> I have downloaded the source code for Jmol, and currently am in the
> process of looking through how you render all the elements, as well as =
> importing and exporting features of Jmol. I am very interested in being=
> able to work on this project to further enhance the capabilities of Jmo=
> However, I would like to understand more about the project first. If yo=
> don't mind, I have a few questions for you.
> 1) I have seen that you clip the diameter of spheres in the Sphere3D.ja=
> part of the g3d package to 128 maximum using the maxSphereCache variabl=
> What unit does the 128 refer to? Is it angstroms?
This '128' refers to on-screen diameter of a sphere, measured in pixels.
The job of Sphere3D.java is to render spheres centered at screen
coordinates in 3D with the specified diameter.
For spheres >=3D maxSphereCache a different rendering routine is used.
The cache is one quarter of a sphere, with an int value representing each=
pixel. So, a sphere of diameter 128 has a radius of 64. 64=5E2 =3D 4096 *=
* 4 =3D 12K bytes used that cache entry.
> Do you have everything
> defined in a single file or are these definitions spread over each of t=
> individual functions?
I don't understand the question.
In this particular case, the value 128 is local to the Sphere3D
implementation. Noone else needs to know about this value. There is reall=
no reason why anyone would ever use it.
If we didn't need to worry about old 1.1 JVMs, and we could assume lots o=
memory, then we might bump this value up.
> 2) How does your unit notation and RasMol's unit notation compare?
> Meaning, for 100 units in RasMol, how many angstroms or units do you ap=
> in Jmol?
I'm not sure I fully understand your question.
RasMol defines a special 'RasMol unit' ... there are 250 rasmol units per=
angstrom ... an unfortunate design error in my opinion.
Jmol fully supports RasMol units, so if you say 'spacefill 250' you will
get a 1.0 angstrom radius in RasMol, Chime, and Jmol.
I recommend that people writing modern scripts always specify values in
decimal values of angstroms, rather than propogating this unfortunate
> 3) This is a question about software, rather than Jmol code. Do you hav=
> preference for using a specific Java development API such as Eclipse?
I use emacs.
> Is your source code setup to work with one of these environments?
Some other developers (nico + ?) use Eclipse.
> Thank you very much for your time,
Thanks for your interest in Jmol.
Open Source Molecular Visualization