darryl ramm's blog

This is some very obvious Monday morning quarterbacking. But I am extremely negative about Sun Microsystems. I think Sun’s board is crazy for not taking the IBM $7B acquisition offer.  Maybe they think they can pull something better off, but it is hard to think who that suitor could be and why they’d value Sun at more than $7B. Sun has been churning though money and has been undergoing endless changes for ages. And while there have been occasional better days; the slumping economy must be hammering them and overall they are on a downward spiral to irrelevance. A sad outcome for this once strong Silicon Valley company. And now that an acquisition has been so publically aired that’s got to further hurt potential customer willingness to buy into Sun’s future.

The shit storm that Sun is about to be dealt from stockholders and their lawyers is going to be interesting to watch.

IBM is clearly best positioned to leverage Sun’s technology and customer base. Sun’s got some things worth having, but many of those things also turn into boat anchors quickly. For me the highlights include Solaris, ZFS, MySQL, core Java, and one or two discrete products, but not much more. Sure you get access to additional customers but competitors like HP will also be pretty aggressive about going after those customers.

While Solaris is a dying franchise, I expect it still has lots of life left, so an acquirer has to work out how to leverage that ongoing business and how long to keep that platform alive while migrating the user base to Linux. IBM or HP have to do the same with AIX and HP/UX and other operating systems so maybe that’s not so hard for them to do.

In an IBM acquisition MySQL would join DB/2 and Informix and other databases in IBM’s portfolio. It would be a harder acquisition for HP or other acquirers. If anybody can leverage MySQL and try to pull together different open source MySQL related development projects its IBM. And I especially expect IBM Global Services to be able to capture MySQL related revenue better than Sun services ever could.

ZFS is impressive as file systems go, but Sun does not obtain revenue from users of the open source code base, and is more and more effectively forced to give away ZFS software for free. The Linux open source community will pick off more and more ZFS capabilities over time. ZFS is also not well intergraded with MySQL, and while doing so could offer some interesting capabilities, doing so this would be a distraction for the MySQL team who have enough problems shipping their base product into a maturing but still fast growing market of diverse needs and expectations.

Companies like BEA with WebLogic, IBM with WebSphere and the Eclipse IDE, and even Oracle (who now owns BEA) have extracted serious value and strategic mindshare from Sun’s Java franchise.  Sun has seen some revenue amongst all this expense. But examples like Java Mobile, are under threat by Apple and their iPhone (why should Apple help commoditize applications on their platform?). And I certainly do not believe any “give Java away for free and customers will buy the hardware” dreaming. I’d hope an IBM or similar acquirer would be pragmatic about cherry picking what parts of Java to keep in-house and open sourcing the rest. I suspect IBM is pretty much there already with a business model around WebSphere, they could float off most of the Java stuff and work to make it succeed in a more open development environment. Making Java successful as a community developed technology is an important anti-Microsoft .Net strategy required for the ongoing WebSphere etc. business.

The thing that sticks out with Sun Microsystems is the litany of failed and mismanaged acquisitions over many years. Somebody must have done a business school study on this. At times it is almost as if after Sun does an acquisition different internal groups all get to gave their own shots at destroying the acquisition, especially if the aquired company was perceived to compte with some existing Sun product or futureware.  I mean how can a company screw up so many acquisitions? It can’t be by accident can it?

So here are some Sun acquisitions that I have looked at more closely than others, or just know people who went through the acquisition.

NetDynamics — Sun acquired NetDynamics and early Java app server company . Even if the software needed major re-architecting Sun should have leveraged the early NetDynamics momentum. They seemed to just let it drop. And by way of disclosure one of my sillier career decisions was turning down Zack Rinat on joining Spider Technologies, that became NetDynamics.

Cobalt Networks — Sun obviously overpaid for Cobalt but then the political anti-bodies inside Sun took over and shut down Cobalt’s highly successful but relatively low end reseller channel. Forcing them through Sun’s reseller program, which casued many smaller Cobalt resellers to drop out and killed much of the Cobalt revenue.

Terraspring – Another company acquired as a part of the Sun N1 systems management “strategy”, except it was hard to see much strategy, or tactics for that matter. And Sun ends up with a product they are giving away. I looked at Terraspring while at VMware, not seriously, more to understand the lay of the provisioning and management landscape. I was underwhelmed. Terraspring appeared to have little technology or traction and was going to get steam-rolled by new generation virtualization enabled approaches. And I never understood Sun doing so many acquisitions with potentially overlapping management technology including Cobalt, CenterRun, Terraspring, etc.

CenterRun –  It seemed one month the word on the street was that the investors wanted to shake up CenterRun’s product strategy and then a month or so later they managed to pull off selling it to Sun. I’ve got to hand it to Sequoia Capital on pulling that one off.

Afara Websystems — Durign this aquisition Sun locked in some core staff but treated most of the staff like they were definitely not wanted at Sun. McNeally pissed many of them off in a “Welcome to Sun” address that  apparently went very badly, with people literally calling recruiters straight after McNeally’s presentation. Some of the Afara folks went on to contribute to the SPARC products but the acquisition was badly handled.

MySQL — At $1B for MySQL, Sun overpaid greatly, including $800M in that hard to come by cash stuff, and that should have been obvious. The MySQL team seems pulled in different directions and maybe they are struggling within the Sun culture. The departure of Monty Widenius must have been a blow internally; it certainly was a public black eye. The evolving mixture of third party storage engines and other open source MySQL related projects, while a fascinating example of open source development, raises concerns that Sun’s leadership of the MySQL market is going to get fairly nebulous. Sun needs a strategy to try to pull this together if they want to continue to grow their ability to make money out of MySQL. Otherwise these other projects, and people around them are going to cherry pick interesting sweet spots of the market.  Like Drizzle related activates for Web infrastructure. See some discussion on that ecosystem recently in Jeremy Zawodny’s The Real or Official MySQL? Does Not Matter! (except it does matter if you are Sun trying to make money on that $1B MySQL aquisition).

Some acquisitions, notably Kealia worked much better, maybe because of the cachet of Andy Bechtolsheim and his ability to drive a product to market from within Sun, even if it did take longer to get those systems to maket than initially seemed to be claimed (and those boxes were pretty darn nice servers).

—

Sun you shoudl have sold to IBM,  but hang in there a few more years and Rackspace might acquire you, and can mount your logo on their mantel piece right next to that SGI logo.