r.buffer problem/musing

[Malcolm Williamson]

Bonnie,
I have to generally agree with Simon's observations re r.buffer. The added 
ability to create different buffer widths from different categories would
require a set of rules to deal with problematic cases, and you can be sure
that the rules would never be correct for your particular needs :-). For example
I have a raster layer with categories 1 and 2 on it. I want to buffer 1/2 mile
from category 1, and 1 mile from catergory 2. Cell size is 500 feet. What 
would I do in the following case?

               111
             11221
            11221
             12211
             1121
              111

Would I only buffer from the outer layer, resulting in a 1/2 mile buffer from
the 1's? Would I only keep the larger buffer zone, which would be 1 mile from 
the 2's? Or would I assign the different buffer zones different values, and, if
so, what would I do in areas of overlap? These are questions that you can 
readily answer for yourself in r.mapcalc; would you really want a buffer module
that decided the answers for you? (or, alternatively, was complex enough for
you to dictate all the rules to it? I think that would not simplify your life
very much.) 

Just my $.02
-- 
Malcolm D. Williamson - Research Assistant       E-mail: malcolm at cast.uark.edu
Center for Advanced Spatial Technologies      Telephone: (501) 575-6159
Ozark Rm. 12                                        Fax: (501) 575-3846 
University of Arkansas              
Fayetteville, AR 72701