The other night I commiserated with a friend over Sun's sale to Oracle, with both of us agreeing gloomily that this most likely means the end of Java as we know it. Actually, it may mean a whole new beginning for Java -- or a whole bunch of new beginnings -- and that's exactly the problem.
My big beef with Oracle acquiring Sun is that it will mean the death knell for the software development teams currently there -- that all the people who are synonymous with the development of Java (and MySQL, and GlassFish, and OpenSolaris, and all the rest of them) will be pared down, dismissed, leave of their own accord, or just plain give up. It's not as if Java development will stop completely, but it'll be bad enough to damage work on increasingly important things like fine-grained threading.
I speculated that what might happen is a forking of Java into at least two branches: the official branch maintained (or, rather, not maintained) by Oracle, and an unofficial branch under different sponsorship. Who would take the weight for a project like that except for ... IBM? They did, after all, have their own open source Java compiler, Jikes, and even though it's since been superseded by other work done since.
Frankly, I'd rather see IBM take up custody of Java completely and for real -- not merely as a response to what will probably happen to it in Oracle's hammy hands (cheery pronouncements notwithstanding), but as a way to keep it on track for real. I don't want to see Java devolve into something that is wholly part of the "Oracle stack"; it deserves to be part of everybody's.
"Just fork it" is not an answer. The people doing the forking have to know what they're doing, and the number of people I can think of up to such a job is vanishingly small. And until proven otherwise, I reserve the right to doubt they work for Oracle.
InformationWeek has published an in-depth report on the use of business software on smartphones. Download the report here (registration required).