From @ashgti

$ perl6

&infix​:<Xxx>
Could not find sub &infix​:<Xxx>
1 Xxx 2
1 1

<TimToady> so Xxx really means something like crosswith(&infix​:<xx>)
<TimToady> however, a p6 implementation should be able to generate
&infix​:<Xxx> at need

So, there should be some &infix​:<Xxx> function generated, somehow...

I am not sure how to generate those, because Meta-ops can be repeated, but
just filling the bug report so people know about the issue.

--
John Harrison

From @coke

No change in last year.

--
Will "Coke" Coleda

@coke - Status changed from 'new' to 'open'

From @usev6

On Sun Oct 03 10​:35​:17 2010, ash_gti wrote​:

$ perl6

&infix​:<Xxx>
Could not find sub &infix​:<Xxx>
1 Xxx 2
1 1

<TimToady> so Xxx really means something like crosswith(&infix​:<xx>)
<TimToady> however, a p6 implementation should be able to generate
&infix​:<Xxx> at need

So, there should be some &infix​:<Xxx> function generated, somehow...

I am not sure how to generate those, because Meta-ops can be repeated, but
just filling the bug report so people know about the issue.

--
John Harrison

I'm not sure I understand this correctly, but doesn't TimToady++'s explanation mean, that the current implementation (which is unchanged) is correct and this is not a bug?

It's possible to define an infix​:<Xxx> with rakudo, like

$ perl6-m -e 'sub infix​:<Xxx> ($a, *@​b) { return 42 }; say 2 Xxx <foo bar>'
42

From @FROGGS

I understand TimToady in a way which would imply that rakudo would autogenerate an &infix​:<Xxx> if it is requested by the user...
Though I'm not sure I'd like to see that happen :o)

Here is a workaround though​:

m​: my &infix​:<Xxx> = &infix​:<X>;
&[Xxx].assuming(&[xx]);
say ((1, 2) Xxx (3, 4)).perl
rakudo-moar f185e0​: OUTPUT«((1, 3), (1, 4), (2, 3), (2, 4))␤»

From @zoffixznet

This now works as described.

Tests added​: Raku/roast@8e585a4365

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