Re: geomview comments

[slevy]

> 1) I only found out about geomview by word of mouth. Which newsgroups or
> mailing lists should I follow for announcements of new versions,
> add-ons, etc? (This could go in the docs!)

We haven't been doing this regularly -- I think there was one announcement
on comp.graphics and comp.sys.sgi.  This is a good point; we should start
a geomview users' mailing list.  We wouldn't want to broadcast to some
large group every time we release a new gadget, but it'd make sense
for a geomview interest group list.


> 2) Do you know if there's an interface to Mathematica in the works
> anywhere? (I have seen an X interface to Mathematica with `live'
> rotation & scaling somewhere, so it should be possible).

Yes -- it exists now.  Pick up pub/geomview/mathoogl.tar.Z from geom.umn.edu.

> 3) For geometric models with a definite `up' axis, the current
> mouse-transform interface is rather inconvenient as you keep tilting the
> model from the vertical. It would be nice if the rotations could be
> constrained to keep the model-vertical aligned with the display-vertical
> (model y-axis in display y-z plane, I presume). Is this possible or
> easy-to-implement?

We've talked about a couple of things along this line -- "constrained"
motion along just the display axes (so e.g. an approximate horizontal drag
would be taken as exactly horizontal), and motion about the object's own
axes, so (say) a pure horizontal motion would amount to rotation about the
object's own Y axis, no matter how that was oriented on the screen.
Would that do?  Both would be pretty easy to implement; the problem
would be choosing a convenient way to specify when you wanted them to apply.


Thanks for your comments & for the "-stdin" script!

    Stuart Levy, Geometry Center, University of Minnesota
    slevy at geom.umn.edu