WinFS - operating system to operating table
»
It was touted as the next big thing from Microsoft. But is Microsoft’s storage strategy in tatters?
If ever we needed a demonstration of how corporate blogging can, when done wrongly, be a very bad idea, then the announcements of the changes to the delivery schedule for WinFS should fit the bill pretty well.
Even several weeks after the debacle, I’m not quite sure what possessed Quentin Clark, head of the WinFS team, to say what he said in the way he said it. What quickly became clear was that the rest of the world didn’t understand the story, so newswires, blogs and commentary websites soon filled up with the news that WinFS was dead.
When I read this news, I was so appalled that I went straight to Clark’s blog posting and read it for myself. Then I went off to get a strong black coffee, because it was written in such a bizarre dialect of marketeer-speak that it made no sense at all, and it certainly didn’t sound like the Clark I’d met on many occasions. He always came across as articulate, quietly spoken and quite precise in his utterances in those days.
How were we supposed to respond to these various announcements? Clark said ‘Is WinFS dead? Yes and No.’ What are we supposed to make of that? Does it have anything to do with Schršdinger’s Cat? It got better still when looking at the timescales involved. This bombshell was dropped on Friday 23 June, just a few weeks after WinFS technology had been touted so highly at the TechEd conference in Boston. Indeed, in a blog posting of 22 May, Shan Sinha, a program manager on the WinFS team, posted the following breathless testimony:
‘I have the good fortune to spend my time working with partners, customers and developers to build out our early adopter ecosystem for Microsoft. It’s one of most fun jobs anyone can have! First off, I get to work with people like you, investigating new opportunities for innovative applications on WinFS and dealing with the challenges of delivering those ideas. Second, I’m part of a team at Microsoft building the most innovative technology our company is currently working on (or at least we think so :-)).
‘Imagine a world where data storage and retrieval just work the way they should – no need to create clumsy mappings between objects, relational tables, and byte streams being stored in files. We finally can realise a world that simplifies the persistence, manipulation and retrieval of data, giving us an opportunity to create unique new applications based on those new capabilities. If you are reading this, it would appear that you believe in this vision too. It could not be more exciting!’
It turns out that it could. Clark’s first blog posting raised the excitement level to a fever, leaving more questions unanswered than answered, and his second one, attempting to patch up the mess, raised it to boiling point. Let’s look at what was actually announced. The current vehicle for delivering WinFS was an OS add-on patch that tapped into a background SQL Server-alike engine, then created the illusion of a new file system on top. This was going to be one of the three main pillars of Vista, the other two being the presentation services and XML middle-tier data services.
Some time ago, it was announced that WinFS wasn’t going to be shipped with Windows Vista, because a decision had been taken to make WinFS much richer in design. There’s a simplified storage engine concept in Vista itself, which is fixed in function and aimed at displaying and handling objects like diary items and so on. But it was deemed that WinFS needed to be bigger and more all-encompassing than this. At first the claim was made that WinFS would arrive some six months after the release of Vista.
Then Beta 1 was released, which I’ve covered in a previous column. It has extensive functionality, especially the ability to define queries and then store them as objects, which can then be queried themselves. In other words, it made it possible to define a search along the lines of ‘tell me which are our top 10 customers by sales in the last three months’, then build a derivative query that said ‘for those top 10 customers, highlight any incoming email in orange’. The clever part here is that the top 10 list changes dynamically over time as different customers do different amounts of business, but the query definition stays the same and delivers automatically updated results. Suddenly, we were presented with the possibility of building highly complex sets of queries based on real business needs, then to build rules engines and triggers based upon their result sets. Best of all, it would all be self managing and self defining.
 |
| So no more 'WinFS Stores' in our filing system. Has Microsoft walked away from this for good? |
»
Copyright © 2008 Dennis Publishing
This article appeared in the November, 2006 issue of PC Authority.
Thoughts on this article? Add a comment below.
Be the first to comment on this article.