Re: XObject (was was was)
Peter Murray-Rust (peter@ursus.demon.co.uk)
Thu, 08 Oct 1998 19:47:47
At 14:30 08/10/98 -0400, Bill la Forge wrote:
>There seem to be a lot of problems with DOM:
> - an incomplete set of interfaces
> - excessive overhead for some applications.
>
>But from the perspective of a component programming
>environment, where each application object has a
>peer member (or is a member) of the DOM tree,
>there seems to be a big advantage:
> a standard navigation api.
>
>So having a standard for converting an application-specific
>document reference into application-specific elements of
>a DOM tree means that we have decoupled the
>application code from the tool that builds the DOM tree.
>
>It also then creates a standard new kind of component--
>an XObject, which could be a really good thing indeed
>if they are easy to write.
Fully agreed. If it helps anyone, they are welcome to download my code from
xml-cml.org. The simple subclassed nodes are in jumbo.xml.data.*Node and
are things like FloatNode, EnumerationNode, etc. More fun things are
jumbo.xml.graphics.*Node with some simple graphics primitives. [I think
this isn't the latest version mounted]. Then there are
jumbo.cmlxml.MoleculeNode, etc. Most inherit methods, often overridden,
from jumbo.xml.data.DataTypeNode.
The public methods at least give some idea of my mental struggles and they
may be helpful...
P.
Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg