Re: estimating # of distinct values

Dne 7.1.2011 20:56, Jim Nasby napsal(a):
> On Jan 7, 2011, at 5:32 AM, tv@fuzzy.cz wrote:
>> Another thing I'm not sure about is where to store those intermediate
>> stats (used to get the current estimate, updated incrementally). I was
>> thinking about pg_stats but I'm not sure it's the right place - depending
>> on the algorithm, this may be a fet kilobytes up to several megabytes. And
>> it's not needed except when updating it. Any ideas?
> 
> If you're using it essentially as a queue, maybe a resource fork makes sense?

A resource fork? Not sure what you mean, could you describe it in more
detail?

regards
Tomas