»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.
00:08 astj joined 00:19 Cabanossi left 00:20 Cabanossi joined, ovechkin left 00:24 astj left
timotimo i'll go to bed now; hopefully going to bed a bit earlier will result in better sleep and better getting-out-of-bed 00:33
00:49 Cabanossi left 00:50 Actualeyes left, Cabanossi joined 00:56 protium joined 00:57 geekosaur left 00:58 geekosaur joined 01:04 noganex left, astj joined 01:07 astj left 01:17 Redrield left 01:18 Cabanossi left 01:20 Cabanossi joined 01:23 mcmillhj joined 01:25 astj joined 01:28 mcmillhj left 01:34 gdonald left, gdonald joined 01:38 mxco86 left 01:52 Actualeyes joined 02:02 Cabanossi left 02:05 Cabanossi joined 02:10 noganex joined 02:11 Cabanossi left 02:16 awwaiid left 02:17 labster left 02:18 Cabanossi joined 02:23 mcmillhj joined 02:28 mcmillhj left 02:32 labster joined
skids m: my %h is default(slip()); [%h{2,}].perl.say 02:39
camelia []
skids m: my %h is default(slip()); [%h{2}].perl.say
camelia [Empty,]
02:39 kurahaupo left, mcmillhj joined
skids m: [slip()].perl.say 02:40
camelia []
skids hmm... was there a reason for nonslice indexing to be protective? 02:43
02:44 mcmillhj left
gfldex m: my %h is default(|()); [%h{2}].perl.say 02:45
camelia [slip(),]
skids m: my %h is default(slip()); (%h{2},).perl.say
camelia ()
skids maybe some interaction with array element itemization... 02:46
02:46 Cabanossi left 02:53 gdonald left, gdonald joined 02:54 Cabanossi joined 02:59 perigrin_ joined, perigrin_ left 03:00 perigrin_ joined
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $p {state %a; %a{$p.key} += $p.value}; 03:01
camelia Too many positionals passed; expected 1 argument but got 2
in block <unit> at <tmp> line 1
03:01 wamba joined 03:02 perigrin_ left, perigrin_ joined
skids reduce func needs parms for the new value and the accumulator. 03:04
03:04 perigrin_ left
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $p, $ {state %a; %a{$p.key} += $p.value}; 03:04
camelia No such method 'key' for invocant of type 'Int'
in block <unit> at <tmp> line 1
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> %reduced, Pair $p {%reduced{$p.key} += $p.value; %reduced}
camelia Cannot assign to an immutable value
in block <unit> at <tmp> line 1
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: <-> %reduced, Pair $p {%reduced{$p.key} += $p.value; %reduced} 03:05
camelia Cannot assign to an immutable value
in block <unit> at <tmp> line 1
SmokeMachine skids: is the accumulator the first or the last one?
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> %reduced is rw, Pair $p {%reduced{$p.key} += $p.value; %reduced}
camelia 5===SORRY!5=== Error while compiling <tmp>
For parameter '%reduced', '%' sigil containers don't need 'is rw' to be writable
Can only use 'is rw' on a scalar ('$' sigil) parameter, not '%reduced'
at <tmp>:1
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $b, $a { say "$a $b"; $a + $b }; 03:06
camelia a 2 a 1
Cannot resolve caller Numeric(Pair: ); none of these signatures match:
(Mu:U \v: *%_)
in block <unit> at <tmp> line 1
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $b, $a { say "$a $b"; $a.key + $b };
camelia a 2 a 1
Cannot resolve caller Numeric(Pair: ); none of these signatures match:
(Mu:U \v: *%_)
in block <unit> at <tmp> line 1
skids m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $b, $a { say "$a $b"; $a + $b.value }; 03:07
camelia a 2 a 1
Cannot resolve caller Numeric(Pair: ); none of these signatures match:
(Mu:U \v: *%_)
in block <unit> at <tmp> line 1
skids hrm.
m: my @a = 1,2,3; say @a.reduce: -> $a, $b { say "$a $b"; $a + $b }; 03:08
camelia 1 2
3 3
6
03:08 perigrin_ joined
skids m: my @a = 1,2,3,4; say @a.reduce: -> $a, $b { say "$a $b"; $a + $b }; 03:09
camelia 1 2
3 3
6 4
10
skids The first then.
03:09 perigrin left
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say @a.reduce: -> $d, Pair $p {my BagHash $data; if $d ~~ Pair {$data{$d.key} = $d.value} else {$data = $d}; $data{$p.key} += $p.value; $data} 03:10
camelia BagHash.new(a(3), c(3))
SmokeMachine there is no way to set the initial value of the accumulator on a reduce? 03:11
like: @a.reduce: :initial(BagHash), -> BagHash $acc, Pair $p {$acc{$p.key} += $p.value} 03:13
skids Add it at the front of the list? 03:14
SmokeMachine m: (BagHash, |@a).reduce: -> BagHash $acc, Pair $p {$acc{$p.key} += $p.value}
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '@a' is not declared
at <tmp>:1
------> 3(BagHash, |7⏏5@a).reduce: -> BagHash $acc, Pair $p {$a		 03:15
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say (BagHash, |@a).reduce: -> BagHash $acc, Pair $p {$acc{$p.key} += $p.value}
camelia Cannot modify an immutable BagHash
in block <unit> at <tmp> line 1
skids .new.
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say (BagHash, |@a).reduce: -> BagHash $acc is rw, Pair $p {$acc{$p.key} += $p.value}
camelia Parameter '$acc' expected a writable container, but got BagHash value
in block <unit> at <tmp> line 1
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say (BagHash.new, |@a).reduce: -> BagHash $acc is rw, Pair $p {$acc{$p.key} += $p.value} 03:16
camelia Parameter '$acc' expected a writable container, but got BagHash value
in block <unit> at <tmp> line 1
skids m: my @a = a => 1, a => 2, c => 3; say ($ = BagHash.new, |@a).reduce: -> BagHash $acc is rw, Pair $p {$acc{$p.key} += $p.value}
camelia Type check failed in binding to parameter '$acc'; expected BagHash but got Int (1)
in block <unit> at <tmp> line 1
skids m: my @a = a => 1, a => 2, c => 3; say ($ = BagHash.new, |@a).reduce: -> BagHash $acc is rw, Pair $p {$acc{$p.key} += $p.value; $acc;} 03:17
camelia BagHash.new(a(3), c(3))
03:17 mcmillhj joined 03:18 Cabanossi left 03:20 Cabanossi joined
SmokeMachine m: my @a = a => 1, a => 2, c => 3; say (Bag.new, |@a).reduce: -> Bag $acc, Pair $p {$acc ⊎ $p.key xx $p.value} 03:21
camelia bag(a(3), c(3))
03:22 mcmillhj left
skids xx probably not the most efficient way to go about that. 03:22
SmokeMachine skids: worse than $acc ⊎ $p.key for ^$p.value ? 03:23
skids m: my @a = a => 1, a => 2, c => 3; say (Bag.new, |@a).reduce: -> Bag $acc, Pair $p {$acc ⊎ $p} 03:28
camelia bag(a(3), c(3))
SmokeMachine :) 03:29
03:32 Actualeyes left 03:38 mcmillhj joined 03:42 mcmillhj left 03:46 Cabanossi left 03:49 Cabanossi joined 04:01 AlexDaniel joined 04:06 mcmillhj joined, AlexDaniel left 04:10 mcmillhj left 04:12 wamba left 04:21 skids left 04:23 mcmillhj joined
ugexe timotimo: zef reporter branch + `zef install Net::HTTP` + ZEF_REPORTER_P6C=1 enables test reporting and the reports show up on testers.p6c.org now 04:27
04:27 mcmillhj left
ugexe I havent merged it into master yet because I have to multiplex output from TAP.pm6 somehow. For reporting to work it doesn't show any test output until the tests are all finished. 04:28
testers.p6c.org/reports/145800.html
fwiw I think reporting quit working for panda (which uses a raw socket) around 2-28, so maybe some newline change? 04:30
04:30 BenGoldberg left, vike left
ugexe if you can figure it out let me know... i'd like to remove the Net::HTTP dependency 04:31
04:35 vike joined 04:45 Actualeyes joined 04:47 Cabanossi left 04:49 Cabanossi joined 04:57 itaipu left 04:58 perigrin_ left 05:00 AlexDaniel joined 05:24 wamba joined 05:31 perigrin joined 05:46 Cabanossi left 05:48 wamba left 05:49 Cabanossi joined 05:52 RabidGravy joined, lowbro joined, mr-foobar joined, lowbro left, lowbro joined 05:55 lizmat_ is now known as lizmat
samcv well my install of every module. it hung somewhere. not sure why 06:01
06:01 xtreak joined
samcv at testing NCurses module 06:01
moritz samcv: putting stuff on hack in ~/public_html/ should work 06:03
samcv and then it's accecible where?
moritz hack.p6c.org/~samcv/ 06:04
samcv yay
moritz I can get an SSL cert for that domain if desired
samcv hack.p6c.org/~samcv/install_log.txt gonna have to figure out how to fix hanges 06:05
i'm using `run` atm to run zef commands 06:06
moritz I believe you have to use Proc::Async to work with timeouts
samcv best way is to use proc::async? yeah
moritz my book has an example :-)
samcv i've done it before. i wish it were less work though
moritz too 06:07
time for a high-level module that abstracts this out, I think 06:08
samcv hmm how long to make the timeout 06:13
5 minutes maybe?
idk. what's the max time a module and every single dependency could take?
it should be a leave it alone thing. so i don't want it timing out too early
06:28 lizmat_ joined 06:31 lizmat left 06:34 lizmat_ left 06:35 movl left 06:40 rindolf joined 06:47 wamba joined 06:53 ChoHag joined
moritz 15min maybe 06:58
samcv moritz, ok going for try 2 now. set a timeout at 10 minutes
we'll see how it goes
moritz, why does my utf-8 file show up not in the right encoding? 07:00
moritz Content-Type: text/plain
samcv not sure what the server settings are hack.p6c.org/~samcv/install_log_live.txt this. in firefox and chrome show wrong. though i can manually select Unicode and shows fine
moritz let me try something... 07:01
samcv thanks :)
moritz Content-Type: text/plain; charset=utf-8 07:02
better.
should work now.
samcv \o/
very yes. much good
moritz AddDefaultCharset utf-8 07:03
was the secret apache sauce
samcv :) thought so 07:05
up to at least 19 failing so far 07:06
also think i found a zef problem with dependencies. though could be the packages to blame 07:09
HTTP::Tiny fails because it can't find HTTP::Tinyish during testing. though it tests HTTP::Tinyish before it 07:10
07:10 wamba left
samcv hmm gist.github.com/samcv/1030286eb95f...xt-L38-L56 07:11
moritz are those two in separate distributions?
samcv not sure 07:12
moritz looking at modules.perl6.org: yes 07:13
samcv that's what i had guessed
i can install HTTP::Tiny with zef myself. but it could be the long dependency chain messed it up somewhere 07:14
07:14 wamba joined 07:15 abraxxa joined 07:16 Cabanossi left 07:19 Cabanossi joined 07:27 darutoko joined
Geth ecosystem: IanTayler++ created pull request #319:
Added MinG.		 07:28
07:30 xtreak left 07:31 xtreak joined 07:35 xtreak left
moritz what's our policy for ecosystem pull requests with a broken source URL in META6.json? 07:40
wait for a fix? or merge straight away?
eater moritz: doesn't matter 07:43
Geth ecosystem: b8e1972ab9 | (Ian Tayler)++ (committed by Moritz Lenz) | META.list
Added MinG.

MinG is a module for working with Minimalist Grammars: a mildly-context-sensitive formalism first described by Edward Stabler in 1996. MinG::S13 is an implementation of Stabler's 2012/2013 parser for Minimalist Grammars.
eater its fixed now :d
moritz aye 07:45
07:45 TEttinger left
moritz just invited a few of the last ecosystem contributors to the perl6 org 07:45
07:46 azawawi joined
azawawi Good morning #perl6 07:46
Bug of the day github.com/azawawi/perl6-ncurses/issues/11 :)
cglobal does not work as expected when use-d from another library that uses that cglobal (e.g. NCurses) 07:47
Commenting out the cglobal section fixes the problem github.com/azawawi/perl6-ncurses/b...es.pm6#L30 07:48
should cglobal's generally be called later? 07:49
like github.com/azawawi/perl6-ncurses/b...es.pm6#L40 that is
07:54 zakharyas joined 08:01 jonas1 joined 08:03 Cabanossi left 08:04 Cabanossi joined 08:06 xtreak joined, AlexDaniel left 08:20 CIAvash joined, bjz joined 08:21 dakkar joined, g5 joined 08:31 Actualeyes left 08:38 xtreak left 08:42 nadim joined 08:47 kent\n left 08:48 salva joined, kent\n joined 08:52 azawawi left
nadim morning P6! \o/ 08:52
who is the king of pod? 08:53
I am thinking about porting my module search.cpan.org/~nkh/POD-Tested-0.0.../Tested.pm and wonder if that is not already done in some way. I read Synopsis but found nothig, want to check. 08:54
08:56 xtreak joined
gfldex nadim: not as a module, we got a test in the docs tho 08:56
nadim do you mean inlining the test in the pod? 08:58
08:58 CIAvash left, xtreak left
gfldex we can extract examples from the docs and compile them, what helps a bit to detect bitrot 08:59
08:59 astj_ joined, astj left, xtreak joined
nadim What tool does that? can Pod be injected as result of code defined in the pod? 09:00
My module is very much about avoiding documentation rot but also as a tool to write cookbooks and design a module with test first 09:01
09:01 Cabanossi left 09:03 Cabanossi joined 09:04 xtreak left
gfldex nadim: in that case you might have to patch rakudo itself because ambient nodes are not implemented yet 09:06
09:12 dogbert17_ joined
dogbert17_ o/ 09:12
is the document 'Learn Perl 6 in Y minutes' available in our repos? It seems to be hosted on a different server. 09:15
it contains some errors which should be fixed
nadim gfldex: do you feel the functionality should be in rakudo or in a module. I am not sure a tool to write cookbooks that are tested are best in rakudo but you know better I guess. 09:20
gfldex nadim: right now the pod parser (that is part of Rakudo) will skip any non-pod parts. If you need them to do facy stuff with the code and pod nodes, you will either have to implement your own parser (very welcome, we don't got that yet) or fix the planned but never implemented Pod::Block::Ambient in Rakudo. 09:26
09:28 gdonald left 09:29 gdonald joined 09:33 xtreak joined 09:34 kurahaupo joined
timotimo github.com/adambard/learnxinyminutes-docs - dogbert17_ 09:34
moritz dogbert17_: github.com/adambard/learnxinyminutes-docs/
github.com/adambard/learnxinyminut...l.markdown 09:35
09:41 mxco86 joined
timotimo m: use nqp; <1 2 3>.map: { when * > 5 { "test {nqp::unbox_i($_ // 0)}" } } 09:42
camelia ( no output )
timotimo m: sub test { use nqp; <1 2 3>.map: { when * > 5 { "test {nqp::unbox_i($_ // 0)}" } } }
camelia ( no output )
timotimo ... ?!
09:44 robertle joined 09:45 bjz left 09:46 bjz_ joined
gfldex lolibloggedtoday: gfldex.wordpress.com/2017/04/19/de...h-fallout/ 09:52
timotimo i'm mildly surprised that try has looser precedence than // does 09:57
09:58 jonas1 left 10:00 labster left 10:01 xtreak left
u-ou m: role A { has $.x }; role B does A { }; role C does A { }; role D does B does C { }; 10:06
camelia ( no output )
gfldex .seen retupmoca 10:07
yoleaux I saw retupmoca 8 Jun 2016 21:09Z in #perl6: <retupmoca> m: say ('a'..'d').rotor(2 => -1).perl;
u-ou m: role A { has $.x }; role B does A { }; role C does A { }; class D does B does C { }; 10:08
camelia 5===SORRY!5=== Error while compiling <tmp>
Attribute '$!x' conflicts in role composition
at <tmp>:1
u-ou m: role A { has $!x }; role B does A { }; role C does A { }; class D does B does C { }; 10:10
camelia 5===SORRY!5=== Error while compiling <tmp>
Attribute '$!x' conflicts in role composition
at <tmp>:1
10:15 lowbro left, mr-foobar left 10:16 mr-foobar joined 10:17 Cabanossi left
timotimo gfldex: i think the code in your blog post is actually slightly wrong 10:18
m: my $result = try "oh lord".&die // "default"; say $result
camelia (Any)
10:18 Cabanossi joined
timotimo well, that's not failure 10:19
m: my $result = try "oh lord".&fail // "default"; say $result
camelia default
timotimo oh, interesting
m: my $result = try "oh lord".&fail.butthismethdoesntexist // "default"; say $result 10:20
camelia (Any)
gfldex timotimo: already fixed
timotimo OK! 10:21
gfldex I noticed because I'm writing a 2nd blog post right nao! 10:22
timotimo that's good bloggage
10:28 kurahaupo left, mr-foobar left 10:29 mr-foobar joined
gfldex m: sub picky(\c:($,$)) {say &c.arity}; picky(* - *); 10:30
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
c used at line 1
gfldex m: sub picky(\c:($,$)) {say c.arity}; picky(* - *); 10:31
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
c used at line 1
timotimo wouldn't you want |c there?
oh
gfldex m: sub picky(&c:($,$)) {say &c.arity}; picky
camelia Too few positionals passed; expected 1 argument but got 0
in sub picky at <tmp> line 1
in block <unit> at <tmp> line 1
timotimo sorry, that's different 10:32
gfldex m: sub picky(&c:($,$)) {say &c.arity}; picky({$^a, $^b})
camelia 2
gfldex m: sub picky(&c:($,$)) {say &c.arity}; picky(* - *)
camelia Constraint type check failed in binding to parameter '&c'; expected anonymous constraint to be met but got WhateverCode ({ ... })
in sub picky at <tmp> line 1
in block <unit> at <tmp> line 1
gfldex is there a way to say "I want a WhateverCode but it must have 2 params!"?
timotimo well, you can have a where clause for that, but it should have worked with what you had there, i think 10:33
gfldex m: my &c = * - *; &c.arity.say
camelia 2
timotimo yeah
gfldex so the compiler knows
timotimo maybe it's expecting Any, but gets Mu?
moritz m: sub picky(WhateverCode \c where .arity == 2) { c(1, 2) }; say *-*
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '&c' is not declared
at <tmp>:1
------> 3ky(WhateverCode \c where .arity == 2) { 7⏏5c(1, 2) }; say *-*
moritz m: sub picky(WhateverCode \c where .arity == 2) { c.(1, 2) }; say *-* 10:34
camelia { ... }
timotimo m: my &c = * - *; &c.signature.perl.say
camelia :(;; $whatevercode_arg_1 is raw, $whatevercode_arg_2 is raw)
gfldex moritz: thx
moritz though in most cases I'd restrict to Callable, not to WhateverCode
so picky(&c where .arity == 2) { } 10:35
10:37 wamba left 10:38 Luna521 joined 10:39 mr-foobar left 10:41 mr-foobar joined 10:42 wamba joined
nadim Playing with the profiler, It seems that quite some time is spend as sink at SETTING::src/core/Seq.pm:183, is that usual? 10:46
10:46 lowbro joined, lowbro left, lowbro joined, wamba left
jnthn nadim: Inclusive time, or exclusive time? 10:47
But yes, it's very normal if your program has, say, a for loop
gfldex m: sub picky(WhateverCode $c) { $c.(1) }; say (try picky( return * )) // $!.^name
camelia X::ControlFlow::Return
nadim it Data::Dump::Tree, lots of loops, but and that is a gut feeling, it seems that most of the time is spend there 10:48
jnthn Since for @a { body } compiles into @a.map({ body }).sink
nadim ah!
jnthn The .sink being what actually makes the work happen since map is lazy
10:48 Luna521 left
nadim ok, good to know 10:48
10:53 itaipu joined 10:54 domidumont joined 10:59 domidumont left 11:00 domidumont joined
gfldex lolibloggedagain! gfldex.wordpress.com/2017/04/19/yo...-you-like/ 11:01
11:04 wamba joined 11:17 g5 left, Cabanossi left, mr-foobar left 11:19 Cabanossi joined, mr-foobar joined
grondilu m: say grammar { rule TOP { <ident> | <TOP>+ % \+ } }.parse: 'foo+bar' 11:25
camelia Nil
grondilu was hoping that would work
jnthn No, the LTM is terminated by recursion 11:26
So the <ident> ends up winning
grondilu oh yeah
11:29 AlexDaniel joined, wamba left 11:30 iH2O joined
grondilu wait but then why doesn't it return "foo"? 11:30
11:32 iH2O left
jnthn Because parse anchors at the end of the string 11:35
Presumably .subparse would give back foo
11:38 xtreak joined 11:41 mr-foobar left 11:43 mr-foobar joined 11:46 ChoHag left, Cabanossi left 11:49 Cabanossi joined 12:00 wamba joined 12:06 mr-foobar left, itaipu left 12:07 mr-foobar joined 12:08 ChoHag joined
Geth doc: 555bb55d87 | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Type/Any.pod6
Document Any.unique
 12:12
12:13 itaipu joined
dogbert17_ moritz: thx for the link, sorry for the late reply 12:15
moritz dogbert17_: no problem 12:16
12:17 Cabanossi left 12:19 Cabanossi joined, mr-foobar left 12:20 mr-foobar joined
Geth doc: af1c4ddc4b | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Type/List.pod6
Add advice to reword .unique :with using :as when possible

The :with has O(n²/2) complexity and basically slows to a crawl with 2000+ element lists.
my @p = 1..3000; @ = @p.unique: :with(&[eqv]); say now - INIT now; # 25.42728348 my @p = 1..3000; @ = @p.unique: :as(*.Str); say now - INIT now; # 0.0185597 say 25.427/.0186; # 1367.043011		 12:20
12:23 jeek left 12:31 bjz_ left, AlexDaniel left 12:33 bjz joined 12:41 mcmillhj joined 12:44 kyan left, mcmillhj left 12:45 mr-foobar left 12:46 mr-foobar joined, Cabanossi left 12:47 mcmillhj joined 12:49 Cabanossi joined 12:52 itaipu left, domidumont left 12:55 mr-foobar left 12:57 mr-foobar joined 13:00 xtreak left 13:03 jonas1 joined, jonas1 left 13:04 jonas1 joined 13:09 mr-foobar left 13:10 mr-foobar joined 13:12 ChoHag left 13:22 jonas1 left 13:23 jonas1 joined 13:24 Stamm joined 13:29 ChoHag joined 13:42 jeek joined
Stamm hello 13:43
i have a small question about hashes
what is the syntax to add an element with a non-string key?
13:43 skids joined
timotimo you just put it in with %thehash{$object} = "hi" 13:46
but the hash must be an object hash, or else it'll stringify $object for you
docs.perl6.org/type/Hash#index-ent...bject_hash
moritz m: my %h{Any}; my $key = []; %h{$key} = 'foo'; say %h{$key}; say %h{[]}
camelia foo
()
moritz Stamm: ^^ that's a way to work with an object hash 13:47
MasterDuke_ m: my %h{Int}; my $a = "a"; %h{$a}++ 13:48
camelia Type check failed in binding to parameter 'key'; expected Int but got Str ("a")
in block <unit> at <tmp> line 1
13:48 Stamm_ joined
Stamm_ ok, thank you 13:48
13:50 Stamm left 13:56 Stamm_ left 13:59 sufrostico left 14:01 Cabanossi left
ugexe samcv: the reason HTTP::Server::Tinyish fails is because it does some-test-that-does-run($my-code-with-dependency) with the assumption that all their dependencies are fullfilled by CURI and not one of the other possible CURs 14:02
14:04 Cabanossi joined 14:05 isBEKaml joined
isBEKaml OHHAI 14:05
timotimo greetings 14:06
14:10 sufrostico joined, sufrostico left, sufrostico joined
isBEKaml Were there some changes to moarvm's configure script recently? 14:11
My daily builds don't work anymore -- I'll try to fetch a detailed trace output in a bit 14:12
14:12 dmaestro left 14:13 gregf_ left, cdg joined
timotimo it could very well be that it's choking on libtommath 14:14
i turned it from a folder full of files into a submodule
git really, really, really, really, really, really didn't like that
so just nuke your moar and all should be good
isBEKaml yeah, that's what it looks like.... it failed on a submodule update and silently screamed at me for those existing folders!
timotimo maximum ugh. 14:15
isBEKaml I wiped 3rdparty, let's see if that works
timotimo might be bad. now there's stuff missing in there
isBEKaml *shrug* -- I'll know soon
\o/ 14:16
it works!
timotimo nice
isBEKaml urr, 3rdparty/uthash.c? Not cool, man 14:17
timotimo it'd be cool if we had a piece of code in the configure script that could figure out what's happening with the submodules
and mv the 3rdparty/libtommath folder out of the way
(so that no local changes, if any, get destroyed)
isBEKaml this 3rdparty folder is trucked 14:20
I'm downloading a fresh zip from Github and copying over missing folders 14:21
Okay, it's all good now 14:25
14:28 wamba left 14:29 nicq20 joined
nicq20 Hello \o 14:29
yoleaux 9 Feb 2017 17:34Z <tony-o> nicq20: you may have to do zef update -
timotimo why not just git reset --hard? taht'll give you all files back, too 14:30
14:31 lowbro left
isBEKaml timotimo: I don't have any local changes. I usually do: 'git clean -xdf' and 'git pull' for taking off on git HEAD 14:32
timotimo you had local changes, you rm -rf'd 3rdparty/ 14:33
3rdparty contains files under gits control
so reset --hard would get them back
isBEKaml Oh, right -- totally forgot about that. Anyway, I've now fully built perl6 14:34
nicq20 Is there a way to get `my $var = 'PLAYING CARD TEN OF DIAMONDS'; say "\c[$var]";` to work? Is interpolation disabled for unicode names?
timotimo we have something for that now 14:35
14:35 sufrostico left
ilmari m: my $var = 'PLAYING CARD TEN OF DIAMONDS'; say $var.parse-names 14:36
camelia 🃊
timotimo m: say parse-names('PLAYING CARD TEN OF DIAMONDS')
camelia 🃊
timotimo i don't remember when that name was decided upon?
nicq20 Yeah, I was going to say it seems like an odd choice... 14:37
Oh, well. :)
ilmari Author: Zoffix Znet [email@hidden.address]
Date: 2017-03-02 19:22:23 +0000
Implement Str.parse-names
Naming: the name follows Str.parse-base that parses a string containing numerals in some base; whereas this routine parses a string of character names 14:38
nicq20 Huh, ok.
14:39 sufrostico joined
timotimo ah 14:41
skids maybe parse-uninames to match Str.uninames? 14:45
14:45 moray` joined 14:48 cdg left
nicq20 skids: That would make more sense to me at least. 14:50
skids wonders what zoffix handle-of-the-day is. 14:51
isBEKaml parse-uninames definitely makes more sense
14:55 nicq20 left
TreyHarris Is a Junction::Utils with constructs to construct complex junctions so you can say "two", "three", "n or more", "n or less" a silly idea? I just needed a "2 or more" junction and wondered if it was worth generalizing 14:58
15:03 Cabanossi left, bjz left, sufrostico left 15:04 Cabanossi joined
perlpilot TreyHarris: what's a "2 or more" junction? 15:05
15:05 sufrostico joined
perlpilot TreyHarris: you mean like "any(@foo) but at least 2" ? 15:05
timotimo yeah 15:06
15:06 Zoffix joined
Zoffix skids: the handle is Zoffix 15:06
skids: what'd you want?
15:06 gregf_ joined
skids Oh somehow I missed seeing you in the user list. 15:07
The suggestion was that Str.parse-names might be better Str.parse-uninames
Zoffix I wasn't
[Coke] I wouldn't use it, but if it's something you needed, abstracting it out into a module might help someone else, sure.
Zoffix skids: I don't follow this channel, but my bot pinged me because of name mention 15:08
skids ZofBot?
Zoffix Yup
skids ZofBot: .botsnack
ZofBot skids, om nom nom nom
synopsebot6 om nom nom
Zoffix .parse-names has been in two releases already. The name bikeshedding kinda concluded when the PR with it got merged. 15:09
You can join #perl6-dev if you wish to keep current to development discussions.
... such as what to name .parse-names :) 15:10
Zoffix reads back 15:12
TreyHarris timotimo: exactly. like "one(...)", only "two(...)" or "two-or-more(...)" (obviously, parameterized, so actually probably "exactly(2, ...)" or "at-least(2, ...)" or "at-most(2, ...)" ... is that the best signature or is there a cleaner one?
perlpilot TreyHarris: are you *sure* junctions are the right answer? Feels more like sets and counting to me. 15:13
Zoffix "don't remember when that name was decided upon?". I proposed the name on March 2nd EST morning, then in afternoon made a PR. Lizmat merged it later in the day, after a +1 from [Coke]: github.com/rakudo/rakudo/pull/1031 15:14
moritz TreyHarris: IMHO junctions are neat, but if they don't do what you want, they aren't worth tweaking
I tend to use junctions only as boolean matchers
perlpilot too
moritz in that context you can typically fall back to a callback instead
Zoffix skids: so yeah, .parse-uninames sounds better if a bit too much typing. But I don't want to change something that's been in a release already. 15:15
TreyHarris moritz: ah, I haven't looked into how to run authothreading manually, but that's why I was using them, because "so at-least-one(...)" was quite useful to me and embarrassingly parallell
Zoffix in 2 releases
15:15 Zoffix left
TreyHarris sorry, "at-least-two" 15:16
"at-least-one" we have, "any"
(er, how to do *threading* manually, I meant, obviously) 15:18
moritz junction don't actually evaluate in parallel yet 15:19
skids ah, well, bummer. Unfortunately I barely have the attention-time to keep current on everything.
15:20 mr-fooba_ joined, domidumont joined
skids s/barely/rarely/ 15:20
15:20 mr-foobar left
TreyHarris moritz: I know, but I like maintaining the fiction for future-proofing. In this case I'm using junctions of attributes of machines in a datacenter, so it's an obvious case 15:20
moritz TreyHarris: I can't follow that line of reasoning 15:21
but you don't actually have to convince me 15:22
putting stuff into a junction makes it harder to actually use the stuff, unless you only ever use them as boolean matchers 15:23
TreyHarris moritz: if autothreading would likely come to Sets at the same time (or soon after) it came to Junctions, then you've convinced me.
moritz: (you don't understand why I'd want to future-proof so that the code would autothread later on? I know you don't want me to convince you, so I won't continue arguing, I just want to be sure I know what you don't understand so will know what to reconsider my thinking on) 15:25
moritz TreyHarris: you're basically saying you rely on behavior that might change in future, in order to future-proof your code 15:27
TreyHarris moritz: oh, i thought the one point of junctions was this thing--if I can't rely on it, then I should be doing map-so-length checks instead, no? 15:29
because maps can be parallelized easily in most languages.
(sorry, I said I wasn't going to argue. So nm, but I apparently don't get it....) 15:30
15:30 Zoffix joined
moritz TreyHarris: the point of junctions is to express ideas more idiomatic 15:30
or more direct
TreyHarris right, and map-so-length >= 2 is idiomatically, "at-most 2" 15:31
but yeah, I'll just use sets and lenghts. 15:32
Zoffix Wanted to clarify "bikeshedding kinda concluded when the PR with it got merged". I meant the name was proposed in a PR first; I didn't just commit. It's very easy to change a name for a feature added, but *before it's in a release*. Telling people to go use something and then pulling the carpet from under them is LTA. And on more official level: .parse-name isn't even part of Perl 6 language yet, since it's 15:33
only in the `master` spec that's yet to be reviewed and approved for 6.d. In the future, we'll likely want to improve this process, so there's some grace period or somethign for features that need time to solidify. OTOH, Slangs are not official either, but people are using them :/
15:34 robertle left
Zoffix Telling people to go use something as in a release changelog/docs I mean 15:35
15:35 sufrostico left
TreyHarris Zoffix: are you saying that bikeshedding names _before_ that _is_ a good use of time? Because I seems to remember a certain TimToady giving a certain State of the Onion on the importance of naming.... ;-) 15:35
15:36 moray` left
Zoffix TreyHarris: that question is ambiguous since "bikeshedding" can mean "trying to decide on a name in a group" or "pointlessly wasting time on irrelevant minutae" 15:37
moritz investing time in names that appear in a public API sounds like a good investment, if used productively
skids slinks to a discrete corner of the room and tries to pretend he didn't start this.
TreyHarris skids: lol 15:38
Zoffix :)
perlpilot not to bikeshed, but I probably would have called it unichars(). uninames() is given some chars it gives you names, unichars() is given some names it gives you chars.
moritz skids: I prefer continuous corners over discrete corners :-) 15:39
TreyHarris Zoffix: maybe "speak up if a name seems objectively BAD (i.e., reads as ungrammatical English in the way it's typically used, so will lead to frustrating and repeatedspelling mistakes) or if you have a name you think is so much obviously better, but don't waste time iterating on names?" 15:40
Zoffix skids: you did bring up an issue in the process, I think
15:40 nadim left
Zoffix That public API changes should probably given some X amount of time for dev/community feedback. 15:41
15:41 sufrostico joined
Zoffix m: "foo.txt" 15:42
camelia WARNINGS for <tmp>:
Useless use of constant string "foo.txt" in sink context (line 1)
skids I am reminded to bother TimToady or anyone else for bikeshedding "insist role" or "insist method" for my noisy roles branch.
Zoffix m: "bar/foo.txt".IO.sibling("meow.txt").say
camelia "bar/meow.txt".IO
Zoffix Like that was added after a brief mention on IRC channel.
skids (gist.github.com/skids/18fa6fb1de77...e82e9fcc2)
Zoffix skids: I liked the old thing that was proposed. Some sort of prefix to method 15:43
forget what it was
skids claim.