Hi Adrian,

> The resource polyhedra are created at the end of base/std_polys.cc
> if you want to look at what I am doing with the others.

I didn't mean to be confusing about values to map indexes and vice versa. Except
that the subject ended up going both ways!

I was concerned about map indexes created by sym_repeat and how they'd be
resolved to values without being able to specify a map. I see in the std_polys
code now that a random map is created. Then that would be resolved! This would
allow pairs of constituents to be colored the same (when this type of symmetry
occured) and sym_repeat is the only way to produce that effect so I would
probably chose that for the UC's.

The only thing I see is that with the UC's it wouldn't work to use average angle
as the coloring method. It seems like it would just use apply_cmap in this case.

Then there was also the case that I would color compounds in iso_delta such that
each constituent had a different color. This could be done via off_color, or it
could be done internally. If done internally then only color values are created
(since averaging face colors values, not indexes). Then the question was how to
be able to create map indexes in that case. The value to map index function
would resolve that.

Hopefully these things are not confusing!

In regards to your note about changing e.g. v_coloring to coloring, the function
color_vef which would be effected by this and is called from multiple programs
is still in lattice_grid.cc. Ultimately some of the lattice_grid code could be
split out into more appropriate modules.

Roger