»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.
00:00 BenGoldberg joined
bazzaar well simplistically I could eliminate case, and filter away 'prefixed-by-sigil' occurrences, it might be gnarly, but it could be useful, and perhaps be less of a source of frustration than the poorly populated utility tags collection is 00:00
ugexe right, but the simplistic case is already served better by the simple regex it would require 00:02
your idea is indefinitely better, but to achieve it is much more difficult (regex idea breaks down when its not so simple to come up with and type as a query) 00:03
00:04 kyan left 00:05 Kyo91_ left
bazzaar agree that it might be a poor man's egrep, but it would be a first pass summary index of coding 'examples' embedded within the modules code, where the user only need to click to get some results 00:08
00:13 Kyo91_ joined
ugexe you could always write a microservice (with Cro or Bailador) to proof of concept, and then simply provide the module website API access to it should you decide it is polished enough 00:14
bazzaar was just an idea that has sprung to my mind a few times, wanted to post it, appreciated folks taking the time to consider
ugexe come up with a list of keywords, and `git grep $_` your way to a database
probably less than 100 lines of code, assuming you have something to clone the repos 00:15
00:16 wamba left
ugexe or write a bot to come in here and send `greppable6: $_` for all your keywords and build your database from that hehe 00:17
bazzaar ugexe: thanks for the insight, I will pursue that line of approach :-)
00:19 ryn1x joined 00:20 kyan joined 00:23 mempko left 00:24 ryn1x left, Herby__ left 00:27 nadim left 00:40 bazzaar left 00:42 Kyo91_ left 00:44 Kyo91_ joined 00:49 Kyo91_ left 00:56 Herby__ joined
Herby__ Also: how can I use a variable as a regex pattern? my @list-of-patterns[0;1] = '(\w+)'; say $/ if "some string" ~~ / @list-of-patterns[0;1] /; 01:00
rather, how do i interpolate a variable in a regex pattern 01:01
MasterDuke Herby__: just the variable if you want it interpreted as a string, <$var> if you want it interpreted as a regex, <{ code }> to execute the code and interpolate the results into the regex 01:03
01:03 bitrauser_ joined
MasterDuke m: my $r = "b"; say "abc" ~~ / $r / 01:05
camelia ｢b｣
MasterDuke m: my $r = "a.c"; say "abc" ~~ / <$r> /
camelia ｢abc｣
Herby__ what if you have capture groups in the variable?
m: my $r = "a(.)c"; say "azc" ~~ / $r /; 01:06
camelia Nil
Herby__ m: my $r = "a(.)c"; say "azc" ~~ / <$r> /;
01:06 bitrauser left
camelia ｢azc｣ 01:06
Herby__ m: my $r = "a(.)c"; say $/ "azc" ~~ / <$r> /;
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3my $r = "a(.)c"; say $/7⏏5 "azc" ~~ / <$r> /;
expecting any of:
infix
infix stopper
postfix
statement end…
Herby__ m: my $r = "a(.)c"; say ~$/0 if "azc" ~~ / <$r> /;
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3my $r = "a(.)c"; say ~$/7⏏050 if "azc" ~~ / <$r> /;
expecting any of:
infix
infix stopper
postfix
statement…
Herby__ m: my $r = "a(.)c"; say ~$/[0] if "azc" ~~ / <$r> /;
camelia Use of Nil in string context

in block <unit> at <tmp> line 1
Herby__ sorry 01:07
MasterDuke Herby__: you might need to use EVAL, not 100% sure
Herby__ hmmm
MasterDuke i.e., create the "regex" as a string and then EVAL that
geekosaur it is known that subcaptures are lost in that case, I don't think anyone has sorted out why yet 01:08
MasterDuke m: my $r = "/a(.)c/".EVAL; say "abc" ~~ $r
camelia ｢abc｣
0 => ｢b｣
Herby__ maybe there is a better way to accomplish my goal. I have a list of regex patterns (that have capture groups) and I want to apply each pattern to a string 01:09
MasterDuke: thanks 01:12
MasterDuke np
01:15 Cabanossi left 01:28 ryn1x joined 01:30 Herby__ left 01:46 ilbot3 left 01:53 ch3ck3r joined 01:55 ilbot3 joined, ChanServ sets mode: +v ilbot3 01:57 napo1eon left 02:02 ryn1x left 02:18 raschipi left 02:21 noganex_ joined 02:24 noganex left 02:28 Herby__ joined, Cabanossi joined 02:30 BenGoldberg left 02:46 ryn1x joined 02:50 ryn1x left 03:10 evalable6 left 03:11 evalable6 joined 03:17 ryn1x joined 03:29 llfourn left 03:33 llfourn joined 03:39 troys_ is now known as troys, troys is now known as troys_, troys_ is now known as troys 03:40 troys is now known as troys_, troys_ is now known as troys 03:41 Cabanoss- joined 03:44 Cabanossi left, Cabanoss- is now known as Cabanossi 03:55 ryn1x left 03:56 Herby__ left 04:00 sena_kun joined 04:14 R0b0t1_ left 04:17 R0b0t1_ joined 04:22 ryn1x joined 04:23 MasterDuke left 04:26 ryn1x left 04:34 sena_kun left 04:36 Herby_ left 04:39 khw left
Geth doc: lefth++ created pull request #1618:
Small changes: wrong link, missing :kv, grammar error		 04:39
04:43 Aaronepower left 04:45 mempko joined 04:55 wamba joined, jeek left 04:59 ryn1x joined
zostay m: dd \(a => 1, b => <a b c>); dd Capture.new(:hash(%(a => 1, b => <a b c>))); 05:00
camelia \(:a(1), :b(("a", "b", "c")))
\(:a(1), :b($("a", "b", "c")))
05:00 troys left
zostay m: dd \(a => 1, b => <a b c>); dd Capture.new(:hash(%(a => 1, b => <a b c>.list))); 05:00
camelia \(:a(1), :b(("a", "b", "c")))
\(:a(1), :b($("a", "b", "c")))		 05:01
zostay I keep banging my head on this and can't figure out a way to make a Capture programmatically that behaves the same as the syntactic one.
05:03 xtreak joined 05:11 ch3ck3r left 05:25 pilne left 05:33 ryn1x left 05:44 nhywyll joined
Geth doc: jstuder-gh++ created pull request #1619:
Document List cmp infix and sorting by multiple values (via List)		 05:52
06:04 evalable6 left, evalable6 joined, ChanServ sets mode: +v evalable6 06:16 mempko left 06:19 rba joined 06:20 konsolebox joined 06:22 rba left 06:23 rba joined 06:26 virtualsue left 06:27 virtualsue joined, wamba left, lowbro joined, lowbro left, lowbro joined 06:35 darutoko joined 06:38 ryn1x joined 06:41 cdg joined 06:44 domidumont joined 06:46 cdg left 06:47 xtreak left, nhywyll left 06:49 xtreak joined 06:59 abraxxa joined 07:02 rba_ joined 07:05 jonas1 joined, rba left 07:08 domidumont left 07:09 domidumont joined 07:12 ryn1x left
lizmat clickbaits p6weekly.wordpress.com/2017/10/23/...ds-racing/ 07:17
07:18 cdg joined 07:23 cdg left, rindolf joined
AlexDaniel` that's a very nice weekly. lizmat++ 07:26
07:30 xtreak left, xtreak joined 07:35 xtreak left, xtreak joined
moritz indeed, lizmat++ 07:43
07:43 dakkar joined 07:50 ryn1x joined 07:59 xtreak left 08:19 evalable6 left, evalable6 joined, jc_ joined 08:20 jc_ left, jc_atikon joined 08:21 jc_atikon left, jc_atikon joined 08:22 jc_atikon left 08:24 jcarty joined 08:26 ryn1x left 08:35 rba_ left 08:38 rba joined 08:40 ChoHag left 08:44 bartolin left, bartolin joined 08:46 wamba joined 08:50 sproctor joined 08:52 rba left, nadim joined 08:54 rba joined 08:55 cpage_ left
AlexDaniel` oh gosh, pascal's triangle in 26 characters… I'm out :) 08:55
( code-golf.io/ )
moritz happy to see so many camels and butterflies on that site 08:57
09:02 itaipu left, parv joined 09:06 rba left 09:09 rba joined 09:20 zakharyas joined 09:22 rba_ joined 09:24 rba left 09:26 xtreak joined 09:28 ryn1x joined 09:34 cdg joined 09:38 rba joined 09:39 cdg left 09:41 rba_ left 09:45 cpage_ joined 09:52 Grimy joined 10:02 ryn1x left 10:03 rba left 10:10 wander joined
wander m: grammar G { rule TOP { <number>+ { dd $<number>; die "wrong" if $<number>[0].made !~~ /\w+/; dd $<number>; } }; rule number { \d+ { make ~$/; }} }; G.parse("1 2 3 4"); 10:12
camelia [Match.new(list => (), made => "1 ", pos => 2, hash => Map.new(()), orig => "1 2 3 4", from => 0), Match.new(list => (), made => "2 ", pos => 4, hash => Map.new(()), orig => "1 2 3 4", from => 2), Match.new(list => (), made => "3 ", pos => 6, hash => …
wander grammar G { rule TOP { <number>+ { dd $<number>; die "wrong" if $<number>[0].made !~~ /\w+/; dd $<number>; } }; rule number { \d+ { make ~$/; }} }; G.parse("1");
m: grammar G { rule TOP { <number>+ { dd $<number>; die "wrong" if $<number>[0].made !~~ /\w+/; dd $<number>; } }; rule number { \d+ { make ~$/; }} }; G.parse("1");
camelia [Match.new(list => (), made => "1", pos => 1, hash => Map.new(()), orig => "1", from => 0)]
Nil
wander interesting
mienaikage I'm lost on this one, anyone know why every 4th test fails here? I was up until 2am scratching my head over it 😕 glot.io/snippets/euu75yqs7d
wander after a smart-match, I lose $<numver>
AlexDaniel` wander: well, matching a regex will reset $/ 10:15
wander: or set it to whatever you get from matching 10:16
wander aha
right, I forget what $/ is
AlexDaniel` $<number> is same as $/<number> :) 10:17
docs.perl6.org/syntax/$$SOLIDUS
Grimy p6: sub infix:<!~>($a, $b) { say "$a, $b" }; 3 !~ 4 10:18
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of !~ to do negated pattern matching; in Perl 6 please use !~~
at <tmp>:1
------> 3fix:<!~>($a, $b) { say "$a, $b" }; 3 !~ 7⏏054		 10:19
Grimy Shouldn’t it be possible to override infix:<!~> ? After all, you can override infix:<+>.
DrForr infix:<<!~>> maybe? 10:20
Grimy Nnope, same result.
Looks like infix:<!~> is special-cased to always give this error message.
DrForr Huh. 10:21
10:28 ChoHag joined 10:31 wander left
AlexDaniel` Grimy: I think you should submit a bug report for that. We can definitely do better. 10:31
10:32 astj left 10:33 astj joined
Grimy Thanks for the suggestion AlexDaniel, I will 10:34
Just noticed that infix << and >> have the same behavior, too
AlexDaniel` and =~ also, and possibly many other things 10:35
DrForr «..» might work around those.
Though it arguably shouldn't need characters outside ASCII. 10:36
Grimy Don’t see how.
10:36 mr-foobar left
Grimy p6: sub infix:«<<»($a, $b) { say "$a, $b" }; 3 << 4 10:36
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of << to do left shift; in Perl 6 please use +< or ~<
at <tmp>:1
------> 3nfix:«<<»($a, $b) { say "$a, $b" }; 3 <<7⏏5 4
10:37 mr-foobar joined
mienaikage re: The above snippet, even more confusingly, it works if I change Int:D, Str:D to Int, Str 10:41
10:47 ryn1x joined 10:52 ryn1x left
timotimo mienaikage: looks like the two Hash[Int:D,Str:D] types end up not being the same object 10:53
oh, no, that's not true 10:54
Hash[Int,Str] also gives two distinct types, but they are considered equal
mienaikage is-deeply would be eqv and not === right? 10:55
timotimo so something's not checking equivalence right for definite type constraint thingies
not sure
will be afk for a bit
10:59 Actualeyes left, Actualeyes joined, Actualeyes left 11:00 Actualeyes joined 11:01 eliasr joined
AlexDaniel` m: $_ = ‘5 6 7 8’; dd «$_» Z+ 1 11:03
camelia (27,).Seq
AlexDaniel` why 27?
I do see that it sum-ed all elements on the left, but why?
Grimy p6: slip(5, 6, 7, 8) Z+ 1 11:07
camelia Potential difficulties:
Useless use of Z+ in sink context
at <tmp>:1
------> 3slip(5, 6, 7, 8) 7⏏5Z+ 1
Grimy p6: dd slip(5, 6, 7, 8) Z+ 1
camelia (27,).Seq
11:07 mr-foobar left
AlexDaniel` hmmm huh 11:08
11:08 lowbro left, mr-foobar joined
Grimy circumfix «» gives a slip, and slips do weird things in argument lists. slip(5, 6, 7, 8) Z 1 gives ((1 5 6 7 8)), not ((1 5)). The behavior of Z+ is consistent with that. 11:10
11:11 kyan left
AlexDaniel` ok, but why does «» give a slip in this particular case? 11:12
m: $_ = ‘5 6 7 8’; dd «$_»
camelia slip(IntStr.new(5, "5"), IntStr.new(6, "6"), IntStr.new(7, "7"), IntStr.new(8, "8"))
AlexDaniel` m: dd «5 6 7 8»
camelia (IntStr.new(5, "5"), IntStr.new(6, "6"), IntStr.new(7, "7"), IntStr.new(8, "8"))
11:14 Actualey` joined
Grimy m: $_ = '5 6 7'; dd «$_ 8» 11:16
camelia (IntStr.new(5, "5"), IntStr.new(6, "6"), IntStr.new(7, "7"), IntStr.new(8, "8"))
11:17 Actualey` left, Actualey` joined 11:20 Actualey` left 11:22 Zoffix joined
Zoffix it's 'cause the variable may contain more than one value 11:24
and with a slip it all ends up one with the rest of the things; that's my understanding of it 11:25
11:25 ryn1x joined
Zoffix m: $_ = ‘5 6 7 8’; dd «"$_"» 11:28
camelia "5 6 7 8"
11:29 margeas joined 11:30 Actualey` joined
AlexDaniel` to me it feels like the slip should not leak outside 11:30
Zoffix m: $_ = ‘5 6 7 8’; dd «$_ 1»
camelia (IntStr.new(5, "5"), IntStr.new(6, "6"), IntStr.new(7, "7"), IntStr.new(8, "8"), IntStr.new(1, "1"))
Zoffix Yeah, to me too
mienaikage m: $_ = ‘5 6 7 8’; dd «1$_»
camelia (IntStr.new(1, "1"), IntStr.new(5, "5"), IntStr.new(6, "6"), IntStr.new(7, "7"), IntStr.new(8, "8"))
Zoffix it's probably made to handle ^ that case and the case where it's just a variable isn't handled 11:31
AlexDaniel` yea
I'll submit a ticket then
11:32 bdmatatu joined
Zoffix m: dd «x».^name 11:32
camelia "Str"
lizmat is running a stresstest for a fix for issue 1220 11:33
11:33 sena_kun joined
lizmat alas, the fix doesn't take away the instability completely :-( 11:33
11:33 itaipu joined 11:36 mr-foobar left, parv left
AlexDaniel` lizmat: does it ever hang for you? 11:37
lizmat no, not so far 11:38
11:38 raschipi joined, xtreak left 11:40 nadim left, mr-foobar joined
timotimo mienaikage: it turns out glot.io has version 2017.01 of rakudo installed; the bug with is-deeply with Blah:D seems to already be fixed 11:41
raschipi Is code-golf.io not working? 11:42
11:42 xtreak joined
lizmat *1202 11:43
AlexDaniel` raschipi: opens just fine here 11:44
mienaikage timotimo: I have the same behaviour on my machine with Rakudo version 2017.09-133-g5ad2fffed built on MoarVM version 2017.09.1-49-gb3dd812a
timotimo oh!
ilmari ls 11:45
mienaikage I'd just put it on glot for the sake of sharing
ilmari EWIN
raschipi working now, sorry for bothering
11:48 itaipu left 11:50 sena_kun left 11:51 HTTP_____GK1wmSU joined 11:52 HTTP_____GK1wmSU left
mienaikage Hmm, I broke it even more... If you use require instead of use it doesn't like the typing at all glot.io/snippets/euuaa6wm8p 11:53
11:54 Actualey` left 11:55 Actualey` joined 11:57 ryn1x left 11:58 lowbro joined, lowbro left, lowbro joined
Zoffix mienaikage: what's the bug? 11:59
oh /me notices the glot.io snippety
mienaikage: where do I get ETL from? 12:03
Oh, there's a second tab, nm
mienaikage :)
Grimy m: say circumfix:<[ ]>(3) 12:04
camelia WARNINGS for <tmp>:
[3]
Useless use of constant value [ ] in sink context (lines 1, 1)
Zoffix :S my perl6 refuses to find it 12:05
Grimy This warning seems wrong, isn’t it?
Zoffix Grimy: yeah. It's ticketed: rt.perl.org/Ticket/Display.html?id...et-history
Grimy Thanks!
Zoffix mienaikage: reproed the problem. Gimme a sec to golf it a bit. 12:08
mienaikage No worries. Was trying to put together a new exercise for exercism. Wasn't sure if I broke something or was just being silly. 12:09
raschipi But can you do it in 26 characters?
Zoffix mienaikage: there's definitely something wrong :) 12:10
12:17 synopsebot left, synopsebot joined, ChanServ sets mode: +v synopsebot 12:18 xtreak left 12:21 xtreak joined
Grimy m: $_ = 42; say $_ &= 3 12:22
camelia MoarVM panic: Memory allocation failed; could not allocate 108304 bytes
Zoffix heh
Grimy Should I make a ticket for this one?
Zoffix Grimy: yeah
12:22 nadim joined
Grimy Alright 12:22
Zoffix Grimy: you can make one here: github.com/rakudo/rakudo/issues
Grimy: note that `&` is an 'all' Junction, not a bit twiddling op 12:23
Grimy Thanks. Is this the recommended method? I’ve been mailing to perl6-bugs
Zoffix Grimy: it's just something new we're experimenting with. Emailing is fine too.
Grimy Alright, I’ll use GH issues since it’s more convenient imo 12:24
Geth doc: a5dee808bc | (Dan Zwell)++ (committed by Dan) | doc/Type/Hash.pod6
Document Hash's :kv adverb

Also fix a small grammar error.		 12:25
synopsebot Link: doc.perl6.org/type/Hash
Geth doc: 932d9f6801 | (Dan Zwell)++ (committed by Dan) | doc/Type/Signature.pod6
Fix the destructuring docs linking to the "Capture" type instead of method.
synopsebot Link: doc.perl6.org/type/Signature 12:26
doc: e4cce1e799 | (Rafael Schipiura)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1618 from lefth/master

Small changes: wrong link, missing :kv, grammar error
doc: 92927dfa18 | (Jeremy Studer)++ | doc/Type/List.pod6
Document the List variant of the cmp operator
 12:28
doc: 191db0bd5a | (Jeremy Studer)++ | 2 files
Document sorting by multiple criteria

This is a powerful way of sorting, but it's not currently apparent that it can be done this way in the documentation.
Also, in this section have infix:<cmp> link to a page documenting the various instances of the cmp operator.
Also, added myself to credits.
synopsebot Link: doc.perl6.org/type/List
12:29 sproctor left
doc: 26ab76444d | (Rafael Schipiura)++ (committed using GitHub Web editor) | 2 files
Merge pull request #1619 from jstuder-gh/sort_multiple

Document List cmp infix and sorting by multiple values (via List)
12:30 xtreak left 12:32 sproctor joined 12:37 pmurias joined
Zoffix mienaikage: field an Issue for it. It's a bug in Rakudo; something's busted with hashes parametired with Int:D types: github.com/rakudo/rakudo/issues/1207 12:38
mienaikage: you can workaround it by not using the `:D` in your types or sticking `no precompilation` at the top of your module (note that'd cause it to be re-compiled each time which can make things very much slower)
.oO( you know you're too tired when you write 'parametirized' as 'parametired' ... )
 12:39
12:39 wander joined
Zoffix *filed too 12:39
12:42 ryn1x joined 12:46 ryn1x left
perlpilot huh ... I was just reading the weekly and the code for running "once in a batch" surprised me. 12:46
lizmat perlpilot: why ? 12:48
[Coke] do we want doc tickets to get announced her?
*here
raschipi yes 12:49
perlpilot lizmat: because "once {...}* run only once, suppressing additional evaluations" That batching is a side-effect of .hyper/.race should factor in to that IMHO. At least not without something more explicit. 12:50
*shouldn't 12:51
[Coke] I would not expect once to mean "once per batch" 12:52
Zoffix perlpilot: then which batch should it be run in?
The phasers will also get special treatment
12:53 mson joined
Zoffix [Coke]: it'd just need to add an event handler here github.com/perl6/geth/blob/master/...#L114-L121 and IRC responder here: github.com/perl6/geth/blob/master/...b.pm6#L123 AlexDaniel` probably already has code for that from the squashable6 bot 12:54
lizmat my proposal would be to *not* handle phasers in hypers/races differently
but set a $*BATCH dynamic variable indicating that it's being run in a batch
perlpilot Or put another way, if I have a .map that calls a function to do work and that function calls `once { }` somewhere down the line, will I accidentally change the semantics of the once block by adding .hyper or .race?
Zoffix lizmat: but would it be per-batch?
lizmat which could have additiional information, like batch number
Zoffix perlpilot: yes 12:55
perlpilot: same as if you have that function do something thread-unsafe
lizmat Zoffix: yes, a FIRST / LAST would be per batch
Zoffix Yeah, sounds like what I was thinking 12:56
lizmat *and* before the very first batch
(FIRST that is)
and after the very last batch (LAST)
those two cases would *not* have $*BATCH set
Zoffix Not that I know much about this part of language :)
lizmat NEXT/REDO/ENTER/LEAVE phasers should work as normal 12:57
12:57 piojo joined
lizmat well, that would be my suggestion 12:57
piojo Is rakudo failing to build for everyone, or just me? I'm seeing this error: While looking for 'Perl6/ModuleLoader.moarvm': no such file or directory 12:59
Geth doc: cd197c1fe6 | (Alex Chen)++ (committed using GitHub Web editor) | doc/Type/Whatever.pod6
Delete incorrect doc of Exception of Whatever-star

  `1 ~~ *` will return a WhateverCode instead of return true
Example:
  ```
  > map (1 ~~ *), <1 2 3>
  (True False False)
  ```
synopsebot Link: doc.perl6.org/type/Whatever
perlpilot piojo: probably just you as I built mine just fine mere moments ago.
lizmat piojo: not seeing that
piojo The build target that's failing is blib/Perl6/World.moarvm
Zoffix had no problems last night
piojo I should say it's on Windows
I haven't tested Linux, though I should do that
lizmat ah, k, can't help then
Zoffix had no problems a couple weeks ago
piojo: is that with a fresh checkout? 13:00
e.g. like these steps: github.com/zoffixznet/r#windows
That's how I built it a couple weeks ago
piojo Zoffix: no, it's not. Thanks! I'll try that
perlpilot anyway ... let's just say that I continue to be uncomfortable with the action-at-a-distance for once (and other phasers) 13:01
mienaikage Zoffix: Thanks. I've added a little extra info on GH re: require as well
lizmat perlpilot: what would be your suggestion?
a complete new set of phasers ?
or just no phasers (change the NYI into a real error) 13:02
afk again&
perlpilot I dunno, I'm not awake enough yet :-) Maybe I'd be more comfortable if the documentation for .hyper/.race called these things out explicitly? Right now, the weekly reads like ".hyper/.race makes things faster" (and I think that's how many people will think of them too) but really it's more like ".hyper/.race will make things faster, but beware the semantic differences they impose upon certain phasers and phaser-like constructs" 13:03
13:03 ggoebel left
Zoffix perlpilot: my impression of watching this work from a distance is people seem to be expecting a magic "make it faster" button that will perform work on all the cores, when in reality there's gonna be a bunch of race conditions people will be struggling with or, like you point out, receiving new semantics for stuff like `once` because they expected the program to know when once has been executed already on one 13:04
of the threads. gfldex already had a bug in their code that cached a Seq twice from a hyper and it took even jnthn some time to figure out what the crash was about. I fully expect more of that stuff to turn up and possibly people eventually realizing .hyper/.race aren't a magical fire-and-forget make-it-faster solution
13:05 cdg joined
perlpilot Zoffix: agreed. 13:06
Zoffix m: hyper for ^4 { sleep 1 }; say now - INIT now 13:08
camelia 4.0099630
perlpilot I think it's just we need more words out there (in documentation, blogs, etc.) talking about the intersection of these features and I'm uncomfortable right now because I've been operating from a mindset that didn't consider this intersection.
Zoffix m: hyper for ^4 .hyper { sleep 1 }; say now - INIT now
camelia 4.00964309
Zoffix was expecting 1s in both cases :/
m: ^4 .hyper.map: { sleep 1 }; say now - INIT now 13:09
camelia 4.0106535
Zoffix :/
wander m: say (42 ~~ *) 13:10
camelia { ... }
wander m: say (* ~~ 42)
camelia { ... }
AlexDaniel` Zoffix: isn't it ok? you didn't specify the batch size
wander m: 42.ACCEPTS(*) 13:11
camelia ( no output )
Zoffix AlexDaniel`: oh, duh
wander m: say 42.ACCEPTS(*)
camelia False
AlexDaniel` m: ^4 .hyper(:1batch).map: { sleep 1 }; say now - INIT now
camelia 1.051253
Zoffix m: say (42 ~~ (*))
camelia True
Zoffix wander: your variant made a WhateverCode closure
m: hyper for ^4 .hyper: :batch { sleep 1 }; say now - INIT now 13:12
camelia 1.0510388
Zoffix \o/
13:12 cdg_ joined
Zoffix m: for ^4 .hyper: :batch { sleep 1 }; say now - INIT now 13:12
camelia 4.00893443
wander WhateverCode in `* ~~ 42`, while Whatever in `42.ACCEPTS(*)`, is it?
ilmari m: hyper for ^4 { sleep 1 }; say now - INIT now 13:13
camelia 4.0116992
piojo I can build rakudo against an older moar/nqp, but not against the newest
Zoffix perlpilot: ^ (my last eval) another semantic difference. `for` is no longer just a `map` under the hood; it's `.serial.map` and a `for` without a `hyper` or `race` prefix iterates over HyperSeq and RaceSeq serially
wander: right
piojo never mind, I thought it was working but it wasn't 13:14
Zoffix piojo: can you create an Issue ( github.com/rakudo/rakudo/issues/ ) with full output of the error, along with Windows version and what type of C compiler/perl you got installed?
piojo Zoffix: I'll do that now 13:15
Zoffix Thanks.
13:16 cdg left
perlpilot Zoffix: yeah, but that's a difference I was expecting, whereas "once" running more than once wasn't :-) 13:16
13:17 ryn1x joined
Zoffix perlpilot: dunno, for me I'd be more surprised with ^4 .hyper(:batch).map: { my $x; once $x = Foo.new; $x.do-things }; crashing, just because the `once` block ran in only one of the batches 13:18
Or something like that. Text::CSV uses this setup
13:20 darkmorph joined
wander m: Whatever.new.Capture 13:21
camelia Cannot unpack or Capture `*`.
To create a Capture, add parentheses: \(...)
If unpacking in a signature, perhaps you needlessly used parentheses? -> ($x) {} vs. -> $x {}
in block <unit> at <tmp> line 1
Zoffix wander: that's by design. Is that creating a problem in some code? 13:22
13:23 khw joined
Zoffix m: my atomicint $v = 0; ^4 .race(:batch).map: { once $v⚛++ }; say $v 13:23
camelia 4
Zoffix m: my atomicint $v = 0; ^4 .race(:2batch).map: { once $v⚛++ }; say $v
camelia 2
Zoffix perlpilot: yeah, kinda surprising behaviour. 13:24
wander on my PC, it doesn't throw Exception
This is Rakudo version 2017.07 built on MoarVM version 2017.07 implementing Perl 6.c.
Zoffix wander: it's a new feature. Implemented like 2 weeks ago.
wander got it
Zoffix
.oO( `hyper once` )
 13:27
perlpilot lizmat++ also because I learned something new in the p6weekly. I'm an "episodic volunteer" :-) (Ann Barcomb)++ too 13:28
AlexDaniel` what new feature? atomicint? Isn't it like 2 months ago now?
Zoffix AlexDaniel`: you're mixing up conversation threads :)
piojo Zoffix: I get a different error with --gen-moar --gen-nqp. Should I file a separate bug report for that? 13:29
DrForr Might want to watch the talks from the Open Source Summit in Prague - she's giving a similar talk there.
Zoffix "m: Whatever.new.Capture" " wander: that's by design." "on my PC, it doesn't throw Exception" "it's a new feature"
AlexDaniel` Zoffix: not only that, but I'm also blind /o\ Somehow skipped “doesn't” 13:30
alright, back to release stuff then
13:30 ggoebel joined
Zoffix piojo: I think adding that to the same bug report is fine 13:30
piojo Okay, thanks
perlpilot DrForr: thanks, I'll check it out
DrForr Kind of annoying - the summit is at the Hilton just a few blcoks from here but she's too busy to hang :/ 13:31
*hang out. 13:32
timotimo m: for ^4 { for ^3 { once say "hi" } } 13:33
camelia hi
hi
hi
hi
Zoffix :)
timotimo for your consideration, perlpilot, Zoffix
perlpilot hmm 13:34
Makes me think "once" needs a new name :) 13:36
moritz roughly-once
perlpilot I would guess that falls out of the "once per clone of the containing object" 13:37
DrForr 'just' :)
Zoffix once-in-a-blue-moon 13:38
13:38 wander left
perlpilot once-upon-a-time 13:39
once-and-future-once
once-more-into-the-breach
once-in-a-while
once-twice-three-times-a-once 13:40
sjn once-twice-often
timotimo once upon a closure in the west
ilmari once-for-the-money
sjn once-for-the-road 13:41
perlpilot once-or-maybe-more is a probably accurate. :)
timotimo first-time?
raschipi change 'once' to 'nonce'
Zoffix 'monce' for 'mostly-once' 13:42
perlpilot once-per-clone -> clonce 13:44
once-per-batch -> bounce (kinda :)
[Coke] to be clear, I am fine with once meaning something slightly different based on hyper, map, just wasn't expecting it at first. 13:46
perlpilot [Coke]: did you expect timotimo's example either? :) 13:47
13:48 napo1eon joined 13:49 wander joined 13:51 ryn1x left
AlexDaniel` is anybody using a 32-bit system here? 13:51
Zoffix I may have one handy.. why? 13:53
AlexDaniel` Zoffix: I'm wondering if github.com/perl6/nqp/issues/377 is related to that 13:54
this is just a wild guess
pmurias bisect: my $b = {"a" => 1000000000000000000}.BagHash; $b.grab(100); say $b.total;
bisectable6 pmurias, On both starting points (old=2015.12 new=b19e352) the exit code is 0 and the output is identical as well
pmurias, Output on both points: «999999999999999900»
Zoffix AlexDaniel`: well, I got 2017.09-223-gd565ded built on it 13:55
AlexDaniel` Zoffix: that's great but kinda 277 commits behind :) 13:58
13:58 callyalater joined
Zoffix Statistically irrelevant :) dogbert17 also got a 32-bit system IIRC 14:00
s/irrelevant/insignificant/;
I noticed some edited paths in that Issue's output; I suspect something was messed up with --prefix or whatever 14:02
That or the nqp-lib thing was messed up and there's some conditon where it causes extra args given
(missing quoting on paths with spaces?)
AlexDaniel` could be that yeah
oh, he responded 14:03
Zoffix
.oO( "he"? )
AlexDaniel` they 14:04
Zoffix ZofBot: what's your preferred pronoun?
ZofBot Zoffix, '"Down with Big Brother!" Yes, I said that! Said it over and over again, it seems
Zoffix hah
AlexDaniel` “Cloning into '.'...” 14:05
Altreus Did you forget that bonce is a word?
Zoffix AlexDaniel`: well, I see what it's complaining about: "/tmp/tmp.o2BIF8otPY/install/bin/moar --libpath= [...] gen/moar/stage1/nqpmo.nqp" I guess that line needs to be found in make file and the `gen/moar/stage1/nqpmo.nqp` stuff put before the named args? Wonder what that line is like on working builds.. 14:07
14:07 mr-foobar left
Zoffix AlexDaniel`: Cloning into '.' is the rakudo clone from my command that person ran 14:07
AlexDaniel` ah
14:07 mempko joined
Zoffix Altreus: didn't know it was :D 14:08
14:08 mr-foobar joined, zakharyas left
Grimy ^... and ^...^ are NIY, but aren’t marked as such, and give really confusing error messages instead. 14:08
14:09 zakharyas joined
[Coke] apparently pigsconce is also a word. (pig-sconce) 14:09
AlexDaniel` m: say 1,20^...30 14:10
camelia 120^..0.3
AlexDaniel` well
Zoffix Grimy: I thought they weren't meant to be implemented 14:11
But I guess they could be. The ...^ is already done and ^... can be done with .tail now
Grimy Maybe? They’re listed in S03.
How do I know when an idea in the Synopses has been abandonned? 14:12
Zoffix Ah, OK. Just vaugely recall TimToady talking about these
Grimy Also ...^ is implemented but isn’t listed in docs.perl6.org/routine.html
Zoffix Grimy: no real way. Synopses are historical documents. Roast is the current spec. Some of the design desisions are archived ( e.g. github.com/rakudo/rakudo/blob/nom/...on-Plan.md ) 14:13
14:13 zakharyas left
AlexDaniel` Grimy: that's a doc issue that should be filed here: github.com/perl6/doc/issues/ 14:13
Grimy: it seems that it's linked but there's no special section for it docs.perl6.org/language/operators#..._operators
[Coke] (spec) there's also a load of those enshrined in NYI tickets in rt that may never be I. 14:14
Zoffix Grimy: looks like ... is in that list, but only the stub op shows up when clicked, not the sequence op 14:15
14:15 dakkar left
Zoffix And … ain't in the list 14:15
Grimy Well UTF-8 operators aren’t in the list, right? 14:16
Wait, some of them are.
14:17 dakkar joined
Grimy Thanks for the links, roast looks interesting. Is there a NYI ticket for prefix:<~^>? It seems simple enough to me and I’m wondering why it’s not there yet. 14:18
Altreus Zoffix: I thought perlpilot should ... maybe not though, perhaps
14:19 aindilis left, ryn1x joined
Zoffix huggable: tag nyi 14:20
huggable Zoffix, nothing found
Zoffix huggable: tag niy
huggable Zoffix, nothing found
Zoffix buggable: tag nyi 14:21
c'mon, robot. I ain't got all day
Grimy: if there is, it'd be one of these here: fail.rakudo.party/t/NYI
14:22 buggable joined, ChanServ sets mode: +v buggable
buggable New CPAN upload: JSON-Tiny-1.0.tar.gz by MORITZ cpan.metacpan.org/authors/id/M/MO/...1.0.tar.gz 14:22
[Coke] Zoffix: what is [MOLD] ?
seems to be on .party only, not on original ticket. 14:23
AlexDaniel` [Coke]: old tickets that were not touched for more than a year or so
Zoffix FWIW, I closed a couple of NYI last week 'cause they were not applicable anymore. So I'd say just 'cause there's an NYI ticket doesn't mean it's accepted that it should be implemented that way or implemented at all.
Grimy: we also have #perl6-dev where more dev eyes would see your questions
[Coke] anything that is NYI or RFC at this point needs some kind of sanity check, yes, zoffix++ 14:24
Grimy Alright, thanks!
[Coke] (even if there's a roast test; some of those tests predate rakudo)
Zoffix [Coke]: tickets that were last updated more than 60*60*24*365*2 seconds ago
Geth ecosystem: 974a161733 | (Moritz Lenz)++ | META.list
JSON::Tiny now lives on CPAN
[Coke] I mean, why not 365.24? :) 14:25
Zoffix :)
piojo Thanks for the help, everyone--I found the problem was basically "git checkout master"--that does nothing, since there is no "master" branch in rakudo ;) 14:29
Zoffix lol
piojo so I had an old commit checked out the whole time, and didn't notice.
Zoffix :)
piojo :D
AlexDaniel` well… maybe we should finally rename it
Grimy Why is the branch named «nom», anyway?
Zoffix Grimy: historical reasons. it started as "New Object Model" and eventually became the "master" 14:30
Grimy Ooh ok
piojo The funny thing is that it was in the first line of my bug report. It feels like I accidentally made a riddle
Zoffix I like "nom" as name, though I admit to on occasion typing "git checkout master" only to be told there ain't no master... 14:31
AlexDaniel` I'm doing that all the time
especially when working with moar or nqp
piojo I'm surprised I didn't notice, but building moar then nqp then rakudo, the process gets a little robotic 14:32
moritz I think we had an agreement that basically all core devs were fine with renaming nom to master
AlexDaniel` OK
moritz just nobody done it so far
AlexDaniel` then will do it after the release
moritz ++AlexDaniel`
piojo good idea
stmuk I think there may have been talk of using nom as a dev branch and master the release (or was it the other way round?) 14:33
Zoffix Yeah, something like that. My pushback against that is people will be more loosy with pushing breaking changes and the commits have to be sequential anyway due to reliance on MoarVM/nqp/roast 14:34
piojo That reminds me, do --gen-moar and --gen-nqp only use the latest release, or are there stable versions in between releases? 14:35
AlexDaniel` I expect it to be consistent across the trio, so if it's “master” for nqp and moar, then let's please have “master” in rakudo
Zoffix So "release" branch would be blocked anyway. The current system where people make scary changes in branches kinda more preferable IMO.
moritz piojo: they use whatever is in tools/build/NQP_REVISION
or MOAR_REVISION in nqp
piojo moritz: Is that bumped more frequently than the montly releases? 14:36
AlexDaniel` piojo: yes
[Coke] only slightly
piojo oh, great
pmurias Zoffix: if I remember correctly master was meant to be the nom commit that passes automated testing
Zoffix pmurias: yeah. So what happens when latest nom commit doesn't pass stuff? 14:37
The devs won't care as much, since bleed users are using master and aren't affected, and the rest of the devs now have a breaking commit in their workflow
[Coke] There's no rush to move it immediately post release; why don't we plan on having it hashed out in time to do so immediately after the next release?
stmuk talking of "release" .. is there going to be one? :) 14:38
moritz kinda agrees with Zoffix
at $work we had some kind of stabilization branch, and it didn't work out too well
Zoffix BTW, that branch proposal is documented here: github.com/rakudo/rakudo/blob/nom/...o-branches 14:39
moritz people did stuff there that wasn't tested by the same mechanism as the trunk, and so we had untested releases
and the "oh, people can just use master" mentality also took hold, at least somewhat 14:40
Geth mu: 05001f3b99 | (Zoffix Znet)++ | misc/perl6advent-2017/schedule
Start Advent 2017 Schedule

Call dibs on 1st and 23rd		 14:41
stmuk wouldn't that proposal also have to specify *which* roast branches pass?
14:41 lowbro left
moritz oh, it's this time of the year again? :) 14:42
Zoffix New blog post: "Rakudo Perl 6 Advent Calendar 2017 Call for Authors": rakudo.party/post/Rakudo-Perl-6-Ad...or-Authors
moritz Zoffix: do you plan to cover Perl 6 books in your review post, or should I do a separate one for that? 14:45
Zoffix moritz: wasn't planning. If you can do it, it'd be great, as despite owning many preview copies I haven't had time to read much of the available books... 14:46
Geth mu: 2a4cf374b5 | (Moritz Lenz)++ | misc/perl6advent-2017/schedule
Claim Dec 02 to talk about books
 14:47
Zoffix \o/ 14:48
.tell TimToady I called dibs on Advent post for the 1st. If you wanted to post something there instead, just take that spot: github.com/perl6/mu/blob/master/mi...chedule#L3 14:50
yoleaux Zoffix: I'll pass your message to TimToady.
AlexDaniel` stmuk: yes, when it's ready. We're putting some extra effort into making sure it's really good
Zoffix & 14:51
moritz traditionally, our benevolent dictator has been happy to take the last spot in the Advent series, if any
14:51 ryn1x left
Zoffix AlexDaniel`: stmuk Gonna take a crack at using `isatty` for the $*IN.getc macos issue in 8hr. If the fix fixes the things and doesn't negatively affect perf of my spectest, I'll merge it and do the bumps and I think that's the last thing blocking the release, no? There's RT#132343 ... looks really big 14:54
synopsebot RT#132343 [new]: rt.perl.org/Ticket/Display.html?id=132343 [REGRESSION] better-sched and other async improvement ecosystem fallout
Zoffix & for 8hrs :)
14:54 Zoffix left
AlexDaniel` Zoffix: yes. I'm looking into RT #132343 right now (but I've decided that most of the stuff in that ticket is a non-issue for the release) 14:55
14:59 abraxxa left
callyalater How can you use matching adverbs with grammars? 15:06
Like the :exhaustive adverb.
15:06 mr-foobar left
perlpilot Zoffix++ for jump starting the advent calendar this year. 15:07
15:07 zakharyas joined 15:08 mr-foobar joined
perlpilot callyalater: I don't think that you can use that particular one inside of a grammar. (I could be wrong though) 15:09
callyalater: as for "how", normally you just put the modifier in the grammar at the beginning of a rule or grouping construct 15:11
callyalater: token foo { :i foo } # match "foo" case insensitively
15:13 jcarty left 15:19 stux|RC-only left, wamba left 15:26 cgfbee left
Geth mu: 82e64bbd8a | (Nigel Hamilton)++ (committed using GitHub Web editor) | misc/perl6advent-2017/schedule
Update schedule
 15:26
15:27 lostinfog joined
callyalater perlpilot: Dang it. I was hoping that I could pass it in to the parse method... 15:28
15:33 cgfbee joined 15:35 domidumont left 15:37 mr-foobar left 15:38 stux|RC-only joined, mr-foobar joined 15:50 llfourn left, piojo_ joined
Geth doc/master: 5 commits pushed by (Patrick Spek)++, (Rafael Schipiura)++ 15:50
15:52 ryn1x joined 15:55 Grimy left 15:57 wander left 16:01 Actualeyes left 16:07 zakharyas left, mr-foobar left 16:08 setty1 joined, mr-foobar joined 16:09 zakharyas joined 16:12 domidumont joined 16:18 kyan joined 16:20 Kyo91_ joined 16:22 nightfrog left 16:26 ryn1x left, zakharyas left 16:27 wamba joined 16:29 domidumont left 16:30 domidumont joined 16:32 sproctor left
[Coke] gets annoyed at perl6/docs mix of make & perl6 site build again. 16:35
[Coke] suspects with new concurrency stuff, we could shove everything into p6 at this point, and do a better job of enforcing build deps than we do now. 16:36
16:36 mr-foobar left
Geth doc: 42e2bc0603 | (Will "Coke" Coleda)++ | xt/words.pws
Learn new word

Closes #1620		 16:37
16:37 ryn1x joined 16:38 Kyo91 joined 16:40 Kyo91_ left 16:41 dakkar left
Geth doc: a0afc5ce15 | (Will "Coke" Coleda)++ | htmlify.p6
use pod for long form comment;minor cleanups
 16:41
16:42 mr-foobar joined, ryn1x left 16:43 mson left 16:44 callyalater left 16:48 piojo_ left, jonas1 left 16:51 knight__ joined 16:53 napo1eon left
knight__ sssddddddssss 16:55
ops 16:56
Sorry
16:56 telex left
knight__ Hello, I am newbie. I have question about calling a function/sub.In mython I can do something liek def fun(a,b,c)... and call it fun(10, b=100, c=300) (for example) 16:57
How can I mix positional argument in perl6?
16:58 telex joined
knight__ i want call function for example fun($a, b=>100, c=>300). Of course I can do it with ($a, $:b, $:c), but what if I want something like fun(20, 30, c=>300)? 17:00
17:00 itaipu joined
timotimo you can have three optional positional parameters and the three nameds 17:00
knight__ How can I do it without definition in signature?
timotimo then depending on how many positionals were actually passed, you fill those into the variables you want 17:01
17:01 ChoHag left
timotimo m: sub fun($one?, $two?, $three?, :$a, :$b, :$c) { say $one // $a; say $two // $b; say $three // $c }; fun(20, 30, c => 300); fun(10, b => 100, c => 100) 17:02
camelia 20
30
300
10
100
100
timotimo m: sub fun($one?, $two?, $three?, :$a, :$b, :$c) { say $one // $a; say $two // $b; say $three // $c }; fun(10, b => 100, c => 300)
camelia 10
100
300
timotimo that was your actual example
it should even work to use defaults in your signature for this purpose
m: sub fun($one = $a, $two = $b, $three = $c, :$a, :$b, :$c) { say $one; say $two; say $three }; fun(10, b => 100, c => 300)
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$a' is not declared
at <tmp>:1
------> 3sub fun($one = 7⏏5$a, $two = $b, $three = $c, :$a, :$b, :$
timotimo nope, they have to be available in order 17:03
raschipi [15:02] <timotimo> m: sub fun($one?, $two?, $three?, :$a = $one, :$b = $two, :$c = $three) { say $a; say $b; say $c }; fun(10, b => 100, c => 300) 17:04
m: sub fun($one?, $two?, $three?, :$a = $one, :$b = $two, :$c = $three) { say $a; say $b; say $c }; fun(10, b => 100, c => 300)
camelia 10
100
300
raschipi m: sub fun($one?, $two?, $three?, :$a = $one, :$b = $two, :$c = $three) { say $a; say $b; say $c }; fun(10, 100, c => 300)
camelia 10
100
300
knight__ Thank you 17:05
raschipi defaults work, just need to rmember they are only available to the right of where they were declared.
m: sub fun(:$a, :$b, :$c, $one = $a, $two = $b, $three = $c ) { say $one; say $two; say $three }; fun(10, b => 100, c => 300) 17:07
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot put optional positional parameter $one after variadic parameters
at <tmp>:1
------> 3sub fun(:$a, :$b, :$c, $one = $a7⏏5, $two = $b, $three = $c ) { say $one; s
expecting any of:…
17:07 mr-foobar left 17:08 mr-foobar joined 17:09 ryn1x joined
Geth doc: 55a077d429 | (Will "Coke" Coleda)++ | htmlify.p6
Fix issue with proc async usage hanging on mac

mac's TMPDIR ends with a /, so we ended up with paths like ...//... ; the external program normalized the filenames to a single slash, so nothing matched.
Remove the command line option that allowed us to force it on the mac; it's now the default.
Fixes #1129		 17:11
synopsebot GH#1129 [closed]: github.com/rakudo/rakudo/pull/1129 Fix comment 17:12
17:13 ryn1x left 17:14 darkmorph left 17:22 raschipi left, ryn1x joined 17:26 virtualsue left 17:28 virtualsue joined 17:33 gdonald left 17:34 gdonald joined 17:35 domidumont left 17:37 domidumont joined 17:43 ChoHag joined 17:45 troys joined, domidumont left 17:47 llfourn joined 17:51 llfourn left 17:55 itaipu left 17:56 ryn1x left 17:57 Rawriful joined 18:01 domidumont joined, greppable6 left
bdmatatu In this section, is '.unique' necessary (after .invert)? docs.perl6.org/type/Hash#method_push 18:02
18:02 greppable6 joined, ChanServ sets mode: +v greppable6, itaipu joined
timotimo m: my %wc = 'hash' => 323, 'pair' => 322, 'pipe' => 323; say %wc.invert.unique.perl; 18:03
camelia (322 => "pair", 323 => "pipe", 323 => "hash").Seq
18:07 espadrine joined
bdmatatu m: my %wc = 'hash' => 323, 'pair' => 322, 'pipe' => 323; say %wc.invert.perl; 18:07
camelia (322 => "pair", 323 => "pipe", 323 => "hash").Seq
18:08 mr-foobar left 18:10 mr-foobar joined
timotimo i find that usage of invert rather strange 18:10
er 18:11
the usage of unique i mean
bdmatatu ok, me too. I'll go ahead and remove it. 18:14
Geth doc: b5a907758b | (Brian Duggan)++ | doc/Type/Hash.pod6
Remove extraneous .unique call
 18:17
synopsebot Link: doc.perl6.org/type/Hash
18:24 piojo_ joined 18:28 espadrine left 18:32 cdg joined 18:35 cdg_ left 18:36 darutoko left, cdg left, mr-foobar left 18:38 mr-foobar joined, troys is now known as troys_ 18:39 synopsebot left 18:40 synopsebot joined, ChanServ sets mode: +v synopsebot
knight__ ok, this works ideone.com/sa8j79 ,but I do not understand why this not ideone.com/UOcLOj Can somebody explain me... 18:40
thanks :-X 18:41
ugexe what is `Str $currency_?` supposed to mean? 18:42
i see 18:43
knight__ I am calling method fun
ugexe because your new method returns the result of say("..."), which is true 18:44
knight__ when 'new' method is not overrided It is works, but when I override it, fun method does not work
Ahm
18:44 aindilis joined
knight__ I understand 18:44
ugexe if you use .new you need to bless attributes
knight__ thank you!
I must check, what method/sub bless does :-) 18:45
18:47 domidumont left 18:50 ryn1x joined 18:52 synopsebot left, synopsebot joined, ChanServ sets mode: +v synopsebot, virtualsue left 18:54 virtualsue joined, ryn1x left
knight__ ugexe: Thank you! Now, understands what is going - I guess. So bless returns my class/objects 18:56
And I have pointer to my object and I can call method/function
I guess? 18:57
TLDR; Is something life self reference?
18:57 piojo_ left, darkmorph joined
knight__ *like 18:59
19:01 zakharyas joined 19:02 wamba1 joined, wamba left
AlexDaniel` knight__: like “self” ? 19:04
docs.perl6.org/language/terms#term_self 19:05
19:07 mr-foobar left
knight__ I mean it returns pointer = like pointer to self object 19:07
method bless
19:08 mr-foobar joined 19:17 darkmorph left
tyil anyone using vim willing te try out github.com/Tyil/vim-perl6/blob/add...l6.vim#L83 19:17
19:22 ChoHag left
AlexDaniel` tyil: is there any way to have **32 converted to ³²? 19:22
tyil not that I know, unless I add abbrev for **32
AlexDaniel` aww 19:23
tyil: π τ and 𝑒 seem to be missing 19:24
tyil good ones
I was considering to add an (atom) to turn into the atom icon for use with atomicints 19:25
AlexDaniel` yeah, although “atomic” is probably more expectable
tyil what are the ascii ways to represent the 2nd and 3rd char you gave" 19:26
?*
AlexDaniel` pi tau e
m: say (pi, tau, e)
camelia (3.14159265358979 6.28318530717959 2.71828182845905)
AlexDaniel` m: say (π, τ, 𝑒)
camelia (3.14159265358979 6.28318530717959 2.71828182845905)
tyil cool
AlexDaniel` tyil: maybe consider adding unicode minus 19:27
or maybe not 19:28
19:28 ryn1x joined
tyil www.fileformat.info/info/unicode/ch.../index.htm this one? 19:28
AlexDaniel` tyil: wait, you know about this page, right? docs.perl6.org/language/unicode_ascii
tyil (i actually stole this list from the atom plugin) 19:29
AlexDaniel` because I realize that some things are missing
like ∘ 19:30
tyil: is it possible to make ‘ "’ (space, double quote) convert to ‘ “’ and just " convert to ” ?
tyil you cant match spaces for iabbrev afaik, since spaces indicate the end of an abbreviation 19:31
19:33 ChoHag joined
tyil about that page, on the atomic ops 19:35
the last entry in the table isnt part of the table
AlexDaniel` the table seems to be broken yeah
tyil I can try to fix it after I add the ops to this list and push it 19:36
AlexDaniel` tyil++
19:36 mr-foobar left
tyil is there a atomic-dec-fetch missing from the table? 19:37
19:38 mr-foobar joined 19:39 itaipu left 19:41 konsolebox left
knight__ m: say (π**π) 19:42
camelia 36.4621596072079
tyil oh, its not particularly missing
TreyHarris AlexDaniel`: just noticed the term doc you just linked to has a definition for Rat terms that includes numbers with a dot, but not for <1/3> forms. So what's happening there? 19:43
m: say (<3>.WHAT, <1/3>.WHAT, <3/1>.WHAT, <3.2>.WHAT) 19:44
camelia ((IntStr) (Rat) (Rat) (RatStr))
TreyHarris (I'm asking, how does the constructor get dispatched?) 19:45
19:47 troys_ is now known as troys, llfourn joined
AlexDaniel` TreyHarris: it's not really dispatched, it's parsed like this directly 19:47
19:47 konsolebox joined
AlexDaniel` TreyHarris: for example: github.com/rakudo/rakudo/blob/nom/....nqp#L3637 19:47
TreyHarris AlexDaniel`: ah, that makes sense. Then it doesn't really have a proper place in the docs except in Rat's own, does it? Unfortunate, as it surfaces as a higher-level feature. 19:49
m: my $x = 1 / 3; say $x.WHAT, $x.nude
camelia (Rat)(1 3)
AlexDaniel` it's a little bit different for things like < 1/3 > 19:50
when you get RatStr
TreyHarris my $x = <1 /3); say $x.WHAT, $x.nude;
m: my $x = <1 /3); say $x.WHAT, $x.nude;
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in quote words; couldn't find final '>' (corresponding starter was at line 1)
at <tmp>:1
------> 3my $x = <1 /3); say $x.WHAT, $x.nude;7⏏5<EOL>
expecting any o…
AlexDaniel` m:
evalable6
TreyHarris m: my $x = <1 /3>; say $x.WHAT, $x.nude; 19:51
camelia No such method 'nude' for invocant of type 'List'. Did you mean any of these?
none
note

in block <unit> at <tmp> line 1
19:51 setty1 left
AlexDaniel` m: my $x = < 1/3 >; say $x.WHAT, $x.nude; 19:51
TreyHarris m: my $x = <1/3>; say $x.WHAT, $x.nude;
camelia (RatStr)(1 3)
(Rat)(1 3)
TreyHarris gah
AlexDaniel` :D
TreyHarris that's too weird
19:52 setty1 joined 19:53 llfourn left
TreyHarris m: my ($x, $y) = <1/3>, < 1/3 >; say $x == $y; say $x === $y; 19:53
camelia True
False
TreyHarris m: my ($x, $y) = <1/3>, < 1/3 >; say "$x and $y"
camelia 0.333333 and 1/3
TreyHarris ooooh, my brain hurts now
< n/d > form is clearly better than <n/d> for most uses where you'd use a literal, but it's not obvious unless you generally put spaces inside bracketing constructs (which in my experience most don't if the contents are very simple like 1/3) 19:54
19:56 ChoHag left 20:01 cdg joined 20:02 ryn1x left
geekosaur it's known we need a better/mrore obvious syntactic construct there, but coming up with one that isn't excessively verbose is difficult :( 20:04
AlexDaniel` well 20:05
m: say ⅓
camelia 0.333333
AlexDaniel` doesn't give a RatStr, but maybe it should?
20:05 cdg_ joined 20:07 mr-foobar left 20:08 cdg left, mr-foobar joined 20:11 okl joined, st_elmo joined
[Coke] m: say ⅓.^name; 20:13
camelia Rat
20:13 ChoHag joined
[Coke] It's a numerical literal, Rat seems fine. 20:13
*numeric
moritz m: say <⅓>.^name 20:14
camelia Str
[Coke] m: say <⅓/3>.^name
camelia Str
20:20 st_elmo left
TreyHarris m: say ³/₁₆ 20:24
camelia 5===SORRY!5=== Error while compiling <tmp>
Bogus postfix
at <tmp>:1
------> 3say ³/₁7⏏5₆
expecting any of:
infix
infix stopper
postfix
statement end
statement modifier
…
TreyHarris heh
so it only knows vulgar fractions 20:25
er, vulgar fraction ligatures
20:35 okl left 20:37 mr-foobar left 20:40 zakharyas left
[Coke] is anyone using --parallel on the doc build? seems like it's very likely to conflict with the highlighter. 20:41
20:41 mr-foobar joined 20:42 pecastro left
gfldex [Coke]: precomp is not threadsafe either 20:43
and that's used for caching
TreyHarris from #perl6 logs, it looks like moritz and timeless were looking at this back in June, but didn't get anywhere. I don't see anything since then. 20:46
timeless looks up
TreyHarris (the doc htmlifying parallelization, I mean)
timeless m: say ³/(₁₆) 20:47
camelia 5===SORRY!5=== Error while compiling <tmp>
Bogus postfix
at <tmp>:1
------> 3say ³/(₁7⏏5₆)
expecting any of:
infix
infix stopper
statement end
statement modifier
statement mo…
timeless [Coke]: anyway, TreyHarris is right, that sounds familiar 20:48
and i don't think i got very far w/ it. I suspect I still have access to the vm that hosts it
but it definitely isn't something i'm spending time on
(tomorrow's adventures are fighting ldap and jenkins)
[Coke] TreyHarris: ok. tempted to rip it out, but will try to fix it instead.
TreyHarris timeless: yeah, you mentioned trying to run it on 100 cores 20:49
timeless yep, that's gcc112, the box i still use :)
TreyHarris but Moore's law is still in effect, right? So between June and now ew've had a 100x speedup anyway. Just on chips that have had since June to evolve though, so if you rebooted a lot in the meantime you may not have as much speedup. ;-) 20:50
TreyHarris thinks this is how that works 20:51
timeless chuckles
actually there was a recent argument about moore's law
some pointed out that mostly we're getting more of a coore's law -- we can double computing power by doubling die count or something 20:52
TreyHarris "recent" meaning every year since it was proposed? yes.
timeless no
an argument about whether we're still properly shrinking, or whether we're just gaining by adding cores
i think intel came back and showed they could still gain by shrinking
20:52 mempko left
timeless > gcc112 2TB 2x10x8x3.42 GHz IBM POWER8 / 256 GB RAM / IBM POWER System S822 / CentOS 7 ppc64le 20:53
-- it's actually 160 cores, but to be polite, I try to avoid using more than 100 :-) 20:54
[Coke] (docs) anyone mind if I remove the highlight path that doesn't use proc::async? i seem to recall we only needed it for the mac.
timeless back then, when I wasn't on gcc112, i was probably on a mac ;-)
20:54 nadim left
timeless (these days i have a w10 laptop at home, and an mbp at work) 20:55
20:55 perlpilot left
timeless anyway, i'm off, happy iterating 20:55
TreyHarris sorry, was kidding again. No, in all seriousness, I think Moore's Law "informal" form, about the speed of a general-purpose consumer CPU to carry out however many ops/sec, has already died. but we are still cramming transistor density at predicted rates right now--we won't hit fundamental issues with unknown physics for at least a few more years, I think. 20:56
s/however many/however many serial/ # important distinction 20:57
20:59 espadrine joined 21:04 setty1 left 21:05 pecastro joined, ryn1x joined 21:08 mr-foobar left 21:11 mr-foobar joined 21:12 pecastro left, pecastro joined 21:14 nadim joined 21:18 notostraca is now known as TEttinger 21:21 raiph left 21:28 wamba1 left 21:31 Hotkeys joined 21:32 troys is now known as troys_ 21:37 mr-foobar left 21:38 margeas left 21:39 ryn1x left, mr-foobar joined 21:43 ryn1x joined, margeas joined 21:46 kyan left, lostinfog left 21:47 ryn1x left 21:50 llfourn joined 21:55 llfourn left 21:59 troys_ is now known as troys 22:05 MasterDuke joined 22:08 mr-foobar left 22:11 mr-foobar joined, Rawriful left 22:17 ryn1x joined 22:21 ryn1x left 22:31 pmurias left 22:36 MasterDuke left, callyalater joined 22:37 rindolf left 22:38 mr-foobar left 22:39 mr-foobar joined
tyil AlexDaniel`: I've looked at the broken table on the atomic ops, but in the source it looks correct 22:39
callyalater m: sub infix:['<<'] (Str $s, Int $n) { say "$s, $n" }; "Name" << 42; 22:42
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of << to do left shift; in Perl 6 please use +< or ~<
at <tmp>:1
------> 3 $s, Int $n) { say "$s, $n" }; "Name" <<7⏏5 42;
callyalater m: sub infix:['<<'] (Str $s, Str $n) { say "$s, $n" }; "Name" << "42";
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of << to do left shift; in Perl 6 please use +< or ~<
at <tmp>:1
------> 3 $s, Str $n) { say "$s, $n" }; "Name" <<7⏏5 "42";
callyalater m: sub infix:['<>'] (Str $s, Int $n) { say "$s, $n" }; "Name" <> 42; 22:43
camelia Name, 42
callyalater m: sub infix:['!~'] (Str $s, Int $n) { say "$s, $n" }; "Name" !~ 42;
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of !~ to do negated pattern matching; in Perl 6 please use !~~
at <tmp>:1
------> 3$s, Int $n) { say "$s, $n" }; "Name" !~ 7⏏0542;
geekosaur gah 22:48
22:48 ryn1x joined 22:52 mempko joined 22:54 raschipi joined 22:55 nadim left, nadim_ joined 22:59 Cabanossi left 23:01 Cabanossi joined 23:05 knight__ left 23:09 ryn1x left 23:17 darkmorph joined 23:37 Herby_ joined
Herby_ o/ 23:37
APic \o 23:38
23:38 mr-foobar left 23:39 raiph joined 23:40 cdg joined, cdg left 23:41 mr-foobar joined 23:44 cdg_ left
raiph .tell moritz comment for you at stackoverflow.com/questions/468978...1_46920109 thanks 23:44
yoleaux raiph: I'll pass your message to moritz.
23:44 troys is now known as troys_ 23:54 callyalater left 23:56 MasterDuke joined 23:57 ryn1x joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!