Re: [Jmol-developers] Re: jmol cvs tip status

[E.L. W. <eg...@sc...>]

On Thursday 01 May 2003 12:51, Peter Murray-Rust wrote:
> At 11:33 01/05/2003 +0200, Miguel wrote:
> > > The best option is probably to renumber the atoms when a new frame is
> > > added to  the ChemFile (and if needed, convert all CDK atoms into Jmol
> > > atoms on the  fly...)
> >
> >I don't like the idea of converting them on the fly.
> >Since we are committing ourselves to CDK we should adopt its model.
> >And, if there is a problem then we should fix it in CDK.
>
> I don't know the exact details, so please ignore this msg if I'm wrong...
>
> Using atom numbers is likely to be fragile. 

Agreed, but note Miguels comment on performance.

> There is not only the 0/1
> dichotomy but you have to make sure that all places through the program are
> aware that atoms have been renumbered. I assume that is why designs such as
> Iterator are recommended.
>
> The CML approach is to use permanent IDs for each atom, and I know this was
> being fitted to CDK (by Egon?). What is the current status of these? Is
> this a useful approach for Jmol?

The ID system should work. Though there is no explicit test for it at this 
moment. Don't think this is usefull for Jmol though... because there is no 
"quick" algorithm to get atoms by ID...

Egon