On Fri, Sep 30, 2011 at 2:12 PM, Merlin Moncure <mmoncure(at)gmail(dot)com> wrote:
>> FWIW, speaking as somebody who has no need of this function, "array_xor" is a pretty clear name that indicates what's going to happen.
>
> +1 on this -- was going to suggest until you beat me to it. I also
> for the record really think the array_ prefix on array handling
> functions is pretty silly since we support overloading -- greatly
> prefer unnest() to array_unnest() etc. So, how about xor()?
Makes sense, in light of your comment about overloading.