Sun & IBM: Speculation

Posted on March 20, 2009

The UNIX world is ablaze with speculation about the news that IBM wants to acquire Sun. Rumors of acquisition have been floating around for a long time, but the fact that institutional investors help almost double JAVA shares means we should take this one very seriously.

To be frank, I’m not sure how to feel about it. I’ll try and play both sides therefore.

On one hand, IBM’s hardware business has a great many advantages over Sun’s. Pairing the hardware lines of the two companies, especially if it potentially meant bringing Solaris to POWER, could be a very big win. One of the things I miss about Sun is the big and beefy midrange systems of yesteryear… running Solaris on an IBM 595 would be amazing. However, IBM is a company that knows how to manage multiple independent product lines, such as it has with OS/400 and the z/OS lines running along with the Windows and AIX lines. They may simply slot SPARC as another parallel line and not bother replacing AIX with Solaris.

IBM certainly would be happy to not just get its hands on Sun’s in-house engineering but also on the variety of acquisitions its made, such as MySQL, Lustre, etc. Obviously Java is the great prize and probly more interesting to IBM than the hardware. As for middleware, I’d think they’d gut the stack, taking gems such as Directory Server and Glassfish, then tossing other bits aside.

On the other hand, the corporate cultures couldn’t be more different. Sun’s internal management has systemic problems that no RIF can seem to shake loose but I don’t know the management structure at IBM and it may get worse not better. Would IBM continue to embrace the liberal try-and-buy model Sun is using or honor all its various communities? Given so many diverse efforts there are a lot of people bound to get crunched as the two giants collide.

Sun and IBM have simply competed too long for there not to be pain. NetBeans vs Eclipse. Solaris vs AIX. SPARC vs POWER. There is a very long list of competing technologies and just because we (as Sun enthusiasts) prefer one technology over another doesn’t mean IBM will agree on all counts.

Besides… from Sun purple to IBM blue? Ick.

Somewhat naturally, I’m not happy with the proposition, but as I alluded to, I can’t rule it out entirely. On one extreme I see Sun ending up like SGI, on a long slow death march… but frankly, Sun has way too much going for it, if they really had to get super lean and mean they would be a stronger player than ever, so thats out. On the other extreme is Sun being acquired and being raped in the way that Cray was by (ironically) SGI…. IBM takes everything good and useful and then discards the corpse. But, again, Sun has too much of value in too many areas, I don’t see that happening either.

I think the most interesting potential outcome is that of Sun becoming a wholly owned subsidiary of IBM. So Sun stays Sun but under the wings of IBM. That is not such a bad future, especially because it wouldn’t necessarily lead to the technology conflicts… IBM could play the game from both sides of the table. Whats more true is that Sun is a valuable brand… the Sun brand plus the Java business alone are worth more than they are offering. If they were to re-brand everything IBM it would really be a huge mistake. Sun could be the hip and edgy side of IBM.

The single most concerning aspect for me would be OpenSolaris’s status as open source. Our board members have continuously shrugged off authority and placed our fate into Sun’s trusting hands… but this raises the question, what if Sun wasn’t in charge anymore? Many of the original leaders in the OpenSolaris effort were interested not just in including the community in Solaris, but also ensuring their own access to Solaris regardless of what Sun did in the future.

So… we’ll see. Please add your own thoughts so we can build up a time-capsule of opinions while we’re all in the dark.