If you have been living in a cave this week, you may have missed the announcement about the demise of WinFS.  On the face of it, this post seems reasonable but like some of the respondents I'm a tad disappointed and worried.

We have been hearing for many years about the new file system based on SQL that would let us do all sorts of cool things.  It never seemed to happen until WinFS came along.  Now that WinFS is gone(ish) the future seems very uncertain.

At least it looks like ObjectSpaces might actually happen in the form of ADO.Net Entities but Microsoft are very late with something that has been done by many others already - e.g., WORM, nHibernate and Base4.  I think Entities will succeed only because of LINQ - without that, it's no better than the other players in the ORM game.

Vista & Longhorn were supposed to deliver the fantastic new file system.  After using Vista for a few weeks I'm now inclined to think that Vista is just a fancy XP.  I'm sure MS will sell lots of Vista to home users via the OEM channel but I'm yet to see one valid business reason for Vista and if there is to be no WinFS then what will be the point of Longhorn Server?  Security? - sure but it's not like XP/2003 will stop being secure when Vista/Longhorn ship.  Performance? - maybe, but at a hardware cost.  Wizbang? - it has plenty of that but apps won't look much better until they have been redsigned for Vista, a simple recompile won't make them look like Vista apps. WinFX / .Net 3?  - well I can do all that on XP/2003 and as most users are still going to be using legacy OS's I'm probably not going to be writing apps for Vista/Longhorn any time soon.

I think Microsoft have reached a cross-road in the development cycle.  Products are getting pretty solid, features locked down, integration is coming together.  The final picture is starting to become clear.  I haven't lost faith in Microsoft - far from it - but they are going to have to pull something convincing out of the marketing hat to reassure the congregation.