Re: find-tag and duplicate tags

[Richard Stallman]

    I think the best approach would be for find-tag to always choose the
    "nearest" available version of a tag, for some suitable definition of
    "near".  For instance, the currently visited file should always be
    nearer than any other file; files in the same directory might be
    nearer than files in another directory; etc.

That is a good idea.  Would you (or someone) like to implement it?