Am Mittwoch, 13. September 2006 10:50 schrieb Nicolas Modrzyk:
> Result of our work with John today is in here: (actually john's work,
> I am just taking the snapshots).
>
> I feel that what has been achieved so far is the following:
> - reuse of the architecture of the xml encoder, with extended testing
> and proper use of reflection to encode/decode java beans

I had thought the 'me' field of the MrBean class would be coded as node data 
named 'me' instead of 'Me'.  Is the capitalization indended/typical?

> - implementation of a wrapper layer to be able to use both jackrabbit
> nodes or magnolia nodes, (or as a proof of a concept mock nodes ...)

By 'magnolia nodes' do you mean magnolia node types?  Or do you mean that the 
magnolia Content class (is/can be) involved in node creation?

> - encoding with the jackrabbit nodes looks like it's all working
> (need a bit more testing)
> - encoding simple beans on magnolia nodes is working
> - encoding of maps on magnolia nodes behaving a bit strange but
> getting there.
>
> Now with this wrapping layer, I wonder if we need to still care about
> node types ... for a demo implementation at the geek meet, I would
> say, being able to encode/decode in any of the magnolia workspace
> (independent of the node type, thanks to the magnolia API) would be a
> fair good start.

Personally I think it would be enough to produce magnolia node types, since 
they work best with the rest of the system.  For example they show up in the 
jcr browser while other node types don't.  And subtypes of the magnolia nodes 
don't work as plug and play alternatives to them.  It is best to have content 
that simply interacts well with the rest of the system.


----------------------------------------------------------------
for list details see
http://www.magnolia.info/en/magnolia/developer.html
----------------------------------------------------------------

Reply via email to