Access Languages are Tied to Schemas

[Mark Baker]

On Thu, 20 Nov 1997, Joe Lapp wrote:
> I think we are in agreement (I disagree, we agree).  An XML document is
> capable of representing any object and all aspects of that object.  But
> an XML document isn't the object it represents.  You have to deserialize
> that document back into an object before you have the fully featured
> object again.

That's one way of doing it of course, and very useful for some 
applications, such as dynamic binding of data to behaviour ala compound 
document frameworks (and the new beans activation framework).  Think of 
this as serializing a class.

But in *many* cases, you just want to make the *object* persist simply, 
perhaps even on the machine with the browser.  This is especially 
suitable for agent systems; you bring the ability to persist along with 
you instead of attempting to store it "behind" you.  It's a move away from 
TP-monitor style ACID transactions, and towards a more "make forward 
progress" means of distributed computing.  Object groups are a good 
example of this.

Certainly though, both tools should be available to us.  We shouldn't 
try to shoehorn everything into a single solution when that solution 
isn't general enough for all of our needs.

But, I've got the feeling that we'll be doing a lot more of one than the 
other before too long.  YMMV.  8-)

>An XML repository could store those objects (in their
> XML document representation) and even keep the relationships among those
> objects, but it does not animate those objects.  The objects are alive
> when they are deserialized on the clients.  To get a repository to
> animate the objects you'd have to make the repository a bit more than
> just a repository.  For one thing, you'd also need a JVM.

Which isn't too difficult nowadays, especially when so much is being done 
with the browser (as it should).  And a JVM is no different than requiring a 
script interpreter.

> As a side note, you mention that in JDK 1.1 every object is a bean.  I
> thought beans had to be serializable.  Are you saying that in JDK 1.1
> every Java object that ever gets created is serializable?

You're right of course.  But you'll find that anything that "makes sense" 
to serialize, can be.

MB
--
Mark Baker, Ottawa Ontario CANADA.                Java, CORBA, XML, Beans
http://www.iosphere.net/~markb               distobj at acm.org  ICQ:5100069

   Will distribute business objects for food.


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)