Hi All
 
I'll be writing my own "editwizard architecture" just as soon as I free up enough time to do so - although I haven't contemplated the front-end yet.  My rationale is that I'm not going to be able to influence the new developments to include what I want and my view on what an editwizard should be is my own.  If we can, however, agree on the architecture, then it may be possible to have the future editwizards as a pluggable front-end into my architecture.
 
Effectively what I'm looking at doing is along the lines of an extension to Node/NodeManager, using java beans on the spring platform.  The beans will provide generic logic to handle complex objects (and relations between other complex objects) and each instance will be created as a new complex object - driven by it's corresponding meta-data tables.  A transactional component will perform validation on the object relations and "low-level" node updates - driven by actions specified by the client (add, update, delete, cache in memory, whatever).  Certain relations could possibly invoke actions - somewhere to put business logic.
 
The philosophy for me is that getting the data in is only one aspect - there should be a facility to use the neatly-defined relation definitions that exist in the editwizard, for playing back content as well, without re-inventing the wheel.  Apart from this - I tend to deal in nothing but complex objects, therefore complex code.  Specifying keys for data retrieval, helper html wrappers for lists (through something a bit more dynamic than xsl) for embedded content will also be useful - some sort of templating engine.  Standard websites would then become more of a configuration exercise than the massive coding effort that is currently required.
 
I haven't started the serious thinking yet, so any idea contributions at this stage would be welcome. 
 
cheers
Emile
 
_______________________________________________
Developers mailing list
[email protected]
http://lists.mmbase.org/mailman/listinfo/developers

Reply via email to