From my limited point of view, such tools even by grid transaction hype can't replace the exiting enterprise development environments. And also we already have clustering as vertical grid against ActiveGrids horizontal grid. They just mean money saving by horizontal grid (and the question is how could we make cheap grid with the logically 1024 machines on the top of L part?! I mean Linux of LAMP). But in the mission critical and enterprise systems the first problem is not money, that's nature complexity of the system. And I think scripts languages such as Robby and Python that are exist in LAMP environment are unable to create true business logics.

I wonder why some folks choose "web 2.0 application" name for ActiveGrid apps just for its SOA pattern implementation and if web 2.0 applications are just about fast view creation I choose model 1 applications and JSPs!

On the other hand, we need reliable platform in the E part of JEE when we work with it. But open source tools always lack reliability (it's just my idea). They have no valuable documentation and support at start point(look at ActiveGrid docs!) and these two are really essential for an enterprise software development team. Open source tools are really helpful as toolset in enterprise projects but they can't be the platform itself. they're just a part of my framework. As I remember there is more than 200 billion dollar invest on JEE by Sun and the other companies like IBM, Oracle, BEA, …. and i guess LAMP will loose enterprise software market without money!

There are some other questions about performance (esp when all the things are web services) and security and so on. So, I don't change my mind and tell you that's just "open standard vs. open source" war! Don't worry about it. We had passed the .NET waves ;)