Bored with computers
Hi, my name is Peter Spiro. I’m the General Manager in SQL Server responsible for the WinFS team and I’ve been bored with computers and software in general.
Now before I continue with that thread, some background info seems appropriate.
I was at DEC for 9 years working on database systems (Rdb and DBMS) before I joined Microsoft in 1994. I was hired into Microsoft to be a “change agent” in the database space. Microsoft had hired Dave Cutler and a bunch of folks from DEC West to make a big advance in the operating systems space. Afterwards they realized to truly be an enterprise player they also needed a bunch of systems technologies that layered on top of the operating system such as database systems, a tp monitor, language integration, data repository, etc. So in 1994 David Vaskevitch (CTO at Microsoft), enabled a series of senior level hires from other systems companies: IBM, DEC, HP, Tandem, etc. I was part of that wave of senior level systems guys. The interesting thing was at the time Microsoft didn’t really know how to interview/hire/deploy these types of senior engineers (but perhaps that’s a topic for another post). Anyway a bunch of us showed up, adjusted to being at a pure software and PC-focused company, and then got to work on building teams and products. In my case, this culminated in Microsoft SQL Server 7.0 and its subsequent releases. In the Spring of 2002, we started incubating the WinFS project within the database organization.
Ok, back to being bored with PCs and computers in general.
So there’s no question we’ve made amazing progress in the last 30-40 years. I started out with Fortran and punch cards in the 1970s. But the experience was so painful I stayed as far away from computers as possible. In the early 1980s I took another look as microcomputers started becoming more mainstream. I also got hooked into minicomputers from DEC. In my view, the 1980s and 1990s were wonderful periods for software advances; fantastic innovation and invention. Operating systems and databases systems stabilized and moved towards commoditization. A whole slew of user and business products such as spreadsheets, word processing software, mail systems, and discussion forums became ubiquitous. The Web happened and we had another boom with online commerce/information for businesses and individuals.
Just think about all the innovation from: IBM/Lotus, Ebay, Microsoft, Oracle, SAP, Amazon, Google, Adobe……
So what’s the problem?
Well what’s really new and different? We’ve had all these types of products in some shape or form for 5, 10, even 20 years in some instances. Where are the new types of applications we haven’t seen before? The Web did create a bunch of new types of products over the past 10 years but this seems to be stabilizing also. Amazon is over 10 years old now!
Which brings me back to WinFS and an illustrative anecdote:
A few years back Ray Ozzie came to Microsoft to speak at an internal workshop we have called WHiPS (Workshop on High Performance Systems). Ray talked about Groove. It was a pretty cool app trying to be a platform (tough to do for a small company). As we were talking after his presentation he described how he had to first build a platform on top of the operating system, which then enabled him to build the app that he had envisioned. He said he needed storage beyond just a file system; he needed synchronization capabilities to share information between different machines/users, he needed advanced security mechanisms, he needed a way to model his particular schemas in the storage system, he needed to search on the information. So he built a richer storage system that incorporated these types of concepts and then he built the Groove app on top of his platform. His comment to me was that Microsoft should build such a platform and stop selling the same old operating system and storage capabilities that we as an industry have been polishing for 20-30 years. He felt this richer platform would create a new ecosystem that allowed a variety of new apps to be developed.
He was right; and I told him we had already started on it, and that it was called WinFS.
Author: Peter Spiro