Re: Oracle vs Sybase

Tim Uckun wrote:

> > >
> > > "As a final note, when NLS_COMP is set to ANSI, a linguistic index must
> > > exist on the column where the linguistic order is desired."
> > >
 

> > > This means I have to build a index on every single text field? why isn't
> > >
> >
> > No you don't have to, you could ;-)
>
> I guess I keyed in on the word MUST in that snip I posted. Rereading the
> documentation I still think I MUST build an index on the column where
> the linguistic order is desired. I still don't know if I have to
> specifically use this index or not.
>

That's really interesting. Can give me some test data which demonstrates this behavior. I would like to test it myself.

>
> > If you have to rewrite all/most of your queries in pass through it'll be no problem to add
> > the upper function to your varchar's statements...
> >
>
> You are right about everything you said of course and I relize that.
> Unfortunately in the beginning anyway I will have to let things slide as
> they are. Even if access is pulling the entire tables over to query the
> database I will have to take the hit. Little by little though I intend
> on moving my queries over to the database and writing pl/sql. I just
> can not stop this database from being used while I upsize it would just
> kill me.

If your performance is acceptable you could start with this issue. You should know that the performance decrease grows exponentially with the data increase.

> Besides I figure oracle can send the data just as fast as jet
> can right?

I wish you were right;-) Oracle can send the data as fast as the network allows it . Suppose you have a table with 1000 recs and a record length of 1000 bytes =1 MB, how long do you need to copy a 1mb file from your file server ? ...;-)) I wish you luck. If you have any questions by the migrating you can ask me. I made a migration project access-> oracle8/access a year ago. There are some very interesting bugs in acces (at least 97) Received on Thu Mar 30 2000 - 00:00:00 CST