Windows v.next: less lame than ever

Storage Bits speaks and Microsoft listens Right.

In Why is Vista lame? I discussed why a 5 year development cycle killed Vista:

Vista's 5 year development cycle is the problem Microsoft needs to get back to 24 month development cycles. There is really no choice. No one can predict what the customer hot buttons will be in 60 months, or what Google, Apple, Linux or the European Union will have done.

I'm not the only one who thought so.

Microsoft's VP of Windows v.next development agrees In this morning's New York Times, John Markoff writes (irritating registration required)

In the battle between Apple and Microsoft, Bertrand Serlet and Steven Sinofsky are the field generals in charge of competing efforts to ensure that the PC's basic software stays relevant in an increasingly Web-centered world.

The two men are marshaling their software engineers for the next encounter, sometime in 2009, when a new generation of Macintosh and Windows operating systems is due. Their challenge will be to avoid refighting the last war — and to prevent finding themselves outflanked by new competitors.

I don't buy the whole PC OS fading to irrelevance schtick It is encouraging that Steven Sinofsky, the Microsoft SVP for the Windows and Windows Live engineering group, wants the next version out in two years. It means a more focused product and less wasted effort. Corporate IT will hate it, but IT managers hate anything new.

In fact, the best way to get IT to upgrade to Vista is to introduce its successor. Bwa-ha-ha!

Will web services kill the desktop/laptop OS? No.

For two reasons:

• Stripped to its essentials, this is the old argument between thin clients and fat clients. Last time I checked the score on that was about 750 to 1.
• The "why" of it is this: people like predictability. A lot. Web-services are and will remain, inherently less predictable. Network latency, demand peaks, network quality all conspire against predictability. There is no easy way around that truth.

Windows will be effectively modularized I doubt they'll call it that, but as the article notes, Microsoft needs a way to make incremental component improvements without the hassle of a major service pack release. The downside for Microsoft is that it requires them to develop and adhere to - let's call them module APIs for lack of a better term - which will make it easier for third parties to develop competitive products that Microsoft won't be able to shut down with a couple of code tweaks.

I think that's good for Microsoft, but they are control freaks and won't like it. Too bad.

What about all that multi-core stuff? There's some chatter about adapting Windows v.next to run on dozens of cores. It sounds sexy, but the real benefit to users is getting applications to run on dozens of cores. If you are spending so much time in the OS that you notice the difference between 8 cores and 24, you need to examine your workflow.

The Storage Bits take Microsoft can never take another 5-6 years on a major Windows release and they know it. If they get their game on it will help keep everyone on their toes, which is good for every computer user. They have the people and the resources to lead. With shorter development cycles they'll actually be able to.

Comments welcome, of course.