WinFS no longer a product

I still say the glass is half full and can't understand why all of you say it is half empty.

We learned long ago that WinFS would not be part of Windows Vista. After TechEd I thought it was just 6 months behind Vista, so now it's going to be another 6 months.

I can find all of my data because of a trick I learned back in the days of DOS, create Directories.

eagle wrote:



I still say the glass is half full and can't understand why all of you say it is half empty.

It's not a matter of the glass being half-full or half-empty. It's whether the glass exists at this point.

Microsoft should get better at either

a) not announcing vaporware
or
b) completing announced projects

eagle wrote:



I still say the glass is half full and can't understand why all of you say it is half empty.

We learned long ago that WinFS would not be part of Windows Vista. After TechEd I thought it was just 6 months behind Vista, so now it's going to be another 6 months.

I can find all of my data because of a trick I learned back in the days of DOS, create Directories.

In what fevered imagination of an alternate reality do you think WinFS is going to ship 1 year after Vista?

The parts being pulled out to roll into Orcas and Katmai might make it out by then, but anything else being incubated in Project "Orange" is probably 2 years out past that. IOW, too long to care about right now.

WinFS is dead Orange Chicken

Shipping this product just doesn't make sense for microsoft right now.  Where's the pressure in this market space?  Microsoft can let someone else prove that the market exists.  Remember the internet.  There's a good chance that shipping WinFS now would provide very little lift to the general acceptance of Windows for users or developers.  But a few developers would embrace it, putting pressure on microsoft to keep supporting it.  The response from the larger market would probably be to propose a set of competing technologies that microsoft would then have to adapt to, while still maintaining compatibility with WinFS 1.x for internal and external applications that embraced it.

Nevertheless, it may make perfect sense for microsoft to keep teasing the market.  There is clearly this idea that some Web-like capability could be placed between apps and storage - some document based but still relational something for everyone that would allow general purpose applications to browse your computer and intranet.  Microsoft clearly want to master this space once it shows itself, but as I said before, LINQ is much more strategic at this point, and database-entities need to align with it so that LINQ + SQL Server applications can capture maximum market share, in what is a real and growing market that microsoft already have a huge stake in.

davidofmorris wrote:

Shipping this product just doesn't make sense for microsoft right now.  Where's the pressure in this market space?  Microsoft can let someone else prove that the market exists.  Remember the internet.  There's a good chance that shipping WinFS now would provide very little lift to the general acceptance of Windows for users or developers.  But a few developers would embrace it, putting pressure on microsoft to keep supporting it. 

Nevertheless, it may make perfect sense for microsoft to keep teasing the market. 

That's just crying sour grapes. Microsoft is doing this "teasing" that you talk about at the expense of their own credibility. Microsoft isn't Google. Those tricks don't work.

You make parallels with the web, but is this a web service that we're talking about? No, it's a client product. There's no point in waiting for others.

The problem isn't about general acceptance of Windows. It's about general acceptance of Microsoft.

eagle wrote:

 I still say the glass is half full and can't understand why all of you say it is half empty.

We learned long ago that WinFS would not be part of Windows Vista. After TechEd I thought it was just 6 months behind Vista, so now it's going to be another 6 months.

At first the glass was full. Then we were told the glass is half full -- it won't make it in Vista. Nonetheless, it was still half full, so we were happy. Now we're being told that the glass is empty.

Did you actually read the WinFS blog entry? It doesn't say WinFS isn't making it into Vista; it says WinFS isn't making it at all.

reinux wrote:

The problem isn't about general acceptance of Windows. It's about general acceptance of Microsoft.

Oh my -  I had no idea the stakes were so high!  I stand corrected.

blatzcoder wrote:

Perhaps Microsoft should go after a bunch of Oracle devs to see how a large scale database is supposed to be built and then have them implement it.

That's not the problem. SQL Server is a very good (large scale) database. The problem with WinFS is how to effectively implement a store for semi-structured data (i.e. to store something that's between an opaque stream of bytes and a relational model in 3rd normal form).

WinFS Update

From the WinFS Team blog: Update to the Update.

Is WinFS dead?
Yes and No. Yes, we are not going to ship WinFS as a separate, monolithic software component. But the answer is also No - the vision remains alive and we are moving the technology forward. A lot of the technology really was database stuff – and we’re putting that into SQL and ADO. But some of the technology, especially the end user value points, are not ready, and we’re going to continue to work on that

That's still as much of a yes as it was yesterday.

The fact that WinFS will not be available as a file system for Vista ruins it completely.

Not a surprise. Everyone knew WinFS was vaporware anyway.

Having a superficial understanding of what WinFs was going to be.. and to me that meant some meta-data layer on top of NTFS, i have to wonder what all the fuss is about that it’s being dropped or whatnot.

I mean how many people fill out the Metadata options we have now for all the files we can make. I have never once entered document information for Office docs. I simply dont have time. Nor am I going to OCDesquely meta-label every media file I have.

So being the lazy content-creator that I am, what am I missing by WinFs not coming to life?

pacelvi wrote:

Having a superficial understanding of what WinFs was going to be.. and to me that meant some meta-data layer on top of NTFS, i have to wonder what all the fuss is about that it’s being dropped or whatnot.

I mean how many people fill out the Metadata options we have now for all the files we can make. I have never once entered document information for Office docs. I simply dont have time. Nor am I going to OCDesquely meta-label every media file I have.

So being the lazy content-creator that I am, what am I missing by WinFs not coming to life?

I filled it out. Or more precisely started to, but the "tools" are far from being easy and user-friendly. I was even considering writing an easier meta-data editor/organizer before WinFS, but I decided to wait for WinFS. Then it was dropped from Vista and we saw the disappearance of virtual folders.

What was cool about WinFS is that data becomes meaningful and you can have relationships between separate chunks of data. That's when you don't have to "meta-label" every file, but associate other data to them. For example associate a contact to a picture as in picture of X.Y. What's exciting about this is that looking at the picture you (and maybe the computer too) can immediately get the information about people in the photo. WinFS is not good because it adds meta-data (file, properties, summary). It's not good because it can search by meta-data (desktop search). The trick is (or rather was) in the  relations.

And yes. WinFS is dead now, but I'm sure another OS will have it in 2 years. And Microsoft can play catch-up again like with mapping solutions (longhorn+keyhole demo springs to mind). Great

Zaki wrote:

... I'm sure another OS will have it in 2 years. And Microsoft can play catch-up again ...

It would be GREAT if someone else were to gain some traction in this space.  It would be particularly wonderful if an OFS layer gained acceptance in Linux or OS-X!  Microsoft are very good at catch-up and me-too.

You can't play catch up on something you're already developing before anyone else. And even if the alternative OSes get an OFS layer, doesn't mean much. Gnome Storage was the closest thing to WinFS and it held only up to it in demostrations in a controlled setting, otherwise it was a trainwreck. So much for that.

And for the n-th time, WinFS is/was more than a metadata store.

davidofmorris wrote:



Zaki wrote: ... I'm sure another OS will have it in 2 years. And Microsoft can play catch-up again ...

It would be GREAT if someone else were to gain some traction in this space.  It would be particularly wonderful if an OFS layer gained acceptance in Linux or OS-X!  Microsoft are very good at catch-up and me-too.

I think Apple could do something similar to the WinFS "vision". They've got some of the building blocks in Spotlight, CoreData and Sync Services.

I doubt Linux has a chance. Judging from the extreme ignorance in the Linux community of what WinFS was promising, they wouldn't even know where to start.

DCMonkey wrote:



davidofmorris wrote:  Zaki wrote: ... I'm sure another OS will have it in 2 years. And Microsoft can play catch-up again ... It would be GREAT if someone else were to gain some traction in this space.  It would be particularly wonderful if an OFS layer gained acceptance in Linux or OS-X!  Microsoft are very good at catch-up and me-too.

I think Apple could do something similar to the WinFS "vision". They've got some of the building blocks in Spotlight, CoreData and Sync Services.

I doubt Linux has a chance. Judging from the extreme ignorance in the Linux community of what WinFS was promising, they wouldn't even know where to start.

What makes you say that?

fdisk wrote:



DCMonkey wrote:  I doubt Linux has a chance. Judging from the extreme ignorance in the Linux community of what WinFS was promising, they wouldn't even know where to start.

What makes you say that?

Because other than the long dead GNOME Storage project, and Novell's Simias (has that gone anywhere?), everyone in the Linux community seems to think that ReiserFS or ZFS does everything WinFS would have done and more. Those that don't seem to think:

1) Spotlight did it first
2) ???*
3) Linux is better than Windows

* You could probably throw in Beagle here somewhere.

There are a few other projects out there with interesting ideas on search and organiziation beyond basic metadata indexing, but nothing quite like what WinFS could have been. And IMHO there's too much graybeardedness running through Linux for anything so radical to come between the file system and the UI in a mainstream Linux desktop.

But it would be nice if they could prove me wrong.

DCMonkey wrote:


And IMHO there's too much graybeardedness running through Linux for anything so radical to come between the file system and the UI in a mainstream Linux desktop.

Normally I'd agree, but right now I'm worried that Microsoft's suffering the same problem too. I was pretty shocked to find out even Scoble has no real idea why WinFS was ditched.

The only real reason those blog entries give for killing WinFS s that there aren't enough end user value points, i.e., we aren't ready for it.

But that's a stupid excuse coming from Microsoft. If they can be the driving force to convert every Average Joe to a Joe User, surely they can be the driving force to move the world forward from a hierarchial file system to a relational one, even if it'd be gradual.

Gotta start somewhere no?