> > I have no idea what you base this on.. I've done your little
> experiment and
> > I find absolutely no difference between the two openings.
> > I placed a datagrid/datasource on the form just to be sure.
> > I assume that you are not confusing the speed of clientdataset
> opening once
> > with connection is made OTOH with a connection not made on the other.
> > I did this over and over first one way then the other, each time
> ensuring
> > that both the clientdataset AND the connection/database components
> were both
> > set back to not active/not connected.
> > No matter which one I did first followed by the other - both came up
> almost
> > instantaneously.
>
> Measured or observed?
>

Glenn used the expression "counted to 21" versus "counted to 1".
You said "5 times slower"
I can't count to 1 in eiher case.
I see no reason to put a timer on it. but go ahead and I'll try your code
Alan