|
|
David W. Fenton wrote:
> "Larry Linson" <bouncer@localhost.not> wrote in
> news:IhoZf.15911$e11.5525@trnddc02:
>
>
>
>
> Yes, that's what he's doing. The scenario is comparing two very
> stupid things, and lo and behold, SEEK wins in the test of this very
> stupid task.
>
> My guess is that with a more realistic scenario, the difference
> between the two will be sufficiently smaller to have no end-user
> noticable impact. I have never used SEEK myself in the few times
> I've jumped around recordsets and there certainly was not 3-second
> lag involved. This is even the case with one of my apps where I use
> FindFirst to pull in data from a recordset of grouped totals.
> Initializing the recordset takes a long time, but the FindFirst
> operation is not slow at all. And that's a summary on a table with
> about 500K records (the summary has c. 200K records, I'm guessing).
>
> I can't log onto that terminal server right now, but I'll check when
> it comes online again.
Yes, the test was to compare the speed of the three methods, not to be a
realistic scenario. It was never presented as such.
--
Bri
|
|