Dude, don't worry. It was clear what you meant. That doesn't mean people will not seize the opportunity to make a joke, though. Personally, I thought mine was pretty funny. See, a robot might prefer the places where he doesn't get treated like a... oh never mind.
What you're describing with the sticky pedal is a classic problem with control systems. Because friction acts more strongly on a stationary object than a moving one, you often need a big force to start something moving, which can cause you to push the system way too hard (this is called integrator windup, a reference is here).
There are definitely techniques to overcome this (the gist of it is, you tell the controller to increase the force slowly until you're out of the 'sticky' range). I agree that it is hard to make a machine "learn" this kind of behavior in advance, and so it is at a disadvantadge in that respect. On the other hand, the machine probably has a quicker response time than a person, and it generally has access to more precise inputs. For instance, the machine can use a sensor (I think they are using ultrasound in this project) to determine its precise distance from the other cars, whereas a human just has a general sense. So even if it gets itself into trouble by pushing the pedal to hard, it can hit the brake much more quickly. Of course, stamping the gas and then the brake is not necessarily a desirable way to park your car. The solution? Spray some WD-40 on the accelerator I guess.
Maybe the Pennsy cops have quotas, but I think you're being a little bold to speak about all police departments everywhere.
Also, speed limits save gas. That's why U.S. Interstate speed limits dropped from 75 to 55 during the gas crisis in the 70s. This happens in several ways. Since you know so much about drag, you no doubt know that it's proportional to velocity squared. Thus, even on an empty highway, slowing down reduces drag. Second, your engine is less efficient at higher RPMs, largely due to increased frictional losses. So, driving slower saves you gas that way too.
"Speeding" does not kill, this is a fallacy that is perpetuated by the auto insurance industry and uninformed public officials.
Pardon my ignorance, but what incentive does the auto insurance industry to perpetuate this fallacy? I tend to believe that the actuarial tables do not lie.
2) How does the system deal with engine/linkage issues. Cars don't provide smooth power/steering at all times. If the engine is out of tune or has a catchy throttle, can the system deal with that as well as/better than a human?
The same way a human does it, feedback control. i.e. You measure your acceleration, and you adjust the throttle until you get the value you want. With the right control system, the computer should be able to do a better job than a person. This is actually a famous problem in controls, and has been much studied. I was a bit disappointed that the article talked more about the mechanics of interfacing with the car, but maybe they consider the controller a "solved problem"
3) How is it told where to park? It would have been nice if it was clear in the video what the driver did to tell it that. The article alludes to some sort of analysis system for this, but I like pretty pictures.;)
It looks like it just scans to the car's right for a space large enough to hold the car. It alerts you when it finds one, and you probably just hit a button to "go to most recent available space". Since this is a prototype, it probably doesn't have a very sophisticated UI.
I have to say, I found that incredibly obvious sentiment much more "insightful" when I read it in your comment than when I first saw it, several long minutes ago, in the article summary. The addition of terrible grammar and excessive punctuation really brought the point home for me.
If I understand the license correctly, there is a key difference that you're missing:
In the first passage you cite, it need only be made available under the CPL if released in source code form. So, you could distribute binaries of the code under any license you want. The satement that the new license "complies with the terms and conditions of this Agreement" isn't the same as saying you have to release it under the same license. It just means you can't violate any of the conditions of that license (say, about warantees or whatever).
In the 90s GM did this with the high-end Camaro and the base model Corvette. There may have been some small differences, but it was basically the same engine with different tuning. IIRC, the Corvette needed premium while the Camaro could run on regular.
Assuming they can sell the crippled item at a lower cost and still make a profit, there's no reason they couldn't sell the full-featured version at the same price and still make a profit.
Well, yes and no. Clearly Cannon can still make a per-unit profit on selling the less expensive camera, but they still have to recover their engineering costs. For cutting edge technology, those are probably non-trivial.
Cannon no doubt developed a strategy for repaying them based on selling X number of full-featured units at a high price, and Y crippled units at a low price.
The whole reason for releasing products with intentionally crippled functionality (which in the electronics world goes back at least to Intel's 486SX, and probably much further) is to minimize engineering overhead by not having to design a new product. While it seems ridiculous to us to sell a product with important features disabled, its a very natural side effect of the economies of scale that the electronics industry has created. That is, it's cheaper to make a million of the same chip, and break half of them, then to make half a million of two separate designs.
Put yet another way: You can bitch all you want, but if Cannon hadn't been able to release the crippled camera, they might not have made the thing in the first place.
Do you have a source for that 60% number? I did a search, and the most current numbers I found were for 2002, when ESA had 41% of commercial space launches (down from 50% in 2001).
Well, since Tenzing Norgay's been dead for 20 years, I think you'll need something more advanced than Wi-Fi to talk to him. Perhaps his son has something to say on the matter, though.
Unless by killed you mean alive and well, I think you're a bit mistaken. In fact, here is an article published today about someone reselling the service. Perhaps you are confused by the fact that the originial Iridium went bankrupt. That doesn't mean that their 66 satellites stopped working, though. They were sold to a new group of investors.
The problem with Iridium in this situation is that, AFAIK, rates are on the order of $1/minute. That is probably out of the reach of your average yak farmer. While Wi-Fi has higher initial costs, it is basically free to operate once you have all the equipment.
Well, maybe they don't need to, but personally, I have a hard time thinking that even a ruminant can get enough nutrition from cellulose alone to get by.
Regardless, your point is irrelevant because we certainly do feed the "edible" part of the corn to cows and other animals, whether they "need" to eat it or not. I don't know if you've ever been to America's beautiful corn-growing heartland, but of all the corn they raise out there, a very small percentage is the "Sweet Corn" that humans eat. In fact, over 50% of the corn produced in the US does go to domestic animal feed, and that number may get closer to 70% when you include the corn we export to feed cows in Japan and elsewhere. You can read about it here and here.
IANA Farmer but as I understand it the green parts of the plant are usually fed to dairy cows as silage, not to beef cattle.
That statment is technically true, in that MSFT stock was trading in the mid 20s in 1998, and is trading in the mid 20s right now. Then again, if you consider there have been 3 2:1 splits since Febuary of 1998, the stock is actually worth 8 times as much. That's an increase of over 100% per year... hardly a bummer.
As with Gates, most of Paul Allen's money was made in options to buy Microsoft stock. I suspect he still has considerable holdings of MS stock (and possibly options as well), thus his wealth increases as MS stock goes up. So, yeah, his wealth is probably still significantly tied to MS's revenue stream.
The laws of chemistry don't care whether I do it in a hybrid, petro-electric, big-block V8 or average suburban runabout.
The reaction that goes on in your car's engine is not as simple as what you learned in general chem. What you're assuming is that the conditions are identical in the engines of all these vehicles. They are not. Things like air/fuel mix and combustion temperature have everything to do with the reaction products (particularly the nasty ones that result from incomplete combustion). Consider that the hybrid's gas engine can run at a much more consistent rate (in terms of RPM), because it can allow the electric motor to soak up the small spikes in input. It needs to perform over a much smaller range of conditions, and so can be better tuned for optimal emissions performance.
I don't know if anyone will read down this far, but I've been itching to point out how the article tries to legitimize this claim by the use of the plural "Scientists" when in fact it seems to be just one dude who is saying this.
Maybe I'm insufficiently skeptical, but when I read that "Scientists" think something, that tends to imply things, like peer review, or even concensus, which are clearly not present here.
Origins aside, the term has, as far as I can tell, been embraced by the group to which it refers.
You can invoke the "P.C." label if you want, but there are perfectly practical reasons not to use the same name for two distinct groups of people, especially given the increasing number of Indian Indians from India in the US. I'd call them Indian-Americans, but I wouldn't want to get you all upset.
I'm not sure what your last sentence means, but are you trying to say that using the name "Indians" to refer to aboriginal Americans has some historical value?
And you can compare the pay with other US companies. Other companies can learn from google here.
I tend to agree with you that corporate executives are overpaid. In this case, however, I expect that all three were taking limited (although still substantial) compensation with the understanding that they would all become filthy, filthy rich once the IPO hit.
What is your source on the Warren Buffett info? It would seem to serve as a pretty definitive refutation to the "this IPO is a sham" claims above, but seems unlikely due to his general avoidance of technology shares.
Re: your sig: Matt Groening explains this in a Fresh Air interview. The idea was that Bart didn't respect his father, but worshipped a clown who looked exactly like him.
I believe this is the link. But he's been on several times, so you may have to listen to a couple to find the right one.
Slashdot Mirror
Dude, don't worry. It was clear what you meant. That doesn't mean people will not seize the opportunity to make a joke, though. Personally, I thought mine was pretty funny. See, a robot might prefer the places where he doesn't get treated like a... oh never mind.
Well, maybe this guy is a robot.
In most places, it is. In my state, insurance companies have rate modifiers based on your zip code.
What you're describing with the sticky pedal is a classic problem with control systems. Because friction acts more strongly on a stationary object than a moving one, you often need a big force to start something moving, which can cause you to push the system way too hard (this is called integrator windup, a reference is here).
There are definitely techniques to overcome this (the gist of it is, you tell the controller to increase the force slowly until you're out of the 'sticky' range). I agree that it is hard to make a machine "learn" this kind of behavior in advance, and so it is at a disadvantadge in that respect. On the other hand, the machine probably has a quicker response time than a person, and it generally has access to more precise inputs. For instance, the machine can use a sensor (I think they are using ultrasound in this project) to determine its precise distance from the other cars, whereas a human just has a general sense. So even if it gets itself into trouble by pushing the pedal to hard, it can hit the brake much more quickly. Of course, stamping the gas and then the brake is not necessarily a desirable way to park your car. The solution? Spray some WD-40 on the accelerator I guess.
Maybe the Pennsy cops have quotas, but I think you're being a little bold to speak about all police departments everywhere.
Also, speed limits save gas. That's why U.S. Interstate speed limits dropped from 75 to 55 during the gas crisis in the 70s. This happens in several ways. Since you know so much about drag, you no doubt know that it's proportional to velocity squared. Thus, even on an empty highway, slowing down reduces drag. Second, your engine is less efficient at higher RPMs, largely due to increased frictional losses. So, driving slower saves you gas that way too.
"Speeding" does not kill, this is a fallacy that is perpetuated by the auto insurance industry and uninformed public officials.
Pardon my ignorance, but what incentive does the auto insurance industry to perpetuate this fallacy? I tend to believe that the actuarial tables do not lie.
2) How does the system deal with engine/linkage issues. Cars don't provide smooth power/steering at all times. If the engine is out of tune or has a catchy throttle, can the system deal with that as well as/better than a human?
;)
The same way a human does it, feedback control. i.e. You measure your acceleration, and you adjust the throttle until you get the value you want. With the right control system, the computer should be able to do a better job than a person. This is actually a famous problem in controls, and has been much studied. I was a bit disappointed that the article talked more about the mechanics of interfacing with the car, but maybe they consider the controller a "solved problem"
3) How is it told where to park? It would have been nice if it was clear in the video what the driver did to tell it that. The article alludes to some sort of analysis system for this, but I like pretty pictures.
It looks like it just scans to the car's right for a space large enough to hold the car. It alerts you when it finds one, and you probably just hit a button to "go to most recent available space". Since this is a prototype, it probably doesn't have a very sophisticated UI.
Well I can't read whatever language the web page is in.
Clearly it's in Hollish.
I have to say, I found that incredibly obvious sentiment much more "insightful" when I read it in your comment than when I first saw it, several long minutes ago, in the article summary. The addition of terrible grammar and excessive punctuation really brought the point home for me.
If I understand the license correctly, there is a key difference that you're missing:
In the first passage you cite, it need only be made available under the CPL if released in source code form. So, you could distribute binaries of the code under any license you want. The satement that the new license "complies with the terms and conditions of this Agreement" isn't the same as saying you have to release it under the same license. It just means you can't violate any of the conditions of that license (say, about warantees or whatever).
In the 90s GM did this with the high-end Camaro and the base model Corvette. There may have been some small differences, but it was basically the same engine with different tuning. IIRC, the Corvette needed premium while the Camaro could run on regular.
Assuming they can sell the crippled item at a lower cost and still make a profit, there's no reason they couldn't sell the full-featured version at the same price and still make a profit.
Well, yes and no. Clearly Cannon can still make a per-unit profit on selling the less expensive camera, but they still have to recover their engineering costs. For cutting edge technology, those are probably non-trivial.
Cannon no doubt developed a strategy for repaying them based on selling X number of full-featured units at a high price, and Y crippled units at a low price.
The whole reason for releasing products with intentionally crippled functionality (which in the electronics world goes back at least to Intel's 486SX, and probably much further) is to minimize engineering overhead by not having to design a new product. While it seems ridiculous to us to sell a product with important features disabled, its a very natural side effect of the economies of scale that the electronics industry has created. That is, it's cheaper to make a million of the same chip, and break half of them, then to make half a million of two separate designs.
Put yet another way: You can bitch all you want, but if Cannon hadn't been able to release the crippled camera, they might not have made the thing in the first place.
Do you have a source for that 60% number? I did a search, and the most current numbers I found were for 2002, when ESA had 41% of commercial space launches (down from 50% in 2001).
Well, since Tenzing Norgay's been dead for 20 years, I think you'll need something more advanced than Wi-Fi to talk to him. Perhaps his son has something to say on the matter, though.
Unless by killed you mean alive and well, I think you're a bit mistaken. In fact, here is an article published today about someone reselling the service. Perhaps you are confused by the fact that the originial Iridium went bankrupt. That doesn't mean that their 66 satellites stopped working, though. They were sold to a new group of investors.
The problem with Iridium in this situation is that, AFAIK, rates are on the order of $1/minute. That is probably out of the reach of your average yak farmer. While Wi-Fi has higher initial costs, it is basically free to operate once you have all the equipment.
So you're not counting GLONASS? As far as I know, the Russian system is a perfectly acceptable alternative to GPS.
Well, maybe they don't need to, but personally, I have a hard time thinking that even a ruminant can get enough nutrition from cellulose alone to get by.
Regardless, your point is irrelevant because we certainly do feed the "edible" part of the corn to cows and other animals, whether they "need" to eat it or not. I don't know if you've ever been to America's beautiful corn-growing heartland, but of all the corn they raise out there, a very small percentage is the "Sweet Corn" that humans eat. In fact, over 50% of the corn produced in the US does go to domestic animal feed, and that number may get closer to 70% when you include the corn we export to feed cows in Japan and elsewhere. You can read about it here and here.
IANA Farmer but as I understand it the green parts of the plant are usually fed to dairy cows as silage, not to beef cattle.
I assume you're referring to this sentence:
Its stock price has stalled at 1998 levels.
That statment is technically true, in that MSFT stock was trading in the mid 20s in 1998, and is trading in the mid 20s right now. Then again, if you consider there have been 3 2:1 splits since Febuary of 1998, the stock is actually worth 8 times as much. That's an increase of over 100% per year... hardly a bummer.
As with Gates, most of Paul Allen's money was made in options to buy Microsoft stock. I suspect he still has considerable holdings of MS stock (and possibly options as well), thus his wealth increases as MS stock goes up. So, yeah, his wealth is probably still significantly tied to MS's revenue stream.
The laws of chemistry don't care whether I do it in a hybrid, petro-electric, big-block V8 or average suburban runabout.
The reaction that goes on in your car's engine is not as simple as what you learned in general chem. What you're assuming is that the conditions are identical in the engines of all these vehicles. They are not. Things like air/fuel mix and combustion temperature have everything to do with the reaction products (particularly the nasty ones that result from incomplete combustion). Consider that the hybrid's gas engine can run at a much more consistent rate (in terms of RPM), because it can allow the electric motor to soak up the small spikes in input. It needs to perform over a much smaller range of conditions, and so can be better tuned for optimal emissions performance.
These people are your ISP. Don't all your packets have to go through their routers to get anywhere?
Can't they monitor all your traffic there just as well as they can at the cable modem?
Couldn't they disable any system on your network by MAC filtering?
I don't know if anyone will read down this far, but I've been itching to point out how the article tries to legitimize this claim by the use of the plural "Scientists" when in fact it seems to be just one dude who is saying this.
Maybe I'm insufficiently skeptical, but when I read that "Scientists" think something, that tends to imply things, like peer review, or even concensus, which are clearly not present here.
So, umm, who came up with it then?
Origins aside, the term has, as far as I can tell, been embraced by the group to which it refers.
You can invoke the "P.C." label if you want, but there are perfectly practical reasons not to use the same name for two distinct groups of people, especially given the increasing number of Indian Indians from India in the US. I'd call them Indian-Americans, but I wouldn't want to get you all upset.
I'm not sure what your last sentence means, but are you trying to say that using the name "Indians" to refer to aboriginal Americans has some historical value?
And you can compare the pay with other US companies. Other companies can learn from google here.
I tend to agree with you that corporate executives are overpaid. In this case, however, I expect that all three were taking limited (although still substantial) compensation with the understanding that they would all become filthy, filthy rich once the IPO hit.
What is your source on the Warren Buffett info? It would seem to serve as a pretty definitive refutation to the "this IPO is a sham" claims above, but seems unlikely due to his general avoidance of technology shares.
Re: your sig:
Matt Groening explains this in a Fresh Air interview. The idea was that Bart didn't respect his father, but worshipped a clown who looked exactly like him.
I believe this is the link. But he's been on several times, so you may have to listen to a couple to find the right one.