A journal that focuses on Flash Platform development, and a little bit about what I am up to on any given day.

Wednesday, July 06, 2005

Flex, Avalon, the oncoming storm

I was just reading this post by Mike Rankin:


He discusses the oncoming battle between Avalon, and Macromedia's Flex product. In summary, he states that Flex is aimed at the enterprise (2% of business), as it's price point is so high per processor. He worries that Macromedia must change their pricing policy as Longhorn and Avalon start to appear on the scene.

I have spent alot of time considering Flex for RIA development. We did consider it for the SNL Merger Model application for SNL financial, but chose not to go with it. The SNL Merger Model would have certainly been a candidate for Flex deployment but the issues that made the client and their in-house development team turn it down were:

1/ Pricing, they could not justify the high cost per processor. They thought it was much too high. This became a big issue because of the next point:

2/ Flex shouldn't be a server. Some may really disagree here, but their thought was: "Why pay 10k per processor for a component assembler?"

Even if we had used Flex, 90% of the development was in creating custom components and skinning. This means that 90% of the job would be done in Flash anyways.

The Flex model certainly would have sped the process in terms of data binding, component assembly, as well as integration with source control and team development, but still not worth the cost as a server. The java developers I was working with thought that the tool should work exactly as is, but compile on the client instead of at the server.

So in the end we developed it all in the Flash IDE, and Eclipse. They were already running JRUN on the back-end meaning we had Flash Remoting at our disposal as well.

So to conclude, I am certain that Macromedia's Zorn project will like address this. Macromedia's original strategy was to create a new entry point into the Flash Platform for application developers coming from other backgrounds. There is really no reason why Flex should be a server in my mind. If it must stay a server, then I would tend to agree with the original poster. The price point is too high.