Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> ... the division between functions and operators is quite unnatural.
Agreed --- especially seeing as how every operator has an underlying
function.
> I'd envision merging the information into one chapter with the following
> outline:
Seems reasonable to me.
BTW, this may be more work than you want to do right now, but ISTM that
the "example" columns in those tables would be lots more useful if the
result of the example were shown too. I did that for the network
functions already, but haven't got round to working on the rest.
regards, tom lane