Integrate with XText/JET/Xpand: Drop Freemarker as generatio

[luxspes]

Hi!
I like Freemarker, I really do, my first web application was built with Velocity, and I enjoyed it a lot back then, and Freemarker is even better,
but support for Freemarker in Eclipse is very weak, and improving it is just not a priority for JBossTools development team

Why not use something designed precisely for code generation and DSL definition, and that is being supported and enhanced by a team that is specially interested in improving the facilities for DSL definition and code generation inside Eclipse?

Have you guys thought about replacing the current Freemarker based templates with something like JET or Xpand? Or perhaps even creating a "Seam DSL" based on XText?

That way you could leave the problem or maintaining and improving the code generation platform to the XText/JET/Xpand team (and of course contribute back) and at the same time concentrate in improving the features of the generated application (more functionality, support of other languages, user-friendly and concise Seam DSL, etc)

What do you think about this?

[maxandersen]

Does it run from ant without massive dependencies and overhead ? If not, then no.

And are you willing to migrate all our templates to JET or Xpand ?

Stuff like this does not happen without someone with the time to do it. Plus the fixes we got waiting in freemarker component did not work when I applied them last (too many changes in one patch).

Care to make them smaller so we can apply them ?

[luxspes]

Does it run from ant without massive dependencies and overhead ? If not, then no.

Not sure...

And are you willing to migrate all our templates to JET or Xpand ?

Maybe, I know them pretty well (having done extensive modifications to them) , but I am not sure I will have the free time :'(

Stuff like this does not happen without someone with the time to do it. Plus the fixes we got waiting in freemarker component did not work when I applied them last (too many changes in one patch).

I did not implemented those, I have not experience in that kind of programming (that is why I like the idea of using Xtext: that work is already being done by someone else with interest in creating the ultimate code generation tool)

Care to make them smaller so we can apply them ?

If I had implemented them, or had the level of knowledge required, I would try... :'(. But that is precisely the point, nobody seems to care-about/have-spare-time for improving support for Freemarker in Eclipse: Why not leave that job to a team of people that actually have the creation of such a tool as their main goal?

[maxandersen]

Because freemarker works and it runs from eclipse and ant and doesn't require fancy tooling to be useful.

Does that mean I don't think we need better freemarker tooling ? no of course not, I would love it to be improved.

Moving to xtext or something else would require much more effort than fixing a few issues in the freemarker editor AND it would affect *all* users of hibernate tools so i'm really not convinced that doing such a move is a Good thing.

Note: Hibernate tools exporters is not in any way tied to use freemarker - you can create your own exporter that calls out to some other template framework.