Has Microsoft changed? WinFS post getting questioned internally
Interesting threads on an internal Microsoft alias today. Employees are questioning why we (Microsoft employees) can't just own up to the truth and stop spinning when we have bad news to report.
Yeah, we're talking about Charles Miller's post about the death of WinFS where he took Microsoft to task.
It's not human nature to admit that you tried to do something and failed at it.
And, it's very hard to share where failure's lessons will be applied in the future because that'd be tipping your hand to your competitors about what you're doing in the future.
What happened to WinFS?
The Web killed it. *
The Web is how people use computers now and will in the future. Steve Gillmor loves to say Office is dead. Well, that has yet to be seen (whether Office dies or not is in that team's hands) but WinFS was a technology that'd have a hard time getting going because it's anti Web. It'll be interesting to see if this stuff comes back in a Web way. Ray Ozzie to the rescue!
* Update: this is my theory. I don't really know why it was killed but when teams and companies don't come clean and give us some transparency into why things get done then people will make stuff up. And, it is interesting timing that this came a week after Bill Gates announced he was going to be less involved in Microsoft and went on summer vacation.
Update 2: Shishir Mehrotra of the WinFS team wrote me and other bloggers who are talking about this internally and said my theory is wrong and that WinFS hasn't died at all, but is actually being rolled into SQL Server and a new project that's under development.
Powered By 
June 25th, 2006 at 5:09 pm
WinFS never was a product it’s a technology and a cool one at that.
It never was a desktop search product; it’s a way of organizing your stuff.
WinFS will never die, it will be the inspiration for the next new thing.
June 25th, 2006 at 5:20 pm
“What happened to WinFS?
The Web killed it.”
What a crock of shit. You make this totally absurd claim with little to no proof. I’ve heard some pretty crazy bullshit excuses in my time, but this takes the cake,
June 25th, 2006 at 5:29 pm
Tetra: hit a nerve?
June 25th, 2006 at 6:01 pm
Ok, so how did the web kill WinFS? I thought they were fighting different battles…
June 25th, 2006 at 6:07 pm
“The web killed it” sounds like misdirection to me, too. Nothing’s changed about the way we look at our own data to make the promises of WinFS any less relevant. PC hard drive sizes are getting insanely huge, and people are still filling them up with information, using a storage metaphor that was really cool back in the 1970’s.
I’ve been trying to work out why I found the announcement of WinFS’s cancellation so disappointing. After all, I haven’t used Windows as my primary operating system for over a decade, so I wasn’t exactly waiting for the next version of Windows with bated breath.
I think what WinFS represented to me, as a technically-minded outside observer, is one of the few things you could point to as a significant attempt by Microsoft to innovate. If Microsoft got WinFS right, it had the potential to change the way we thought about how information was stored on our computers, and how all that information could work together. Even if they got it _partly_ right, it would have been pretty cool.
And the fact it was going to come from Microsoft, instead of just being adopted by them after the fact, was a neat piece of cognitive dissonance for those of us who thought such a thing wasn’t possible.
It’s very rare that any company has the courage to try to tear down one of the major pillars of personal computing, and replace it with something better. And it’s sad when such an attempt fails.
June 25th, 2006 at 6:11 pm
Sure…whatever. “The Web killed it”
And what the hell are the webservers running on? A computer has to have a FILESYSTEM. Duh. Dumbass. And what about all the other data on the local system?
You should Google Filesystem, Scoble, because you have no idea what it means.
June 25th, 2006 at 6:12 pm
Kevin: almost all the heat and investment in the world of software is on the Web. Usage patterns are seeing huge shifts. Also, the funding for software is seeing a huge shift to advertising funded rather than user funded (do you buy anything from Google or Yahoo or MSN? No.)
WinFS was going to make a new kind of Windows application possible. Very cool stuff, but the world is moving away from Windows apps and toward Web-based ones.
In such a world WinFS is irrelevant and unneeded.
At least on client machines. I bet we see this tech come back as server-based stuff.
Watch what’s happening inside Microsoft’s leadership. Bill Gates is on the way out. He was the guy who funded WinFS. Who’s on the way in? Ray Ozzie. He is funding Web-based stuff.
This is my theory at least. I’d love to hear other theories, or facts. This is what happens when you try to spin and obfuscate people come up with theories. Right or wrong.
June 25th, 2006 at 6:16 pm
Q: WHY ISN’T VISTA OUT YET?
A: DURR IT’S THE WEB’S FAULT!
Hit a nerve? Yeah, the stupidity detection cluster.
June 25th, 2006 at 6:59 pm
Tetra: WinFS wasn’t part of Vista for two years.
June 25th, 2006 at 7:09 pm
And, Tetra, if you work on the team and actually know something, why don’t you share? Otherwise you know less than I do.
Cody: WinFS was NOT a filesystem. It was a layer that sat on top of NTFS. NTFS is a file system. WinFS was going to be a database that would add lots of features to NTFS, but was NOT going to replace it. Maybe you should watch the videos I did with the WinFS team before you attack.
June 25th, 2006 at 7:11 pm
Hey, Scoble, I use OpenBSD 3.9 for my server. I need you to tell me right now why I should switch to Vista Server.
Oh, and the machine I run BSD on is an Intel Celeron @ 533MHz, with 96MB of RAM. I can successfully run SSHD, HTTPD, sendmail, and all kinds of other services. Windows Vista can’t do that. Oh, and I don’t use X11, either, so I run in a pure command-line environment.
Beat that.
June 25th, 2006 at 7:11 pm
Robert,
There were a lot of things wrong with WinFS, none of them had anything to do with the Web.
PS: It’s quite provocative to claim this is Ray Ozzie killing BillG’s pet projects he didn’t like without proof.
June 25th, 2006 at 7:12 pm
Charles:
>Nothing’s changed about the way we look at our own data to make the promises of WinFS any less relevant.
That’s bull. EVERYTHING has changed about the way we look at our own data. You really need to come out and look around. Today Amazon has a file system, Google is giving away 2GB of email space that can be turned into a file storage space.
When I watch normal users use their computers they are switching to Web-based approaches very quickly.
Anyone who doesn’t notice that trend hasn’t been paying attention. You need to look at what Google is doing to the usage model. Ray Ozzie is.
June 25th, 2006 at 7:16 pm
Dare: you’re right. I’ll mark that as a theory. I don’t know either way. But, it is interesting timing that this came out a week after Gates said he was stepping down.
If something else happened, it sure would be nice for the team to come out and explain. This was pure spin. When you don’t apologize for leading the industry down a dead-end path you can’t expect people to take it without coming up with theories.
I notice everyone is attacking my theories but no one is providing a plausible alternative.
Cody: if what you have is working for you why would you change?
June 25th, 2006 at 7:29 pm
Robert, if you think that file storage system == file system, then it’s very good that you’re leaving Microsoft, because that would show a simply appalling level of ignorance for someone making comments on what killed WinFS.
Web servers have to have a file system, you know that thing that Gmail uses to actually store the files you keep in it? A database as a file system for a data retrival system is an outstanding concept, and why the AS/400 aka iSeries has been doing it for almost 20 years. Yeah. Think about that one Robert. Note that AS/400s tend to have uptimes measured in decades too.
WinFS wasn’t first, it wasn’t even particularly innovative in that space. But for data storage and retrieval systems, it would have been outstanding, as it would have allowed a lot of the basics to be handled by the OS, and then you build tools on top of that.
Of course, if every copy of Windows comes with a solid database file system, then maybe the Access and SQL Server teams start feeling nervous. Hmm.
But to say the Web killed WinFS is just stupid, and pandering for hit counts. It’s being Dvorak is what it is, and I thought that, at least in theory, that kind of thing bothered you.
June 25th, 2006 at 7:31 pm
As far as leading the industry down a dead - end path, that’s what Microsoft does really well. Gets people to put off all kinds of things waiting for SUPERCAPTAINCOOLGUYPRODUCT that aww, never shows up. But it’s folded into THESE products. Well, one day. But if you buy the current version and stay with Microsoft just a little bit longer, well, you’ll get some of this thing you’ve been waiting for since NT 5.
Dude, enough. That’s not rain in our faces, and the handjob stopped working years ago.
You want to know why no one trusts a thing Microsoft says?
Well, WinFS is a STELLAR example.
June 25th, 2006 at 7:44 pm
Well, ReiserFS is better than NFTS anyday.
June 25th, 2006 at 7:53 pm
Is it fair to say that XML killed WinFS?
I am not an expert on WinFS but my undestanding is that XML and WinFS were both trying to solve more or less the same problem: the notion of structured data and metadata on top of the filesystem.
XML, XML schema, XSL and all the related XML technologies are already doing that in a text format.
Since MSOffice annonced to support XML by default starting with Office 12 maybe that made WinFS less relevant.
As a developper, I saw a lot of overlapping between XML and WinFS and thought XML was the way to go.
Anyway MS has been trying to develop an object file system for like 15 years now. First with ofs then with WinFS. Both failed.
June 25th, 2006 at 8:06 pm
In such a world WinFS is irrelevant and unneeded.
If you truly believe that, it’s over for Microsoft. But the ‘web killed it’, ‘blame it on Ozzie’ oh oh…amazingly drecked on the sea-bottom. You haven’t even got into the databasesy Enterpriseish uses of file-systems.
You know, when you jump to these instant-microwaved philosophical leaps about software, I am sure at least half of the real programmers at Microsoft are glad to see you go. All the attention, none of the smarts.
Basically fitting, when Microsoft went most off-track, it had Marketing robots and bloggers as it’s frontline force.
So what’s your Microsoft legacy? Humanizing them? Well so? It killed them. Your Microsoft legacy is the PDC 2003 overhype euphoria. If being hated, allows Microsoft to produce good code and ship it on time, follow the roadmaps, not frauding with Software Assurance, then I hope they are eternally hated. If humanizing them and filming them, produces Vista vaporware, then screw that. End result is all that matters, not blogger-pretending-to-work feel-good happy Valley-hippieish emotions.
You are a great guy, but becoming Microsoft’s Apologist in Chief, well least you had the thick-skin for it. :) But I think podcasting, blogging and all that more up thy line. Plus, will reduce the virtol. Why fuel on the fire as parting shots?
June 25th, 2006 at 8:30 pm
Robert,
Web services are going to become more important but the idea that they’re going to obsolete desktop applications is a bit wild-eyed futurist–a lot of what you see in trendy tech magazines like Wired. Years ago Oracle/Sun/Netscape proposed that thin clients, network computers, would take over. But, I never liked this idea, and I don’t think an average casual user would like this–not because they don’t find this a neat idea–but because people don’t like the idea that you are forced to be connected to a network in order to use your computer. Especially when you have a less rich client. People also don’t like the idea that you are forced to be connected to a network to access your files.
I use services like Gmail, I like the fact that I can have my megabytes of email on store and access it from any public computer and I don’t have to worry about transferring them if I get a new computer or saving them if my computer breaks down. There are benefits to public storage. But I also wan’t to be able to store my files off of the Internet, and there are some things I will always have on a hard drive even if I back them up on the Internet.
And even putting the average home user aside, corporations and other organizations have the need for local storage systems. As I remember, one of the main things delaying WinFS was operation across a server system–which is not important for a home user.
And, as others have pointed out, a server which hosts web storage itself needs to have a storage system, and WinFS is a good solution. WinFS would have its place doing that; plus it would be available for local, desktop storage, and not leave local storage dry and in a primitive state compared to the web just because its ‘untrendy’.
A lot of visions of how the Internet would take over and NCs would displace PCs became popular among the Wired futurist crowd in the 90s; and now with the whole notion of ‘Web 2.0′ this idea that web services will displace desktop applications is becoming popular among the same crowd. Futurists are not pragmatic in the way they think about technology.
Web services will have growing importance, but it doesn’t obsolete the desktop or rich/thick client in any way–and in fact the same basic OS technology that aides web services will aide desktop applications, like WinFS which could be used for web servers.
I think you understand all of this. If it is true that to some people in Microsoft, ‘the web killed WinFS’ then they are making a mistake.
Franco, WinFS heavily used XML in how it operated
June 25th, 2006 at 8:40 pm
I’m amused at the claim that Microsoft was “innovating” this WinFS. IBM’s VM/ESA’s had a database layer on top of a file system for ages now, aka the Shared File System. I’ve been wondering when they’ll incorporate a decent logical database layer as well.
As far as web-based versus filesystem based applications go, most web-based applications turn out to be database-based, using the memo type or its equivalent; so WinFS is merely making that the standard practice for desktop application as well, blurring the lines between web-based and desktop applications.
I think Microsoft blew it. Microsoft doesn’t need to show and tell anyone where this failure’s lessons will be applied - anyone with any nous knows that already. You can tell, can’t you?
As far as “innovation” goes, database layers on top of file systems is old hat.
June 25th, 2006 at 9:17 pm
John: I wasn't the one that brought up FileSystems. Cody was. I was reacting to him. I know that WinFS isn't a FileSystem.
Christopher: if half of Microsoft employees feel that way they aren't paying attention. I listen to what our leaders say. Kevin Johnson, Microsoft's co-President has met with our group twice in the past few months. He said that Microsoft's growth was coming from Advertising technologies and that they are investing in Web technologies.
But, don't take his word for it. Here, let's do a little investigative reporting. Go and visit all the parking lots on campus. There's one building that is almost impossible to park in: the one housing Windows Live. Go in and workers are tripled up because they've been hiring so fast.
Or, just look at where the smart people are working. Dare? He left the XML team which worked closely with the SQL Server team to go to Windows Live. Niall? Windows Live. Ray Ozzie? Windows Live.
It's pretty clear where Microsoft is spending its $2 billion that the investors weren't planning on.
Brian: you're right. But wrong too. Many normal people don't use files anymore. They use Web sites. Go and watch them. Or, look at where the investment is being done in Silicon Valley. Or, go and look at what technologies are seeing rapid adoption and growth. Hint: it isn't Windows apps. WinFS was an attempt to jump start Windows app growth again. I think the fact that it's being rolled into server-based stuff makes it pretty clear. Microsoft is investing in Web technologies in a big way. That's why I say Web killed WinFS.
June 25th, 2006 at 9:27 pm
>If humanizing them and filming them, produces Vista vaporware, then screw that.
I saw Windows Vista, er Longhorn, before I was a Microsoft employee. The hype machine was in place long before I was. Ever since Longhorn went back to the drawing board two years ago I’ve been very careful not to hype it up. If I should be blamed for anything it should be for not explaining why Vista went back to the drawing board.
>You haven’t even got into the databasesy Enterpriseish uses of file-systems.
The Enterprise is moving to the Web too. Did you miss Salesforce.com? When I visited companies I saw more and more Web applications and services.
>If being hated, allows Microsoft to produce good code and ship it on time, follow the roadmaps, not frauding with Software Assurance, then I hope they are eternally hated.
Hey, they have some evangelism jobs open. You’re free to join and mold Microsoft in your vision.
>people don’t like the idea that you are forced to be connected to a network in order to use your computer.
Brian, I know quite a few geeks who feel that way. Problem is the usage model proves you wrong. Google requires you to be connected to a network to use it and there isn’t a huge outcry that says “I don’t want to be connected to use that.”
Your cable TV requires you to be connected to use it and there isn’t a huge outcry that says “I don’t want to be connected to use that.”
Your Xbox (and, soon, your Playstation) requires you to be connected to use it fully and there isn’t a huge outcry that says “I don’t want to be connected to use that.”
To tell you the truth, it would be a good thing for Microsoft if people believed the way you did, but they don’t. The growth in business will come from connected systems, particularly ones that work in a Web browser.
Oh, and watch Apple — I predict its market share will take a sizeable jump this quarter. That’s another reason WinFS wasn’t going to fly. But, if I said that then I really WOULD be pimping to the Dvorak-side of the house.
June 25th, 2006 at 9:34 pm
Robert,
If the main delay behind WinFS being incorporated into Windows was that the server-side operations weren’t refined—and Microsoft is continuing to refine them as WinFS technologies are being roled into server-based applications, what is to stop Microsoft from putting it in the OS once they get all of this polished down? It would benefit the user and there’s no need to throw away the concept or the work.
June 25th, 2006 at 9:36 pm
Brian: nothing is, but by the time that all gets done the customers won’t care.
I tried to resist Steve Gillmor’s points. But, he’s right. More and more of my files are going online. In three years (there’s no way Microsoft can get out a new version of Windows before then) I really won’t care about a local file system.
Oh, and Windows Live Desktop Search delivered a good chunk of what WinFS promised anyway.
So, maybe I should have said desktop search AND Web killed WinFS? I am actually suprised no one pointed that out.
June 25th, 2006 at 9:48 pm
Normal people don’t use files anymore? Do you know any normal people? Gee, hmmm, what exactly were those pictures I downloaded to my MAC and printed for my wife’s scrapbook made up of, then? I thought they were files.. silly me.
What was it that my daughter was typing in when she was doing here research papers for college then sending off to her professor? I could sworn they were word docs.
Where did I get the data to fill our my tax return this year? I could have sworn they were stored in some Quicken files. How could I have been so confused?
What is that data I have to clear off my Mac via my browser?
What is my wife creating when she does the flyers for our school? This is all so baffling now. Normal people don’t use files anymore?
And when did Kevin Johnson get promoted to CEO? Did the street miss the announcement that Ballmer was on his way out? The Microsoft web site says he’s Co-President of Platforms and Services. Must be outdated.
June 25th, 2006 at 9:55 pm
Dmad: yes, I meet lots of normal people. And they show me how they are using Internet technologies to do more and more of their work. I was typing too fast, though.
Oh, crap. I’ll fix the Kevin Johnson thing. I meant President.
June 25th, 2006 at 9:56 pm
Dmad: it’s funny that you use a Mac. You wouldn’t have liked WinFS anyway.
June 25th, 2006 at 11:03 pm
Interesting, the WinFS team wrote me and gave me more details and explained why I was wrong. I wish they could just come here and to other blogs and say that, though.
June 26th, 2006 at 12:21 am
Robert,
On Don Box’s blog Michael Herman left what I think is a very insightful comment:
“The strategic importance and success of a technology at Microsoft is almost totally dependent the extent to which it is adopted and integrated into other Microsoft products”, Michael Herman (about 1990)
When Windows stopped believing in the WinFS vision (at least for the next round of o/s releases) and Microsoft Business Solutions didn’t pick it, the list of interesting internal WinFS customers was almost non-existent.”
That sounds on the money too me.
Personally I think that Microsoft have just questioned their initial assumption that WinFS had to be a clientside file system, with all the inherent security, performance and compatibility issues that entails.
By focusing on making a server offering I think they will get something out a lot quicker.
June 26th, 2006 at 4:54 am
I agree on your comments, we are really moving toward a distributed file system on the Internet.
Doing some quick maths at home - almost all of my disk usage is application and system files. Pictures, music, calendar, e-mail and movies are all stored in different Internet-based services. The only important stuff left in my computer is source code since I haven’t signed up on a web service yet, though I expect this will happen during the next year. However, they are still all nicely indexed using a desktop search application which renders WinFS unnecessary.
That said, all this is from a consumer’s view. We mustn’t forget the server side! I do believe that we need a better way of storing and indexing massive amounts of data there, which is where I found WinFS an interesting technology. I hope WinFS makes it to the Longhorn Server.
June 26th, 2006 at 5:45 am
[...] scobleizer always has the inside edge on all things MS. He weighs in on the WinFS announcement as well. [...]
June 26th, 2006 at 6:01 am
[...] Scoble thinks the web killed WinFS. More opinion can be found on TechMeme. Personally, I think Charles nailed it (pun fully intended). Comments | Blog This | E-mail This | Print This | Permalink [...]
June 26th, 2006 at 7:13 am
[...] Scoble has an interesting, but IMO, completely offbase opinion on why WinFS is dead. At least I hope it is off base, if it isn’t Microsoft is even more far off in their thinking than I ever thought and I need to start spending even more time on FreeBSD. [...]
June 26th, 2006 at 7:45 am
@27. Where did I say people weren’t using the internet? You are making this a two-valued proposition and it’s not. Sure people are using the internet, but they still use files, too. How do those pictures get up on flickr? How are your vidcasts going to get recorded? Not saying the internet is not getting more use from a sharing and distribution perspective for normal people, but to say normal people don’t use files? It takes quite a while for the other end of the adoption curve to catch up to where you want to believe they are.
June 26th, 2006 at 8:11 am
I think ‘more and more people are using web to store their files’ is just wishful thinking on Robert’s behalf. Maybe some consumers, yes. But you won’t see this happening in the business sector or anywhere else.
Robert is totally mixing up what the ‘web’ is. Storing files remotely and locally is totally compatible with WinFS and windows applications. Using HTML as the UI is what most people think the web is - and I wouldn’t bet on this being an area of growth over the next 5 years. I think we’re pretty much at the limit as to what HTML/AJAX can do. Look at WPF, Flash, Apollo as the future UI platforms. These applications will need to work in connected and disconnected modes with local file stores and remote web service calls.
In 5 or 10 years time the ‘web’ as we know it now will be how we consider an HTML3 page now.
June 26th, 2006 at 8:35 am
I think this post from Dare is quite good on this topic: http://www.25hoursaday.com/weblog/PermaLink.aspx?guid=f27302c4-b898-4cec-9c46-249106ea76c8
June 26th, 2006 at 8:39 am
The main reason that a lot of corporate environments are angry at Microsoft pulling features from Vista is that the companies were sold on the new features when they signed up for Windows Software Assurance. One by one, the features have been pulled from the Windows Vista final product. The most requested feature and the tipping point for many companies was WinFS.
June 26th, 2006 at 8:41 am
Perhaps WinFs no longer being seen as improving the windows consumer filesystem this will open the door for windows desktop search to be integrated in the filesystem so as a home user WinFS-light?
WDS is a tool that people can use without having to know where there files are, plugin based so it can understand some information of the files etc, and it runs on top of NTFS.
Yes it’s not the same and it’s not perfect but probably in the way it can be used it can become similar?
June 26th, 2006 at 9:00 am
I knew you’d give have more interesting Microsoft news once you left the company. You have all the sources and no longer feel any reason to toe the company line.
Yes I know, no one ever told you what to do, but it’s a lot easier deciding to write something controversial when you’re paycheck isn’t on the line. And, I know you took some unpopular positions, but not very many, and only ones you felt strongly about.
June 26th, 2006 at 9:01 am
From what my users experience, they tell me that they liked the Windows 98 search feature better because it was more accurate than the Windows XP search (searching for text within documents.) Microsoft approached us like they did everyone else with their vow that they would “surely” release “Longhorn” within three years (3 years from March 2002) and asked us to sign up for Software Assurance. We did not sign up for Software Assurance and I do wonder what happened to those who did? Are they just out of the $$$ that they spent for a “free” upgrade to the next version of Windows? I can’t understand why Microsoft realized the importance of finding files quickly, told their users that the next version of Windows would allow them to find files anywhere - that they didn’t need to use folders, and then ran off to compete with Google in the web search arena. I guess since we have no alternative, we just buy it at retail price after Vista SP1 is released…along with lots of sticks of RAM.
June 26th, 2006 at 9:19 am
The web killed my brain.
June 26th, 2006 at 9:22 am
Isn’t WinFS already here in the form of Google Desktop? Or, Spotlight on that other platform that dare not speak it’s name?
Also, most people can find all their pictures by looking in the “My Pictures” folder… :)
June 26th, 2006 at 9:28 am
The light flickered, went out, plunging the room into inky darkness.
Suddenly, a shot rang out.
The lights came back on.
WinFS was in the middle of the room, lying face down in a pool of its own blood.
A woman screamed. The murmuring crowd instinctively recoiled in horror, backing away from the corpse.
The double doors flew open. A man strode into the room, and locked the doors behind him. “I must ask everyone to remain the room,” he exclaimed. “Someone here has committed a murder.”
“Who the f–k are you?” asked Tetra.
“My name is Inspector Scoble,” he said, doffing his deerstalker hat. With his other hand, he lifted a Meerschaum pipe to his lips, and puffed vigorously.
“Surely I’m not a suspect,” Cody said. “I don’t even use Windows. I’m a Linux troll!”
“You might have had the motive, but I do not consider you a suspect,” Scoble said. “I believe… THE WEB killed WinFS.”
A weighty silence hung in the air.
“That is the stupidest f–king g–damned thing I’ve ever heard in my f–king life,” replied Tetra. “You’ve said some amazingly stupid bulls–t, but that just takes the f–king cake.”
Scoble lit a wooden match and relighted his pipe. “It sounds like someone has a case of the Mondays,” he said.
“Your hypothesis is nothing but wild speculation,” said Christopher Coulter. “I bet all the other Police Inspectors think you’re a retard.”
“The web couldn’t have killed WinFS,” said Brian Shapiro. “It was a trendy thing to say a few years back — the Web killed this, the Web killed that — but the reality is that desktop file systems are still with us today.”
“WinFS wasn’t a file system,” said a pedant in the back.
Dare Obasanjo crossed the room to a crystal decanter of spirits, uncorked it, and poured a crystal tumbler to the rim. “WinFS had a lot of… shall we say… ‘health issues.’” He raised the tumbler to his lips and sipped.
The crowd stared at him.
“Not cryptic enough?” asked Dare, taking another sip.
“Saying the Web killed WinFS is just stupid,” said John Welsh. “It’s something you’d expect from Inspector Clouseau.”
A small, wet cough emanated from the body on the floor. The crowd gasped. WinFS stirred.
“I… I… can’t believe you’re all so angry at Inspector Scoble and his theory… instead of being angry at… the one who pulled the trigger.”
Scoble sprang to the center of the room, and knelt next to WinFS. “Who shot you, WinFS? Who pulled the trigger?”
WinFS made an effort to speak, but could only murmur.
Scoble leaned in closely, placing his ear next to the whispered words.
“I see,” Scoble said to WinFS. “Well why don’t your developers just put that on their blogs? Why the mystery?”
WinFS coughed, a bright red sputter of arterial blood appearing on its lips… and smiled.
The smile was frozen on its face, as the WinFS process terminated.
Shishir Mehrotra stepped forward and raised his hands delicately. “Do not think of this as the death of WinFS,” he said to the crowd. “For this is merely part of the coding cycle. No code ever truly dies. It merely appears in another form.”
A woman sobbed quietly.
“The soul of WinFS will appear in other products,” Shishir continued. “One day, you may see a database, or an application, or a tree structure, its leaves and nodes glittering in its memory space, and you will know: truly, that is the soul of WinFS, reborn into another product experience, part of the greater Data Platform Vision.”
The crowd pondered his words.
“That is the biggest f–king piece of bulls–t I’ve ever f–king heard in my whole f–king life,” said Tetra.
Scoble grinned wryly and puffed on his pipe.
June 26th, 2006 at 9:35 am
Robert, you think because all the investment in Silicon Valley is going into tiny web startups means that it’s the future of computing? Have you not learned your lesson from the last time?
This “free world where advertisers foot the bill” didn’t work in 1998 and it isn’t going to work in 2008. The investment back then was in companies that did grocery store delivery and scooters that were going to change transportation as we know it.
Why do you web geek types think that it’s an all or nothing desktop and web world? Do you remember mainframe computing? Do you remember BBSes? Do you remember CGI-generated web sites loaded with Java applets? This prophecy that the desktop is dead is really getting old. How long will people tolerate being confined to the borders of their browser windows? Or bogged down by standards bodies? How can the web make any progress if Microsoft and Mozilla can’t agree on what hot new feature to add to their browser?
The answer is of course that they won’t tolerate it. Just like they didn’t tolerate the slow feature crawl of terminal emulators and “web 1.0″ browser features. Eventually we come back to the client. Let the network be what it was designed for… moving data.
June 26th, 2006 at 10:22 am
I have a Mac at home that I tinker with and I love Spotlight. I have users at work who have tinkered with Google Desktop Search and find that it works very well. I just wish that Microsoft would include some useful features (for everyday users) in the new OS rather than just stuffing it with eye candy that eats up RAM. And here I sit with a brand new stack of Software Assurance Agreement Summaries in my inbox. Office Software Assurance yes! Windows Software Assurance no!
LOL @ Karim
June 26th, 2006 at 11:26 am
You’ve been very careful not to hype it up? Man. You live in some alternate universe? You do know blogs have archives, doncha? What 3 months of non-stop PDC 2003 hype and the off-the-chart post hype beyond that, Team 99 and all, until that big brick-wall code redo. The “unlike any event” and “never see another one like it again” was that part of the careful not hype up part? ;)
http://scoble.weblogs.com/2003/10/10.html
Also, keep in mind that this is unlike any event Microsoft has ever done. I’ve never seen an event like this. We’ll probably never see another one like it again. A new OS. A new SQL Server. A new Visual Studio. All being shown off by the techies who are building these things. 15 days and counting.
And Salesforce outages are legendary, and Salesforce is mainly for Small Biz, you think all the Fortune 1000 Oracle and DB2 tables chuck full of proprietary and confidential data is gonna go out on some external web app? Dreaming, bucko. Doncha member the failure of Hailstorm and Passport? As if we don’t have enough of a problem already with massive data disclosures. The tech utopia, prodcuing id theft nightmares.
Anyways, end prog.
June 26th, 2006 at 11:31 am
Karim bravo. Get that into Final Draft format and sell to Hollywood. :)
June 26th, 2006 at 12:43 pm
The whole “it’s not dead, it’s just rolled into SQL Server” argument is not going to wash unless we see SQL Server appearing as manifestly an integrated part of Windows.
Remember that WinFS was going to change the way we interacted with and organised information on Windows.
That’s not going to be achieved by any number of bells and whistles added to SQL Server.
Of course, the rolled-into-SQL-Server excuse might be sustainable from the point of view of whatever the scope of WinFS was reduced to before it was er, not killed - but that’s not what we were originally sold. Saying WinFS is not dead on that basis is like saying the Amiga isn’t dead because it’s name is still used by a company that makes mobile games. Whoopee.
That announcement was, let’s face it, creepily disingenuous, and eerily reminiscent of the way the Scientologists spun LRH “not needing his body anymore”.
June 26th, 2006 at 12:51 pm
You know, i find it interesting that Robert is saying out of one side that local storage is dead, and it’s all about online storage.
But then he says the web can’t replace fat client applications. Well, you need local storage for fat client applications, or you need an always on, gig-E level connection with guaranteed uptime no matter where you are in the world.
I don’t quite see how one can have both without massive changes somewhere.
June 26th, 2006 at 12:53 pm
But kevin, if it’s part of SQL Server, then they can charge you for what you would have gotten for free.
June 26th, 2006 at 1:01 pm
PS. Karim, that was brilliant.
June 26th, 2006 at 1:58 pm
Karim, please warn me when you write something so good and funny. I almost spit Diet Coke all over my monitor.
>This “free world where advertisers foot the bill” didn’t work in 1998 and it isn’t going to work in 2008.
Really? Google came out of that period. If that’s “not working” then I wanna “not work” too! :-)
June 26th, 2006 at 2:21 pm
Personally, and I don’t know about anyone else, but I don’t want various web silos to be the sole owners of my personal data. Maybe there are some photos I’d rather not have up on Flickr, or my blog or whatever. Maybe I’d rather not have Google as the sole custodian of my spreadsheets and documents.
Some of the arguments here are utterly facetious. I don’t want Google while I’m not connected because in that situation being able to search web pages seems pretty redundant. But I sure want to still be able to read and write documents, email (even if I can’t send it) read feeds I downloaded while I was online, play games, listen to music etc. while I’m disconnected.
My Xbox 360 works just fine disconnected. Sure, I don’t get the value-added stuff that’s coming from Live, but you can bet there’d be a hideous outcry if being online was a requirement. People are, mark my words, not going to like the day they wake up and find their computers are little more than dumb terminals to ad infested/supported web apps.
Anyways, personally, I wanted WinFS to be a nice object-orientated relational data model that I could dump data into in fantastic structured ways and they query later. For developers it would be gold dust. Imagine a world in which every program logically connects together every single bit of data in a consistent way. Your instant messenger contacts are connected to your e-mail to your chat logs to the music they’ve sent you, to the documents you’ve shared with them. That’s really great compelling stuff.
I get your Web 2.0 enthusiasm, but you live in rarefied air. Most of the world doesn’t know what RSS/Web 2.0 is, and they care less. Whilst connectivity is important, the really interesting applications in the future are the ones which leverage the client AND the network. I mean, your big HD craze of late is just incompatible with the “web” alone. It’s relying on client tech like BitTorrent to get it done.
The future is peer-to-peer, I think, maybe backed up by central services. Think of a vision like WinFS, but one that connects with other people you know. So your queries will touch stuff they have too. Maybe. I don’t know. I think the future is in interconnectivity between things out there in the cloud, people out there, and things I have. I want my phone to sync contacts properly with my computer. I want my photos and text messages from my phone to come up to my computer and be associated with the contacts who sent them. I want to not have totally seperate contact lists in Outlook / Windows Live / MySpace / whatever. I want to have stuff that’s mine and stuff that’s out there seamlessly. So there’s a role for the web, there’s a role for my own data to be collected and organised.
The value is in the plurality of the experience. It’s going to suck if we have to use Google everything to get any value out of it. The only way to make best use of data is if everyone buys into a single way of describing and organising data. I mean, one of my pet peeves at the moment is José González. If I want to find every song on my computer right now that has him in, I have the first problem that not every song with him in is labelled as such, and the second that some of them miss the accents, or spell his surname with an s. If instead they referenced some standard way of representing him as an artist, that’d be better. I don’t see how Google is going to go about solving that problem! Are they going to give me a Google music store with non-broken metadata that I can search across? What about the music I’ve already bought?
With WinFS, this kinda music data would be broken out of the silos it’s in currently, then I could use some general-purpose querying tool to go in and fix up that data. It gives the user back ownership of these things. It would mean different programs could all seperately agree on some of these basic concepts, and then start to do some sensible things with them.
WinFS is dead because it was over-ambitious, and they tried to do too much at once. So Microsoft have cut their losses and reassigned the bits that could be salvaged into other other products. It was always a hard proposition. It’s hard because it’s worth doing. But oh well.
June 26th, 2006 at 3:34 pm
[...] Even Robert Scoble, who usually has the inside scoop on everything newsworthy happening insde the software giant had a hard time making heads or tails of the situation. Interesting threads on an internal Microsoft alias today. Employees are questioning why we (Microsoft employees) can’t just own up to the truth and stop spinning when we have bad news to report….What happened to WinFS? The Web killed it…Update 2: Shishir Mehrotra of the WinFS team wrote me and other bloggers who are talking about this internally and said my theory is wrong and that WinFS hasn’t died at all, but is actually being rolled into SQL Server and a new project that’s under development. [...]
June 26th, 2006 at 4:29 pm
Vista odds and ends
(Jim Thompson is blogsitting TechBlog while Dwight Silverman takes a 2-week unplugged vacation. Dwight will resume his usual TechBlog duties on July 1.) Here’s a summary of recent news from the world of Windows Vista: Ed Bott notes that a…
June 26th, 2006 at 8:37 pm
June 27th, 2006 at 8:20 am
Digg 3.0 launching…
(Jim Thompson is blogsitting TechBlog while Dwight Silverman takes a 2-week unplugged vacation. Dwight will resume his usual TechBlog duties on July 1.) As TechCrunch reported last week, a new version of Digg, 3.0, is due to launch today. Digg……
June 27th, 2006 at 8:22 am
Scoble’s right, and I’m glad to see him sticking to it.
Our work stuff is all on servers, our home stuff is all online. I don’t need a database to run Warcraft, and I don’t need one to organize photos on flickr. People are writing stuff online, not in Word, and Google can find it.
Until someone does real image search (here’s a picture of my dog, find all the other pictures with my dog in it) we’re not going to get much better locally. What would the database have done that tagging can’t do? That folders can’t do? That date sorting can’t do?
I don’t think Picasa can be significantly improved.
I don’t think Itunes can be significantly improved.
Oh, eventually, but not in the near future, and not with some database built into an OS.
The real issue is where is MS going to improve Windows? What exactly can they do to challenge the Internet as a platform? And how can they have so many more developers and so much more money than Apple and just be treading water?
1. They’re duplicating Unix (Least Privileged Mode) instead of adopting and subverting. Dumb.
2. They’re trying to make IE the predominant platform and losing. Dumb.
3. They obsess over backwards compatibility, even in the consumer version of Vista. Dumb.
I mean, what are these people doing?:
http://research.microsoft.com/os/singularity/
Just plan a path, use virtual technology like Apple for compatibility, and move on with life. MS could have created an entirely new OS in the time it’s taken Vista to not come out. If you’re going to have 9 versions of Windows, quit trying to make them identical underneath. Plenty of companies use Linux and Windows side by side, I’m sure completely different versions of Vista home and business could co-exist.
June 27th, 2006 at 12:15 pm
I think a more accurate comment would be:
“Microsoft’s incompetence killed WinFS. The Web renders it’s death inconsequential”.
June 27th, 2006 at 12:29 pm
solomonrx,
well, as many people have said, the same winfs database technology that can be used on the desktop can be used on web servers; and it seems microsoft is continuing to develop this path as it is incorporating the winfs technology into server products. my point was there is no reason to divorce it from local storage technology and leave the desktop deprived of this; the implementation of a file store on a local machine isn’t the problem delaying WinFS its apparently both implementing network service and creating an intuitive end-user interface for accessing the store.
Whether or not more and more files are on the web, all files will not be on the web; and the amount will differ per user. An home user might have more on the web than a developer, a photographer, a designer, a writer, etc. But most people have some sort of job, or some sort of project they’re working on. And like I said, people don’t want to be forced to connect to the Internet to do anything.
Even so I still don’t agree that as of now people are no longer putting everything online. You really really believe people are no longer using word processors?
A WinFS data store is a great improvement even though, just like Avalon it may not seem like much to an end user who already feels like he can do everything he needs. But development of rich and adaptable user interfaces and controls is leagues easier and things render much smoother with Avalon. We should expect better applications as a result, but users won’t really see something in-your-face that on technical terms couldnt have been done in XP. Users don’t ‘need’ WinFS but it is an improvement. It does more than what tagging and indexing alone can do; it makes the metadata for files self-described, so you don’t need to have to create a DLL or dowload a programed filter to access the data associated with a file–it also makes the line between file content and header info unimportant. This is a great improvement and there are numerous ways within my personal projects that this would make the OS a bit more helpful.
It would also make it helpful for others who also will continue to work on projects more complex than a PTA flier, letter to the editor, or blog entry—but even if everything is done at work through servers, those servers could use a WinFS file store, ande even if at home done through the web they web servers could use a WinFS file store.
And this hasn’t been given up on… as you see Microsoft wants to use WinFS technology in server applications. The implementation of WinFS in servers had been cited as one of the reasons for its delays. So it isn’t the web, in that case, that has caused its ‘death’. The second reason cited has been problems implementing a good user interface for home users. This will probably mature and there will be no issue adding it later.
Whether people will care when its finally in the OS, as Scoble puts it, maybe isn’t the most relevant thing, and it certainly doesn’t mean WinFS is ‘dead’ or that the web killed WinFS. WinFS was always less important than observers made it out to be–people are fine with how they access their files now. It still is important, but in the way Avalon, Indigo, etc. are important; in a less in-your-face way in that they revolutionize the core OS functions–the improvements they make to the end-user experience are substantial but subtle and users should just expect to see a better product. People for some reason seem to expect every major OS release to look completely different from the one before it. Windows will look the same, just better.— even though one can see a double standard when talking about Apple–because even when they implement very minor things (that Microsoft may implement also), everything is seen as revolutionary whether it is or not.
But anyway, this is why people have reacted the way they did to Scoble’s idea. Not everything is an issue of web vs desktop; and I don’t believe this is.
June 27th, 2006 at 1:18 pm
What Does the Death of WinFS Mean?
First of all, on the WinFS blog there is an update to the update on WINFS’s demise.
You may have noticed that storage has become a commodity over the past decade or so and that we’re all creating more and more data these days, taking photos…
June 27th, 2006 at 5:43 pm
Just a guess. But I doubt the web was the primary issue. Most likely, the needed work was not getting done rapidly enough. Plus the case for developing a relational file system may not be that strong. You can always make SQL Sever Express a standard part of the operating system and develop the tools to make it easy to use for an object relational mapping. That should basically accomplish the same thing without requiring as much new work. It also may take more steps than can be done for Vista. But the Linq project may still make a significant start.
June 27th, 2006 at 8:45 pm
You know what? If the Web can kill WinFS just because it’s not a web service, the Web can kill Windows and it can kill Microsoft. Will it? No.
Scoble, please investigate why WinFS was abandoned. Please. I can’t believe even you don’t know the truth.
And its use in SQL and ADO.NET is NOT WinFS.
June 27th, 2006 at 10:47 pm
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?
None of the people bitching here have stated just what the negative impact to them is going to be. They just oppurtunitistically use this to get out one more anti-MS complaint. you people are tiring.
June 28th, 2006 at 5:12 pm
MS catches flack for their announcement that WinFS as a standalone product is dropped
I think many of the reactions to today’s announcement from Microsoft that they are scrapping WinFS as a separate product is a good evidence of the challenges and risks big business has to deal with in the blogging era I talked about earlier.
The …
June 28th, 2006 at 10:14 pm
“Well, ReiserFS is better than NFTS anyday ” Its true
June 29th, 2006 at 8:08 am
“None of the people bitching here have stated just what the negative impact to them is going to be.”
James,
The added feature of being able to do a “Search” and actually find what we are looking for within a document. We have been waiting for WinFS to help solve this problem, and I’ve looked at some alternative solutions a small bit. When Microsoft said that they would include WinFS for Windows Longhorn, I almost signed up for Software Assurance in 2002 when they wanted me to. For now, it’s back to the drawing board looking at third-party applications.
That’s one example of a real world letdown from Microsoft’s lack of implementing WinFS.
July 4th, 2006 at 6:52 pm
Let’s just roll this back a bit to some more interesting debate like “users don’t use files anymore they use websites”
That’s partly the problem - security - there is none left. Website are for oozing out opinions that aren’t worth the enough to be put into a file stored on a personal drive; unless you’re a researcher by trade there isn’t much need for putting anything into a documentative format.
Such flame posts as 60% of those above don’t add anything to the debate. They simply inflate the value of opinions.
Around the age of 25, a person today, must think of re-investing in personal education which will grow them beyond mere opinions read here and there.
For all the “paperless” technologies we’re trying to invent as a society, books are still the #1 source of educative and documentative forms of scholarly knowledge.
It’s very very clear by most of the posts here, which of you have classical educations in datastructures and computational theory, since 80% of the data posted contains no information.
[Recommended Reading: Practical File System Design: With The Be File System by Dominic Giampaolo ISBN:1-55860-497-9]
As one of my database design projects in college, I developed a fully functional database using only txt files stored on a FAT12 Bootable Partition which booted directly into the db management environment. It wasn’t complex at all since it was mostly batch files that poled for input at the prompt and did a variety of comps, pulls, pushes, and appends +++. I scored 100 on that project because the idea was so unconventional to Prof, but followed perfectly all of the guidelines established in our database design textbook, I simply applied an alternate approach to how we view a database environment should be. We’d all become too accustomed to ‘infrastuctures of our generation’ lol.
Well, WinFS? Never noticed to begin with… I’ve been busy engineering custom solutions to real industry problems for the past 10 years without ever noticing that Microsoft’s new technologies were going to affect what I did, and still do today.
Boohoo, no WinFS. My real concern is with Tech Companies WoWing the kiddies into believing that “Personal Computers” are not important or relavent anymore because “the web” does it better. Hmm, well my private information and documents are oddly enough… MINE, and not stored on an Internet connected PC.
Ciao.
July 8th, 2006 at 10:45 am
There is no reason WinFS couldnt be used behind the scenes of a web app as well as a winform app, so I do not necessarilly agree that the web killed WinFS.