I was disappointed by this article. At the most basic level, it requires serious editing. Take as an example the ending, where the author used both ‘JWS’ (the name of the standard) and ‘JSW’, by mistake. I understand that keeping Java three letter acronyms straight is difficult, but surely it can be done within the context of a single article.
On a deeper level this article was long on marketing hyperbole, and short on technical detail. For example, we are asked to believe this marketing drivel; “…robust, maintainable, and interoperable Web services”, which is left generally unsubstantiated by both the article and the JCP specification. While on the technical level being left with brief, vague, technical examples that fall well short of the standard required to assure any trained professional that the _standard_ is any more than vaporware.
The article gives a database example of an insert, without any regard to transactions or rollback. It also provides no mention of database query protocols, or high-throughput issues such as how this relates to caching, or standards issues about how this relates to the four other Java database access standards.
Which brings me to my final point, which is that it is deplorable that BEA should be throwing into the mix yet another database access protocol. When it is the very lack of an acknowledged database access production standard that is the foremost serious threat to the Java server market. BEA, which represents the leader in the Java application server community, needs to decide on and evangelize a single database access standard for production systems. It needs to do this today or else both BEA, and the Java web server community at large, will lose against the united front of Microsoft.
BEA, if you are listening, here is some simple advice:
1. Decide on a SINGLE database standard for production servers; JDBC, EJB, JDO, whatever. You are losing customers because with Microsoft there are no technical disputes and they can get on with the job of writing software that solves customer problems.
2. Proselytize it.
3. Then put something together, like the novel code generation scheme presented in this article, to make using that standard easier.
Use whatever marketing cruft you like to describe this ‘three step plan for success’. Just do it! Microsoft already has. They are months ahead with their technical and marketing efforts, so it is catch-up time.