This issue has come up a lot recently at MIT. More and more research here is being sponsored by corporations instead of the government or university. Additionally, the entrepreneurial and engineering spirit bred here tends to encourage research into practical applications and not just research for research's sake.
Here's a story about potential conflict's of interests for professors. It involves Prof. Leighton, an MIT prof who's research led to Akamai.
The issue of how patents are handled at universities is also an issue at MIT. Currently, as I understand it, MIT's patent department assists students in obtaining patents in return for a percentage of any income derived from those patents. A view on this was expressed by Prof. Bose (yes, that Bose =) in a guest column in the school newspaper.
A recent partnership with Microsoft (!)brought up more discussions about MIT's role and how the many partnerships with industry affect MIT's goals. (I can't seem to find the articles that popped up about it, though:(
Here's an old one by "Drexel University professor Noble, formerly an assistant professor in MIT's Program in Science, Technology and Society".
IMHO, I don't think industry ties are entirely bad. Like anything else in the world, there are pros and cons. Obviously, as these ties become closer, ethical issues arise, and there isn't a clear resolution to this problem. Unless you make the professors and students choose between academia and industry. The Media Lab is an example of how industry and universities can dovetail successfully. AFAIK, corporations provide funding to the Media Lab as a whole. Then, if they see anything they like, they can take it in house and develop it to their own desires. As a result, the Media Lab receives a huge influx of funding, and is thus able to research things that may or may not have an obvious commercial value. But cool things come out of there all the time. (i.e. Lego Mindstorm was developed there).
On a personal note, I think if Apple had only put a couple of more buttons on and a scroll wheel, I probably would have forsaken my Intellimouse Explorer to get one. While the functionality is still there with a one-button mouse (you have always been able to use keyboard modifiers: Ctrl-, Shift-, Command-, Option-), it's not as convenient. Plus that scroll wheel, and the ability to map macros to buttons is a huge plus in the Intellimouse.
<wishful thinking>When's there gonna be an apple-branded multi-button mouse?</wishful thinking>
We're getting away from the main discussion... =) But let's assume Linux has had a 4% market share every year since 1993. Even that would only yield 28 million users. And, it's obviously false that Linux has not held that market share since it's inception.
The problem here is drawing conclusions from very few data points.
it's kind of absurd to use those low numbers as evidence of consistent doubling. If I made a penny today, and find a quarter tomorrow, should I assume I'm going to be increasing my profits everyday by 25 times? It's just wrong statistically to draw the conclusions from the "facts" presented. Especially since the article links are from all over the place chronologically. Having 28 Million in May 98 and only 7 million in Oct 98 doesn't seem consistent. Look at 1993-1994. Going from 100,000 to 500,000 users. Should we extrapolate that Linux is growing 500% every year? or even 400% from the 1993-1995 range? When you're dealing with low numbers it doesn't take much to make big percentage gains. Absolute numbers give a bit clearer picture here.
Those 1993-1996 data points could also be used to show linear growth of 1 million users per year. Obviously, Linux usage has taken off dramatically with all the recent press coverage, but I still highly doubt that 50 million figure.
hmmm... interesting math =) Any reason for deducting 2 million from the low end and 4 million from the high end of the range...?
Anyways, I highly doubt that upper figure. Especially considering this recent report by IDC at MacCentral about Linux overtaking the server market....current numbers that show Windows having 87 percent share of a 98.5 million shipment market...and Linux paid copies having just under four percent.
Even if you were to say every Windows user in the "shipment market" had Linux on their machine, you still wouldn't hit that high number. And the low number indicates a 50% market share, also unlikely.
I can't see why a company wouldn't want to port their product to an untapped market. The cost of MacOS is under $100, free with a Mac, which can cost much less than a UNIX box.
If the major stumbling block is reconnecting the GUI, I for one would prefer the excessively-candied Aqua over X anyday. (Something about X's aesthetics just bothers me... =)
MacOS X definitely represents an opportunity for UNIX vendors to gain access to an even larger base of users. And this could work both ways, with programs developed for the Mac being ported to UNIX platforms.
At Siggraph, there have already been some statements (albeit lukewarm) about porting high-end graphics applications to OS X.
Apple has put a ton of stuff up: press releases, promotional material, pics, and QTVR movies. Apple usually uses Akamai (check the source code on Apple's sites and search for "Akamai") to effectively do global load balancing, so the slowdown is probably more due to Apple's web team than the traffic they're getting. Granted, Akamai's not magic, but it helps a lot. Especially, when you've got a ton of people going after the eye candy =)
"it costs these people the $125k they'd be making at work"
The $125K is called "opportunity cost." It's the same thing you're doing by going to college, or entering a graduate program: giving up the chance to earn wages in the expectation that you'll be able to recover those loses later with your increased earning ability.
But that's not the issue here. As philg mentioned in his post, and in his 6.916 lectures, (paraphrasing here) people should be using the full potential of the web to provide useful services. He's looking to train people who might make a difference.
Also, isn't the point of learning to take advantage of the time others have spent? Think about it: when someone teaches you something, you're benefitting from all of the time they've sacrificed to gain that valuable experience/knowledge.
"...being in school 12 hours a day..."
That's what going to MIT is like - a ridiculously intense regimen of work, but we still find plenty of time to play.;o)
It's Marathon Day!
Re:Fit four years into one? Thats going to be toug
on
ArsDigita University
·
· Score: 1
Differential and Integral calculus are covered in the most basic math course (18.01) at MIT. Most students have finished Multivariable (18.02) and Differential Equations (18.03) by their freshmen year. Both 18.01 and 18.02 are "General Institute Requirements" (necessary for graduation regardless of your major).
I don't think the ArsDigita course intends to go past the material present in 18.01. Fitting single-variable calculus into a month isn't as bad as you might think. Remember, they're going after the best and brightest grads, people who have learned how to learn.
Slashdot Mirror
Here's a commercial use of lisp for ya. If you've ever searched for airfares on Orbitz, Kayak, etc, you've been using a Lisp-based engine.
I found this article to be a better overview of the whole story.
Here's a story about potential conflict's of interests for professors. It involves Prof. Leighton, an MIT prof who's research led to Akamai.
The issue of how patents are handled at universities is also an issue at MIT. Currently, as I understand it, MIT's patent department assists students in obtaining patents in return for a percentage of any income derived from those patents. A view on this was expressed by Prof. Bose (yes, that Bose =) in a guest column in the school newspaper.
A recent partnership with Microsoft (!)brought up more discussions about MIT's role and how the many partnerships with industry affect MIT's goals. (I can't seem to find the articles that popped up about it, though :(
Here's an old one by "Drexel University professor Noble, formerly an assistant professor in MIT's Program in Science, Technology and Society".
IMHO, I don't think industry ties are entirely bad. Like anything else in the world, there are pros and cons. Obviously, as these ties become closer, ethical issues arise, and there isn't a clear resolution to this problem. Unless you make the professors and students choose between academia and industry. The Media Lab is an example of how industry and universities can dovetail successfully. AFAIK, corporations provide funding to the Media Lab as a whole. Then, if they see anything they like, they can take it in house and develop it to their own desires. As a result, the Media Lab receives a huge influx of funding, and is thus able to research things that may or may not have an obvious commercial value. But cool things come out of there all the time. (i.e. Lego Mindstorm was developed there).
On a personal note, I think if Apple had only put a couple of more buttons on and a scroll wheel, I probably would have forsaken my Intellimouse Explorer to get one. While the functionality is still there with a one-button mouse (you have always been able to use keyboard modifiers: Ctrl-, Shift-, Command-, Option-), it's not as convenient. Plus that scroll wheel, and the ability to map macros to buttons is a huge plus in the Intellimouse.
<wishful thinking>When's there gonna be an apple-branded multi-button mouse?</wishful thinking>
yeah, but distinguishing the +/-10 meters/day in the rover's movement isn't really worth the unnecessary words in a press release.
But let's assume Linux has had a 4% market share every year since 1993. Even that would only yield 28 million users. And, it's obviously false that Linux has not held that market share since it's inception.
The problem here is drawing conclusions from very few data points.
it's kind of absurd to use those low numbers as evidence of consistent doubling. If I made a penny today, and find a quarter tomorrow, should I assume I'm going to be increasing my profits everyday by 25 times? It's just wrong statistically to draw the conclusions from the "facts" presented. Especially since the article links are from all over the place chronologically. Having 28 Million in May 98 and only 7 million in Oct 98 doesn't seem consistent. Look at 1993-1994. Going from 100,000 to 500,000 users. Should we extrapolate that Linux is growing 500% every year? or even 400% from the 1993-1995 range? When you're dealing with low numbers it doesn't take much to make big percentage gains. Absolute numbers give a bit clearer picture here.
Those 1993-1996 data points could also be used to show linear growth of 1 million users per year. Obviously, Linux usage has taken off dramatically with all the recent press coverage, but I still highly doubt that 50 million figure.
Anyways, I highly doubt that upper figure. Especially considering this recent report by IDC at MacCentral about Linux overtaking the server market. ...current numbers that show Windows having 87 percent share of a 98.5 million shipment market...and Linux paid copies having just under four percent.
Even if you were to say every Windows user in the "shipment market" had Linux on their machine, you still wouldn't hit that high number. And the low number indicates a 50% market share, also unlikely.
Here's an example showing tar and gzip in use.
If the major stumbling block is reconnecting the GUI, I for one would prefer the excessively-candied Aqua over X anyday. (Something about X's aesthetics just bothers me... =)
MacOS X definitely represents an opportunity for UNIX vendors to gain access to an even larger base of users. And this could work both ways, with programs developed for the Mac being ported to UNIX platforms.
At Siggraph, there have already been some statements (albeit lukewarm) about porting high-end graphics applications to OS X.
There's a pop up handle. Look at the quicktimeVR movies at http://apple.com/hardware/gallery/
Apple has put a ton of stuff up: press releases, promotional material, pics, and QTVR movies. Apple usually uses Akamai (check the source code on Apple's sites and search for "Akamai") to effectively do global load balancing, so the slowdown is probably more due to Apple's web team than the traffic they're getting. Granted, Akamai's not magic, but it helps a lot. Especially, when you've got a ton of people going after the eye candy =)
Check it out
MIT museum
But that's not the issue here. As philg mentioned in his post, and in his 6.916 lectures, (paraphrasing here) people should be using the full potential of the web to provide useful services. He's looking to train people who might make a difference.
Also, isn't the point of learning to take advantage of the time others have spent? Think about it: when someone teaches you something, you're benefitting from all of the time they've sacrificed to gain that valuable experience/knowledge.
That's what going to MIT is like - a ridiculously intense regimen of work, but we still find plenty of time to play.It's Marathon Day!
Differential and Integral calculus are covered in the most basic math course (18.01) at MIT. Most students have finished Multivariable (18.02) and Differential Equations (18.03) by their freshmen year. Both 18.01 and 18.02 are "General Institute Requirements" (necessary for graduation regardless of your major).
I don't think the ArsDigita course intends to go past the material present in 18.01. Fitting single-variable calculus into a month isn't as bad as you might think. Remember, they're going after the best and brightest grads, people who have learned how to learn.