From: Miguel <mi...@jm...> - 2007-12-25 13:33:56
|
William wrote: > Thoughts? Very interesting. The org.jmol.g3d classes were designed to stand separately and to be usable apart from Jmol. Frankly, I expected gamers to adopt/adapt it ... but so far that hasn't happened. > 1. What is the relation between javaview and jmol? None that I am aware of. For my own part, I am not in the academic world and am not familiar with Javaview. I did not encounter javaview back in 2002 when I conducted my survey of java-based 3d graphics engines to use in Jmol. > 2. What is the relation between java3d and jmol? None ... by design. At the time that I started writing the Jmol g3d graphics package it was pretty clear that java3d was losing mindshare ... if it ever had any. There was limited hardware/driver support and (it seemed) only one or two developers working on it within Sun. More importantly, java3d placed constraints on the client that were unacceptable. My goal for Jmol was that it could run as an applet on *any= * java-enabled web browser without requiring a special client install. The fact that java3d required client installation of libraries made it unacceptable for Jmol's target audience of casual users and students. OpenGL had similiar issues back in 2002 ... and, to the best of my knowledge, many of those java-related issues still remain. > 3. Are you aware of jmol being used for applications outside of > chemistry? E.g., applications > to general mathematical visualization? Not that I am aware of. Bob Hanson and the current Jmol development team may know more. As I said above, I expected someone to pick it up and use it for interactive java-based games, but so far that hasn't happened. Actually, I *am* aware of at least one project where they are interested in using org.jmol.g3d within their own application ... but their bio-chemistry so they are not looking at it from a mathematics or compute= r science perspective. > 4. Do you have any advice as to how jmol could be best used for more > general > mathematical visualization? We can use/abuse the existing > molecular modeling > interface, but perhaps we should somehow coordinate with you and > modify > jmol itself to have more general functionality? My time is quite limited, but I would be interested in this. I am sure that Bob would be very interested in pursuing this. > 5. Viewing examples here: > http://www.javaview.de/demo/index.html > Do you think we can do as good -- or better -- then they do, but > using jmol? I don't see any reason why not. In all of the samples shown at http://www.javaview.de/demo/index.html the number of discrete points involved is quite small ... relative to the ten= s of thousands needed for viewing proteins. I have always been concerned about Jmol's ability to render triangles. Fo= r rendering molecules I put my focus on fast rendering of spheres and lines= . Relative to other 3d APIs, there is limited little support for rendering triangles. There is, for example, no display-list mechanism. I think that would be a problem for potential gaming and virtual-reality applications. (Bob Hanson other jmol-developers might have done more work in this area over the past 18 months that I am not familiar with.) The surfaces that I see at http://www.javaview.de/demo/index.html look simple and clean. I'm sure that the Jmol graphics engine would work just fine. > MANY MANY thanks for the huge amount of work you > have all put into creating a very > powerful easy-to-use and flexible 3d visualization toolkit! I am exceedingly pleased that you have found the org.jmol.g3d Trying to maintain an objective perspective ... JavaView looks very powerful and quite mature. It seems a shame that you can't build upon the seemingly outstanding work of that project. You said that licensing constraints were too onerous ... Can you talk about that? Having another _user_ for org.jmol.g3d would certainly be valuable. It would be a big step in helping to round-out the API and to make it stand alone on its own. The development of the org.jmol.g3d API would benefit tremendously from the input of someone with formal OpenGL graphics training and and experience. Your interest in using Jmol for another branch of scientific education is the best holiday present I received this year! Miguel > -- > William Stein > Associate Professor of Mathematics > University of Washington > http://wstein.org |