Re: [Cdk-devel] RFC: list of core classes

[Egon W. <eg...@sc...>]

On Thursday 23 May 2002 11:25, E.L. Willighagen wrote:
> On Thursday 23 May 2002 09:23, Christoph Steinbeck wrote:
> > Anyway, beside the AtomicNumbers class (which, if you think it further,
> > would logically bring AtomicWeight.class and classes for other atomic
> > constants with it) I think the list is ok. I've done a lot of
> > algorithmic work with the CDK now and the existing core classes were
> > always sufficient. How about the CDKConstants?
>
> I would say that would be not in the core classes.... I would like to
> devote the core classes just for data storage, not for actual
> information... But, that is arguable...

With my GCJ efforts I realized that several core classes violate the proposed 
RFC. Some of them rely on non-cdk-core classes to get compiled.
I found out that these classes are also needed to get things compiled:
org.openscience.cdk.CDKConstants
org.openscience.cdk.ChemObjectListener
org.openscience.cdk.event.ChemObjectChangeEvent
org.openscience.cdk.exception.CDKException
org.openscience.cdk.exception.NoSuchAtomException
org.openscience.cdk.tools.ConnectivityChecker
org.openscience.cdk.tools.PathTools

Either these need to be included or we should file bugs. Esspecially, the
need for the cdk.tools classes seems to be faulty...

Egon