Jon Udell has started a series of blog posts about the pillars of Longhorn. So far he has written Questions about Longhorn, part 1: WinFS and Questions about Longhorn, part 2: WinFS and semantics which ask the key question "If the software industry and significant parts of Microsoft such as Office and Indigo have decided on XML as the data interchange format, why is the next generation file system for Windows basically an object oriented database instead of an XML-centric database?"
I'd be very interested in what the WinFS folks like Mike Deem would say in response to Jon if they read his blog. Personally, I worry less about how well WinFS supports XML and more about whether it will be fast, secure and failure resistant. After all, at worst WinFS will support XML as well as a regular file system does today which is good enough for me to locate and query documents with my favorite XML query language today. On the other hand, if WinFS doesn't perform well or shows the same good-idea-but-poorly-implemented nature of the Windows registry then it'll be a non-starter or much worse a widely used but often cursed aspect of Windows development (just like the Windows registry).
As Jon Udell points out the core scenarios touted for the encouraging the creation of WinFS (i.e search and adding metadata to files) don't really need a solution as complex or as intrusive to the operating system as WinFS. The only justification for something as radical and complex as WinFS is if Windows application developers end up utilizing it to meet their needs. However as an application developer on the Windows platform I primarily worry about three major aspects of WinFS. The first is performance, I definitely think having a query language over an optimized store in the file system is all good but wouldn't use it if the performance wasn't up to snuff. Secondly I worry about security, Longhorn evangelists like talking up what a wonderful world it would be if all my apps could share their data but ignore the fact that in reality this can lead to disasters. Having multiple applications share the same data store where one badly written application can corrupt the entire store is worrisome. This is the fundamental problem with the Windows registry and to a lesser extent the cause of DLL hell in Windows. The third thing I worry about is that the programming model will suck. An easy to use programming model often trumps almost any problem. Developers prefer building distributed applications using XML Web Services in .NET to the alternatives even though in some cases this choice leads to lower performance. The same developers would rather store information in the registry than come up with a robust alternative on their own because the programming model for the registry is fairly straightforward.
All things said, I think WinFS is an interesting idea. I'm still not sure it is a good idea but it is definitely interesting. Then again given that WinFS assimilated and thus delayed a very good idea from shipping, I may just be a biased SOB.
PS: I just saw that Jeremy Mazner posted a followup to Jon Udell's post entitled Jon Udell questions the value and direction of WinFS where he wrote
XML formats with well-defined, licensed schemas, are certainly a great step towards a world of open data interchange. But XML files alone don't make it easier for users to find, relate and act on their information. Jon's contention is that full text search over XML files is good enough, but is it really? I did a series of blog entries on WinFS scenarios back in February, and I don't think's Jon full text search approach would really enable these things.
Jeremy mostly misses Jon's point which is aptly reduced to a single question at the beginning of this post. Jon isn't comparing full text search over random XML files on your file system to WinFS. He is asking why couldn't WinFS be based on XML instead of being an object oriented database.