity, .count] for -> *@ {}, -> $, $? {}, {;} 02:25
camelia [0, Inf]
[1, 2]
[0, 1]
Geth rakudo: b212fc5e20 | (Zoffix Znet)++ | src/core/Cool.pm6
Remove Cool.path; cfa++ for pointing it out

Looks to be a vestigial leftover from when there were plans for qp// and the like. We already have Cool.IO, so there's no point for .path. It's unspecced and undocumented.
03:10
rakudo: 76eff07dd6 | (Zoffix Znet)++ | src/core/Cool.pm6
Revert "Remove Cool.path; cfa++ for pointing it out"

This reverts commit b212fc5e20b5f4989d2120ec3c10c1f755c2f2be.
zef was using it. Going to push through full deprecation cycle instead.
04:33
6.d-prep: 87d4fbacde | (Zoffix Znet)++ (committed using GitHub Web editor) | TODO/FEATURES.md
List deprecation of Cool.path
04:35
rakudo: a918028e05 | (Zoffix Znet)++ | docs/archive/2018-03-04--Polishing-Rationals.md
Add "Polishing Rationals" proposal
06:22
Zoffix New proposal: "Polishing Rationals": github.com/rakudo/rakudo/blob/mast...tionals.md 06:23
Fixes a bunch of bugs, and based on my preliminary testing also makes stuff faster
.tell jnthn a proposal for some tweaks to Rationals to fix a number of bugs and the race condition, if you wanted to review: github.com/rakudo/rakudo/blob/mast...tionals.md 06:24
yoleaux Zoffix: I'll pass your message to jnthn.
Zoffix .tell TimToady a proposal for some tweaks to Rationals to fix a number of bugs including the issue of some Rats being >64-bit denominators, if you wanted to review: github.com/rakudo/rakudo/blob/mast...tionals.md
yoleaux Zoffix: I'll pass your message to TimToady.
Zoffix And I think it'll make things slower to make Rat default to native `uint64` denominator, as there ain't many places where'd you'd work with it in native: ops work in full Int, as you could have a FatRat or an operation that blows the 64-bit limit, so it'd be constantly being boxed up 06:36
m: use nqp; my int $n = 10.rand.ceiling; class Foo { has uint64 $.foo =10 }; my $o := Foo.new; { for ^20_000_000 { nqp::add_I($o.foo, 10, Int); Nil }; say now - ENTER now } 16:17
camelia 1.2885429
Zoffix m: use nqp; my int $n = 10.rand.ceiling; class Foo { has Int $.foo =10 }; my $o := Foo.new; { for ^20_000_000 { nqp::add_I($o.foo, 10, Int); Nil }; say now - ENTER now }
camelia 1.4201100
Zoffix I think I'll add nativizing Rat denominator to my Rat Polish proposal after all.
m: use nqp; my int $n = 10.rand.ceiling; class Foo { has Int $.foo =10 }; my $o := Foo.new; { for ^20_000_000 { Nil }; say now - ENTER now } 16:41
camelia 1.0447117
Zoffix m: use nqp; my int $n = 10.rand.ceiling; class Foo { has Int $.foo =10 }; my $o := Foo.new; { for ^20_000_000 { Nil }; say now - ENTER now }
camelia 1.046860
Zoffix m: use nqp; my int $n = 10.rand.ceiling; class Foo { has Int $.foo =10 }; my $o := Foo.new; { for ^20_000_000 { Nil }; say now - ENTER now }
camelia 1.06424856
Zoffix oops, wc
dogbert17 AlexDaniel: you there? 16:42
AlexDaniel dogbert17: yes, more or less
dogbert17 didn't you have a bot like evalable where I could add env vars? 16:43
AlexDaniel committable yes 16:45
dogbert17 does it have any modules installed, like say HTTP::UserAgent?
AlexDaniel c: MVM_SPESH_DISABLE=1 HEAD say 42
committable6 AlexDaniel, ¦HEAD(a918028): «42␤»
AlexDaniel no but we can make it work, what are you trying to do?
dogbert17 oneliner to show a bug 16:46
this code use HTTP::UserAgent; my $cl = HTTP::UserAgent.new; loop { my $res = $cl.get("rosettacode.org"); }
with MVM_SPESH_NODELAY=1
AlexDaniel c: MVM_SPESH_NODELAY=1 HEAD gist.githubusercontent.com/AlexDan...thingie.p6 16:47
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦HEAD(a918028): ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
AlexDaniel c: MVM_SPESH_NODELAY=1 HEAD gist.githubusercontent.com/AlexDan...thingie.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, gist.github.com/da931ac542a3daea64...76a67f249e
AlexDaniel dogbert17: this bug? ↑
dogbert17 yup, aka github.com/rakudo/rakudo/issues/1573 16:48
AlexDaniel c: MVM_SPESH_NODELAY=1 2016.01 gist.githubusercontent.com/AlexDan...thingie.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
dogbert17 my first golf was longer as you can see
committable6 AlexDaniel, ¦2016.01: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
AlexDaniel c: MVM_SPESH_NODELAY=1 2016.01 gist.githubusercontent.com/AlexDan...thingie.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦2016.01: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
AlexDaniel heh… well… 16:49
dogbert17 hmm
AlexDaniel c: MVM_SPESH_NODELAY=1 2016.01 gist.githubusercontent.com/AlexDan...thingie.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦2016.01: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
dogbert17 the bug is 'relatively' new
AlexDaniel oh, that's because it works there, right?
c: MVM_SPESH_NODELAY=1 2017.12 gist.githubusercontent.com/AlexDan...thingie.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, gist.github.com/efdb2d496f08b9fd2a...f571814e00
dogbert17 could be of interest to jnthn
AlexDaniel c: MVM_SPESH_NODELAY=1 2017.10 gist.githubusercontent.com/AlexDan...thingie.p6 16:50
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦2017.10: ««timed out after 10 seconds» «exit signal = SIGHUP (1)»»
AlexDaniel dogbert17: well, you can probably golf it a layer deeper…
dogbert17 possibly 16:51
AlexDaniel so it happened between 2017.11 and 2017.12?
was it bisected?
dogbert17 "AlexDaniel: thundergnat: see, I bisected it: gist.github.com/Whateverable/5fb5b...5e797f959" 16:52
AlexDaniel … okay…
dogbert17 I think that was a MoarVM bump
AlexDaniel dogbert17: ok yes, same result for f78fd7c17^,f78fd7c17 with your golf 16:53
k
Geth rakudo: 207313be5a | (Zoffix Znet)++ | src/core/REPL.pm6
Do not assume Junction guarantees execution order
18:43
bartolin could someone explain to me, what's the purpose of this line: github.com/perl6/nqp/blob/41b2f055...O.nqp#L225 20:15
Was it added so that the return value of nqp::closefh is not exposed to rakudo?
If so, would it make sense to do the same for method seek? github.com/perl6/nqp/blob/41b2f055...O.nqp#L199
I'm asking because .seek on a IOHandle returns a BOOTIO object on rakudo-j and Mu on rakudo-m. The former leads to some failing tests in S32-io/seek.t, because method 'sink' is not found for invocant of class 'BOOTIO' 20:17
./perl6-j -e 'my $fh = "foo.txt".IO.open; { $fh.seek(0) }; $fh.close' ## fails with the above error 20:19
Zoffix I plan to submit a TPF grant proposal "Perl 6: Performance and Bugfixing of Rationals and Constraints on Constants". The draft is available at temp.perl6.party/Constant-Rat-TPF-Grant.html Let me know your comments. 20:40
MasterDuke Zoffix++ only $1499? i'd double that at least 20:46
AlexDaniel for others: “Performance is one of the bigger downsides of Perl 6” – this is no longer true lately. Reliability is the biggest issue now 20:47
there's no point having great performance when the damn thing crashes after a few seconds of excercising this available performance
Zoffix Yeah, I don't think perf is THE biggest issue anymore, but I think it's still "one of the" biggest.
AlexDaniel or starts leaking badly 20:48
I get that, the text is fine
MasterDuke eh, depends what you're doing. look at rindolf's code. doesn't crash, but takes 10m compared to 20s for Perl5, Python, etc
AlexDaniel what's in rindolf's code again? rosettacode stuff? 20:49
I think I missed that actually
dogbert17 is this the same Zoffix who was burned out a few days ago
AlexDaniel so? 20:50
jnthn Zoffix: I'd support that (and yeah, would support it at a higher price too :))
MasterDuke i think the *ables exercise much more of the code that's tougher to get correct and more likely to crash if buggy (instead of just being slow or give wrong results)
dogbert17 thought he was going for some RaR
Zoffix dogbert17: well, I think the next grant voting will be like at the end of March and I won't hear of the result by April, so plenty of time still :)
dogbert17 Zoffix++ 20:51
bisect: uptime 20:52
bisectable6 dogbert17, 1 day, 2 hours, 29 minutes, and 56 seconds, 218.488281MiB maxrss. This is Rakudo version 2018.01-209-gf74890550 built on MoarVM version 2018.01-97-g22d2db5e0 implementing Perl 6.c.
AlexDaniel dogbert17: yes, no longer leaks and also no longer passes its tests :P
dogbert17 the memory problems seem to be a thing of the past no ?
AlexDaniel: details details :-) 20:53
did you try the ugly SEGV workaround ? 20:54
AlexDaniel not really, no
dogbert17: memory usage for github.com/MoarVM/MoarVM/issues/680 climbs up also. It's not really leaking, it's just converging to some unreasonably high goal :) 20:57
dogbert17 AlexDaniel: you're on 64 bit right? 20:59
AlexDaniel yes
dogbert17 so my 32 bit system should get roughly half the memory usage then 21:00
dogbert17 tests
hmm, last line was: {:data("235.120 kB"), :dirty("0 kB"), :lib("0 kB"), :resident("203.448 kB"), :share("26.796 kB"), :size("264.732 kB"), :text("8 kB")} 21:02
AlexDaniel I think the dot is misleading? 21:03
dogbert17 indeed 21:04
but didn't it eat gigs on 64-bit initially
AlexDaniel I think so, yea
dogbert17 I wonder what makes this program 10 times slower than Perl 5 21:06
MasterDuke dogbert17: infix:<%> 21:35
oh, i think you were talking about github.com/MoarVM/MoarVM/issues/680, nevermind what i just said 21:37