On Saturday 30 August 2003 23:26, Miguel wrote:
> I sat down yesterday with intensions of getting Jmol to work better with
> the CDK libraries. Per my previous message, I ran into some difficulties.
> Thanks for the responses.
>
> In addition to the class hierarchy 'impedence mismatch' between jmol and
> cdk, there is still a problem with CDK and the atom attributes. That is,
> the vdw/covalent radii and color are not being assigned by cdk. This
> problem was fixed in CdkJmolModelAdapter for the app, but still doesn't
> work for the applet. Presumably there is a problem locating the
> appropriate .txt file ... this has been a recurring problem.

What about putting this text file inline in the class... i.e. as a String... 
that would solve all possible security things... The build.xml could put the 
current txt file form CDK in the appropriate class...

> Since I could not work on the cdk stuff, I decided to take a different
> tack. I switched the applet to use a SimpleModelAdapter which only reads
> .xyz, .mol, and .pdb files. The result was an applet which was
> significantly smaller. Previously the applet was >650K ... the reduced
> applet is 256K.
>
> For web-based applications where end-users will need to download the
> applet this 400K is a tremendous difference.

Which libraries made the difference, i.e. which one could you remove?

> We need to talk about this some more to see what can be done. Otherwise, I
> will feel compelled to argue for splitting the applet into two jars so
> that people who are running simple applications will not have such a heavy
> cost in download time.

Yes, we need to have a close look at that for the new code base...

Egon



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Jmol-developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jmol-developers

Reply via email to