firstly databases don't store queries - your application has the queries.
second, a database can be upgraded very easily using a script component and
tracking a version number stored by you in your own version table data.
There are many protocols you could follow to achieve this.
third, you don't copy gdb files, you back them up and restore them (that's
just good practice)
fourth, if data security is an issue and data needs to be distributed (i.e.
if you can't have one client getting access to another client's data), then
yes, your application will need to create, upgrade etc it's own database for
the client it is working with, otherwise, it would be much better C/S
methodology to keep all data from all clients in the one database where
table data has a field called e.g. client number. Your queries (in your
application) would always ask, inter alia, for only "where clientno=999" for
each request for data.
I'm assuming you want each of your clients to use this application and not
you (the accountant) to use it for all clients.

Alan

-----Original Message-----
From: Cassandra Harley [mailto:cass.harley@...]
Sent: Sunday, 23 February 2003 9:01 PM
To: ib-support@yahoogroups.com
Subject: [ib-support] Non-technical database question


I am writing an application to store staff/wage details for an employer.

Generally the employer will only have one database.

However I would like the program to be able to handle more then one
database, for in the case of accountants and bookkeepers who use the
application for more then one client.

Therefore, the application needs to be capable of creating new instances of
the database.

I have created a metadata file, one thought is to have the application
create a new database from the metadata.

The problem with this is, if I do this, then will each instance of the
database need to store its own queries. Which seems silly, as each database
will have need of the same queries. Also if a later application upgrade
wanted different queries, then each database would need upgrading to be
used.

Should each 'new employer' not have an individual database, but instead
share one database, with their own tables?? That way the queries are shared.
I think this is worse then the first idea, as copying one employers data
would need a function, instead of just being able to copy the .gdb file.
Also each 'employer's dataset will contain more then one table.

I know I probably sound way off track. But I am a newbie to this game
(recent uni. grad...all theory no practice as of yet). But I am wanting to
know the best way to do these things.

All thoughts/comments etc very much appreciated.

Regards,
Cassandra.



To unsubscribe from this group, send an email to:
ib-support-unsubscribe@egroups.com



Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/