From @masak

<TimToady> perl6​: .say for ([1, 0 xx 5-1], *.rotate(-1) ... *)[^5]
<p6eval> niecza v15-6-gefda208​: OUTPUT«1 0 0 0 0␤0 1 0 0 0␤0 0 1 0 0␤0
0 0 1 0␤0 0 0 0 1␤»
<p6eval> ..rakudo 4373f0​: OUTPUT«1 0 0 0 0␤0␤1␤0␤0␤»
<TimToady> um...
<TimToady> rakudobug?
<masak> seemingly.
* masak submits it

From @FROGGS

are both wrong now?

<FROGGS> perl6​: .say for ([1, 0 xx 5-1], *.rotate(-1) ... *)[^5]
<p6eval> rakudo bf472b, niecza v22-36-g19fa03a​: OUTPUT«1 0 0 0 0␤0␤1␤0␤0␤»

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

From @perlpilot

On Wed Nov 21 11​:45​:19 2012, FROGGS.de wrote​:

are both wrong now?

<FROGGS> perl6​: .say for ([1, 0 xx 5-1], *.rotate(-1) ... *)[^5]
<p6eval> rakudo bf472b, niecza v22-36-g19fa03a​: OUTPUT«1 0 0 0 0␤0␤1␤0␤0␤»

I think the GLR should fix this as itemizing seems to do the right thing​:

.say for ([1, 0 xx 5-1], *.rotate(-1).item ... *)[^5]
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 0 1 0
0 0 0 0 1

-Scott

From @niner

The working post-GLR version is​:
.say for ($[flat 1, 0 xx 5-1], *.rotate(-1).item ... *)[^5]

I don't see any way this could become simpler without losing features.
http://design.perl6.org/S03.html#line_1925 gives the use case for multiple return values of the generator function as interleaving unrelated sequences. To support that, we have to incorporate multiple values into the return list. Itemization is the way around that.
Thus I'm inclined to reject this ticket.

@niner - Status changed from 'open' to 'rejected'