About a week ago, JBoss’ Marc Fleury posted the following to the JBoss group blog:
But I want to save the best for last But actually just a teaser because I will not say anything except that we GOT SOME DYNAMITE OF AN ANNOUNCEMENT COMING OUT ON TUESDAY MORNING. That’s right, I am chuckling under my breath when I think about it. The big-ness of it, the beauty of it, the surprise of it.
It will get the attention of mr Mills at’s for sure!…. Buaaahahahahahahaah
That led a lot of people, including Denver’s own Matt Raible, to speculate on just what the announcement might be. Here’s what Matt said:
Any idea what the JBoss Announcement will be next Tuesday? Since he’s laughing at Oracle, I’m guessing they hired a prominent MyFaces developer – or they’ve “acquired” a prominent database developer (i.e. PostgreSQL). What’s your guess? Will it really be that impressive, or just impressive for JBoss employees?
Matt had one of the right product areas, as it turns out, the database – but the real target here in my mind isn’t Oracle but IBM. But in case you haven’t seen the actual news yet, here it is in the words of Microsoft’s open source point man, Jason Matusow:
Microsoft and JBoss are announcing our intentions to work together on enhancing the interoperability between the JBoss Enterprise Middleware System (JEMS) products and the Microsoft Windows Server products.
You’ll notice he didn’t mention IBM once in that announcement, nor would I have expected him to. But to me this deal is similar to the Microsoft/Sun detente, in the sense that it’s an enemy of my enemy is my friend deal. A couple of other quick questions and answers:
Q: Let’s start with the basics: what’s in this for each player?
A: Despite the apparent incongruity of an open source J2EE stack player and the purveyor of the world’s most popular non-open source operating system – which incidentally includes a popular Java alternative – the deal actually makes sense for both players. From the JBoss side, the opportunity is to optimize their product for the platform that a sizable portion of their customer base chooses to run on. If large industry players such as IBM and Sun have felt it necessary to align with Microsoft on some issues, it shouldn’t come as a surprise that JBoss does as well. It’s the old “environment in which you compete” argument. More specifically, while J2EE servers have typically had more of an affinity for *nix operating systems, be that AIX, HP-UX, Linux or Solaris, there’s no denying that a huge number of application servers are run on top of Windows. This is attributable to a variety of factors, but probably most commonly can be explained by the enterprise preference for J2EE for certain workloads. Many of these same customers, however, choose to run these transactional platforms on the Windows operating system. Hence a large customer base that could be better served by a JBoss that runs better on Windows. In addition, a partnership with Microsoft can provide it with momentum and mindshare – at the risk of F/OSS community goodwill, of course. This could become more important as IBM’s acquisition of Gluecode sinks in and some enterprises might question the ongoing viability of an upstart like JBoss; the Microsoft partnership is an easy answer to such questions.
Q: Ok, how about Microsoft? What’s in it for them?
A: Well, many things. First, and most obviously, the success of an open source JBoss can be portrayed as a direct strike at some of the higher end J2EE commercial players such as BEA and IBM. To that extent, this partnership is, IMO, as much about trying to weaken its competitors as it is about furthering its own ends. But it’s also true that JBoss is a very popular platform amongst Java developers, and a tighter relationship with JBoss may prevent Windows from being replaced with Linux as the underlying OS in some accounts. Key to that is the ability to manage JBoss instantiations as just another asset within the Microsoft Operations Manager, as was also announced. And last but not least, there’s the possibility of opening new markets for SQL Server in JBoss/Hibernate projects.
Q: Well, if the deal is at least partially about IBM – as seems to be the case given Fleury’s mention of “mr Mills,” which one assumes is IBM’s head of SWG Steve Mills – what does it mean for the folks from Armonk?
A: Well, the high level implications are mostly negative, as a stronger JBoss is certainly not something Big Blue’s likely to be blissfully happy about. When JBoss is competing with IBM in accounts now, it will undoubtedly be touting this deal as proof that they’re serious and around for the long term. But on a more detailed level, I think the impact to IBM will actually be relatively minimal. Through its acquisition of Gluecode and the injection of new EII/ESB/SOA functionality into the WebSphere brand, IBM’s pretty well equipped to compete with JBoss, whether that’s on a no-barrier-to-entry/”good enough” basis, or in a more traditional high end, feature/function slugfest. They can even out-simple JBoss by pitching to the scripting audience via their Zend partnership around PHP. The JBoss/Microsoft partnership makes JBoss a more capable and credible player in the Windows world, but it doesn’t introduce any dramatic new features which would seem to change those competitive landscapes.
Q: Some people are asking whether or not this signals a sea change in Microsoft’s attitude towards open source? What do you think?
A: Well, that to me is like asking whether or not this signals a decommitment to .NET, or whether IBM’s Zend partnership signalled a departure from Java. The answer is: absolutely not. This is ultimately about a business opportunity. Here’s how Matusow answered the question:
Weve had a colorful past on the issue of OSS. Does this announcement mean that we are stepping away from .NET and Windows to embrace Linux and Java? Are we going to throw open the doors to the collective source base of the past few decades of development?
First of all, this is no way means that we will ease off in our competition with J2EE, we still believe .NET is the better technology with more upside for our customers and partners. We are still going to compete vigorously with Linux and constantly expand and improve upon the Windows technologies. And Shared Source will continue to grow as collaborative development and increased transparency continue to be increased. But we will not be open sourcing anything and everything. Our position on this remains the same.
That’s pretty much how I see it. I don’t see this partnership as much of a signal of anything, other than a confirmation that Microsoft is striving to find new opportunities to be competitive. It’s not a “if you can’t beat them, join them,” it’s a, “let’s keep trying to beat them, and make money off them at the same time.” If this had been 2 or 3 years ago, during Microsoft’s “open source is a cancer” phase, that’d be one thing, but things are far more pragmatic now - IMO. Microsoft may never be totally comfortable with open source and vice versa, but I certainly don’t see the JBoss partnership as an indication that they are a.) radically changing their view on OSS or b.) jumping ship from .NET to Java.
Q: You mentioned earlier that JBoss was “risking goodwill” – could you explain that?
A: Sure. In some quarters of the the free and open source software world, there are a few commercial entities that are regarded as “the enemy.” SCO is one example, Microsoft is another. What we’re seeing from a variety of the commercial software organizations with substantial open source commitments, however, is that emotion is not slamming the door on business opportunities. We saw this in the MySQL partnership with SCO, and also when longtime Microsoft nemesis Sun turned over a new leaf with their old arch enemy. So too have we seen this with JBoss. Given the current climate within open source, which is increasingly pragmatic, this is less of a risk than it once was, but according to Dana Blankenhorn, Fleury, “worried a bit about how “the community” might feel about the deal, about whether he was selling out.” From where I sit, I think JBoss is right to have some concerns, but I personally don’t expect them to amount to much.
Q: Any last thoughts?
A: Not really, other than I personally think it’s great to see companies being pragmatic and trying to find ways to work together while competing. Hope to see more of that.
Update: Forgot the disclaimer: of the above, BEA, IBM, Microsoft and Sun are clients, JBoss is not.