The Enterprisey one thinks that businesses should be rewriting their Cobol apps in Java. Hmm. I wrote a long missive on why rewrites are usually a bad idea here; in particular, I think taking a group of Cobol developers and trying to move them to Java will be an "interesting" experience.

Here's the question a serious architect would ask himself: Assuming you have working code in Cobol, compare the cost of adding functionality in Cobol to the cost of a rewrite plus the cost of adding functionality. The non-clueless ones will train someone in Cobol....