I have to admit that I was one of those people who sat through the entire 5 hours of the Oracle/Sun presentations the other day (it seemed longer!) In a way it's sad to see Sun finally set over the horizon, but in another way it has been inevitable for a while and the whole process of the acquisition really couldn't have been drawn out much longer. So there we are: gone is Sun and in its place is Snorcle (or is that Oracun?) But where does this leave the industry as a whole? Well Sun had quite a portfolio of hardware and software, so unlike the BEA or PeopleSoft acquisitions this has potential wider ramifications. But if you listened to the presentations then it's almost as if Sun hasn't really gone away but Oracle is just injecting a lot more cash (and people) into the business.
Well not quite. As many had predicted, the Sun middleware stack (Glassfish and family) has pretty much been banished back to the Reference Implementation plane, because it really wouldn't do to have two competing solutions. Furthermore, in a Geronimo-like approach, it seems that Glassfish will be less capable than Web Logic (read: crippled) because the "departmental" solutions to which it will be directed don't need them (read: bait-and-switch). There were no strong statements about the future of the JCP ("... making the Java community process a more participatory process to people from a variety of different organizations." is too hand-wavy for my liking.) And I think when the various presenters referred to Oracle's "open" platforms they really meant "standards based", since when we talk about "open" in JBoss it tends to me so much more than that.
So at first glance it seems as though this announcement leaves the Java landscape relatively unchanged, if not a little stronger (if you're a Sun customer). But the reality is a lot different, particularly if we look at open source middleware. Before the deal closed there were 4 main EE application server vendors, Red Hat/JBoss, IBM, Oracle and Sun. IBM uses Geronimo in a bait-and-switch arrangement, with their mainstay solution closed source. Oracle, with WebLogic, was obviously closed source as well, leaving the open source space to Red Hat and Sun. JBossAS has remained the number 1 application server for many years, acting as a strong alternative to the closed source equivalents and although Glassfish was obviously a competitor in the community to JBossAS, it rarely was to EAP. But with Glassfish returning to Reference Implementation status, and remaining free of what we consider basic capabilities such as clustering, Oracle is saying that open source is not a core part of their middleware business (it'll be interesting to see how MySQL evolves). The Snorcle future is closed, with all that entails.
It would seem therefore that as far as Snorcle is concerned, those communities can continue to develop whatever they want, but if they need 24x7 support on enterprise-grade software then they're going to have to be prepared to migrate their applications to another codebase. Fortunately that's not been something we've entertained. If you develop on community projects and want to get guaranteed SLAs then the move from project to platform (e.g., JBossAS to EAP) is based on using the same code: everything we do is open source (there's that "open" word again) and based on what our communities help us design and develop in an open way. Quite a contrast to where things now seem to stand with Snorcle. So if you're part of their open source community (users and developers) why not consider coming across to Red Hat? The home of open source!