d.histogram problem

Has anyone noticed that histograms produced by d.histogram
can be significantly different when they are drawn on display frames
of different sizes?

For example, given a raster called "dem", create a histogram
using the entire graphics screen. Then, use "d.frame -c" to make
a smaller frame in the top-left margin. Draw another histogram of dem
in the new frame without erasing the first one.

When I do various repetitions of the above, using different-sized
frames, I get vastly different results at times. I'm using 4.1, but
not the latest release.

Has anyone seen this happen? Am I missing something?

Thanks for any information,

Rusty Dodson
dodson@towhee.cor2.epa.gov

an alternative may be to use r.stats and g.gnuplot
GRASS 4.1 > r.stats -cz cellfile > file
GRASS 4.1 > g.gnuplot
gnuplot> plot 'file' w boxes

of course, you could add axis labels and perhaps even
write a script:
GRASS 4.1 > g.gnuplot mycellfile.stats

--Darrell