Yes, I see our telepathy is working just fine. <Knock knock> ;-)

I thought about the DB design and the generic dictionary too.

On May 16, 10:40 pm, Joe Enos <[email protected]> wrote:
> I see all these answers, and none of them really seem to answer the
> original question.  If I understand the question correctly, you just
> need to know which states border a given state.  You can't do this
> with latitude and longitude, or picking the "before" and "after"
> states, or zip codes, or any other calculation, since states can be
> close to others without actually touching, or can be huge, meaning one
> end of one state to another state may be pretty far away, and border
> anywhere from zero to about ten other states.
>
> The only way to reliably do this is manually.  If you're doing it in a
> DB, I'd probably do something like:
>
> ADJACENT_STATE table with two columns.  The table contains records for
> each state, and each state that it borders.  For example:
> AZ CA
> AZ NV
> AZ CO
> AZ UT
> AZ NM
> CA OR
> CA NV
> CA AZ
>
> Column 1 would be indexed.  When you need to look up a state, you'd
> filter by column 1.
>
> The way I figure, it would only take an hour or two to set up - look
> at a map, do the states one at a time - there are only 50 states, and
> it wouldn't take more than a minute or so each.  And the table would
> only have maybe a few hundred records, so performance would be
> ridiculously fast.
>
> I would bet that there's a list of these already out there on the web,
> so you may not even have to do the work yourself.
>
> If you don't have a database in your small app, you could do all of
> this in an XML file, or you could get it into the compiled code using
> something like a Dictionary<string, List<string>> or something
> similar.
>
> On May 13, 6:30 pm, S <[email protected]> wrote:
>
>
>
> > I am developing this small application in which I need to find states
> > adjacent to the one I am looking up. Say for example, if  I am looking
> > up Connecticut, adjacent ones are NY, MA. Now I could potentially
> > store it but I would want to rather compute it and do it very
> > efficiently.
>
> > Any ideas anyone can throw on this ?
>
> > Thanks
> > S- Hide quoted text -
>
> - Show quoted text -

Reply via email to