> Unfortunately Milos an index isn't likely to help on this type of a
> query. It would appear that the value 'UDP' accounts for the vast
> majority of the rows, and so an index loses most of its value. You see,
> it actually takes *longer* to return queries using the indexes if a
> significant portion of the table is being touched, because the database
> has to check both the index and the tuple. One of the new features of
Jason and Neil,
Thank you very much for the detailed answer. I'm relatively new to SQL.
My background is much more inferior system, FoxPro for DOS, which in
ancient times used what they called "patented Rushmore technology". I
think that FoxPro was able to return COUNT immediately, if index was
used, no matter what number of records the condition was true for. I
wonder if this is something that PostgreSQL will eventually be able to
do, or if it is simply not technically possible for some reason.
--
Milos Prudek