Thursday, March 19, 2009

IBM buying Sun, bad for most everyone

I truly hope that another suitor will pick up Sun, or the IBM deal falls through on its own. Can you think of a single technology or software that has been improved after it has been bought by IBM? I can't. I can see why IBM wants to do the deal, and of course, I can see it from a pure economic standpoint for Sun shareholders, but for everyone else this very bad with one notable exception, namely Microsoft.

Microsoft will be the big winner here. Contrary to some thoughts out there that suggest that an IBM/Sun combination will strengthen Java and open technologies, I believe just the opposite. IBM's game is to monopolize the entirety of the IT stack, from hardware through services, and it inexorably moves in that direction. Why is this bad? IBM will stack the deck with proprietary (and often inferior) technologies like WebSphere, DB2, Cognos and more. Companies that loved Sun for it's commitment to openness and compatibility will begin to wonder how long they can stay on a JBoss/Oracle platform when IBM starts providing monetary and support advantages to other proprietary components of the stack. If that happens, it will be a shitty day for SaaS providers and other software developers.

So why does Microsoft win big by this? Because is reduces the stack monopoly providers, really to two companies, Microsoft and IBM. I tend to believe Microsoft will grow its market share substantially in database and .NET framework at the expense of ex-Sun customers that want no part of the big so-called value proposition IBM will drive, namely Blue from top to bottom.

So where will the open systems and Java communities seek shelter? My bet is HP. My hope would be the HP does everything it can to buy Sun in lieu of IBM, but I am doubtful. Perhaps they see some short-term growth in the enterprise server business if IBM gobbles Sun, but over the haul, this could be setting the stage for Microsoft and IBM to carve up the landscape. The nightmare scenario would be for IBM to take Java in a proprietary direction. I hope IBM would not do this, but I trust them not one bit.

Who is the biggest loser? Why Oracle of course. Larry Ellison must be ripping his hairpiece out right now at the thought of his biggest enterprise database competitor controlling the hardware that runs Oracle databases more than any other application. Can you imagine the headlines in 18 months? "DB2 processes on average 34% more queries per second than Oracle 11g on the IBM Sun new I-Sun Series database servers." I am sure Ellison is out of his mind about this.

Truly though, everyone in software will lose because all of those enterprise-class and scalable SaaS applications that utilize innovative open technologies are going to suddenly be marginalized in a market now being dominated by two players, named IBM and Microsoft, who have no interest in open technologies.

My list of potential saviors: HP, Intel, Cisco. I am already praying to their Boards for intervention. I suggest everyone do so as well.