I am still cruising through the Standish Group report that claims "" and I came across one very interesting gem that I think helps to explain some of my previous thoughts on why IBM isn't pushing any open source for SOA.
I theorized thattools as many of them meet the full requirements that enterprises look for.
I still think that's true and I think a big part of that is because open source products have improved so dramatically they set a baseline for functionality and cost where users are not willing to pay the extortion fees that big vendors charge for upfront licensing.
Using open source creates an instant baseline for any project. This baseline not only includes working programs, but also requirements, documentation, support procedures, and education material. In many cases, especially in infrastructure software, the baseline is a fully developed and working system. Many applications and service components are fully functional and can be used immediately. Other applications and components provide a firm baseline around which to develop a more elaborate system. Using DARTS and other data, Standish Group has concluded that 11% of all new commercial software requirements are being satisfied by open source solutions and components. This does not include the application service providers (ASPs) that are using open source software to service their clients, with products ranging from ERP to project management.
This baseline notion is also interesting as products like JBoss used to be considered just for development with BEA for production, but over the last 2 years or that sentiment has changed with lots of JBoss is production.