Wait, what? You can bet that Paul Allen's crew will edit out these segments... um. This article was about Vulcan Ventures (Paul Allen) selling TechTV. Those things you loved about the show were while Paul Allen could have made those edits. It now goes out of his hands. Sheesh.
The group I work for is concerned with this issue (creating useful architectures from basic, hard to manipulate components). It is a compilers and computer architecture research group with a nanotech habit.
The solution the group is working on is that rather than try to synthesize complex chips, as we do now, it is to make large regular reconfigurable structures. No DNA, just straight chemical self-assembly. There is a proposed design for this that we are working with companies like HP and Mitre on. The key technology bits to make it work from our side is a good system for generating circuits from code, and very fault-tolerant place and route (the assumption is 10% structural errors).
Spoken like someone who has never had an accident. It eventually happens, and it changes your attitude. No one, not the finest rally drivers, Possum Bourne for example, can always perform the proper maneuver at the appropriate time in a given situation.
My father learned this -- pristine driving record for over thirty years -- when he recently had his first accident. it wasn't bad, just bad weather and some significant damage to the new truck by way of guardrail. Prior to it, he had built up the confidence that he was a good enough driver to avoid such a fate.
That's not to say that skill plays no role, nor that one should not be responsible for their actions while behind a wheel. But to say that an accident is definitive evidence one shouldn't be on the road is ridiculous.
I suspect they will. The pricing differences are not very substantial (if you actually look at the plans) and they can't win in a direct price-war. They have loyalty -- both consumer and channel (Best Buy will not pitch Wal*Mart); they've already dealt with the growing pains that Wal*Mart is encountering; and they've got a better system. What I haven't seen anyone mention is their recommendation system, which has probably millions of user reviews. The system has consistently gotten better -- and I trust the Netflix userbase far more than WalMart's for good movie selections. The first couple of months of using NetFlix, you go solo on choosing what you want to see... but after a while, (at least for me) the queue has several recommendations I had never heard of but sound good.
It was frustrating when they didn't hit profitability at the Million User mark. I guess it took another hundred thousand. But I think the profit status will remain sustainable (growth, however...)
Along the same point, and illustrating another important principle of the parent post, is distribution. Sure, it may only cost pennies to make carbonated sugar water... but it wouldn't cost me pennies if I want carbonated sugar water (now!) nor would it be generally feasible (here!). Marketing is only part of the Way.
I'm not sure that that is incorrect. There are only 4 timeslices, and a total of 12 slots, with repetition. The (unfortunate) bit is that, at least in the commercial world, the giant computer corporations of the past 50 years are and were US-based. It is these places who fund expensive research labs -- as someone pointed out, Microsoft is up there because they pump incredible budgets at results-free work. I believe Microsoft Research is the largest academic CS environment in the world.
And that the two universities listed are US-based says nothing of the international demographics of the programs. Generally CMU and MIT attract the best of the best in the world (certainly, CMU has more graduate students from IIT than MIT&CMU).
Can you think of who a competitor might be? Perhaps some of the Japanese megacorps... but I don't think Sony has a dedicated HCI research lab. And their HCI impact is still arguably less significant (good or bad) than PARC (windows), Apple or Microsoft.
That said, I prefer the direction ion is going: graphical usable interfaces and whatnot. Overlapping window jockeying sucks.
without a doubt in the top 5 computer education schools in the world.
What exactly does this mean? Without a doubt for a temporal, relative (and on the whole) meaningless value is a rather silly position to hold. Top 5? On what criteria? By whose metrics?
Absolutes and Relatives don't mix.
Within robotics, you have CMU's Nomad (first robot with an autonomous scientific discovery... antarctica, atacama desert)
You also had a popular robotic platform known generally as Nomads (from Nomadic Technology). Those are gone now, though they live on in courses and labs.
you must remember that CS stands for Computer Science, that is IT IS A SCIENCE. Much of what computer science is is to guide you towards a research position.
As a chemistry and a computer geek, I'm going to have to call you on this one. CS, as practiced, is not really a science. It lacks the formality and discipline of just about any field except the social sciences... where it tends to put on less of a show. That academic CS pushes you towards research -- that's certainly true.
research != science
In its academic form, CS resembles academic engineering research or mathematics. Mathematics is also refered to as Mathematical Sciences, but that is a similar stretch. Logic != Science.
Which says nothing of the relative merits of any of it. Computer Science just kind of bugs me for how most think of it (and equate it with computer programming and/or software engineering).
The beauty of this system for Microsoft is that it did not have to spend anything to grant the options, but gained $2.25 billion (shown as "common stock issued" in the cash flow) from what the employees have to pay to exercise the options.
That's not entirely true. Technically, they are diluting ownership by share creation, so that stock options are not just printing money. However, these are the financial games that can be played with a reasonably strong stock -- shareholders who are being diluted don't care much as long as the stock continues to rise. And Microsoft has been hit hard when their stock goes below the strike price of options -- they've had to take various penalties to reduce the strike price.
So yes, its a nice way to make the numbers look better; and yes, its a pretty cheap money option, and keeps the company from paying taxes, etc -- but its not bad or free money. Its just another thing that backfires when/if these companies plateau or start their downward spiral. Especially Microsoft, with their 4.5 year vesting plans and the great extend to which they are overextended with regards to the number of options versus the amount of ownership held by the company.
Many institutional investors are also playing with the money from stable, real-long-term investors like the endowments of academic institutions, foundations, and non-profits. And those funds tend to accumulate, not deteriorate (unless the endowed institute isn't doing so hot).
I suppose that most of that money isn't in volitile tech stocks... but at least some percentage is, and that still represents pretty big figures.
Depends on what you considered a mass of ugly colors. Basically, it is an array of leds turning off/on really quickly. But its array'd like a font, where the x-axis is time.
So, lets say you got
red
blue
red
red
yellow
blue
green
You'd have a red line, a blue line, a thick red... So, if you really wanted to, you could make something like the apple logo.
To do a tiedye, you'd have to do something slightly more interesting electronically. A couple of ideas:
1. array of each color, in an assortment:
red,blue,green
red,blue,green
...
And then do the color changing in software (tricky to get it sync'd)
2. like an electric typewriter, you'd have a ring of lights, and rotate to the appropriate color. Tricky mechanically, high in geek skill value
3. You could at least use tricolor leds
4. array and then a focal lens, so you could turn any particular one on and the "spot" would be that color.
5. blended colors could look weird if you oscillated the on-off phase between the two.
I don't quite see what you mean about the list favoring comfort over communication. Certainly, there is a glut of convenience elements to the list, but they have "radio & television", "telephone", "internet", and "fiber optics". To a degree, it could be argued that "automobile", "highways", and "Airplane" could be lumped into the communications category. What communications did they miss?
I've had a lot of success with the MX-2900. Its a pretty nice camera in the mid-high end (that is, its not your standard party camera). 2.3 MP, 1800x1200; its not got the best of user interfaces, but that's for the non-geeks. For the geeks, its easy to use and has way too many features. I like manual control of all aspects of my 35mm camera, and this one has it.
It uses SmartMedia, which is cheap and cute. There is some basic support under linux for the camera, and the software that comes with it is lame. However, given that the images are typically 300 or 900+k, transfering over serial sucks. Get a smartmedia PC card adapter (assuming a laptop) and you can transfer at 20Mbits/sec, which is much nicer. And linux supports them very well - you just mount the card as a harddrive.
If you want a geek's camera, with gorgeous resolution and images, and like to play with the best... and are willing to spend the money (though cheaper than its comparable units), check out the 2900.
Hmmm. I've been a longtime Be watcher, but I haven't installed the latest Be (4.5). From the FAQ, however:
Q: Will I have to repartition my hard drive to install BeOS 5? A: No. If you're currently using Windows, you'll be able to download BeOS 5 via a Web browser and store it as a file within the Windows file system. Downloading BeOS 5 will be no different than downloading any application or utility. No repartitioning will be necessary, and launching BeOS 5 will be as simple as double-clicking an icon on the desktop. If you decide to delete BeOS 5 -- though we hope you don't -- doing so will be as simple as uninstalling any typical Windows application.
True? Required? Part of the mystique of Be has always been how system efficient it is, and how it is completely new (not backwards compatable) to take advantage of the latest in OS research, not more of the same. I've always been impressed with their 64-bit journalling filesystem. Is this abandoned? Is this still an option?
I've been waiting for them to do encrypted messaging.
Good encrypted messaging takes a lot more protocol work than just slapping it on a present protocol. Check out Caliban for a protocol and program which has been working on this.
well, see, you didn't read cryptonomicon or the review, because this is very relevant for people reeling from the book. half of the book was historical fiction, taking place around Blechley Park and projects ULTRA and MEGA and the Engima cracking.
So for people without a whole lot of historical cryptographic background, anything that discusses this history is seen as relating to the book.
Slashdot Mirror
Imagine a compiler that doesn't just compile code but tapes out the coprocessor need to run it.
Or at least the verilog to reconfigure the coprocessor or build the ASIC? Yeah, I work for one of those research groups.
http://www.cs.cmu.edu/~phoenix/
Wait, what? You can bet that Paul Allen's crew will edit out these segments... um. This article was about Vulcan Ventures (Paul Allen) selling TechTV. Those things you loved about the show were while Paul Allen could have made those edits. It now goes out of his hands. Sheesh.
The group I work for is concerned with this issue (creating useful architectures from basic, hard to manipulate components). It is a compilers and computer architecture research group with a nanotech habit.
The solution the group is working on is that rather than try to synthesize complex chips, as we do now, it is to make large regular reconfigurable structures. No DNA, just straight chemical self-assembly. There is a proposed design for this that we are working with companies like HP and Mitre on. The key technology bits to make it work from our side is a good system for generating circuits from code, and very fault-tolerant place and route (the assumption is 10% structural errors).
Spoken like someone who has never had an accident. It eventually happens, and it changes your attitude. No one, not the finest rally drivers, Possum Bourne for example, can always perform the proper maneuver at the appropriate time in a given situation.
My father learned this -- pristine driving record for over thirty years -- when he recently had his first accident. it wasn't bad, just bad weather and some significant damage to the new truck by way of guardrail. Prior to it, he had built up the confidence that he was a good enough driver to avoid such a fate.
That's not to say that skill plays no role, nor that one should not be responsible for their actions while behind a wheel. But to say that an accident is definitive evidence one shouldn't be on the road is ridiculous.
It was released around then.
I suspect they will. The pricing differences are not very substantial (if you actually look at the plans) and they can't win in a direct price-war. They have loyalty -- both consumer and channel (Best Buy will not pitch Wal*Mart); they've already dealt with the growing pains that Wal*Mart is encountering; and they've got a better system. What I haven't seen anyone mention is their recommendation system, which has probably millions of user reviews. The system has consistently gotten better -- and I trust the Netflix userbase far more than WalMart's for good movie selections. The first couple of months of using NetFlix, you go solo on choosing what you want to see... but after a while, (at least for me) the queue has several recommendations I had never heard of but sound good.
It was frustrating when they didn't hit profitability at the Million User mark. I guess it took another hundred thousand. But I think the profit status will remain sustainable (growth, however...)
Looks like they are profitable (GAAP) in this latest quarter. That's the consensus, at least. Even with expensing stock options.
Along the same point, and illustrating another important principle of the parent post, is distribution. Sure, it may only cost pennies to make carbonated sugar water... but it wouldn't cost me pennies if I want carbonated sugar water (now!) nor would it be generally feasible (here!). Marketing is only part of the Way.
Or a different scenerio, out of /. style:
Buffet wins Microsoft. WTF would happen then?
Gas Molecules? Galium Sulfur? If we are being accurate, naturally, it is molecules in a gas phase.
Hmmm. To make this post mean something, here is a nice clear explaination for students.
And for a bit more than kids will digest on rainbows. Solar position is important.
Ellison does have a MiG or two...
I'm not sure that that is incorrect. There are only 4 timeslices, and a total of 12 slots, with repetition. The (unfortunate) bit is that, at least in the commercial world, the giant computer corporations of the past 50 years are and were US-based. It is these places who fund expensive research labs -- as someone pointed out, Microsoft is up there because they pump incredible budgets at results-free work. I believe Microsoft Research is the largest academic CS environment in the world.
And that the two universities listed are US-based says nothing of the international demographics of the programs. Generally CMU and MIT attract the best of the best in the world (certainly, CMU has more graduate students from IIT than MIT&CMU).
Can you think of who a competitor might be? Perhaps some of the Japanese megacorps... but I don't think Sony has a dedicated HCI research lab. And their HCI impact is still arguably less significant (good or bad) than PARC (windows), Apple or Microsoft.
That said, I prefer the direction ion is going: graphical usable interfaces and whatnot. Overlapping window jockeying sucks.
Within robotics, you have CMU's Nomad (first robot with an autonomous scientific discovery... antarctica, atacama desert)
You also had a popular robotic platform known generally as Nomads (from Nomadic Technology). Those are gone now, though they live on in courses and labs.
What computers in Star Wars.
And I thought the reference was for Tron,
around the same time, which _was_ denied
oscars for FX because computers were considered
cheating...
you must remember that CS stands for Computer Science, that is IT IS A SCIENCE. Much of what computer science is is to guide you towards a research position.
As a chemistry and a computer geek, I'm going to have to call you on this one. CS, as practiced, is not really a science. It lacks the formality and discipline of just about any field except the social sciences... where it tends to put on less of a show. That academic CS pushes you towards research -- that's certainly true.
research != science
In its academic form, CS resembles academic engineering research or mathematics. Mathematics is also refered to as Mathematical Sciences, but that is a similar stretch. Logic != Science.
Which says nothing of the relative merits of any of it. Computer Science just kind of bugs me for how most think of it (and equate it with computer programming and/or software engineering).
The beauty of this system for Microsoft is that it did not have to spend anything to grant the options, but gained $2.25 billion (shown as "common stock issued" in the cash flow) from what the employees have to pay to exercise the options.
That's not entirely true. Technically, they are diluting ownership by share creation, so that stock options are not just printing money. However, these are the financial games that can be played with a reasonably strong stock -- shareholders who are being diluted don't care much as long as the stock continues to rise. And Microsoft has been hit hard when their stock goes below the strike price of options -- they've had to take various penalties to reduce the strike price.
So yes, its a nice way to make the numbers look better; and yes, its a pretty cheap money option, and keeps the company from paying taxes, etc -- but its not bad or free money. Its just another thing that backfires when/if these companies plateau or start their downward spiral. Especially Microsoft, with their 4.5 year vesting plans and the great extend to which they are overextended with regards to the number of options versus the amount of ownership held by the company.
Many institutional investors are also playing with the money from stable, real-long-term investors like the endowments of academic institutions, foundations, and non-profits. And those funds tend to accumulate, not deteriorate (unless the endowed institute isn't doing so hot).
I suppose that most of that money isn't in volitile tech stocks... but at least some percentage is, and that still represents pretty big figures.
Depends on what you considered a mass of ugly colors. Basically, it is an array of leds turning off/on really quickly. But its array'd like a font, where the x-axis is time.
So, lets say you got
red
blue
red
red
yellow
blue
green
You'd have a red line, a blue line, a thick red... So, if you really wanted to, you could make something like the apple logo.
To do a tiedye, you'd have to do something slightly more interesting electronically. A couple of ideas:
1. array of each color, in an assortment:
red,blue,green
red,blue,green
...
And then do the color changing in software (tricky to get it sync'd)
2. like an electric typewriter, you'd have a ring of lights, and rotate to the appropriate color. Tricky mechanically, high in geek skill value
3. You could at least use tricolor leds
4. array and then a focal lens, so you could turn any particular one on and the "spot" would be that color.
5. blended colors could look weird if you oscillated the on-off phase between the two.
On and on
With people with a track record of getting things done (Red Whittaker, etc).
Project 339: Solar Blade Solar Sail
Their homepage
I don't quite see what you mean about the list favoring comfort over communication. Certainly, there is a glut of convenience elements to the list, but they have "radio & television", "telephone", "internet", and "fiber optics". To a degree, it could be argued that "automobile", "highways", and "Airplane" could be lumped into the communications category. What communications did they miss?
I've had a lot of success with the MX-2900. Its a pretty nice camera in the mid-high end (that is, its not your standard party camera). 2.3 MP,
1800x1200; its not got the best of user interfaces, but that's for the non-geeks. For the geeks, its easy to use and has way too many features. I like manual control of all aspects of my 35mm camera, and this one has it.
It uses SmartMedia, which is cheap and cute. There is some basic support under linux for the camera, and the software that comes with it is lame. However, given that the images are typically 300 or 900+k, transfering over serial sucks. Get a smartmedia PC card adapter (assuming a laptop) and you can transfer at 20Mbits/sec, which is much nicer. And linux supports them very well - you just mount the card as a harddrive.
If you want a geek's camera, with gorgeous resolution and images, and like to play with the best... and are willing to spend the money (though cheaper than its comparable units), check out the 2900.
Hmmm. I've been a longtime Be watcher, but I haven't installed the latest Be (4.5). From the FAQ, however:
Q: Will I have to repartition my hard drive to install BeOS 5?
True? Required? Part of the mystique of Be has always been how system efficient it is, and how it is completely new (not backwards compatable) to take advantage of the latest in OS research, not more of the same. I've always been impressed with their 64-bit journalling filesystem. Is this abandoned? Is this still an option?A: No. If you're currently using Windows, you'll be able to download BeOS 5 via a Web browser and store it as a file within the Windows file system. Downloading BeOS 5 will be no different than downloading any application or utility. No repartitioning will be necessary, and launching BeOS 5 will be as simple as double-clicking an icon on the desktop. If you decide to delete BeOS 5 -- though we hope you don't -- doing so will be as simple as uninstalling any typical Windows application.
I've been waiting for them to do encrypted messaging.
Good encrypted messaging takes a lot more protocol work than just slapping it on a present protocol. Check out
Caliban for a protocol and program which has been working on this.
well, see, you didn't read cryptonomicon or the review, because this is very relevant for people reeling from the book. half of the book was historical fiction, taking place around Blechley Park and projects ULTRA and MEGA and the Engima cracking.
So for people without a whole lot of historical cryptographic background, anything that discusses this history is seen as relating to the book.