pharv hey all, i just wanna say thank you for your dedication and commitment to making perl6 a success... back in 2012, when i was doing perl5 development, my coworkers convinced me that perl6 was dead and that switching languages would be a good career move... i then took a job as a php/symfony developer and have been doing that since... although i enjoyed symfony development and was even pleased at the introduction of php7, i felt like 04:26
it was still lacking in some areas... i just recently started reading up on perl6 again and was pleasantly surprised at the tremendous progress that has been made and how modern the language really is... so, thank you
llfourn pharv++ # here,here 05:29
yoleaux 28 Jul 2017 16:54Z <nadim> llfourn: I founf what the problem was in my code; I have unit module X ; use Y ; use Z ; sub want_this_sub{} ; When unit module X is after the 'use's it works, as if require did not put it in the right unit.
AlexDaniel pharv: \o/ thank you
llfourn .tell nadim Thanks for suffering to figure out that problem. Did you manage to make an RT? It still sounds like a bug. 05:32
yoleaux llfourn: I'll pass your message to nadim.
pharv i would love to give back to the community, but C is just way too low level for me, so i'm gonna have to find some packages to help out with... i'm considering Bailador, since it seems to have a lot of contributors and momentum 05:33
AlexDaniel pharv: but it's not just C. Most of Rakudo is written in Perl 6 05:41
pharv oh, then maybe i can help :)
nine m: EVAL q/system "perl6", "-e", q{my $socket = IO::Socket::INET.new(:listen);}/, :lang<Perl5>; 07:04
camelia Invocant of method 'Bridge' must be an object instance of type 'Int', not a type object of type 'Int'. Did you forget a '.new'?? in block <unit> at -e line 1??
nine Now that I call a weird error. Ignore the Perl 5 bits, those are just to get around the restricted setting in camelia.
[Tux] This is Rakudo version 2017.07-86-gf083cfc6a built on MoarVM version 2017.07-263-gb161c851 08:24
csv-ip5xs 2.996
test 14.123
test-t 4.916 - 4.960
csv-parser 15.462
Geth rakudo/nom: 6d3ba60c11 | (Elizabeth Mattijsen)++ | src/core/Any.pm
Improve info on bare "dd"

Before, it would just say the type and name of block being executed, which isn't too useful with multis. It now also adds the signature, so that you can more easily figure out which candidate was called.
12:20
lizmat m: multi sub a() { dd }; multi sub a($a) { dd }; a
camelia sub a?
lizmat m: multi sub a() { dd }; multi sub a($a) { dd }; a 42
camelia sub a?
lizmat will become "sub a()" and "sub a($a)" 12:21
timotimo nice 12:26
samcv i'm still getting a failing 'require' test, seems to have been failing for weeks now 13:38
weird it's not failing when i run it manually 13:39
jnthn samcv: Me also fwiw 13:40
I think it also passed reliably if make install was done prior to make spectest, but I may be misremembering 13:41
lizmat fwiw, I've seen the require test flapping *after* a make install 13:42
samcv oh and i'm adding MVM_string_codes to make .codes faster 13:46
do we want the nqp op called nqp::codes or nqp::codes_s. probably just codes?
it's not currently added to nqp
but is in MVM's oplist even though it just returns the number of graphemes currently
jnthn samcv: just nqp::codes I think 13:48
Geth nqp: 393eb01659 | (Samantha McVey)++ | tools/build/MOAR_REVISION
Bump MoarVM and add new nqp::codes op

Gets the number of codepoints in the string.
Changes brought in by MoarVM bump: abc38137 Fix MVM_string_compare to support deterministic comparing of synthetics b161c851 be safe about marking NULL call graph nodes ... (9 more lines)
14:54
¦ nqp: version bump brought these changes: github.com/MoarVM/MoarVM/compare/2...-gabc38137
ugexe can someone give me higher privs on RT? 15:58
Zoffix ugexe: what's your login? 17:26
ugexe: nv, ugexe cpan.org right? You should have higher privs now 17:31
s/nv/nm/;
m: 42 <= Int 17:37
camelia WARNINGS for <tmp>:?Useless use of "<=" in expression "42 <= Int" in sink context (line 1)?Invocant of method 'Bridge' must be an object instance of type 'Int', not a type object of type 'Int'. Did you forget a '.new'?? in block <unit> at <tmp> l…
Zoffix nine: ^ it's due to that. Compiling a fix
nine ++Zoffix :) 17:39
Zoffix ZOFFLOP: t/spec/S11-modules/require.t 18:18
ZOFFLOP: t/spec/S17-supply/syntax.t
Geth rakudo/nom: 365a1d0862 | (Zoffix Znet)++ | src/core/IO/Socket/INET.pm
Suggest user may be using invalid arguments

The candidate is called even in cases where, say, a port was given as a Str instead of an Int and the current error message is very confusing in such cases.
Include suggestion that the user passed invalid arguments to .new
18:21
rakudo/nom: 998535edf2 | (Zoffix Znet)++ | 2 files
Improve error message when port is not passed

In such a case the $port ends up being a :U after we process and split all the things. Then, we get a confusing error when a Int:U is used in a `<=` op.
Fix by checking definedness. Also, detect failure from port processing routine and return it from .new instead of trying to assign it into the Str variable.
Bug find: irclog.perlgeek.de/perl6-dev/2017-...i_14939293
18:22
dogbert2 Zoffix: Test 51 in t/spec/S17-supply/syntax.t seems to uncover a bug 20:17
valgrind doesn't like it much 20:18
perhaps something for timotimo 20:19
the mandatory gist can be found here: gist.github.com/dogbert17/e23c3340...eda3feb18f 20:23
Geth rakudo/nom: 51e59eeb5e | (Elizabeth Mattijsen)++ | 2 files
Make .unique|repeated|squish return containers if applicable

If you do .unique|repeated|squish on an array, you should be able to assign to the values produced by the iterator.
23:15