[Biojava-l] persistence - and the problems with it

[Gerald Loeffler]

Mark Schreiber wrote:
> 
> On Wed, 3 May 2000, Gerald Loeffler wrote:
...
> 
> Along the 'ML lines has anyone ever used UML to make a object database?
> I've only ever heard vague things about it and I would like to know what
> others experiences are, ie is it worth looking into at all?

The object databases I've worked with (mainly POET and ObjectStore) work
on the level of Java code (source or byte-code) to extract the
information they need (to build up the object schema). In turn, they add
magic to the generated byte-code so that your classes "know" how to talk
to the database.

There probably exist "integrated" tools along the lines you describe
(i've never seen one), i.e. they could generate the object schema not
from Java code but from a UML class diagram. This can easily be imagined
since you can also generate very reasonable Java code from such an UML
class diagram - i.e. the information is there. Anyways, somehow the
"enhanced" byte-code for you persistence-capable Java classes still
needs to be generated...

(In a similar fashion, The Theory Center bundles a plug-in to Rational
Rose with their Enterprise Java Beans that helps generating most of the
EJB-Java-code from UML diagrams.)

	cheers,
	gerald

> 
> Mark
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Mark Schreiber                  Ph: 64 3 4797875
> Rm 218                          email mark_s@sanger.otago.ac.nz
> Department of Biochemistry      email m.schreiber@clear.net.nz
> University of Otago
> PO Box 56
> Dunedin
> New Zealand
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-- 
   Gerald.Loeffler@vienna.at _________________ Software Architect
   http://www.imp.univie.ac.at ____ http://www.daemonstration.com
   OOA&D, Java, J2EE, JSP, Servlets, JavaBeans, ODBMS, RDBMS, XML