> "To the east of" sounds like a simple concept, until one explores
the
> corner cases.

I disagree; "to the East of" is well defined, since it concerns the
directional relation between two locations. "Farthest East", in
contrast, is not well defined.

The problem in this thread is that "East" has two quite different
meanings:
 1) "East" as direction (relation)
 2) "East" as location (absolute)

As mentioned in another posting, "there is no East Pole". Actually, as
far as 2) is concerned, there is; just as the North Pole is "as far
North as there is", meridian 180 is "as far East as there is" (and
also "as far West).

> The specifications that seemed so clear to the user who wrote them
turn
> out to be imprecise or surprising when the corner cases start
appearing."

Exactly the problem. The original question
 "What states are the farthest North, South, East, and West in the
United
States?" is imprecise: does it mean
 1) located at the greatest/least North latitude [East/West-most
longitude] or does it mean
 2) from which no other State can be reached by proceeding in the
indicated direction without going around the world?

Steve

P.S. In case anyone wonders what my "well defined" "to the East of"
looks like:
 Given two points A and B on the surface of the Earth where the
difference in the longitudes of the points is > 0 and < 180, point A
is "to the East of" point B if the arc drawn in the Eastern direction
along the latitude parallel of point A from point A to the longitude
of point B is shorter than the arc drawn in the Western direction
along the latitude parallel of point A from point A to the longitude
of point B.