> >
> >
> > Mike,
> >
> > Normalization does not really affect the ability for multiple users to work on a single record. That has to do with your buffering/record locking strategy and how you separate the data entry tasks. With an optimistic record locking strategy, user1 could work with fields 1 - 5, user2 could work on fields 6 - 13, etc. No conflicts would occur.
>
> I think that is rarely the case. If the data is normalized, the relevant columns only exist in the one table. Therefore the chances of conflict are higher.
>
> By denormalizing, what you describe can happen, but requires much more complexity to resolve conflicts, possibly resulting in a deadlock.
>
> Mike Yearwood
>
>
>
Mike,
The point I am trying to make is that you can work with non-normalized tables in a multi-user environment. How many times have you seen an employee table with
CREATE TABLE Employee (RecordID I AUTOINC, Surname C(25), MiddleName C(25), Surname C(25), ...)
This is a VERY common scenario in multi user environments but one could argue that this table is not normalized. How many employees have the name "Smith" or "John". One could (and I would argue, should) normalize this data but it will nave no impact on the ability for multiple users to work with the data.
Ken
You shall know the truth - and the truth shall set you free. (John 8:33)