Tom Lane wrote:
The existing geometric containment tests seem to be nonstrict, so if we
wanted to leave room to add strict ones later, it might be best to
settle on

        x @>= y              x contains or equals y
        x <=@ y              x is contained in or equals y

reserving @> and <@ for future strict comparison operators.

Since the choice of @> and <@ comes from current ltree operators I'd like to point out that they are non-strict for ltree, and this could add a little bit of inconsistence.

Best regards
Matteo Beccati

---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings

Reply via email to