From @labster

[1​:09pm] labster​: r​: my ($a, $c); ($a,$, $c) = 1..3;
[1​:09pm] camelia​: rakudo 14e7d7​:
OUTPUT«[31m===[0mSORRY![31m===[0m␤Unsupported use of $, variable; in Perl 6
please use .join() method␤at /tmp/OlBWFcsobt​:1␤------> [32mmy ($a, $c);
($a,$,[33m⏏[31m $c) = 1..3;[0m␤»
[1​:09pm] labster​: http://perlcabal.org/syn/S02.html#Names says that line
'will result in the message, "Anonymous variable requires declarator".'
[1​:10pm] labster​: r​: my ($a, $c); ($a, $ , $c) = 1..3; #added space
[1​:10pm] camelia​: rakudo 14e7d7​:
OUTPUT«[31m===[0mSORRY![31m===[0m␤Non-declarative sigil is missing its
name␤at /tmp/eFX9NIpWzs​:1␤------> [32mmy ($a, $c); ($a,$[33m⏏[31m , $c) =
1..3; #added space[0m␤ expecting any of​:␤ postfix␤ infix
stopper␤ infix or meta-infix␤ …
[1​:10pm] masak​: labster​: seems the error message is shadowed by rakudo
thinking it's the $, variable
[1​:11pm] masak​: ...or another error message is. :)
[1​:11pm] labster​: so, theoretically a rakudobug, but do we really want
exact wording for error messages defined by the spec?
[1​:15pm] masak​: I think it's a rakudobug that it considers ($a,$, $c) as
being the '$,' variable.
[1​:15pm] PerlJam​: labster​: The message in the spec at least points them in
the direction that anonymous vars can only be used in declarators
[1​:15pm] masak​: but I don't know what to do about it, really.
[1​:18pm] labster​: Well really, the only place you should ever write $, is
in a subroutine signature.
[1​:20pm] PerlJam​: labster​: You don't think "my ($a,$,$c) =
$string.split(/\t/);" would ever happen?
[1​:20pm] labster​: oh... right
[1​:20pm] labster​: I'm not thinking straight.
[1​:21pm] lizmat​: r​: my ( $a, Mu, $c )= <a b c>; say $a; say $c
[1​:21pm] camelia​: rakudo 14e7d7​: OUTPUT«a␤c␤»
[1​:21pm] labster​: r​: my ($a,$,$c) = "a b c".split(/\s/);
[1​:21pm] camelia​: rakudo 14e7d7​: ( no output )
[1​:22pm] labster​: hm
[1​:27pm] dalek​: std​: 8850393 | larry++ | STD.pm6​:
[1​:27pm] dalek​: std​: only carp about P5 $, variable where it's obvious
[1​:27pm] dalek​: std​: review​: perl6/std@8850393dd7
[1​:34pm] masak​: TimToady++
[1​:35pm] labster​: okay, shall i submit a proper rakudobug then?
[1​:35pm] PerlJam​: labster​: can't hurt
[1​:36pm] jnthn​: plz reference the STD commit TimToady++ just made too, so
the fixer knows where to steal a fix :)

From @FROGGS

Fixed​:
rakudo/rakudo@25bfa5f

Closable with test.

The RT System itself - Status changed from 'new' to 'open'

From @labster

Well, almost fixed. The spec says​:

Attempts to say something like​:
($a,$, $c) = 1..3;
will result in the message, "Anonymous variable requires declarator".

It results in the message​:

Non-declarative sigil is missing its name
which is also what STD generates. So is this now a specsbug or rakudobug?

From @coke

On Thu May 23 14​:54​:13 2013, labster wrote​:

Well, almost fixed. The spec says​:

Attempts to say something like​:
($a,$, $c) = 1..3;
will result in the message, "Anonymous variable requires declarator".

It results in the message​:

Non-declarative sigil is missing its name
which is also what STD generates. So is this now a specsbug or rakudobug?

Lack of test has allowed behavior to revert to original quoted on ticket​:

19​:44 < [Coke]> r​: my $a; my $c; ($a,$, $c) = 1..3;
19​:44 <+camelia> rakudo-jvm 4f66ce​: OUTPUT«===SORRY!=== Error while compiling
  /tmp/_8CZwy0Onq␤Unsupported use of $, variable; in Perl 6
  please use $*OUT.output_field_separator()␤at
  /tmp/_8CZwy0Onq​:1␤------> my $a; my $c; ($a,⏏$, $c) = 1..3;␤
  expecti…»
--
Will "Coke" Coleda

From @usev6

This works now​:

$ perl6-m -e 'my ($a, $c); ($a, $ , $c) = 1..3; say $a; say $c'
1
3

Also, S02 was modified recently (Raku/old-design-docs@aadf058) and the new version explicitly allows the use of "($a,$, $c) = 1..3;"

I added a test to S03-operators/assign.t with the following commit​: Raku/roast@82b0065586

@usev6 - Status changed from 'open' to 'resolved'