Re: [Update REQ 5178]: nd-stuff and tcl/tk

["Tim Rowley" <trowley>]

> 1) the modules that he wrote for coloring and slicing of 4d-objects
> 2) the 4d tetris that he (and others) wrote.
> If possible I would like to be able to run these on Next (and SGI)  
> hardware.  (I do have X windows for the NeXT--CoXist-- and  tcl/tk as  
> well).

Yes, this is possible.  You can run the tcl/tk modules under
coXist.  If you want to try it out, ftp the file
/priv/trowley/ndutils.tar.Z from ftp.geom.umn.edu.  Change
directory to where you have installed Geomview (ie
/LocalApps/Geomview.app) and untar it.  Then run "install.nd".
The modules "NDview", "Colormap", "Slicer", and "3D Snapshot"
should appear in the External Modules list when you start
Geomview.  You should have coXist running before you start any of
these new modules.

> My understanding is that (one) reason for the delay is the problem of  
> how package this stuff since it uses  tcl/tk (email from slevy of  
> Sept21).

Yes, there was a problem with preparing the modules for
distribution.  The modules required some nonstandard extensions
for tcl/tk which would require anyone who wanted to compile
Geomview to recompile tcl/tk first.  Happily, a new version of
tcl/tk has been released which includes these extensions as
standard.  I have converted Brian's modules and NDview/NDdemo to
this version of tcl/tk.  These should appear in the next release
of Geomview.  4D Tetris will probably have to wait, as it needs
work to remove its screen size assumption, its use of many large
GIFs, and to ready it for distribution.

> So my question is:  Is this still the "right thing" to do?  Or are  
> the bundling problems,that seem to be the case, causing second  
> thoughts?

I would still recommend tcl/tk as being an ideal choice for
writing modules.  It's very easy to learn, has a nice interface
with C, and the interpreted language makes for a quick
development cycle.  The bundling problems we experienced were due
to our use of nonstandard extensions.  I doubt you will have this
problem.

-- 
Tim Rowley  --  trowley at geom.umn.edu  --  "Do or do not, there is no try."