»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
MasterDuke heh. i bet this is the culprit github.com/rakudo/rakudo/blob/mast...#L422-L424 00:02
00:08 comborico1611 left
japhb RT#128655 00:28
synopsebot RT#128655 [open]: rt.perl.org/Ticket/Display.html?id=128655 [OPTIMIZER] Mixup in candidates through optimizer
00:30 MasterDuke left
japhb Seems like that one's do for a re-look. 00:30
00:30 aindilis` joined
japhb With jnthn++ working on optimization, I wonder if that's LHF for him now ... 00:30
00:30 cdg joined 00:34 aindilis left 00:35 cdg left 00:42 charsbar left 00:43 charsbar joined 00:48 charsbar left, charsbar joined 01:10 Rawriful left 01:14 xenotrope left 01:16 xenotrope joined 01:20 aindilis` left 01:24 lookatme joined 01:25 MasterDuke joined 01:26 DrForr left 01:27 DrForr joined 01:31 wamba joined 01:32 aindilis joined 01:33 wamba left 01:53 char_var[buffer] left
Geth doc/edits: 5758cc236a | (Antonio Quinonez)++ | doc/Language/operators.pod6
Adjust some grammar
01:57
02:07 cheese__ joined 02:09 cheese__ left 02:14 travis-ci joined
travis-ci Doc build passed. Antonio Quinonez 'Adjust some grammar' 02:14
travis-ci.org/perl6/doc/builds/346107439 github.com/perl6/doc/commit/5758cc236a92
02:14 travis-ci left 02:22 cdg joined 02:26 cdg left 02:33 pierre joined 02:45 ilbot3 left 02:47 wamba joined 02:56 pochi left 02:58 ilbot3 joined, ChanServ sets mode: +v ilbot3 02:59 pochi joined 03:06 kurahaupo left, kurahaupo joined 03:10 aindilis left 03:16 martinp2 joined 03:22 martinp2 left 03:29 shinobi-cl joined, martinp2 joined 03:32 dct_ joined 03:34 Zoffix joined, Zoffix left 03:40 cdg joined 03:42 webstrand left 03:46 webstrand joined, webstrand left 03:47 webstrand joined 03:48 webstrand left, cdg left, webstrand joined 03:49 cdg joined, webstrand left
shinobi-cl Hi all... 03:50
m: class C { has $!empty = "-"; multi sub trait_mod:<is>(Variable:D, :$default!) { $!empty = $default; } has @.data; method new(@data) { self.bless(data=>@data); } }; my $c is default ('N/A') = C.new([1,Nil,3,4]); say $c.perl; 03:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable $!empty used where no 'self' is available
at <tmp>:1
------> 3d:<is>(Variable:D, :$default!) { $!empty7⏏5 = $default; } has @.data; method new(@d
shinobi-cl m: class C { has $empty = "-"; has @.data; method new(@data) { self.bless(data=>@data); } }; my $c is default ('N/A') = C.new([1,Nil,3,4]); say $c.perl;
camelia 5===SORRY!5=== Error while compiling <tmp>
Signatures as constraints on variables not yet implemented. Sorry.
at <tmp>:1
------> 3a=>@data); } }; my $c is default ('N/A')7⏏5 = C.new([1,Nil,3,4]); say $c.perl;
shinobi-cl i understand that the error says that is not implemented. i just wonder if that would be the right way to call is default when creating an object..... 03:52
03:53 cdg left 03:56 martinp2 left 04:01 webstrand joined 04:02 webstrand left 04:06 webstrand joined 04:08 webstrand left 04:09 webstrand joined, webstrand left
shinobi-cl the idea is to be able to pass a default value for empty elements to my class without using a parameter to the constructor but instead using the trait syntax... does that makes sense? 04:11
04:11 webstrand joined, webstrand left 04:12 webstrand joined 04:14 char_var[buffer] joined 04:15 Zoffix joined
Zoffix shinobi-cl: the error is just because you've used too many spaces 04:16
m: class C { has $empty = "-"; has @.data; method new(@data) { self.bless(data=>@data); } }; my $c is default('N/A') = C.new([1,Nil,3,4]); say $c.perl;
camelia C.new(data => [1, Any, 3, 4])
Zoffix m: class C { has $empty = "-"; has @.data; method new(@data) { self.bless(data=>@data); } }; my $c is default('N/A') = C.new([1,Nil,3,4]); say $c.perl; $c = Nil; say $c.perl
camelia C.new(data => [1, Any, 3, 4])
"N/A"
Zoffix shinobi-cl: you want the `is default` that's applied on $c to function as the default for empty cells? So it comes out as `[1, 'N/A', 3, 4]`? I don't think that'll work, as it's constructed before it ever gets to the variable. 04:17
shinobi-cl: and another mistake you're making is `[1,Nil,3,4]`. That never arrives as a `Nil` to your class, but as an `Any` 04:20
It's using the `is default` of the Array and the default default is Any 04:21
m: dd [1,Nil,3,4]
camelia [1, Any, 3, 4]
Zoffix m: dd (1,Nil,3,4)
camelia (1, Nil, 3, 4)
04:21 xi- left 04:24 xi- joined
Zoffix shinobi-cl: what I said on reddit is for you to return Lists with Nils in them so that the receiver of data would be able to use is default on their arrays 04:24
shinobi-cl basically. im tryin to use "is default" to populate a private variable that will be used to fill 'empty' cells. Those empty cells dont come from the original data, but instead from the construction process inside my class. 04:25
Yes, i changed all the Anys to Nils by your suggestion :) 04:26
crudely, the idea is something like this: m: class C { my $empty = "-"; multi sub trait_mod:<is>(:$default!) { $empty = $default; }; has @.data; method new(@d) { my @r; for @d {push @r, ($_ % 2 == 0)??$_!!$empty}; self.bless(data=>@r)} }; my $c = C.new([1,2 ,3,4]); say $c.perl; my $c2 is default('N/A') = C.new([1,2 ,3,4]); say $c2.perl; 04:28
m: class C { my $empty = "-"; multi sub trait_mod:<is>(:$default!) { $empty = $default; }; has @.data; method new(@d) { my @r; for @d {push @r, ($_ % 2 == 0)??$_!!$empty}; self.bless(data=>@r)} }; my $c = C.new([1,2 ,3,4]); say $c.perl; my $c2 is default('N/A') = C.new([1,2 ,3,4]); say $c2.perl;
camelia Cannot allocate memory
shinobi-cl oops
Zoffix Right, the information from `is default` on $c is not available inside your constructor. It doesn't even know whether you're going to store it into a variable
(without hacks)
shinobi-cl oh, i see. well,, just wanted to see it that was possible :) I added more functinality based on your suggestions, btw. github.com/shinobi/Data-StaticTabl...method-new Thanks again for the feedback 04:30
Zoffix m: class C { has $.empty = "-"; has @.data; method new(@data, \SELF) { SELF = self.bless: :@data, :empty(SELF.VAR.default); } }; C.new: [1,Nil,3,4], my $c is default('N/A'); say $c.perl 04:35
camelia C.new(empty => "N/A", data => [1, Any, 3, 4])
Zoffix that works, but is a really weird interface :)
shinobi-cl so this would work for any trait then :) ? what's so special about "trait_mod" ? 04:41
04:42 AlexDaniel joined
Zoffix trait_mod is just a category of subs that let you add new candidates to traits 04:43
shinobi-cl and then what's the use of them if you can't modify internal variables in your own class? how would i then implement my own traits? 04:44
Zoffix You can, you just never called your traits 04:45
m: class C { my $empty = "-"; multi sub trait_mod:<is>(Variable:D, :$default!) is export { $empty = $default; nextsame }; has @.data; method new(@d) { my @r; for @d {push @r, ($_ % 2 == 0)??$_!!$empty}; self.bless(data=>@r)} }; import C; my $c is default('N/A') = C.new([1,2 ,3,4]); say $c.perl 04:46
camelia C.new(data => ["-", 2, "-", 4])
Zoffix Don't use that though. It'd affect `$empty` for all instances of your class and it'd affect them any time someone uses `is default`, even for purposes other than instantiating your class 04:47
Just take a named param to the constructor. 04:48
shinobi-cl i see. thanks Zoffix++ 04:50
04:54 Cabanossi left 04:55 Cabanossi joined 04:56 jeromelanteri joined 05:02 wamba left, pierre left 05:07 pierre joined 05:12 pierre left 05:13 dct_ left, dct_ joined 05:17 eliasr left 05:18 kurahaupo_ joined 05:19 skids left 05:21 kurahaupo_ left, kurahaupo_ joined, kurahaupo left 05:26 xinming left 05:28 xinming joined 05:33 lookatme left 05:34 lookatme joined 05:38 dct_ left 05:42 dct_ joined 05:43 pierre joined 05:46 khw left 06:03 wamba joined 06:05 troys left 06:09 wamba left 06:12 wamba joined 06:13 reportable6 left, reportable6 joined, ChanServ sets mode: +v reportable6 06:14 ufobat joined 06:16 reportable6 left, reportable6 joined 06:21 xenotrope left 06:22 dct_ left, dct_ joined 06:23 xenotrope joined 06:31 eroux left 06:40 wamba left 06:41 reportable6 left, reportable6 joined, jast left 06:53 wamba joined 06:55 Zoffix left 07:04 wamba left 07:14 robertle left 07:15 dct_ left 07:23 wamba joined 07:28 someuser joined 07:31 espadrine left
Geth whateverable: 1c8633d6e7 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 5 files
Added a sneak peek for quotable

It will now print something like:
  “12 messages (2010-01-27⌁2016-01-21): …”
Also removed hardcoded ‘Found nothing!’ message from perl6-grep (and added tests for both greppable and quotable that were using this sub).
07:43
07:44 greppable6 left, quotable6 left, quotable6 joined, greppable6 joined, ChanServ sets mode: +v greppable6 07:45 KDr2 joined 07:47 wamba left 07:54 wamba joined 07:58 rindolf joined 07:59 domidumont joined 08:07 domidumont left 08:08 domidumont joined
Geth doc/master: 9 commits pushed by (Luca Ferrari)++ 08:15
08:20 shinobi-cl left 08:21 shinobi-cl joined, robertle joined 08:26 zakharyas joined 08:29 zakharyas left 08:30 zakharyas joined 08:31 pierre left 08:33 lookatme left, lookatme joined 08:38 pierre joined
Geth doc: ae94bbea71 | (Luca Ferrari)++ | doc/Language/regexes.pod6
Better explaination of a \n meaning in regular expression.

Partially due to the issue #1790. I guess the original author's intent was to specify that \r\n can be either mangled thru a regex or an IO operation.
08:38
synopsebot Link: doc.perl6.org/language/regexes
08:42 pierre left 08:43 pierre joined 08:45 kurahaupo_ is now known as kurahaupo
timotimo MasterDuke: there's an int,int --> int and an Int:D,Int:D --> Int:D candidate for &infix:<%>, though 08:50
08:55 zakharyas left 08:56 zakharyas joined 09:03 shinobi-cl left 09:06 kurahaupo_ joined, pmurias joined 09:09 kurahaupo left 09:11 kurahaupo_ is now known as kurahaupo 09:23 dakkar joined 09:25 robo6 joined 09:26 robo6 left 09:28 pmurias left 10:17 Blackraider joined 10:21 Blackraider left, SyrupThinker joined 10:24 aindilis joined 10:29 cdg joined 10:31 lookatme left, lookatme joined, natrys joined 10:33 cdg left 10:52 jantore left 10:53 jantore joined 10:58 zakharyas left
Geth doc: 912e133a3e | (Luca Ferrari)++ | doc/Type/Mu.pod6
Split sub defined and method defined.

Fix issue #1800.
10:58
synopsebot Link: doc.perl6.org/type/Mu
10:59 agwind left 11:14 aindilis left 11:18 pierre left 11:19 pierre joined, lalitmee joined 11:23 pierre left 11:25 agwind joined 11:26 zakharyas joined 11:40 zakharyas left 11:42 sena_kun joined 12:01 pierre joined 12:07 pierre left, raschipi joined, pierre joined 12:08 aindilis joined 12:12 pierre left
tyil .tell zoffix Thanks for the headsup, I've pushed github.com/scriptkitties/perl6-app...ixt/pull/2 12:33
yoleaux 25 Feb 2018 07:18Z <Zoffix> tyil: you have several bugs in your Appsixt module. Variables inside «» gets split on spaces within their values so, say, spaces in paths will break the path into separate args (`my $source = 'bar ber'.IO; dd «foo {$source.IO.path}»;`): gist.github.com/Whateverable/576bc...e13e4a7f26
tyil: I'll pass your message to zoffix.
AlexDaniel tyil: this doesn't make much sense tbh: say "tar czf \"$output\" {@tar-flags} ."; 12:36
tyil why?
AlexDaniel because paths can contain double quotes :) I know it's just something to show to the user but… eh…
tyil paths can contain pretty much anything, so I have to pick something
AlexDaniel iirc you only have to “escape” single quotes when using single quotes 12:37
tyil I don't want to waste too much effort on something that's just logging output, since it already makes the intent clear
yes, but what if the path contains a single quote then ;) 12:38
jkramer AlexDaniel: If you escape single quotes you'll also need to escape the escape character ;) 12:39
AlexDaniel jkramer: it's not escaping, it's “escaping” :)
you can't escape *anything* in single quotes
jkramer Although it's trivial and I'm sure there's a good shell-quote module :)
AlexDaniel there's non IIRC 12:40
none*
tyil it's in a say, so it's not like the shell is being called there
it still shows the intent, which is what I intended it to do
jkramer Huh weird, I always you could escape single quotes with '\ in a '' in shells 12:42
*thought
ilmari you need to do '\'' 12:43
AlexDaniel tyil: that's k, but I think you can also call .perl on it
ilmari: or '"'"' if you want everyone to freak out
ilmari AlexDaniel: yeah, that's prettier, but more confusing 12:44
tyil hmm, doing a .perl could work
jkramer tyil: I usually use <...> when dumping values for logs, it's visually pleasing and </> are unlikely to show up in most kinds of values :)
tyil but < ... > doesn't show the value of vars in them, do they? 12:45
12:45 dalek left
tyil m: my $foo = "a"; dd < some $foo > 12:46
camelia ("some", "\$foo")
tyil m: my $foo = "a"; dd << some $foo >>
camelia ("some", "a")
jkramer tyil: No I mean in output, like say "foo=<this>, bar=<that>, ..."
masak m: my $file = "name-with-a-quo'te"; say $file.subst(:g, /"'"/, "'\\''") 12:47
camelia name-with-a-quo'\''te
tyil masak: .perl seems simpler then :p 12:48
masak doesn't target the shell-quoting problem though, does it?
mostly I meant to show that a shell quoting library could very well start as that one-liner there ;) 12:49
tyil ah
masak s/target/address/ -- my English is rusty :P
tyil I mean, it's still just output, not a shell command, so I don't really consider is a shell quoting problem
jkramer masak: Needs more quotes though :)
m: my $file = "name-with-a-quo'te"; say "'" ~ $file.subst(:g, /"'"/, "'\\''") ~ "'" 12:50
camelia 'name-with-a-quo'\''te'
masak aye, that's better :)
&safe-single-quote
masak .oO( ship it )
moritz dives into the wonderful world of Angular development 12:51
jkramer m: my $file = "name-with-a-quo'te"; say $file.subst(:g, /"'"/, "'\\''").subst(:g, /^|$/, "'")
camelia 'name-with-a-quo'\''te'
tyil having done angular for school, I must say I find it one of the least nice js frameworks in existence
jkramer m: use MONKEY-TYPING; augment class Str { method quote { .subst(:g, /"'"/, "'\\''").subst(:g, /^|$/, "'") } }; "quo'te".quote.say 12:56
camelia No such method 'subst' for invocant of type 'Any'
in method quote at <tmp> line 1
in block <unit> at <tmp> line 1
jkramer m: use MONKEY-TYPING; augment class Str { method quote { $.subst(:g, /"'"/, "'\\''").subst(:g, /^|$/, "'") } }; "quo'te".quote.say
camelia 'quo'\''te'
12:57 Lynx_ left
moritz it's also quite the mindset shift. "ng new foo; du -sch foo" comes back with nearly 600MB 12:57
jkramer m: use MONKEY-TYPING; augment class Str { method quote { $.subst(:g, /"'"/, "'\\''").subst(:g, /^|$/, "'") } }; augment class Mu { method quote { $.Str.quote } }; IO::Path.new("foo/'bar").quote.say 12:58
camelia IO::Path is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in method new at src/RESTRICTED.setting line 32
in block <unit> at <tmp> line 1
12:58 p6lert left 12:59 Lynx_ joined
jkramer m: use MONKEY-TYPING; augment class Str { method quote { $.subst(:g, /"'"/, "'\\''").subst(:g, /^|$/, "'") } }; augment class Mu { method quote { $.Str.quote } }; Hash.new(lol => 123, bar => 'baz').quote.say 12:59
camelia No such method 'quote' for invocant of type 'Hash'. Did you mean any of these?
note
quote

in block <unit> at <tmp> line 1
13:00 pierre_ joined
jkramer Are augmented methods not inherited? 13:01
lizmat jkramer: they are, sort of 13:05
you would need to actually also augment any child classes
m: use MONKEY; augment class Cool { method foo { "foo" } }; say 42.foo 13:06
camelia No such method 'foo' for invocant of type 'Int'. Did you mean 'foo'?
in block <unit> at <tmp> line 1
lizmat m: use MONKEY; augment class Cool { method foo { "foo" } }; augment class Int {}; say 42.foo 13:07
camelia foo
Geth doc: 15034eedf4 | (Will "Coke" Coleda)++ | doc/Type/IO/CatHandle.pod6
fix typo
synopsebot Link: doc.perl6.org/type/IO::CatHandle
13:07 Lynx_ left
lizmat jkramer: this is basically a result of child classes knowing about their parents, but parent classes not knowing about their children 13:07
jkramer Ah 13:09
13:09 Lynx_ joined
jkramer Hmm, so I guess there's also not Mu.give-me-all-child-classes-so-i-can-augment-them-all 13:09
13:10 zakharyas joined
[Coke] can probably get a list of all classess in the current scope and work up. 13:10
13:10 zakharyas1 joined
jkramer Ah nevermind, I just wanted to play with the quote thing a bit :) 13:11
13:14 zakharyas left
Geth doc/master: 4 commits pushed by (Luca Ferrari)++ 13:20
13:21 tt_ joined
masak moritz: I see what they are aiming for with Angular. I think it's not a bad design per se, but I've yet to work on a big-enough project where investing in all that DI and infrastructure pays off. 13:29
meanwhile, React has kind of convinced me that a few well-chosen primitives and some FP-think can get you very far. 13:30
if I were to go for an opinionated framework, I'd probably try out Ember (which isn't a majority choice, but seems very well-thought-out)
(end of unsolicited opinion)
moritz: what are you building? :) 13:31
moritz masak: so far I'm just trying to learn it :) Then I'll be building a web frontend for our CMDB 13:33
which already exists as a hybrid between angularjs 1.* and angular 4
masak moritz: I've been doing some Angular consulting lately. but I find that to a large extent it consists of teaching JavaScript and TypeScript. ;) 13:35
13:41 aindilis left
moritz masak: I hope I'll be learning that along the way too :-) 13:41
masak TypeScript is perhaps the nicest aspect of the new Angular ;) 13:43
13:46 wamba left 13:50 notable6 left, notable6 joined 13:51 wamba joined 13:53 xinming left, mcmillhj joined 13:54 cdg joined 13:56 cdg left, cdg joined 13:58 lalitmee left 14:08 scimon joined 14:15 someuser left 14:20 vcv joined 14:21 athenot joined 14:22 wamba left 14:25 pierre_ left 14:30 pierre_ joined 14:38 raschipi left 14:41 natrys left 14:43 cdg left 14:45 pierre_ left, zakharyas1 left 14:46 jeromelanteri left 14:47 zakharyas joined 14:50 comborico1611 joined
Geth doc: antquinonez++ created pull request #1801:
Adjust grammar.
14:54
doc: 24a20d44f2 | (Antonio Quinonez)++ (committed by Moritz Lenz) | doc/Language/operators.pod6
Adjust grammar.
14:55
synopsebot Link: doc.perl6.org/language/operators
AlexDaniel I think github.com/perl6/doc/pull/1801 needs a review from a native English sprecher :) 14:59
ah, it was merged already
okay
14:59 skids joined
moritz I checked a few samples, and found the changes to be improvements 15:00
15:05 wamba joined 15:10 natrys joined 15:11 wamba left 15:13 khw joined 15:15 pierre_ joined 15:19 pierre_ left 15:35 wamba joined 15:36 troys joined 15:43 MilkmanDan left 15:44 MilkmanDan joined 15:46 zakharyas left 15:47 zakharyas joined 15:48 wamba left 15:54 konsolebox joined 15:56 cdg joined, wamba joined 15:57 cdg left 15:58 facetious left, domidumont left 15:59 pierre_ joined, cdg joined 16:00 cdg_ joined 16:01 wamba left, khisanth_ left 16:03 pierre_ left 16:04 cdg left 16:06 _fran_ joined 16:08 zakharyas left 16:14 khisanth_ joined
_fran_ I built rakudo-star-2018.01 with gcc on debian on termux on android 6.0.1. Invocation of perl6 or perl6-m in interactive repl-mode gives "Bus Error", but only after modules are built. Rakudo binary from debian is working. I suspect Linenoise to be responsible. Is it autoloaded and how to prevent that? 16:18
jkramer Bus error sounds familiar :) 16:25
github.com/rakudo/rakudo/issues/1554
16:32 mcmillhj left 16:33 mcmillhj joined 16:34 pierre_ joined 16:37 SyrupThinker left 16:39 pierre_ left
stmuk _fran_: bet.github.io/perl6-android-gnuroot.txt may help 16:41
16:42 kaare_ joined
stmuk note I havn't yet tried it yet but I met someone in a restaurant who said his friend built ok on android :) 16:42
16:48 aindilis joined 16:49 vike left 16:51 vike joined
Geth doc: d311d565ab | (Luca Ferrari)++ | doc/Type/Mu.pod6
Change 'defined' to 'declared' when documenting method signatures.

Applies to Mu.defined.
16:55
synopsebot Link: doc.perl6.org/type/Mu
Geth doc: 97e2a9fe92 | (Luca Ferrari)++ | doc/Type/Mu.pod6
Move routine 'defined' after documentation of method Mu.defined.

Could be a fix to issue #1800 ?
doc: 1bf8646a4d | (Luca Ferrari)++ | doc/Type/Mu.pod6
Reword 'Defined as' to 'Declared as' in Mu.Capture.
16:57
16:58 mcmillhj left 17:00 zakharyas joined 17:02 mcmillhj joined, cdg joined 17:05 cdg left 17:06 cdg joined, cdg_ left 17:07 mcmillhj left, eliasr joined 17:08 kurahaupo left 17:11 wamba joined 17:20 cdg_ joined 17:21 mcmillhj joined 17:22 raschipi joined 17:24 cdg left 17:30 dakkar left 17:32 scimon left 17:34 comborico1611 left 17:36 raschipi left 17:38 raschipi joined 17:39 sftp left 17:42 kybr left 17:47 sftp joined 17:49 synopsebot left 17:53 sergot left, SourceBaby left, psch left, Geth left 17:54 Util left, dalek joined, ChanServ sets mode: +v dalek, synopsebot joined, Geth joined, ChanServ sets mode: +v synopsebot, ChanServ sets mode: +v Geth, SourceBaby joined, ChanServ sets mode: +v SourceBaby 17:56 SyrupThinker joined 17:58 cdg_ left, pm5 left 18:01 ab5tract_ left 18:03 domidumont joined 18:05 robertle left, cdg joined 18:20 cdg left 18:26 _fran_ left 18:32 Manifest0 left 18:37 pierre_ joined 18:38 AlexDaniel left, Manifest0 joined 18:39 domidumont left 18:41 xinming joined, pierre_ left 18:43 xenotrope left 18:44 xenotrope joined 18:57 zakharyas left 19:04 Util joined 19:14 comborico1611 joined 19:33 cdg joined, robertle joined 19:37 cdg left 19:38 domidumont joined 19:42 Kaypie left 19:45 Kaiepi joined 19:49 espadrine joined 19:53 sena_kun left 19:56 sena_kun joined 19:57 domidumont left 20:05 sergot joined 20:11 mcmillhj left, mcmillhj joined 20:12 natrys left 20:21 raschipi left 20:26 kurahaupo joined 20:30 rindolf left 20:34 pierre_ joined 20:39 pierre_ left
buggable New CPAN upload: App-Assixt-0.1.4.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...1.4.tar.gz 20:44
20:44 |oLa| joined, novichok joined 20:46 konsolebox left 20:53 konsolebox joined 20:54 freeberz joined 20:55 cdg joined 20:59 cdg left
hahainternet in a code golf stack overflow post from last week, there's a snippet of code: `.rotate(set(0,+?.[0],-?.[4]).pick)` what's going on with `+?.[0]` 21:01
lizmat m: say +"42" 21:02
camelia 42
lizmat m: say +?"42"
camelia 1
lizmat I think it's just that: making sure you get a 0 or 1 ?
hahainternet hmm, i'd like to find it in the docs if you know where lizmat 21:03
ah, symbolic unary precedence section 21:04
cheers :)
lizmat hahainternet++ 21:05
21:06 geospeck joined, wamba left, geospeck left 21:08 SyrupThinker left 21:10 |oLa| left 21:11 SyrupThinker joined 21:14 Rawriful joined 21:19 cdg joined 21:20 SyrupThinker left 21:21 Manifest0 left 21:26 SyrupThinker joined 21:36 Kaiepi left 21:37 Kaypie joined 21:46 Kaypie left 21:48 notable6 left, notable6 joined, ChanServ sets mode: +v notable6, Kaiepi joined
lizmat And another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2018/02/26/...-cheese-d/ 21:50
22:00 someuser joined, someuser left 22:03 SyrupThinker left
MasterDuke timotimo: not sure what you mean. i made all variables ints and the int,int candidate wasn't being called 22:06
22:07 jast joined, mcmillhj left
MasterDuke and the reason i believe is that +< and +&, which were used in all the expressions that also had a %, don't currently have a int,int candidates 22:07
22:08 vcv left
MasterDuke however, what's even odder, is that i uncommented the int,int candidates for +< and +& and the profile is still showing the Int,Int versions of everything 22:10
22:11 cdg left, cdg joined 22:12 skids left 22:30 comborico1611 left 22:34 pierre_ joined 22:35 thundergnat joined 22:36 AlexDaniel joined
thundergnat \o perl6 22:36
I have a perl 6 script for a Rosettacode task that reliably works under 2017.11 and earlier but reliably fails under 2017.12 and later. 22:38
I've spent the last few days feebly trying to figure out what has changed to cause the failure but am not having much luck.
22:39 pierre_ left, sena_kun left
thundergnat The weird thing is it doesn't fail right away, it succeeds for about 200 iterations and _then_ fails. 22:39
(It is downloading revisions for the various rosettacode tasks to find the first revision.) 22:40
Anyway if anyone wants to take a look and see if they may have any suggestions, I would be grateful. 22:41
Gist of the failing script: (cut down a bit) gist.github.com/thundergnat/d70596...ffe0b24639
I'm not online very much but I read the logs pretty regularly. 22:42
El_Che thundergnat: there is a bot that to .tell $nick that can remember the answer until you come back 22:43
thundergnat El_Che Yeah, I know... but thanks.
El_Che thundergnat: I am pretty sure someone will know more, as there is only a difference of 2 releases 22:45
thundergnat: alternatively, you could raise an issue as it may be a regression
thundergnat The script doesn't seem to fail due to the version of HTTP::UserAgent or JSON::Fast loaded. (I tried rolling each back separately) Only the compiler version seems to affect it.
El_Che github.com/rakudo/rakudo/issues 22:46
thundergnat El_Che Yeah, it seems to be a regression, but there is a distinct possibility I am doing something stupid too.
El_Che it would be strange that the stupid part wasn't that stupid 2 releases ago :) 22:48
AlexDaniel hello 22:51
thundergnat \o
AlexDaniel let's bisect it maybe
it depends on HTTP::UserAgent which is not cool, but still can be done :) 22:52
let's see…
thundergnat I tried various versions of HTTP::UserAgent and JSON::Fast and they don't seem to affect it. Only the compiler makes a difference. 22:53
22:54 releasable6 left, releasable6 joined, ChanServ sets mode: +v releasable6
thundergnat Though I should have mentioned... the failure shows up as a fault in HTTP::UserAgent - Too few positionals passed; expected 2 arguments but got 1 in method get-proxy (HTTP::UserAgent) line 384 22:57
Thats with H::U ver 1.1.38. It gets the same fault (though with different line numbers ) with different versions of HTTP:UserAgent. 22:58
Changing the compiler to 2017.11, it works with both UA 1.1.38 and 1.1.37. (That's all I tested.) 22:59
23:04 _fran_ joined
timotimo thundergnat: if it only breaks after a certain number of iterations, try setting the MVM_SPESH_DISABLE in the environment and see if it keeps working 23:08
thundergnat set MVM_SPESH_DIABLE=TRUE or some such? 23:09
^S
AlexDaniel c: 2017.11 gist.githubusercontent.com/AlexDan...authors.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦2017.11: «TERM environment variable not set.␤The spawned command 'clear' exited unsuccessfully (exit code: 1)␤ in block <unit> at /tmp/hfOZ70j3LF line 32␤␤ «exit code = 1»»
AlexDaniel errr?
thundergnat You may need to remove / comment out line 20 23:10
AlexDaniel c: HEAD gist.githubusercontent.com/AlexDan...authors.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦HEAD(838782b): «TERM environment variable not set.␤The spawned command 'clear' exited unsuccessfully (exit code: 1)␤ in block <unit> at /tmp/TAi1LurDZN line 32␤␤ «exit code = 1»»
timotimo thundergnat: can be anything, except the empty string
i usually set it to "yes"
AlexDaniel c: 2017.11 gist.githubusercontent.com/AlexDan...authors.p6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦2017.11: «Retreiving task information...␤1 Task: File input/output␤«timed out after 10 seconds» «exit signal = SIGHUP (1)»» 23:11
AlexDaniel gah!
thundergnat Unfortunately, it takes about 30 seconds to fault
_fran_ RAKUDO_LINE_EDITOR=none perl6 will start without Linenoise or Readline. Without that I get "Bus Error". 23:12
23:13 _fran_ left
MasterDuke timotimo: catch my comments earlier? 23:14
timotimo yeah. don't know what's up with that :( 23:15
23:16 novichok left, freeberz left
MasterDuke i even added prints in the native versions, they're definitely not getting hit 23:16
which seems crazy
AlexDaniel thundergnat: sorry, what's the output when it fails? 23:17
thundergnat AlexDaniel: Setting MVM_SPESH_DISABLE true doesn't seem to change anything.
AlexDaniel gist.github.com/Whateverable/f0acc...d524806477 this? 23:18
thundergnat AlexDaniel: yes
AlexDaniel: It varies failing on task 202 to 209, so it isn't a specific request that does it. 23:20
timotimo that is a big relief 23:21
23:23 athenot left, Actualeyes left 23:25 comborico1611 joined 23:28 robertle left 23:29 g- left 23:48 cdg left 23:49 noganex left 23:51 committable6 left, committable6 joined, ChanServ sets mode: +v committable6
AlexDaniel Geth: ver github.com/rakudo/rakudo/commit/f78fd7c17 23:53
Geth AlexDaniel, version bump brought in these changes: github.com/perl6/nqp/compare/2017....7-gf1b2692
AlexDaniel Geth: ver github.com/perl6/nqp/commit/f1b269...11012303cb
Geth AlexDaniel, version bump brought in these changes: github.com/MoarVM/MoarVM/compare/2...7-gfac2999
23:54 comborico1611 left
AlexDaniel thundergnat: what about MVM_SPESH_INLINE_DISABLE=1 23:56
thundergnat AlexDaniel: Let me give it a try.
AlexDaniel oh, you already tried with MVM_SPESH_DIABLE … then it should be same
thundergnat: see, I bisected it: gist.github.com/Whateverable/5fb5b...05e797f959 23:57
that took a bit of human labour :)
thundergnat: so it's bad on f78fd7c17 but good on its parent 23:58
thundergnat: can you file a ticket please?
thundergnat MVM_SPESH_INLINE_DISABLE=1 results in the same failure mode. (slower, but same failure)
AlexDaniel weird 23:59
but okay
thundergnat AlexDaniel: Sure. Thanks for looking into this for me.