PG Doc comments form writes:
> The ceil, ceiling, and floor mathematical functions in documentation
> versions 9.5 through 12, and probably others is not correct. The example
> output shows the ceil* functions returning the integer below the float value
> while the floor function shows the opposite. Thanks!
No, they're correct as they stand. These examples intentionally use
negative numbers to make the point that the rounding is towards or
away from plus-infinity, not towards or away from zero.
If we had room for more than one example, I'd be inclined to make these
cases be secondary examples rather than the only one. There's some work
afoot to restructure this chapter's function tables in a way that would
make that a bit less painful vertical-space-wise, so maybe it'll get
done soon.
regards, tom lane
