[postgis-devel] State of the postgis hibernate union?

[Chris Hodgson]

> (3) Geoserver/HibernateSpatial/Postgis
> We will starting the investigation on this soon. No results yet. Of course
> the nicest and cleanest solution. My idea is that HibernateSpatial is doing a
> very good job but it is very recent software and not yet used in many
> systems. And I think they might be the first with a proper implementation. 

I don't know how you are going to skip geotools for this approach? 
Presumably you will still need a datastore. This won't change the 
problem you discovered in (2) - all this is, is a standard 
implementation of part of what you would need to do for (2) ie. 
postgis-enable hibernate.

> http://www.foss4g2007.org/presentations/view.php?abstract_id=118
> Not investigated, to me seems less clean than HibernateSpatial because they
> reverse engineered from the database instead of starting from the domain
> model(JTS)

Again, this is exactly what you would have to do for (2), the choice of 
reverse-engineering vs. annotations is really moot once you have decided 
on your model - it's just a question of whether you prefer to write SQL 
or Java, and then generate the other. In truth, we reverse engineered 
primarily because our data model is being accessed by other software 
than just hibernate, so it had to be exactly the way it was. We had 
already designed it with specific goals in mind, including high 
performance for our use cases, and we just wanted to use hibernate to 
access it. So reverse engineering was a very effective way to go.

Chris