> I am about to add to an application the ability to store scanned
> documents. There will be 10-12,000 documents per annum. The documents
> will be held in a database containing only a single table which will
> contain the document stored in a blob field and some other fields to
> index the documents so that they can be retreived. Is this a sensible
> approach? Can FB cope with this and are there any problems or pitfalls
> of which I should be aware? Alternatively, I could store the documnets
> on disk as jpg's or pdf's, with an entry in the database indicating the
> file name. Is this a better approach?
>

IMO the latter approach is better, it makes backups a little more difficult
though but it makes the db much lighter.
You may get acceptable over-the-wire performance though, if you use the
former accompanied by "on-the-fly" document compression (zipping) prior to
storage.
Alan