On Mon, 26 Jun 2000 08:20:13 -0700, Rodney Broom wrote:
>Hi Keith,
>
>It sounds to me like you're not getting alot of help on this one.
>Personally, I tend to agree with John the most. However, I find nothing
>intrinsically wrong with what you want to do. If you really ~need~ to lock a
>given data set against modification, then there are about a kagillion ways
>to handle. My consern is still John's:
>
>John Hughes> What do you do if user A goes away for lunch with the data
>still locked?
>
>So I'd set locks on write, not on read.
>
When I was faced with this problem, I developed a sub-system that "checks out" the
record for
write/edit. Viewing data does not use this system.
When someone checks out a record for edit, they have it, period. If they walk away
for lunch, well ....
they are violating corporate policy <grin>, and some re-education is in order.
However, there is the
ability to have an administrator clear the lock, and of course when the user tries to
save after the
lock has been cleared they are no longer allowed to ...
This is similar to how most source code control systems work. Not anywhere near
perfect, but adequate
for our web data entry needs at least.
Brgds,
Mike.
