Re: width_bucket function for timestamps

Neil Conway <neilc(at)samurai(dot)com> writes:
> On Mon, 2006-10-09 at 12:02 -0400, Tom Lane wrote:
>> It's not clear to me why we have width_bucket operating on numeric and
>> not float8

> I asked about this when I originally implemented width_bucket(), I
> recall[1]. At the time, there was scepticism about whether it was even
> worth implementing width_bucket(), let alone providing multiple
> implementations of it. I'd be happy to provide a float8 implementation
> (I may even have one lying around somewhere...)

It's probably too late for 8.2, unless some other compelling reason to
force an initdb surfaces. But if you can find the code, please stick it
in when 8.3 devel starts. As it stands, one must do an explicit
coercion to numeric to use width_bucket() with floats; which is tedious,
slow, and nowhere required by the spec.

regards, tom lane