WinFS, Was it Really so Good?
If you were in any way interested in Windows Longhorn then WinFS will not be an unfamiliar word to you. This technology was to be part of the revolutionary next OS and to change the way software shared data and interacted with each other.
You will also be aware that WinFS was removed from the feature list when the Longhorn project was reset.
The lack of real discussion and comment from Microsoft seems to have led to a great deal of speculation, rumour and myth concerning the WinFS file system which really has more in common with the “Microsoft should rewrite Windows†movement of thought than actual reality.
I know personally I had a lot of misconceptions about the whole thing, perhaps due to the fact I’m a non-programmer. What confused me the most was the while using Vista I came across a lot of functionality which seemingly resembled the WinFS I had heard so much about, yet I frequently heard the comments about Vista being pointless because it lacked the promised Longhorn features.
However Ed Bott’s article “Why Do You Want WinFS?†explained a lot and after checking it out and reading around I discovered a lot of interesting things. Apologies if this is old news to you, hopefully some of you will find this interesting:
WinFS was never going to replace NTFS:
WinFS was not a new file system designed to replace NTFS, it was a system designed to run on top of NTFS, which potentially provided the capability to be moved between OS’s:
“WinFS is based on SQL Server and its core feature is to provide a data relations mechanism. This means that your pictures are related to dates, events, and persons and so on. This allows you to perform a search such as "All pictures of Joe taken the last month". This query is not possible for a traditional file system such as NTFS, but will be a trivial part of WinFS. In reality, WinFS stores are simply SQL Server database files (.mdf files) with the FILESTREAM feature enabled. These files are stored in secured folder named "System Volume Information" placed into the volume root, in folders under the folder "WinFS" with names of GUIDs of these stores.â€
WinFS was a drag on system performance and potentially had a bad programming model:
Boy, and you complained about Windows Vista. Imagine if Vista had included WinFS, Microsoft wouldn’t have had a chance against the forum haters. (Or, ironically, perhaps people may have actually accepted system performance slow down thinking it was worth it for this great new technology)
According to Dare Obasnjo; “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.â€
Vista can do much of what WinFS was intended for:
From the WinFS development blog (2006: “The vision for a richer storage in Windows is very much alive. With the new tools for searching and organizing information in Windows Vista, we are taking a good step towards that vision.â€
Perhaps what is the most frustrating thing with WinFS is that Vista actually does include much of what WinFS was intended for but without the kind of disadvantages previously mentioned. It clearly doesn’t have WinFS, but most of what was relevant and useful for end users has been included. You’ll find much of those features present in the Windows Indexing and Search technologies.
The irony of this is that if consumers and developers have so far been so unwilling to utilise the current technologies of Vista than what does this say about their willingness to adopt something even more radical which would have been the result of WinFS?
As Paul Thurrott wrote some time ago about WinFS, “the big deal, of course, was desktop search, a capability that had always been part of the plan, but was being promoted because of the sudden rise in Internet-based searches.†This never changed through the Longhorn/Vista development and integrated searching is the single biggest improvement in Windows.
Other WinFS features included metadata sorting, filtering and indexing, virtual folders and shared data between applications. All these are essential aspects of WinFS and can be found in Vista. Here is one more detailed example of just how powerful some of the Vista search tools are.
You can also take a look at Paul Thurrott’s Virtual Folders showcase here.
If I’m wrong about any of these statements, maybe even completely up the wrong tree then let me know!
Advertisement
Hello let me say that WinFS was about relationships and complex metadata model. The metadatamodel from what I can say was not super cool because it tried to have one big model describing all the world where each item in the world would be represented with virtual item in WinFS. The point is that in WinFS it was planned to store items completely in a structured manner. That leaded to concrete schemas for items and the problem team got was that different applications needed different schemas for same items – say “Message” schema can be completely different from Hotmail point of view and Outlook, Exchange point of view and say Outlook Express.
But the benefit of WinFS was in… Shell, the whole Microsoft Max was about that, the new shell that enabled user to navigate through relationships between different items – say navigate in social networks cached to local WinFS and so on. Of course in Microsoft Max, the project didn’t lead to this functionality as it was abondoned before these changes, but the vision was really cool.
This can’t be done using today’s day Windows Search & Windows Explorer functionality, sorry to deliver such news.
And that is the thing I believe (IMHO) that would bring Windows OS to Internet in a way MUCH faster than the whole Windows Live Suite. IMHO of course.