From: Bruce S. <ba...@an...> - 2000-11-03 22:13:33
|
--On Friday, November 03, 2000, 3:38 PM -0500 Jason Cunliffe <ja...@no...> wrote: > Hi Bruce > > Thanks for the reply > >> The obvious advantage of VPython is its extraordinary immediacy, compared >> with generating POVRay image descriptions to be rendered. > > Yes no argument with real-time! I have been playing around with VPython > more today. It is great adn I admire the simplicity of syntax and direct > access to results. Well done! > > A couple of things bother me. Perhaps you can fill me in: > > 1. The low quality of sphere rendering. > Simply how can I make them smoother? > Is there a parameter to increase the quality of GL being called when the > situation demands it? > > I realize there are many applications where speed and interactivity are > most important. Currently the smoothness of a Visual sphere is determined by a heuristic based on how far away the sphere is. We think that the current setting errs a bit too far on the side of speed at the expense of visual quality. There is currently no option for controlling this. > 2. Is transparency possible?. > I though openGL did it, but I don't see a way to control this. > Transparency would be very helpful for all sorts of special visualization > and intersections etc.. Control of opacity is planned but not currently implemented except in the label object. >> > Has anyone done any convertors or tried any import export with VPython? >> >> Others have asked about this, but this is low on our own priority list >> for building into Visual. Obviously you could rather easily create >> routines that would generate such descriptions along with creating >> Visual objects. So you would write a sphere routine that writes out >> POVRay specs for a sphere, then creates a Visual sphere. > > OK. > One format which might make sense is .NFF > There are some useful format descriptions at: > http://www.swin.edu.au/astronomy/pbourke/3dformats/ > > A more general concern is to understand why VPython needs to uninstall an > existing Python? > At least _please_ can you offer an explanation somewhere. Otherwise it > sounds too much a M$oft piece of advice. > > I did not uninstall my existing Python for fear I would loose some work, > time and functionality. I did not want to disturb my present happy > working conditions [if it aint broke dont fix it].. So I installed > VPython in its own folder on another drive to test it out first. Then I > started copying the cvisual.dll into other python installations I have to > seee if I could get it to work. > > I launched it successfully from IDLE but when I closed any VPython window > it would crash the Python version. What is going on here? I know you say > it would break.. but WHY? > > Some more information from you about WHY you recommend VPython only and > what the dependencies are would be really helpful. I would love to get > VPython running from wxPython, BOA, wxDesigner for example. I am not sure of what all the issues are, and we do hope to get around this in the context of Python 2.0. The main problem was that with Python 1.5, Tcl/Tkinter (used by our particularly novice-friendly version of IDLE) was completely separate, and this caused installation and execution problems. To get around that we bundled them together, as has also been done as I understand it in Python 2.0. > On my system I have pythons inside of Blender, Alice, 2 zope installations > [zope versions] and regular Python 1.5.2 + anyday now will have Python 2.x > For zope for example I find there are binaries for libraries which prevent > or slow me down from de-installing or upgrading too quickly. Like most I > want to test it out before I commit. > I usually use PythonWin on Win98 in my Sony laptop, but also have Linux on > another machine. > How can I get PythonWin to run with VPython? I have no idea. > The reason is that PythonWin interactive shell [scintilla] has inline > prompting of class objects, params etc. Very useful. > > Good luck with your work and looking forward to more experimentation with > VPython > - Jason > > > > |