Does anyone in this list is developing or developed a system
involving XML technologies and Adaptive Object Models (ref: http://www.adaptiveobjectmodel.com
). As I said earlier I am working on an XML based language that can precisely
be used to that purpose. It would be useful to share respective experiences and
possible synthesize our respective works.
Up to now, I used PDML (the language I designed)
To move a collection of objects (instances) from a
server to a client. Hence, I broke the actual convention that a browser is a
dummy device which can only “present” information. Instead, I used
PDML to serialize, transport and re-construct the instances and their ontology
on the client. It works well on IE, I am having a hard time to reach the same
results with Mozilla. This is because in IE allows to access three document
models at the same time 1) the XML DOM, 2) the HTML DOM, 3) the XSLT DOM. I
used only the XML and HTML DOMs. The XML document contains the model, this
latter is modified at run time by ECMAScript objects (translated from PDML
using XSLT), then if the model is changed it is returned to the server to
update the permanent storage (i.e. an object database). I did not stored the
objects in Relational DB but I heard that some did. If nayone did that, I would
be interested to ear from your experience.
To encode an ontology and create instances from that
I am having difficulties right now to deal with:
a) When an object is not transported to the client but is
still resident on the server. I want them to communicate through a web service
defined with a WSDL document. The service may be SOAP or REST based (since WSDL
allows both to be defined in WSDL). In that case, no code is generated on the
client, just the object’s signatures to call the object stored on the
b) When an object is transported to different agents and
different code has to be generated.
The difficulty is not to do it but to express it in a
notation that can be as much as possible “usable” and “readable”
. I do not want to replicate the mess of OWL Is there any usability study
somewhere about XML based languages?
Didier PH Martin