This post originated from an RSS feed registered with Java Buzz
by Carlos Perez.
Original Post: The Unexpected Answer to Java's "Data" Standard
Feed Title: .:Manageability:.
Feed URL: http://www.manageability.org/blog/stuff/rssDisabled?portal_status_message=Syndication+is+Disabled
Feed Description: Random thoughts on the manageability of complex software.
One of the gaping holes of the Java platform is the absence of a decent "Data Integration" API. This statement may come as a complete surprise for the proponents of either EJB or JDO. However, I'm going to tell you why both are inadequate and I'm going to tell you what the "powers that be" are really doing about it.
EJB Entity Beans is a whoefully inadequate API for working with corporate data. If you've got money to spend you can buy Rod Johnson's book that'll explain it in detail. Now if you don't have money, then you can read through "EJB's 101 Damnations" and pick out the places where it refers to Entity Beans.
JDO is also a whoefully inadequate API for working with corporate data. JDO proponents claim that its a better API than EJB Entity Beans, that may be true but it's an API that designed by OODBMs experts and not relational database experts. I've ranted before about the flaws of JDO. However, the most telling indicator is the lack of support of this standard by any of the major relational database vendors. Matter of fact, at JavaOne it was painfully obvious how much resentment Oracle had against JDO.
The corporate market being such a big pie you would think the vendors would be scrambling to fill this hole. Well that's precisely what Sun is attempting to do, that's one of the main themes of this year's JavaOne conference.
Sun's solution is Rave. The rave strategy is to provide an IDE based environment and a set of APIs to make it easier for corporate developers to assemble systems. One interesting aspect is its choice of the JDBC Rowsets (i.e. JSR-114) as the mechanism for retrieving data. That's quite telling since it's not EJB and it's somewhat similar to how Microsoft does it.
What about IBM and Oracle, the two biggest players in the corporate space? Well it turns out they've submitted a pretty exciting JSR. That is XQuery API for Java (XQJ), JSR-225 for reference. Initial expert group members consisting of heavyweights like BEA, Sun, DataDirect and Sybase, none of the minor vendors promoting the JDO specification. Matter of fact, I couldn't find any mention of the upcoming "JDO 2.0" spec. at JCP, can someone tell me where it is?
Read the writing on the wall, these are two API's one for quick and dirty access of data and the other for more sophisticated federated access. The big question in my mind, I've asked it before, is "who in the open source world is building a XQuery engine?".
Previous Topic
