At 4:53 PM +0200 6/8/01, Claes MÂnsson wrote:
>Antony T Curtis skrev:
>>
>> Markus Grundmann wrote:
>>
>> Such a feature as this would have to be a "optional" rather than
>> default.
>>
>> I wouldn't want it to open 400+ tables in a database to find out how
>> many million rows there are...
>
>Ehrm. Not that big of an issue. From the docs at
><http://www.mysql.com/documentation/mysql/bychapter/manual_Performance.html#Where_optimizations>
>
>"COUNT(*) on a single table without a WHERE is retrieved directly from
>the table information"
>
>As I understand it, this means that the speed of fetching the number of
>records from a table is pretty near constant. A one-record table would
>take the same time as a million-record table...
This is not true for BDB tables - at least it didn't used to be true, a
full scan was needed. Has that changed?
>
>Regards
>/Claes
--
Paul DuBois, paul@stripped
