Re: Much Ado About COUNT(*)

Jon Jensen wrote:

> If you're willing to do the work, and have the motivation, probably
> the best thing to do is just do it. Then you can use empirical
> measurements of the effect on disk space, speed of various operations,
> etc. to discuss the merits/demerits of your particular implementation.
>
> Then others don't need to feel so threatened by the potential change.
> Either it'll be (1) an obvious win, or (2) a mixed bag, where allowing
> the new way to be specified as an option is a possibility, or (3)
> you'll have to go back to the drawing board if it's an obvious loss.
>
> This problem's been talked about a lot, but seeing some code and
> metrics from someone with a personal interest in solving it would
> really be progress IMHO.

Jon,

This is pretty much where I'm coming from. I'm looking at working on
this and I'd rather discuss suggestions for how to implement this than
whether we should have it, etc. If it turns out better, great; if not,
then I just wasted my time. I really do appreciate everyone's comments
and suggestions.

Thanks!

-Jonah