Red Hat Eats JBoss

Red Hat's acquisition of JBoss made perfect sense to me. Red Hat has far more credibility if it is providing a whole software stack than if it is just offering various Linux packages. Also both Red Hat and JBoss have been targeting the corporate market and, as a single company, there should be significant economies of scale open to it. It also evens up Red Hat's competitive battle with Novell.

I have been a fan of JBoss for quite a while because the technology road map looks well thought out to me—although I feel that JBoss needs to be asking for license fees on some of its products that are currently free. It had enough of a portfolio to do that. Maybe Red Hat will make that move. A viable strategy on how to make a commercial success—and I mean healthy revenue growth—from Open Source needs to be articulated.

The acquisition creates a conundrum for Novell, because JBoss is (or was) a significant partner. A phenomenon of Open Source has been that a single product normally emerges in any given category. Firefox (browser), Linux (OS), Apache (Web Server) and JBoss (middleware stack) are all examples. The only market where more than one product seems to have emerged is database—which is dominated by MySQL, but also served by FireBird, SleepyCat, Ingres and others. The acquisition is not good news for Novell.

Oracle to Market Linux?

Larry J Ellison of Oracle was quick to rain on Red Hat's parade. A few days after the JBoss acquisition, he remarked, in a Financial Times interview, that Oracle might distribute its own version of Linux. Red Hat's shares sank 7% immediately. Ostensibly the motivation is to help Oracle compete with Microsoft's SQL Server stack (middleware plus SQL Server plus Windows). Larry mentioned that he had considered acquiring Novell. Rumour has it that Larry had also considered acquiring JBoss, but backed off.

Larry did, however, acquire the Open Source SleepyCat database about a month ago and picked up two small MySQL partners prior to that. As he indicates in the FT interview, Oracle intends to exploit Open Source rather than compete with it. In that respect that acquisition of Novell would make sense because Novell has a good deal more than a sack-full of Open Source products in its portfolio and some—such as its ID management software and NetWare.

LJE's Stack?

MySQL has prospered well from the acronym LAMP—standing for Linux, Apache, MySQL, P(ython, Perl, PHP). I guess everything in the stack has prospered, but as MySQL has a commercial axe to grind, the success of the LAMP acronym must be sweeter to MySQL AB than the Open Source Orgs in the stack.

SCO recently came out with the idea of the SCAMP stack, and actually branded it with the face of a cartoon bulldog. It's a genuine proposition, exploiting the LAMP acronym both for fun and, it hopes, for profit. SCAMP stands for SCO, Apache, MySQL, P(ython, Perl, PHP). It's a positive turn of events, SCO competing in the market rather than in the court room. I know, the famous litigation continues to continue but there's little to be said about it and there will be little to say until some kind of result, or indication of a result, emerges.

A third stack idea was suggested to me by CA's new CTO, Mark Barrenechea. It's pronounced ba-ran-chay, by the way. Mark's a nice enough guy but the spelling of his surname is a bitch. Anyway Mark suggested the LJE'S stack, poking fun at Oracle's Open Source schizophrenia. This stack consists of Linux, Java, Eclipse and Sleepycat. It is, in truth, quite an acceptable software development stack, a direct alternative to the one offered by Oracle (the LOOO stack). The joke here is on Larry's initials. Larry was Mark Barrenechea's direct boss for about eight years.

Rumour Mongering

Every now and then I hear rumours and usually I don't blog them. But at the recent HP analyst conference, HP virtually announced that it intended to get involved in a “big” software acquisition. Given that HP also declared that it had no interest in extending towards the IBM mainframe (in effect eliminating the idea that it could acquire BMC) there are few complementary companies that it could acquire that would qualify as “large”.

The obvious possibility is Mercury Interactive. Mercury's software portfolio is neatly complementary to HP's OpenView and Mercury has got itself into a mess with the SEC to the point where its stock was actually suspended. It doesn't seem as though Mercury's business is actually suffering much from this by the way, but the SEC would probably be happy to see Mercury acquired by the HP.

I give it a 50/50 chance of happening. That's about how reliable the rumours I hear are.