John,

>I'm not disagreeing with you about disconnected datasets, especially since you used words like "typically" to avoid making unjustifiable absolute statements... ;-)

LOL! That's *exactly* why I used the word "typically"!! <s>

but what if you need to access one of those big iron databases that can take a second or two to open a connection across a WAN?

That shouldn't be an issue with connection pooling.

~~Bonnie




>I guess part of my point is that all database access should use disconnected datasets.
>
>I'm not disagreeing with you about disconnected datasets, especially since you used words like "typically" to avoid making unjustifiable absolute statements... ;-) but what if you need to access one of those big iron databases that can take a second or two to open a connection across a WAN? If you've tied yourself to just one way of doing everything, sometimes you can't meet new challenges.