Say I have items in column A that are not unique. Also, say that if
something is filled in on another column (B), to count that unique item in
column A.

Column A is sorted. Column B may or may not be.

Graphically:

(A)
040000-000001
040000-000002
040000-000003
040000-000003
040000-000003
040000-000004
....

(B)
COMPLETE
ABORT
INVALID
INVALID
COMPLETE
COMPLETE
....


... should return 4. However, if one (or two) of the instances for
040000-000003 are null (blank), it *still* should return 4 (as one of the
instances have a "mark"). In other words, the only time I should see a
decrease in number of uniques in column A is if and only if every occurence
of that unique in column A has a null for column B.

Any ideas?