Sounds like a similar concept to my "averaging" in post #9. However, I wonder how well this could be implemented? It would have to recognize that "Jon" and "John" are misspelled variations while other "words" are in fact different names. Similar issues go with knowing the relationship between "Bill" is frequently short for "William", "Liz" for "Elizabeth", ignore a woman's married names, etc. Not impossible, but a significant challenge. Particularly when we start adding name rules from other cultures. It's something that would probably have to go though a beta test to get working right.GarysTurn wrote: Wouldn't an even better sort change be to add the (Most Submitted) order to the sort. RE: The version of a name in the folder which has the most contribuitors of that version. See Example.
(Personal Submission) if none (LDS Membership Record) if none (Most Submitted) then (A-Z Sort)
Example Sorts:
Current Sort: (When no personal submission or LDS Membership record exists)
Howard Doe (1) Submission
John (1) Submission
John Doe (1) Submission
John H. Doe (5) Submissions
John Howard Doe (45) Submissions
Jon Howard Doe (1) Submission
Proposed Sort:
John Howard Doe (45) Submissions
John H. Doe (5) Submissions
Howard Doe (1) Submission
John (1) Submission
John Doe (1) Submission
Jon Howard Doe (1) Submission
The problem with "(Personal Submission) if none (LDS Membership Record) if none (Most Submitted) then (A-Z Sort) order" is if the "most submitted" algorithm gets it wrong, then we're back to a problem like we have today where everyone has to enter a personal submission to see the correct name.
