Slashdot Mirror
Microsoft Research Turns 10
Alec Muzzy writes: "Did you know that Microsoft Research, the first research laboratory started by a software company, just turned 10 years old? Their website is currently featuring some highlights of their research in the past 10 years and how it is applying to the new products Microsoft is making today - for instance their work in Real-Time Fur will be used in some XBox games, and Speech Recognition may be in future Pocket PC's. Reading these pages gives you a real insight into what new technologies Microsoft is working on."
I thought Apple was much older than 10 years...
;)
--T
http://www.theMediaBunker.com
In the late 1990s the college of engineering at my university would have reps and engineers from Apple's ATG (Advanced Technology Group) visit to judge projects, talk about the industry, and share stories over a BBQ. From what I understand, the ATG was a research group that had free reign to experiment with software and hardware projects, some of which were eventually wrapped into shipping Apple products. In about 1994 I remember a demo/presentation that included some neat webserver CGIs for "intelligent" searching and document organzation (cool for that time period). We were also shown a cool speech recognition + text-to-speech utility that utilized facial recognition as well as displaying a spooky relaistic animated talking face. I also recall a semi-working mockup of a 3D version of the Macintosh Finder (Apple's Macintosh desktop / file manager). One of the coolest things I remember was that not all of their projects were on Apple Macintosh hardware. Most were, but a few were on IBM RS/6000 (AIX) and SGI Indigo (IRIX) workstations.
Cool stuff.
Let us not forgive or forget that. :-)
The sad thing is Microsoft has spent a pretty penny on research, but because of Microsoft's internal structure and development philosophy, the research doesn't get to do more than provide a gimick or two. E.g., Microsoft research spent a lot of time and money to develop a technique using Baysean probability to analyze what a user was doing and figure out what they were trying to do. The end result of that was the mother-#$! Office Paperclip that popped up whenever you typed the words, "Dear John".
Microsoft Research should be figuring out how to improve the performance of NT's Microkernel architecture, improve virtual memory management on multi-media machines and a host of other useful technologies. But they don't. Go figure.
The bitter lessons of a veteran coder: http://bitterprogrammer.blogspot.com
WTF? Real Time Fur?? How about Real Time Stability?
Now as much as everybody would like to deny that Microsoft has come up with anything new and original, you have to give them this:
All of their research on the blue screen of death has paid off. And they obviously know how to allocate their resources, devoting the most effort to the feature that gets seen most often.
I didn't know about this site. But it looks really very interesting, and something that is even more interesting is that they have special licensees for universities and other non profit research institutes to have access to the source code for NT and CE. This is some of the most interesting things I've seen.
Here is a link.
Will work for bandwidth
"Hey, look what those guys are doing!"
A feeling of having made the same mistake before: Deja Foobar
between researching security and implementing it. Microsoft has lots of smart people working for it, at least some of whom I'm sure understand completely the security implications of what they do. They just purposely decide not to do anything about it.
Which brings up the next point in that there is often a difference between doing what's "Right" and doing what's profitable. Easy is what sells to most folks. Secure is not. (talking generalities here...) And making things secure often makes them dramatically less easy. Since the primary purpose of Microsoft is to make money, easy will always win out over secure in their world. Good, bad, or indifferrent, that's the way it is. Follow the money trail and you'll understand why MS acts the way they do.
Limux has the opposite approach. Generally in the *nix world, performance (including stability, speed and options) usually wins out over outright ease of use. That's what the users of it demand. Certainly some things are very easy, but in many cases it's a different kind of easy for a different kind of audience. Whether that is good or not is an excercise left to the reader. (i.e. you)
Digital was not, strictly speaking, a "software company" but had a major research lab a long time ago.
Same for IBM.
CCA (Computer Corporation of America), creator of the venerable Model 204 database system, had an excellent research group. The did some of the classical database research in the 70s and 80s. (In fact, Phil Bernstein, who did this work while at Harvard U. and CCA, is now at Microsoft although not in research, I believe).
In 2006 or so, someone is going to submit to Slashdot about the 10th anniversary of Microsoft inventing the browser.
Microsoft Research Turns 10
Military Intelligence, Honest Politicians, Professional Wrestling, and Sexy Geek are also celebrating birthdays later this month.
I think despite what most people here on /. think about Microsoft as a company, :) you have to admit a lot of their research has created some very good ideas.
For example, the Microsoft Natural Keyboard and the improvements in the design of the Microsoft Mouse came out of this group. And MS Research has done a lot to dramatically improve the look and feel of Windows, especially the placement of menus, icons, etc.
I believe that the Linux supporters and developers should seriously look at creating an Open Source equivalent of Microsoft Research (companies like Dell and IBM could provide the initial seed money for such a lab). Imagine tightly-controlled research that could result in dramatic improvements in the usability of Linux on both the graphical and command line level, and developing keyboards and mouse pointers geared towards the needs of Linux users.
You're an idiot - LOTS of people came up with fur algorithms before R&H. E.g. Jim Kajiya who WORKS at MSResearch and wrote one of the seminal papers about it:
Kajiya, James T. and Timothy L. Kay, ``Rendering fur with three dimensional textures,'' in Proceedings of SIGGRAPH 1989, ACM SIGGRAPH, 1989, pp. 271-280.
Have a clue about the topic before you post. Just look at the list of staff at MSR - it's a 'who's who' of various fields........ CGI included...... Jim Blinn, Hughes Hoppe, Michael Cohen... etc etc etc........
Welcome to Slashdot - blind microsoft bashing. MSResearch is doing some damn good work - look at any set of the conference proceedings from SIGGRAPH for the past 5 years and see the published work.........
Most people who have worked on both research and real world development can tell you that there are always trade-offs to make between what works under limited conditions in a lab and what works in a production system with dozens of variables. Hypothetically, what if the Paperclip algorithm developed by the researchers actually were pretty smart at learning and predicting the user's behavior but would either eat up too much RAM take up too much time do perform their predictions?
What would you do if you were a PM for Office? Scrap the research opr pare it down to where it works in a reasonable amount of time and uses a reasonable amount of resources but isn't as clever asd you'd like? Real managers and real developers make decisions like this everyday.
Microsoft Research should be figuring out how to improve the performance of NT's Microkernel architecture, improve virtual memory management on multi-media machines and a host of other useful technologies. But they don't. Go figure.
I just looked at the MS Research page which lists the current research areas and noticed the following
- Research into Systems and Networking
- Research into Systems and Performance
- Research into Advanced Compiler Technology
- Research into Collaborative and Multimedia Technologies
- Research into Security
These are just the ones that address your immediate questions. There are several dozen more cool and worthwhile research areas at MS Research. Of course, being a typical slashdotter it is easier for you to bash them unthinkingly than do an ounce of research.PS: For those who think Microsoft isn't interested in the work done by MSR, when I was at a presentation at BillG's house this summer he kept on going on and on about the interesting projects being worked on at MSR and about how of all of MSFT that is probably one place where he is familiar with all the projects being worked on.
They made people rediculous offers to lure them away from their universities and other companies.
One example recruitment I heard went like this.
ring ring
Hello?
Hello, this is Microsoft Research, we'd like you to come work with us
Why should I? I'd never work for the great Satan. [thinking that this would make the caller hang up. But, what would Satan say? You got it...]
Well, what are your terms?
Ummm [trying to think of something completely unreasonable] How about $XXX.XXX [twice what he was getting then.]
Fine.
Ok, I want to work three months, then take a month off, work three months, take a month off...
We can't do that. How about this, you work for four years, then you get four years off at that same rate.
uhhhhhhh, well, ok.
When they set up the CG research group, they promised to have half the papers in Siggraph (the premier forum for computer graphics research) in a few years. This was a little scary, but not as scary as what really happened. What really happened is that these people pretty much stopped publishing at all; and stopped interacting with the rest of the graphics community.
I asked a few of the people there about it, and they seemed happy as clams, they weren't worried about it. To me, it appears that their world had shrunk to be just Microsoft. It's more than a pity, it's almost criminal.
thad
I love Mondays. On a Monday, anything is possible.