Re: [Update REQ 5154]: intro and summary

[snelson at canopus.llnl.gov]

Thanks for the info.  We are pretty committed to having a
native viewer on MACs and PCs for speed reasons.  Although
the network performance of PCs is so-so, MACs seem to be 
a little further behind that for some reason.  We can
run GeomView on our PCs now using PCXview from NCD.
Our intent though is to get the a viewer interfaced 
into Mosaic (or netscape) so that the user just points
and clicks and then can manipule their 3D object.  This
works fine through X but on PCs the performance is slow
since it is sending GOBS of info through the network
and our PCs seem tops out at 80kB/sec.  A localbus
ethernet card would probably help that; we are using ISA bus
right now.  Our SGIs get close about 400kB/sec which is pretty 
good for a 10MB/sec cable. 


Thanks for your input

Scott Nelson
LLNL

PS  our web server was down during the holidays.  It
    crashed a couple of days after I left for vacation.
    It's back up now and you can find the viewer test
    page at:
      http://www-dsed.llnl.gov/documents/WWWtest.html



On Dec 28, 12:06am, Tamara Munzner wrote:
> Subject: [Update REQ 5154]: intro and summary
>Date: Wed, 28 Dec 94 00:06:03 -0600
>From: munzner
>Message-Id: <9412280606.AA08554 at hilbert.geom.umn.edu>
>To: software@geom
>Subject: [ REQ 5154]: intro and summary
>
>Hello Scott,
> Thanks for your message, sorry for the delay in replying.
>
>>   1. Simple 3D manipulation:  rotate, zoom, translate, materials on/off
>Geomview can certainly handle this.
>
>>   2. Platforms: UNIX, PC, Mac
>Unfortunately at this time we don't have the resources to port
>Geomview to Windows/DOS or MacOS. Would the X version suffice through
>Linux on a PC and MacX on a Mac?
>
>>   3. data types: vector data, wireframs 3D, polygonal surface, 
>>      volumetric mesh data (structured and unstructured) like finite 
>>      difference and finite element, solids data
>>   The above data types can vary as a function of time. e.g.:
>>      A. weather simulations showing moving vector data
>>      B. solids simulations using Dyna3D (crashing cars into telephone poles
>>	or earthquake simulations with bridges, thus modify mesh structure)
>>      C. eletromagnetic wave propagation
>>      D. chemical model interaction display
>>   Neither VRML/OOGL nor PACT has everything that we need but we would like
>>   to extend OOGL to support the few features that we need (specifically
>>   time variations) that it doesn't seem to have. 
>
>At the moment we're not working on extending OOGL to support
>time-varying data, and don't know of anyone else who is either. While
>we do make the source available, you may have noticed that the OOGL
>libraries are, shall we say, not over-documented. If you decide to try
>extending them, we'll do our best to answer specific questions. 
>
>I do know that the GRAPE designers (from Bonn, Germany) have a system
>that deals with time-varying data. While I don't think GRAPE is
>necessarily what you're looking for, I'm including the README below.
>If that does turn out to be a useful direction, the chief designer of
>GRAPE is now at the Technical University of Berlin, working on a new
>system called OORANGE. I'll know more about it after a seminar here in
>a few weeks.
>
>I briefly checked out your WWWtest page a while ago, but haven't been
>able to get through for a few weeks. I'm definitely interested in what
>you all are doing. 
>
>Tamara Munzner		The Geometry Center	     ((555) 555-5555
>munzner at geom.umn.edu    http://www.geom.umn.edu/people/munzner.html
>
>------------------------------------------------------------------------
>What is GRAPE
>-------------
>
>GRAPE is a GRAphical Programming Environment to use the facilities of
>interactive computer graphics for solving mathematical problems from
>continuums mechanics and differential geometry. It has been developed
>at the Sonderforschungsbereich 256 for `Nonlinear Partial Differential
>Equations` at the University of Bonn.
>
>An essential part of the conception of GRAPE is the natural approach
>to mathematical visualization. GRAPE offers efficient tools for the
>mathematician to work with geometric data, generate graphic output and
>film sequences, and it supports numerical interaction. The environment
>is based on an object oriented kernel and a machine independent
>interface for the graphics. Two main branches have been developed,
>both using the same conceptual ideas and interactive environment:
>Differential Geometry and 3D Adaptive Finite Elements.
>
>Our latest development is the inclusion of time dependent data
>structures into GRAPE. Time is an emphasized parameter, and the data
>is no longer a static model but now a model which exists for a time
>period, e.g.  a deforming surface or a flow in a volume. We now use
>'time objects' in the same natural way as formerly 'static objects'.
>We provide tools for handling time dependent parameters, operations on
>time dependent data structures, interpolation and topology changes.
>
>This environment is designed for dynamic extension, i.e. the user may
>supply new features adjusted to his own needs.
>
>
>What is available
>-----------------
>
>The GRAPE software is non-commercial. It may be obtained by other
>scientific sites on request from the address mentioned below only.  It
>is not allowed to use GRAPE for commercial purposes and software
>developed using GRAPE must be made available to us. See license.ps
>for a more detailed version of the license agreement.
>
>At the moment there are drivers available for nearly all SGI machines,
>Sun workstations, X-Windows, PostScript, Titan and rendering systems
>like Softimage.
>
>The files in this directory are
>
>   README			this file
>   license.ps.Z			license agreement
>   demo.tar.Z			demo files (source, main program
>				and input files)
>   manual.tar.Z			beginners doc and GRAPE manual
>   sgivgxlib.tar.Z		library for SGI	machines like indi, gt etc.
>   rs6000x11lib.tar.Z		library for IBM RS6000 with X11 graphics
>   rs6000gllib.tar.Z		library for IBM RS6000 with gl graphics
>
>
>WARNING: It is useless to copy these files without registering because
>         these archives don't contain the headerfile grape.h which is
>         needed to use GRAPE. Registered user will get this file from
>         the Sonderforschungsbereich 256.
>
>
>How to get GRAPE
>----------------
>
>For more information about GRAPE or if you are interested in installing
>GRAPE on your machine(s) please email to
>
>	grape at sfb256.iam.uni-bonn.de
>
>or write to
>
>	Sonderforschungsbereich 256
>	am Institut fuer Angewandte Mathematik
>	der Universitaet Bonn
>	z.Hd. C. Teitzel
>	Wegelerstr. 6
>	53115 Bonn
>	Germany.
>
>
>Bonn, Dez 12, 1993 (bo)
>
>
>
>-- End of excerpt from Tamara Munzner



-- 

+---------------------------------------------------------+
|Scott D. Nelson   B131  Rm2074   email:nelson18 at llnl.gov |
|Lawrence Livermore National Laboratory                   |
|7000 East Ave., L-153                                    |
|Livermore CA  94550                                      |
+---------------------------------------------------------+