Apple Inc (AAPL)

[Tex]

re Same As It Ever Was

Apple definitely doesn't want to be enslaved to the demands of outsiders. On the other hand, with a commitment to Unix/POSIX and to the future of Cocoa, Apple can provide some points of consistency (e.g., in non-kernel software interfaces; and Apple is improving the consistency of kernel interfaces by developing KPIs that make implementation details irrelevant to programmers who need to use kernel extensions) without giving up on the areas in which genuine innovation is plausible.

With Cocoa and Unix/POSIX, Apple can offer some interesting things to the enterprise. Apple can support lots of enterprise applications (like all the serious database tools -- well, any serious database tools not made exclusively in Redmond), and being able to support these on its OS puts Apple in a position to compete for business it never previously could reach. The question is whether Apple wants to pursue that business, or not.

If Apple wants to be part of the back-office solutions that run whole companies -- which isn't a bad way to get related products sold all the way down the organization -- Apple might benefit from thinking about what it would want to offer enterprises, and how. If not, Apple is going to end up in enterprise only by accident, when someone takes a lot of time and trouble to set it up on an ad-hoc basis. How many System X or COLSA sales can Apple expect to make? Yet, servers back-office solutions are paid for and deployed every day, and by competing with MSFT's costly licenses Apple could take some profitable business and gain some share -- not just among installations, but minds.

Basically, Apple's disinterest in building this kind of business convinces me that it would see no value in a Sun acquisition. Apple doesn't appear to be still interested in Java (which is a great technology if you ignore the wallowing, slow UI tools with which it's been saddled), and some of the most of the cool things in Solaris are available without cost (DTrace, ZFS), so the real question is which engineers Apple should poach to help with kernel stability, performance, size, and other issues (like maybe improved KPIs, to support greater longevity of drivers by making more of what programmers need accessible to be accessible without dependence on particular implementation details).

So I stand by my "I doubt a Sun purchase" statement, but hold out some hope that Apple might make clear that experienced kernel developers and people with enterprise implementation expertise (of which Apple has considerable in-house demand, as we saw with the MobileMe rollout http://jadedconsumer.blogspot.com/2008/08/apple-service-reliability-quantified.html ) should be sending applications to Apple.

Take care,
--Tex.