On Mar 9, 2010, at 4:18 PM, Richard S. Russell wrote:
Make two databases, one database is just an interface, the other is
just the data. Distribute the interface database, put the data
database on a server, you can update it as needed.
That doesn't quite do it, Will.
Perhaps an example will help. Let's say I distribute the 2010
version of the data file, with 700 words and phrases in it.
During the following year, Tutor A adds 80 prepositions to his copy
of the file.
During that same year, Tutor B adds 120 rhyming pairs to her copy of
the file.
In 2011, I want to distribute another data file that includes an
additional 300 words and phrases, above and beyond the original 700.
I want both Tutor A and Tutor B to be able to make use of the 300
new flash cards (as well as any revisions to the old ones) IN
ADDITION TO the records they've already added to their own systems.
But I have no way of knowing in advance how many new cards the
tutors may have added, so I'm not sure where to start numbering the
300 new ones, or if I should even make the attempt. (Still, I really
need each flash-card record to have a unique ID number.)
Richard
Why not simply distribute the additional words and phrases (rather
than the whole set) and import them into the originally distributed
file? Serial numbers can be automatically assigned on import.
It's probably going to be a good idea, also, to maintain some kind of
flag field so you can find out which are the user-added words and
phrases.
Or is there some reason you need each word to have a unique serial
number across all instances of the database?
Steve
