reclassing reclass files with zero data

Doug Hersh writes:

        Using GRASS4.0 I ran into a problem trying to reclassify zero
values in a reclass map to non-zero values. The results were inconsistent.
Some of the zero values were reclassed correctly, but some remained zero.
        When I used r.resample to create a "real" raster map of the
reclass file with the zeroes in it, and then used r.reclass to change these
zero categories to non-zero categories, the reclass worked.
        Am I trying to do something unnatural when I try to reclass zero
values in a reclass file to non-zero values? Should I only be reclassing
"real" raster files? Can one reclassify reclass files ad infinitum?
Or am I trying to do something reasonable, but dealing with a bug? Is
this also a problem in 4.1?

As far as I know, r.reclass can be used indefinitely since it is only a
mapping operation on the original cell file - at least that's what is says
when you do a double reclass. If you reclass cat '0' to some other value,
then obviously all 0's should get changed. Any 0's remaining in your
reclassed map must therefore be due to a MASK that you have set or (less
likely, this) to the fact that your region is larger than your map.

----\_\_\_--\_----\_------\_-------------------- P. Martijn van Leusen --------
     \_ \_ \_\_\_\_ \_ Dept of Pre- and Protohistoric Archaeology
      \_\_\_ \_ \_ \_ \_ University of Amsterdam, The Netherlands
       \_ \_ \_ \_\_ \_ martijn@scanner.frw.uva.nl
--------\_------\_----\_--\_--\_\_\_\_-----------------------------------------