I'm voting for private cloud.
1. They have a strong presence in businesses in both the desktop and server spaces.
2. They have the cash in the bank and plenty of good people to throw into a serious push to move the market.
3. They'd be playing to their established strengths, instead of trying to challenge the likes of Apple and Google in areas where the latter are already far ahead.
4. They'd be playing to today's cloud-happy market sentiment, but also able to avoid some of the inescapable flaws that have been starting to show as more businesses have gained experience with the reality of outsourcing too much.
5. This strategy is compatible with almost any serious strategy for the home user/entertainment market. In fact, with the rise of practices like remote working via VPNs and mobile working with BYOD, there is a lot of potential to bridge the two, again in ways their competitors simply aren't geared up for.
With Ballmer gone, if people like Hanselman and ScottGu gain more influence within the company, I'm sure we'll see great things being open sourced.
It's hard to shift the company to open source with someone like Ballmer at the top, it would be like admitting defeat to the open source movement, and people in such positions of power never like to admit they were wrong in their strategy. It's human nature, I almost don't blame him.
Huh? How do you figure?
Look what phone DIRNSA uses today: a Blackberry.
I'd sell non-core assets (Xbox, Bing, hotmail, consumer stuff in general, etc.), and focus on the core OS/mobile/Azure/Office/MSDN systems.
I'm with you on splitting the enterprise cash cow from the rest of the business.
The underlying OS on the Exchange server doesn't matter. That DoD is (I'm pretty sure) the world's biggest MS deployment does matter, though -- even though I prefer UNIX, if I'm in an environment where 99% of systems are Windows based, I'd build my ancillary systems on Windows, too.
I would grant that there are specialized markets that MS could serve well in the enterprise market, but what do you think the typical enterprise deployment needs out of the box that iOS can't provide right now?
iOS is undeniably a better web browser experience than Blackberry. It also has a viable app ecosystem, but that mostly doesn't matter for enterprise.
The only large OS X Server-managed deployment I've heard of is Apple internally. In Silicon Valley, I'd probably go with iOS-only (I don't believe in BYOD, and crossplatform means LCD and pain) with either a third-party MDM or maybe try OS X server. No way I'd do that in a 100k seat enterprise with high security requirements -- 30-50 person shops aren't really "enterprise".
I am excited about iOS 7, but not at all excited about Apple's lack of any real focus on enterprise. The only reason Apple has stuff like Kerberos support in OSX is that Apple uses it internally.
Apple focuses on consumer (primarily), with iCloud, etc. as the main management tool. There is a little bit more for small businesses and some smaller education deployments, but aside from Apple.com, there isn't much attention given to the enterprise market.
(If I got to pick a job, I'd rather be HHIC of Apple Enterprise Products/Enterprise Security rather than CEO of Microsoft, personally. Either would be a turnaround.)
They really need to either spin off a separate business, sell some divisions or just stop trying to compete in some areas. Experience shows their OS dominance doesn't count for much in the phone, tablet, search or cloud markets.
But, there was a time when Microsoft entering a market would send all the existing players running for the hills. They used their OS dominance to control what software OEMs installed and killed off every desktop software competitor they ever had. Your best hope was to sell your company to them. But they couldn't do that with the phone, tablet, search or cloud markets. They've built some good products and some of them are doing ok, but they haven't had the kind of automatic win in those markets that those of us around in the 90s remember.
Maybe it was the delays in getting products out? Or false starts? Or that they just started playing nicer?
Nobody is close to unseating them on the desktop, but the excitement is everywhere else.