WinFS no longer a product

staceyw wrote:

Vista is built on 2000 SR1 OS.

No, Vista ist built on the Windows Server 2003 code base. That was the main reason for the delay, they switched from the XP code base to the Windows Server 2003 code base.

Worse, they started on Win2k3, then went WinXP SP2 for some dumbass reason and when SP1 hit Win2k3, they went back.

WinFS exists in beta, but yea it’s evolving.

It never was a desktop search product; it’s a way of organizing your stuff.

ISV’s have been working with WinFS for about a year now and they have created applications that will organize your CD collection and your Moms recipes.

eagle wrote:

ISV’s have been working with WinFS for about a year now and they have created applications that will organize your CD collection and your Moms recipes.

Wow -- it's that scalable?

How many CD's do you have?

Roger Lueder's TechEd  talk: Building on WinFS: Developing a Great WinFS App from Scratch.

He had a cool app for a Wine Collection:

Is WinFS, in its current incarnation, like "Core Data"?

mawcc wrote:



staceyw wrote: Vista is built on 2000 SR1 OS.

No, Vista ist built on the Windows Server 2003 code base. That was the main reason for the delay, they switched from the XP code base to the Windows Server 2003 code base.

oops.  I meant 2003 SR1.  Note that is SR1, not SP1.  Many updates to the core went into SR1.

eagle wrote:

How many CD's do you have?

About 260 or so...

eagle wrote:

Roger Lueder's TechEd  talk: Building on WinFS: Developing a Great WinFS App from Scratch.

He had a cool app for a Wine Collection:

I haven't seen that video yet, but why would you need WinFS for that? Seems like a fairly average database app, certainly nothing requiring a mothballed multi-million-dollar project to run off.

Cairo wrote:

Is WinFS, in its current incarnation, like "Core Data"?

While 'm not certain what you're referring to when you say "current incarnation", I can perhaps provide a partial answer. To put it simply: sorta.

Part of what Core Data is trying to solve is eliminating the need to write tedious object persistance code; that's something we have now in .Net via the CLR serialization services. The entity and relationship modeling sounds very similar to what Microsoft's proposing with their Entity Framework, though Apple doesn't seem to have the rich query functionality exposed through LINQ.

A lot of the big differences, I think, are in the services that are layed on top of these models. Core Data has basic change management (undo/redo), but WinFS had that and more. Particularly, WinFS provided a synchronization engine which could connect with other clients, web services or server-based databases (such as SQL Server)--Core Data seems focused entirely on the client from what I can tell. Also, there was an interesting concept called an InfoAgent, which would basically montior the store for particular conditions, and then do some action in response (e.g., "If I get an email from my doctor, alert me immediately even if I'm busy."). It was neat stuff, but I'm not sure what's to become of it.

WinFS was much more ambitious than Core Data--no doubt a consequence of Microsoft's desire to unify the functionality between client and server. Still, Core Data is actually here today, so I'm in no position to fault Apple's evolutionary approach.

HTH (at least somewhat).

BryanF wrote:



HTH (at least somewhat).

It does, thanks. The asynchronous aspects of it sound interesting. And one of Core Data's primary limitations at the moment is that it can't talk to a server-baesd database system. It's a but like a cut down and then extended version of "EOF" (Enterprise Objects Framework) from WebObjects. Perhaps CoreData 2 will add some more enterprise-type features.

i too was very much interested in WinFS. it is really sad when the project you work on is scrapped ,but another truth that i realised today is not all projects reach their target (being a student, all projects were succesful ).

hope all the work u guys put in would be useful in other forms atleast .

eagle wrote:

WinFS never was a Product, it’s a technology and beta 1 is available for ISV’s to build Products and they will.

Well beta 1 has no go-live license that I'm aware of, and there won't be a beta 2 or a release, so unless those ISVs only used WinFS for entites/ORM and/or "unstructured data" (whatever that includes), then they can't build squat.

The WinFS Team is intact and as enthusiastic as every about this cool technology, it’s just not going to be part of Windows Vista next year

Shiv wrote:

hope all the work u guys put in would be useful in other forms atleast .

http://blogs.msdn.com/winfs/

eagle wrote:

The WinFS Team is intact and as enthusiastic as every about this cool technology, it’s just not going to be part of Windows Vista next year

Your kool-aid must be stronger than mine, but I see no reason to be “enthusiastic”.

Let’s look at upcoming releases from Microsoft.  Vista is currently scheduled to RTM out October/November 2006. Assuming that holds, we’ll be at .Net 3.0, which included .Net 2.0 and the former WinFX APIs (WPF, WCF, WWF). “Orcas”, the next Visual Studio, isn’t expected to ship till the 2^ND half of 2007. “Orcas” is also the version of Visual Studio that will support the LINQ technologies as well as the Entities framework being picked from the bones of WinFS, now to be known collectively as ADO.Net 3.0. The self-tuning bits of WinFS would then be rolled into “Katmai”, the next version of SQL Server. This is expected to ship 2007/2008. Judging from Microsoft’s Integrated Innovation kick and the delays SQL Server 2005 (“Yukon”) inflicted on Visual Studio 2005, and the fact that a new Visual Studio always seems to ship after I renew my MSDN subscription, I would guess that both Orcas, and Katmai will ship very late in 2007 at the earliest.

This, my friend, is over a year and a half away from now. Microsoft currently has no credibility in predicting release dates out that far.  Any other technologies “incubating” in the WinFS project that weren’t mature enough to make “Orcas” and/or “Katmai” by then might as well be dead chicken.

WinFS is dead chicken.

Discussions around this topic:

http://www.longhornblogs.com/robert/archive/2006/06/23/16365.aspx.
http://www.hunterstrat.com/news/2006/06/24/say-goodbye-to-winfs/
http://www.25hoursaday.com/weblog/PermaLink.aspx?guid=43532b55-f8f8-4b72-a532-c7343b4e1b9e

Microsoft from now on should not showcase or declare any feature, platform or technology that they are not sure will make the cut. They have lost almost all credibility in my books with this news. I just realize I wasted a year on working and learning a dead technology.... thank you MS!

This is the saddest news I've heard from MS lately. This will be even worse for my friend to hear because he is a big DB guy. I was so looking forward to that UX that allowed me to surf my sea of data in such a natural way...oh well...still a dream.

I don't care if WinFS died or if WinFS was never the super duper NTFS replacement that we all thought it was to begin with.

The fact of the matter is that people feel betrayed, and Microsoft just lost a lot of credibility, even amongst die-hard fans like myself.

Seriously, I've never lost trust in Microsoft so much as now. This is ridiculous.


WPF and Indigo are the only standing pillars (features, rather) of Longhorn. Indigo is invisible to end users. WPF is obviously a lot more than the "eye candy bullsh-t" that people say it is (because they, users and devs alike, don't understand the subtle implications of a real, practical, successful implementation of a truly document/controller-view separated API), but if people are stupid enough to believe that that's all Vista is, then Windows no longer has any foundation.