Appistry, the start-up whose motto is “we scale ugly apps” and claims to offer “Googlization for the masses,” is stepping off into what it calls “open distribution” of a Community Edition of its grid-based application platform, its alternative to Java EE application servers and .NET.

The new initiative offers for free a download of what is currently the company’s core product, Appistry Enterprise Application Fabric (EAF).

In time, the company intends to give its commercial product more enterprise-grade features like enhanced management tools that handle larger, multiple fabrics but as it is now EAF Community is what FedEx, say, is using for its mission-critical Roads Project to dynamically route its trucks rather than have them follow fixed routes, a crisis management solution in the face of skyrocketing gas prices and a weakening US economy. Other commercial clients include Sprint, Northrop Grumman and GeoEye.

The only restrictions are that it can only be used on a maximum five servers or 10 CPU cores. Otherwise there are no strings attached.

The EAF Community Edition can be used for production deployment of applications as well as development and testing. It can also be used in perpetuity.

Appistry CEO Kevin Haar says the Open Distribution program “combines the best of open source and commercial software.”

The company is giving away $20,000 a year worth of software, based on its usual price of $3,750/cpu/year.

The start-up has also set up a P2P portal where developers and architects can go for peer support, RSS feeds, documentation and social networking.

Appistry quietly noised the Community Edition round among its contacts in December and reportedly attracted 150 downloads from the preview. Appistry is expecting the seeding exercise to broaden its base in the mid-market and eventually give it a chance to upsell.

The widgetry gives developers a way of building highly scalable applications in Java, Spring, .NET or C/C++ for extreme transaction processing (XTP), software-as-a-service (SaaS) and cloud computing.

It’s supposed to make it very easy to scale out CPU- and data-intensive apps across virtualized grids or clusters of commodity hardware.

Applications are supposed to simply inherit transparent scalability, software-based reliability and automated management without manual coding.

According to Gartner VP of research Massimo Pezzini, Java EE applications and .NET are “wearing thin when it comes to supporting the growing transactional workloads generated by modern service-oriented and event-driven architectures.”

He figures grid-based application platforms and XTP middleware “will become the norm for most back-end software.”