Re: Why is box <-> point missing, and box <-> box not indexable?

Andrew Gierth <andrew(at)tao11(dot)riddles(dot)org(dot)uk> writes:
> point <-> point, circle <-> point and polygon <-> point all exist as
> orderable-by-operator operators (in fact they are the only ones by
> default). But there's no box <-> point operator at all, and no index
> support for box <-> box.

> Was this intentional, or just a strange oversight?

Seems like an oversight. None of these operators have commutators,
but it seems like they all should:

<->(point,line)
<->(point,lseg)
<->(point,box)
<->(lseg,line)
<->(lseg,box)
<->(point,path)
<->(circle,polygon)
<->(line,box)

Dunno about the index-support angle.

regards, tom lane