Chris, you monumental idiot!
Well, you asked ;-) And you're in good company; count me in. I've
worked with files containing over 1.5 million records, and heard of at
least one national delivery service that hit over 16 million records
( several years ago. ) I'm not saying that all of the data was retrieved
at one time, but, short of the psychic lines ( and yes I know we are
supposed to be, ) it's not clear how to know this will happen. That's
why the lack of scrollable cursors in the first JDBC release was so
frustrating. Add to that using optimistic methods for maximum
concurrency and you can run out of memory PDQ if you don't watch
what you're doing. And that should give some pause to those who
blithely do in-memory sorts.
Jon, I wish you wouldn't be so defensive, As far as I can tell, you do
good work. I have some interest as far as the posting that seemed to
show ( in limited tests ) some slowdown using Town/Village and
would like to see some valid tests. Also, if one should choose when
to use it, I hope you have documented it somewhere.
As to the original question, as Chris and Scott noted, without using
JDBC 2.0 drivers ( few out there -- hooray for the AS/400! ) select
count is about the only realistic choice; and hope it doesn't take too
long.
With 2.0 scrollable cursors, you can use
rs.last();
int myRowNumber=rs.getRow();
Joe Sam Shirah
Autumn Software
-----Original Message-----
From: Chris Pratt <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Monday, June 28, 1999 8:02 PM
Subject: Re: Number of rows a query retrieved?
> Well then call me a monumental idiot...
___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".
Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
