From @masak

<masak> rakudo​: say ([[1, 2], 3] >>+<< [4, [5, 6]]).perl
<p6eval> rakudo 5d23e3​: OUTPUT«Sorry, sides are of uneven length and
not dwimmy. [...]
<masak> I'm thinking whether this error message is LTA.
<masak> the sides are in fact not of 'uneven length'.
<masak> (well, they are both of length 3, which is not an even number,
but...) :)
<moritz_> what could we say instead?
<masak> good question.
<masak> printing out the contents of the sub-arrays might be a Bad
Idea for big arrays.
<moritz_> "the structures on both sides of non-dwimmy hyperop are not congruent"
<masak> I like that better.
* masak submits rakudobug
<masak> moritz_++
<moritz_> left​: 1 element
<moritz_> right​: 2 elements
<moritz_> at nesting level 2
<masak> excellent.
<masak> yes!
* moritz_ has no idea how hard the nesting level will be to obtain
<flussence> wouldn't the code that prints the error know where it went wrong?
<moritz_> flussence​: the code for hyper is spread out over half a
dozen multis, and there are three die()s in there concerning non-equal
length
<flussence> oh, hm.
<moritz_> but that would be nice project to work on
<moritz_> (one could even track the indexes that lead to the mismatch,
and then say (left)[1]<foo>[3] has length 3, (right)[1]<foo>[3] has
length 2)

From @moritz

On Thu Oct 14 04​:25​:27 2010, masak wrote​:

<moritz_> "the structures on both sides of non-dwimmy hyperop are not
congruent"
<masak> I like that better.
* masak submits rakudobug
<masak> moritz_++
<moritz_> left​: 1 element
<moritz_> right​: 2 elements
<moritz_> at nesting level 2
<masak> excellent.
<masak> yes!
...
<moritz_> (one could even track the indexes that lead to the mismatch,
and then say (left)[1]<foo>[3] has length 3, (right)[1]<foo>[3] has
length 2)

That's what happens now​:

say ([[1, 2], 3] >>+<< [4, [5, 6]]).perl
Sorry, lists on both sides of non-dwimmy hyperop are not of same length​:
  left​: 2 elements
  right​: 1 elements
At .[0]

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

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