At 30/08/2002 05:25 AM, you wrote:

>Why? Sure, I can see that it occationally can be useful to join between
>several databases, but mainly in spesial circumstances. And in those
>circumstances the alternatives of pumping the data from one of the
>databases into the other or writing a special program getting potential
>candidates from one database and then checking if they are in the other
>normally isn't too scary. If the requirement to use SELECT across multiple
>databases is more of a regular happening, I'd rather say that there could
>be something wrong with your design and that the databases maybe ought to
>be merged into one.
>
>Basically, I'm just curious as to why you think this is a very important
>feature,

Just my 2 cents here.

I personally don't see this feature as very important, but still good to have.

First of all, I do understand your point about having a good database
design. It doesn't make sense to have multiple databases that are used by
the same application all the time, it's better to have the data all-in-one
database.

I will be in a position where such a feature would be helpful, especially
in the areas of reports. Let me explain my scenario. There are regional
offices and a head office. The stores connect to the regional offices
database. The regional offices databases are replicated to the head office
at least once a day. The head office wants to ba able to generate reports
off of those replicated databases. If the databases are very big, would it
be wise to create another bigger database to enclose all the data of the
replicated databases in order to generate reports or lookup miscellaneous
data. Let say each regional offices DB has approx. 2GB of data each and
you have 4 offices, then you already have 8GB of data that is replicated on
the header offices server. Do you want to another database on that server
that would total close to 8GB in data, making the storage space on the
server grow to 16GB used? I know hard-drives are cheap and 16GB is not
much nowadays for todays hard-drives, but imagine if the regional databases
are 20GB each(for the same scenario the server would have to be able to
handle 160GB of file storage).

But, since cross-db queries are not supported at the moment, I'll have to
bite the bullet and create a program that merges the data from those
regional offices.

Have a nice day.


Daniel Rail
Senior System Engineer
ACCRA Group Inc. (www.accra.ca)
ACCRA Med Software Inc. (www.accramed.ca)