At 11:56 AM -0700 8/18/99, Thimble Smith wrote:
>At 13:41, 19990818, Paul DuBois wrote:
>>> [ me saying fetchrow_arrayref is better than fetchall_arrayref ]
>>
>>Why is it better? I don't think you get any memory savings by fetching
>>a row at a time and working with your data that way. The MySQL DBD uses
>>mysql_store_result() by default, so it's fetching the entire result set
>>into memory anyway. And you're not making an extra copy by using
>>fetchall_arrayref() since all you're getting back is a reference.
>
>Yep. I bet you're right. Anyway, if there is a difference it's probably
>not big enough to notice.
>At 13:41, 19990818, Paul DuBois wrote:
>>> [ me saying fetchrow_arrayref is better than fetchall_arrayref ]
>>
>>Why is it better? I don't think you get any memory savings by fetching
>>a row at a time and working with your data that way. The MySQL DBD uses
>>mysql_store_result() by default, so it's fetching the entire result set
>>into memory anyway. And you're not making an extra copy by using
>>fetchall_arrayref() since all you're getting back is a reference.
>
>Yep. I bet you're right. Anyway, if there is a difference it's probably
>not big enough to notice.
Especially compared to the latency involved in sending the query and
retrieving the rows. I can't imagine that the dereferencing time is
even in the same order of magnitude.
There's something to be said for processing data efficiently, but
in a client/server environment, network latency is bound to swamp
a lost of those efficiency considerations.
--
Paul DuBois, paul@stripped
