bmg next steps

[Seth Teller <seth@graphics.lcs.mit.edu>]

citywalk/bmg folks:

looks like we have a working BMG on both ends.  (the current
state is that it works as it should on UCB floorplans.  the
MIT floorplans are sufficiently different to cause problems.)

i exhort andreas and rick to follow through on the plan to 
document and clarify the structure of the BMG code, including:

  * which pieces are utility routines?
  * ... user interface?
  * ... batch control code?
  * ... documentation of assumptions?
  * ... example inputs and outputs?

etc.  the result should be 1) a nicely separated library of
batch routines for contour closing, extrusion, etc.; 2) a
main program with UI that runs just as BMG runs now; 3) a
batch program that calls routines from 1) to produce the
same effect; and 4) documentation, makefiles, reorganized
CVS structure, etc., so that all of it can be checked out
and built from scratch by a new student.

when will you be meeting to make this happen?  if you need
tele-conferencing facilities, let bryt (at MIT) and ming 
(at UCB) know so they can arrange the room and time.

prof. t.

-------- Original Message --------
From: Andreas F Wehowsky <andreasw@MIT.EDU>
Subject: bmg
To: <bmg@graphics.lcs.mit.edu>

BMG is finally running now and able to extrude a floorplan to 3d !
Next step is to use it on a floorplan from MIT. I tried with
NE43.1.dxf, but without luck, since acad2ug is not able to convert
that dxf file to ug.
Do you (Todd or Keyuan) know how to solve this problem?

- Andreas