Woowar

The essence of the story was:

• Vista may well be the end of the road for Windows as we know it*
• Windows Server 2008 is already modularised to some extent
• A modular Windows would be easier for corporates to deploy and manage, thus reducing cost of ownership and speeding adoption, thus heading off the threat from Linux and OS X, thus bringing more cash into the Microsoft coffers
• The licensing model could get interesting
• The development process itself for Windows could be made more Agile as a result of modularisation
• Different modules could be on different release cycles, thus improving maintenance cycles and reducing the impact of upgrades

If this is true, and if Microsoft successfully make that transition**, then Windows will maintain more of its current market share than it might otherwise do.

However, far be it from me to be the cynical naysayer, but on past record they might struggle to actually achieve this in a sensible timeframe: Windows is currently very tightly coupled, which is to say that it's the very opposite of being modular. In fact, to date the strategy has been to tightly couple it - e.g. the dependence on IE - to lock customers into the Microsoft software stack and the Microsoft view of the world. They deliberately*** sacrificed good software architecture to create that lock-in, and will now pay the price if they want to modularise Windows. They may succeed, of course, but it will feel like trying to untangle 5 balls of string. In the dark. With gloves on.

* Ain't that the truth

** In fact, they may find taking Windows apart and sticking it back together again to be a "paneful" process (groan).

*** I sincerely hope it was deliberate - if that's the best they can do then, well, oh dear.