»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
Michael_S I'm confused by this difference: 01:02
p6: sub x1() { [1, 2, 3] }; my @a = x1(); say @a;
camelia [1 2 3]
Michael_S sub x2($k, %h) { if (%h{$k}) { %h{$k} } else { %h{$k} = [1, 2, 3]; %h{$k} } }; my %h; my @a2 = x2("k", %h); say @a2; 01:03
evalable6 [[1 2 3]]
timotimo is it because you get an extra scalar container around the [1, 2, 3] from getting it back out of the hash? 01:04
m: sub x3() { $[1, 2, 3] }; my @a = x3(); say @a; 01:05
camelia [[1 2 3]]
Michael_S How do I bypass that scalar container? I want the array, not an array with the actual array that I want in the first element.
timotimo you can get it out of the hash with an extra <> at the end
%h{$k}<> 01:06
that will strip the scalar container on the way out
Michael_S sub x2($k, %h) { if (%h{$k}) { %h{$k}<> } else { %h{$k} = [1, 2, 3]; %h{$k}<> } }; my %h; my @a2 = x2("k", %h); say @a2;
evalable6 [1 2 3]
geekosaur or:
sub x2($k, %h) { if (%h{$k}) { %h{$k} } else { %h{$k} = [1, 2, 3]; |%h{$k} } }; my %h; my @a2 = x2("k", %h); say @a2;
evalable6 [1 2 3]
timotimo m: my %h; %h{"with"} = [1, 2, 3]; %h{"without"} := [1, 2, 3]; dd %h; dd %h{"with"}; dd %h{"without"}
camelia Hash %h = {:with($[1, 2, 3]), :without([1, 2, 3])}
Array %h = $[1, 2, 3]
Array element = [1, 2, 3]
timotimo another way to not have the scalar container 01:07
Michael_S The switch from %h{key} = value to %h{key} := value does it? The ':='? Thanks. Thank you to timotimo and geekosaur for the fast responses, I appreciate it. (And thanks to good old camelia and evalable6 :D ) 01:08
timotimo that's right
= assigns into something, := binds "over the existing container"
geekosaur there's more than one way to do it, but each has its own particular behaviors 01:09
the := isn't appropriate if you need to assign to that hash element later, for example
timotimo true 01:10
Michael_S I will keep that in mind too. 01:11
Geth doc: 912fdf79bb | (JJ Merelo)++ | doc/Type/IO.pod6
Fixes link refs #2385
01:25
synopsebot Link: doc.perl6.org/type/IO
Geth doc: d4b482b516 | (JJ Merelo)++ | 2 files
Adds dependence to avoid test error

Also adds links and some explanations. This closes #2383 closes #2385
01:39
Zoffix Michael_S: there's an extensive article on containers if you have any interest: perl6advent.wordpress.com/2017/12/02/ 01:52
And this one, targeted for those with Perl 5 backround: opensource.com/article/18/8/containers-perl-6 01:53
k-man i don't understand the difference between Q and q for quoting strings 03:24
Q is literal and q is escaped strings? but I don't understand what that difference is 03:25
geekosaur m: say Q'x\yz 03:26
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3say Q'x\7⏏5yz
expecting any of:
postfix
geekosaur er
m: say Q'x\yz'
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3say Q'x\7⏏5yz'
expecting any of:
postfix
k-man oh, is it that you can include \ to escape in escaped strings?
geekosaur duh
right
k-man ok, i get it now thanks geekosaur
geekosaur \is literal in Q, allows escapingin q"
k-man thank you geekosaur 03:27
geekosaur and I still type too fast for this damned keyboard :/
ryn1x :m say Q 'fails to escape \' quote' 04:03
m: say Q 'fails to escape \' quote'
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3say Q 'fails to escape \'7⏏5 quote'
expecting any of:
infix
infix stopper
postfix
statement end
s…
ryn1x m: say Q :q 'escape \' quote' 04:04
camelia escape ' quote
ryn1x ^ k-man 04:06
ryn1x :m Q 'x\\yz' 04:30
m: Q 'x\\yz'
camelia WARNINGS for <tmp>:
Useless use of constant string "x\\\\yz" in sink context (line 1)
ryn1x m: say Q 'x\\yz' 04:31
camelia x\\yz
ryn1x m: say Q:q 'x\\yz'
camelia x\yz
buggable New CPAN upload: Perl6-Bayes-Learn-0.1.1.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.1.tar.gz 04:53
holyghost ^-- There's now an Action class for such loss functions and a minimax method for a Strategy class. Everything is based on Populations of probabilities and Distributions of them 04:55
yoleaux 15 Oct 2018 16:59Z <Zoffix> holyghost: you have a problem with your modules: colabti.org/irclogger/irclogger_log...10-15#l556
holyghost .tell Zoffix I'll fixify later on, I have to read up 05:01
yoleaux holyghost: I'll pass your message to Zoffix.
holyghost .m "Vampires do exist" 05:05
an example hypothesis for an RPG with Bayesian inference 05:06
Think classes :-) 05:07
Geth doc: f76a5f1896 | (Richard Hainsworth)++ (committed using GitHub Web editor) | doc/Language/syntax.pod6
Update syntax.pod6

It seemed at first that C<unit> added something, when in fact unit is just telling the compiler that the rest of the file is the body of the package.
06:08
synopsebot Link: doc.perl6.org/language/syntax
Geth Pod-To-HTML: 77f0af09a7 | (Richard Hainsworth)++ (committed using GitHub Web editor) | t/010-basic.t
Update 010-basic.t

Eliminated sub twrap at end of this file. It is not called. The tests all run when this code is eliminated.
06:20
finanalyst I'm working on an issue in Pod::To::HTML and now I'm trying to run the tests using prove -I lib -e ... But Pod::To::HTML is already installed, so the test files seem to be taking use ... from the installed version, not the new one I am debugging. What should be my work flow? 06:31
holyghost .tell Zoffix I have blurb.meta in my CPAN package directory so Mathx::Stat and Bayes::Leanr should be indexed. I am not sure the source-urls are correct. 06:51
yoleaux holyghost: I'll pass your message to Zoffix.
buggable New CPAN upload: Perl6-Mathx-Stat-0.1.2.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.2.tar.gz 06:53
New CPAN upload: Perl6-Bayes-Learn-0.1.2.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.2.tar.gz
lizmat clickbaits p6weekly.wordpress.com/2018/10/15/...tom-for-6/ 09:18
robertle is there a way to distribute and look up files that are not perl modules in some consistent fashion? I want to e.g. bundle a .csv file with data with my application and of course access it at runtime. I can of course just put it in a path somewhere but then I can't distribute with zef anymore, and I guess it would make it non-portable to windows or such 09:35
jnthn robertle: Look up %?RESOURCES 09:37
lizmat that's what docs.perl6.org/language/variables#...FRESOURCES was invented for
although that could well need some extra documentation :-(
holyghost thx, I would probably need %?RESOURCES for graphics 09:39
I didn't know that
nwellnhof p6: $_ = ^4; say .[^2], .[*-2..*]; say .[^2] ~~ .[*-2..*]; 09:59
camelia (0 1)(2 3)
True
nwellnhof p6: $_ = ^4; say .[^2], .[*-2..*]; say .[^2] ~~ .[*-2..3]; 10:00
camelia (0 1)(2 3)
False
nwellnhof p6: $_ = ^4; say .[^2], .[*-2..*]; say .[^2] ~~ .[2..*];
camelia (0 1)(2 3)
False
nwellnhof The result of the first smartmatch looks like a bug. 10:01
tyil can I use a LEAVE block to close DBIish's database handle, but only if the handle was succesfully created? 10:06
also, is LEAVE the best phaser to use for these kinds of tricks to begin with? 10:07
jnthn tyil: Maybe with `LEAVE $dbh.?disconnect`?
nwellnhof p6: $_ = (1, 1); say .[0] == .[1]; say .[0] ~~ .[1]; 10:08
camelia True
False
jnthn tyil: I tend to factor this out into something like a with-db method, so I can say like `self!with-db: -> $dbh { $dbh.do(...) }` and so forth
tyil jnthn: I think `with` should work here too, but I'm not completely sure
.disconnect with $dbh
jnthn tyil: Yeah, that's probably better
tyil I'm not working from within a class in this context, so self! is not available 10:09
nwellnhof p6: my $a = (1, 1); say $a[0] ~~ $a[1]; 10:10
camelia True
tyil I just have a sub, and I thought the code would be nicer if I could just end the sub with $sth.allrows.map(), and make that the implicit return value as well
instead of having to assign that to a var, close/disconnect and then return the var
nwellnhof p6: $_ = (1, 2); say .head ~~ .tail; 10:15
camelia True
nwellnhof my $a = (1, 2); say $a.head ~~ $a.tail;
evalable6 False
nwellnhof p6: my $a = (1, 2); say $a.head ~~ $a.tail;
camelia False
nwellnhof There's something weird going on with smartmatch and the topic variable. 10:16
jnthn Yes, ~~ sets $_ to the result of evaluating the LHS while it evaluates the RHS 10:19
That's why one can do things like $file ~~ .e
tyil: Yeah, sure, can do it as a sub too :) 10:20
nwellnhof I see, thanks for the explanation.
Zoffix .tell holyghost for source URL just set the URL of your GitHub repo. PAUSE rewrites it to proper CPAN URL and IIRC modules.perl6.org uses the original's GitHub URL to figure out the repo URL 10:34
yoleaux 05:01Z <holyghost> Zoffix: I'll fixify later on, I have to read up
Zoffix: I'll pass your message to holyghost.
06:51Z <holyghost> Zoffix: I have blurb.meta in my CPAN package directory so Mathx::Stat and Bayes::Leanr should be indexed. I am not sure the source-urls are correct.
Zoffix .tell holyghost Your `provides` section is still broken. Looks like you listed all keys as "Mathx::Stat::" without the last portions of the names. Also your modules are still not indexed, possibly because you named your archives with "Perl6-" at the start. Check out App::Mi6 module. It has CPAN uploading functionality and it will build the proper archive for you 10:37
yoleaux Zoffix: I'll pass your message to holyghost.
Zoffix .tell finanalyst give -Ilib to the executable prove runs instead of prove itself: prove -e "perl6 -Ilib" -vr t/ 10:39
yoleaux Zoffix: I'll pass your message to finanalyst.
tyil how to denote a sub's return type as being a Positional where each element will be a Str 10:56
moritz have you tried --> Postional[Str] 10:59
?
tyil no
moritz m: sub f(--> Positional[Str]) { Array[Str].new('b', 'b' }; f()
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in argument list; couldn't find final ')' (corresponding starter was at line 1)
at <tmp>:1
------> 3itional[Str]) { Array[Str].new('b', 'b' 7⏏5}; f()
moritz m: sub f(--> Positional[Str]) { Array[Str].new('b', 'b') }; f()
camelia ( no output )
moritz m: sub f(--> Positional[Str]) { Array.new('b', 'b') }; f()
camelia Type check failed for return value; expected Positional[Str] but got Array ($["b", "b"])
in sub f at <tmp> line 1
in block <unit> at <tmp> line 1
moritz might or might not be what you want 11:00
tyil hmm
I'm getting that last error in the context that I want to use it 11:01
holyghost .tell Zoffix I don't have access to App:Mi6 because of botcherous SSL, check again if the source-url is correct
yoleaux 10:34Z <Zoffix> holyghost: for source URL just set the URL of your GitHub repo. PAUSE rewrites it to proper CPAN URL and IIRC modules.perl6.org uses the original's GitHub URL to figure out the repo URL
holyghost: I'll pass your message to Zoffix.
10:37Z <Zoffix> holyghost: Your `provides` section is still broken. Looks like you listed all keys as "Mathx::Stat::" without the last portions of the names. Also your modules are still not indexed, possibly because you named your archives with "Perl6-" at the start. Check out App::Mi6 module. It has CPAN uploading functionality and it will build the proper archive for you
holyghost .tell Zoffix I don't need a github repo, because it's broken sometimes to login, so I want clean CPAN usage, it's more perlwise 11:02
yoleaux holyghost: I'll pass your message to Zoffix.
buggable New CPAN upload: Bayes-Learn-0.1.3.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.3.tar.gz 11:03
New CPAN upload: Mathx-Stat-0.1.3.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.3.tar.gz
holyghost .tell Zoffix check again if you please 11:05
yoleaux holyghost: I'll pass your message to Zoffix.
jkramer gist.github.com/throwaway-30964/39...4b1d873d15 11:15
I'm getting "Cannot assign to an immutable value" for the assignment in line 13, shouldn't that work?
holyghost .tell Zoffix I would really like to update the perl6 CPAN package parser, if you would pass me his URL/email I'd be glad to help, I'd gunzip and parse instead of otherwise 11:21
yoleaux holyghost: I'll pass your message to Zoffix.
scimon There's no plans to remove the Git Based module system from zef are there? I just think it is a great basis for local modules. Whilst I'm a fan of open source code having company specific modules is a useful thing. If we can have an easy to setup and use DarkPan I'd be happy. 11:37
(Especially if it didn't actually require running a local CPAN. 11:38
buggable New CPAN upload: Bayes-Learn-0.1.4.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.4.tar.gz 11:43
New CPAN upload: Mathx-Stat-0.1.4.tar.gz by HOLYGHOST cpan.metacpan.org/authors/id/H/HO/...1.4.tar.gz
leont Filed yesterday's issue 12:04
Zoffix scimon: no, no plans to remove. 12:06
yoleaux 11:01Z <holyghost> Zoffix: I don't have access to App:Mi6 because of botcherous SSL, check again if the source-url is correct
11:02Z <holyghost> Zoffix: I don't need a github repo, because it's broken sometimes to login, so I want clean CPAN usage, it's more perlwise
yoleaux 11:05Z <holyghost> Zoffix: check again if you please 12:06
11:21Z <holyghost> Zoffix: I would really like to update the perl6 CPAN package parser, if you would pass me his URL/email I'd be glad to help, I'd gunzip and parse instead of otherwise
Zoffix holyghost: most modern perl packages include a URL to a repo with the source. Makes it easier to report problems
CIAvash jkramer: $_ is bound to an integer(by `for ^10`). so you are assigning something to 0. 12:07
Zoffix holyghost: I don't have contacts for the CPAN package parser, but mst likely knows how to reach them. lizmat might know too
holyghost Zoffix: I'm sorry about the .tell usage you're online now
I used the tarball in both my packages
I am going to write some package source compilers probably, the thing's too stupid 12:08
If only I find the time with my statistics packages
Zoffix holyghost: yeah, I know. So there's a tarball on CPAN and a repo with code somehwere too, usually. So you put repo as source URL and when you upload to CPAN the PAUSE parser changes the source-url to CPAN URL in the meta file it generates. So there are two separate meta files and modules.perl6.org can use the original one to figure out the repo 12:09
holyghost I understand
Zoffix Like here: modules.perl6.org/dist/Bailador:cpan:UFOBAT it's a CPAN package, but there's a "Dist's repo" button that leads to github 12:10
holyghost I only need a doc for modules.perl6.org, I don;t like test as it's open source
s/test/tests
ok 12:11
Zoffix holyghost: oh you mean modules.perl6.org's CPAN handler?
holyghost of docs yes 12:12
Zoffix "of docs"?
Zoffix doesn't understand
Zoffix The CPAN handling on mp6.org is here github.com/perl6/modules.perl6.org...#L128-L190 and here github.com/perl6/modules.perl6.org...ce/CPAN.pm 12:13
holyghost There's a html page written to modules.perl6.org for some doc files ?
Zoffix doc files of what?
holyghost I don;t remember
pod maybe
Something said there was HTML output in docs.perl6.org/language/modules#Pr...the_module 12:14
jkramer Is there a nicer way to do for ^10 without assigning the number to $_, other than for ^10 -> $ { ... ? 12:15
Zoffix "The README.md file is a markdown-formatted text file, which will later be automatically rendered as HTML by GitHub."
jkramer: no, but you could use $OUTER::_ inside the loop to reference the outer $_ instead of the one used by the loop
holyghost that's what I was talking about 12:16
Zoffix m: $_ = 'meows'; $OUTER::_.say for ^3
camelia meows
meows
meows
jkramer Hmm that doesn't make it nicer though, I wanted to use $_ for convinience :)
holyghost Zoffix: For general usage we should use a README.txt
nor only github but Windows too 12:17
Zoffix holyghost: right, but that's for p6c GitHub-hosted ecosystem. But if you have README.md in your tarball, the modules.perl6.org will render it too like here: modules.perl6.org/dist/Bailador:cpan:UFOBAT
use markdown and save it in README.md
jkramer do { ... } xx 10 seems to work but not sure if that's better or worse than for ^10 -> $ :) 12:18
holyghost README.html, from README.txt
that's also why I want to hack the parser
README.md can still be in use 12:19
Zoffix Doesn't make sense for me to make README.html from README.txt...
holyghost Zoffix: on the docs server 12:20
Zoffix .txt signifies it's plain text, not HTML 12:20
m: $_ = 'meows'; .say xx 10
camelia meows
meows
meows
meows
meows
meows
meows
meows
meows
meows
holyghost It's a div probably in HTML
Zoffix jkramer: ^ do {} is often optional
holyghost IF anything gets lost we can retain it from an htmldump to text 12:21
Geth doc: 38de31936d | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Language/modules.pod6
Document mp6o renders for CPAN
synopsebot Link: doc.perl6.org/language/modules
CIAvash jkramer: what about something like `while %ng{$last} -> $_` 12:23
jkramer CIAvash: I don't think that works for inline-while 12:43
CIAvash jkramer: what do you mean by inline? 12:44
jkramer m: .say while (^2).pick -> $_ 12:45
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected block in infix position (missing statement control word before the expression?)
at <tmp>:1
------> 3.say while (^2).pick7⏏5 -> $_
expecting any of:
infix
i…
tyil trying to use DBIish, I'm getting the following error: DBIish: DBDish::Pg needs 'pq', not found 12:53
zef search pq doesnt yield anything
searching my distro repos doesnt return a match either (substring match that doesnt seem to be related)
moritz tyil: it's probably a system library 12:55
not a perl module
tyil but nothing seems to provide it on my system 12:56
so what is it exactly looking for? the error is very LTA (especially since I would expect it to be a perl module by default)
araraloren tyil what is Pg ? 12:59
q
tyil if I knew I wouldn't have to ask, araraloren
araraloren Your system should already have PostgreSQL server installed and running. 13:01
that's what I found
Maybe a command tools of PgSQL 13:02
CIAvash jkramer: then try `while $_ := %ng{$last}` 13:08
jkramer CIAvash: That works, thanks! 13:10
AlexDaniel holyghost: perl6-compiler is a wrong mailing list :) 13:13
perl6.org/archive/lists/ 13:14
maybe perl6-users is a better choice? Not sure 13:15
holyghost I cannot get onto perl6-users because of majrodomo 13:16
s/majrodomo/majordomo
holyghost Larry told me to put it on perl6-compiler 13:17
holyghost He told me that perl-users is less for packages but for perl6 blurb usage 13:20
holyghost I hope that's ok 13:21
masak I know I don't blog so often these days :/ 13:24
but I just found this, that I wrote a while ago: github.com/masak/007/issues/345
I would say if it weren't an issue on github, it could've been a blog post :)
tyil araraloren: but I dont want to run a server just so I can run a client, that'd be silly 13:43
holyghost My only concern is that you could use Bayesian Inference for conditional compilatin in MoarVM 14:13
So you could generate several parsing techniques
I might write a C file for it and post it 14:14
It'd be an option for the perl6 command line
-opt:bayes
It's more difficult than it sounds. My Bayesian Inference C file should become refactored for such things 14:17
inference == pulsing
buggable New CPAN upload: Sparrowdo-Cordova-OSx-Build-0.0.7.tar.gz by MELEZHIK modules.perl6.org/dist/Sparrowdo::C...n:MELEZHIK 14:30
scimon Zoffix: Good to know. I may get round to writing an article on managing a Perl6 DrakPAN. (In a couple of weeks, after the wedding). 15:05
El_Che scimon: will it work for github modules? 16:14
scimon El_Che: Should do. My plan is to document what you need to do to run your own local github/gitlab based set of packages. 16:32
uzl Hello, #perl6! 16:38
yoleaux 11 Oct 2018 06:33Z <jmerelo> uzl: received, thanks!
uzl p6weekly mentioned a new intro page for perl6.org but it seems it's back to normal. anybody knows what happened? 16:40
AlexDaniel uzl: github.com/perl6/doc/commit/13739c...77948fc77f 16:43
AlexDaniel .tell jmerelo is there a ticket for rerevert of the commit that was reverted in D#2374 ? 16:44
yoleaux AlexDaniel: I'll pass your message to jmerelo.
synopsebot D#2374 [closed]: github.com/perl6/doc/issues/2374 [bug][site] Most search results point to pages that do not exist
uzl AlexDaniel: Oh, I see. Thanks!
AlexDaniel .tell I got to know about it in colabti.org/irclogger/irclogger_log...10-16#l581 and was somewhat surprised 16:45
yoleaux AlexDaniel: I'll pass your message to I.
ryn1x Does rakudo with the JVM backend output class files? 16:48
AlexDaniel ooooooooh 17:05
AlexDaniel (#perl6-dev) 17:08
AlexDaniel .tell jmerelo I got to know about it in colabti.org/irclogger/irclogger_log...10-16#l581 and was somewhat surprised 17:30
yoleaux AlexDaniel: I'll pass your message to jmerelo.
Geth doc: e276c76e2b | (JJ Merelo)++ | META6.json
Eliminates tabs
18:06
Geth doc: 28ffef7fda | (JJ Merelo)++ | util/travis-build.sh
Forces installation of IO::Socket::SSL
18:12
leont wonders why that «--/tap-harness» is necessary, that suggests dealing with invalid TAP 18:17
Geth doc: bce2cc8798 | (JJ Merelo)++ | doc/Language/glossary.pod6
Eliminates outdated list of bots

And points to a single source of truth in the community page. Closes #2380.
18:37
synopsebot Link: doc.perl6.org/language/glossary
Xliff m: say 1024 * 12 18:53
camelia 12288
isBEKaml voice? Is everyone voiced now? 19:13
Ohhai!
leont Anti-spambot measure 19:16
isBEKaml 11,99leont99,99: Ah, okay 19:17
Xliff jnthn? 20:34
jnthn Xliff: o/ 20:40
Geth doc: MorayJ++ created pull request #2386:
Concurrency
20:45
Xliff .tell jnthn I did manage to get signals straightened out in a generic way. Use of Supplies has been preserved, but it's still kinda hackish. See: github.com/Xliff/p6-GtkPlus/commit...e7db5c0b45 20:56
yoleaux Xliff: I'll pass your message to jnthn.
Xliff Doh!
That took me a long time to write.
jnthn: \o 20:57
jnthn Xliff: Left a couple of thoughts on the commit 21:03
yoleaux 20:56Z <Xliff> jnthn: I did manage to get signals straightened out in a generic way. Use of Supplies has been preserved, but it's still kinda hackish. See: github.com/Xliff/p6-GtkPlus/commit...e7db5c0b45
Xliff jnthn: Thanks!
Xliff jnthn: One issue I've been wondering about is how to make it so I don't have to have separate signal hashes for each object type. 21:11
I thought it would work well enough to have a %!signals, and have every descendant use that, however if an attribute is not defined in a role, it isn't useable, even by descendants. 21:12
Is there a way to accomplish that in Perl6?
IE - Have an attribute in one role be used by another (or at least the equivalent behavior). 21:13
jnthn Xliff: Can probably workaround it by writing a private method to get hold of the attribute 21:16
Xliff m: role A { has $!a; }; role B does A { method b { say $!a; }; }; my $b = 1 but B; say $b.b 21:26
camelia 5===SORRY!5=== Error while compiling <tmp>
Attribute $!a not declared in role B
at <tmp>:1
------> 3role B does A { method b { say $!a; }; }7⏏5; my $b = 1 but B; say $b.b
expecting any of:
horizontal whitespace
Xliff m: role A { has $!a; method a { $!a }; }; role B does A { method b { say self.a; }; }; my $b = 1 but B; say $b.b 21:27
camelia (Any)
True
Xliff m: role A { has $!a = 2; method a { $!a }; }; role B does A { method b { say self.a; }; }; my $b = 1 but B; say $b.b
camelia 2
True
Xliff m: role A { has $.a is rw = 2; method a { $!a }; }; role B does A { method b { say self.a; }; }; my $b = 1 but B; say $b.b 21:28
camelia 2
True
Xliff m: role A { has $.a is rw = 2; method a { $!a }; }; role B does A { method b { say self.a; }; }; my $b = 1 but B; say $b.b; $b.a = 3; $b.a.say
camelia 2
Cannot modify an immutable Int (2)
in block <unit> at <tmp> line 1

True
Xliff m: role A { has $.a is rw = 2; method a is rw { $!a }; }; role B does A { method b { say self.a; }; }; my $b = 1 but B; say $b.b; $b.a = 3; $b.a.say
camelia 2
True
3
21:29
Xliff Yeah. I should have thought of that.
Geth doc/master: 10 commits pushed by MorayJ++, timo++
review: github.com/perl6/doc/compare/bce2c...b93ca9c1ab
21:48
buggable New CPAN upload: Cro-H-0.0.1.tar.gz by GARLANDG cpan.metacpan.org/authors/id/G/GA/...0.1.tar.gz 22:50
buggable New CPAN upload: Cro-H-0.0.2.tar.gz by GARLANDG modules.perl6.org/dist/Cro::H:cpan:GARLANDG 23:50