I'm just thinking out loud on this.
There was a recent post on one of the forums ( i dont remember which one and cant find it again) where the proposed solution was using the not in list event. The proposed solution contained a warning that allowing users to add to lookup tables may result in duplication due to mis-spellings and such. I agree thats a valid concern and was curious how I could limit the chances of that happening. In some of my apps where not enterring duplicate people is a must, I use a procedure which uses the levenshtein distance and soundex to check for possible duplicates. I thought I'd try applying it to my NIL procedure.
I dont use the NIL event very often so I'm not sure how this would work across different types of data. In the attached example I used a list of cities. If you mis-spell the name of a city the message box shows the possible duplicates before adding it and allows you to cancel the addition.
Any thoughts?