Microsoft Research Projects Showcased

prostoalex writes "Seattle Times reporter visited the Microsoft Research expo hosted by the company. The inventions of the future include a robot that could attend conferences in your behalf and allow you to communicate via video and audio applications, a software package that translates the sign language into readable English, e-mailable identification documents and some enhancements to Microsoft's operating systems."

Comments

Re:Microsoft Research or Ripoff?

[EnlightenedDuck]

Recognizing letters from ASL also isn't that useful - it would be the equivelent of taking dictation one letter at a time, albeit a bit faster for proficient signers. It is the first step, however, of recognizing any ASL sign, which would allow dictation by deaf people. However, to do that you'd need to know where the glove is in space, not just its configuration with respect to itself, which would require more sophisticated hardware.

Also, imagine the possiblities if you could teach in your own signs, as a method of interacting with the program of your choice.....

MS style innovation

[Jade E. 2]

The idea of having a robot attend meetings for you (or touring and inspecting factories, repairing computers, or any other task that normally requires physical presence but doesn't justify travel) is called telepresence, and people have been working on it for years and years. For example, here's a rather sophisticated 386-based telepresence robot from 1995. That happens to be the very first hit out of 39,100 if you look for telepresence at Google.

(This comment has been stripped of it's MS-bashing nature, because really, if you don't like them you don't need me to explicitly point out that they're reinventing the wheel, and if you do like them you'll ignore it anyways...)

Re:Yes, thank you Microsoft Research

[josh crawley]

But what about clippy! that's a big innovation! it's so hard to think of keywords like "margins" and search for them, I like to type in "how do I change the margins?" instead. It's so much quicker!

Clippy is definitely not for geeks. However, there is a large segment population that wouldn't know what the help menu was if it bit their ass, and who also don't look at things in terms of "input keyword - get results". They think in terms of "ask a question, get an answer." Also, bear in mind that clippy was a combination of two pieces; online help with "natural language" search, as well as a bayesian reasoning piece (the whole "It looks like you're writing a letter..." bit). The suggestion piece also doesn't go very far with geeks, since they generally know (or think they know) what they're doing, whereas that other segment of the population welcomes the help in many cases.

Also, bear in mind that as annoying as Clippy and the pop-up bits are, there are still some people who just can't grasp the concept of "asking Clippy" before they go elsewhere. I'd imagine because it still isn't "personable" enough to engage these novice users in the same way a helpful person would.

There's something about the paradigm of text on a screen, and the psychological experience of using a computer that just hasn't been understood yet in interface design. It's something more than a tool-using experience, but less than a "person to person" experience. Hence, the whole argument about "the only intuitive interface is the nipple".

MS stability not that far from Linux stability

[0x0d0a]

MS stability isn't all that far from Linux stability. I'd pretty safe-feeling with both the NT kernel and the Linux kernel. GNOME software and Explorer -- *application software* both have instabilities.

Granted, so much crap is tied into Explorer that Explorer dying is generally worse than the GNOME panel crashing, but if you compare each chunk to its Linux equivalent, it's not *that* far away.

If MS hadn't made a couple of totally stupid moves, tying functionality into Explorer instead of doing it the right way, in the kernel, Explorer crashing away wouldn't be such a big deal (Explorer simulates symlinks, Explorer works around stupid MS file-locking semantics in XP, Explorer provides the high-level widgets for many other applications...)

Re:Where do ideas come from?

[Bert690]

Apart from the university setting, who else is out there?

Not many others. IBM Research is still going strong, but it's generally more focused on shorter term research goals than PARC/Bell Labs. But that's probably why it hasn't gone the way of PARC & Bell Labs...

Re:MS style innovation....

[josh crawley]

What you say is true - but only in theory - as it requires user intervention to support it. Do you really think your average secretary is going to bother pre-ordering each lift he wants to take?

Umm, you pre-order the elevator already, when you press the button in the waiting area. The only problem is that you're only telling the micro-processor that controls it "up or down". If I'm on the 5th, and going to the 10th, and there's an elevator on the 4th floor which already has 5 people in it who are going to the 10th, and 2 others for the 15th, it would make sense for that elevator to stop, rather than another elevator at the 4th which has 3 people destined for the 11th and 12th. Right now your average elevator just says "people who want to go up should get on elevators that are already going up, and vice versa". Now, we could have the same capability by just having the floor number buttons in the elevator waiting atrium, but the cell-phone capability has two potentials:

1. Since cell phones are nigh-ubiquitous, it replaces any unwieldy "50 buttons" interface needing to be in place in the building, and if adopted across the board, becomes an intuitive act for the user (i'm walking toward the elevator, pull out my cellphone...*beep* bam there's my elevator)
2. The location of the cellphone could be tracked within the elevator; therefore the scheduling doesn't get confused by someone getting off before their floor, or some prankster dialing up 50 random floors in a row