>Thanks for the quick reply Ed. I guess my basic problem is the parameratized views will be faster than filtered tables but I have to have the locking approach as above. My users feel the whole approach of editing a record and then finding out another user edited and saved the same record in the meantime is stupid and ineffecient. I think I'll try the RLOCK() of the underlying table approach. There may be a minor performance penalty but since it will only occur when trying to edit it shouldn't be a big deal. Thanks again.

Colin,

There is a better way. You use your views and set the update where clause to be key and modified. What is the probability that two users changed the same fields on the same record? This gives you all the benefits of using updateable views without the problem you're talking about.

Also, be careful what you tell users. They'll have to 100 hours of programming to prevent a situation that may occur once every 200 years, like two users changing the same fields in the same record at the same time.