Would that increase the accuracy so that both systems together are more effective? I don't really think so. Statistics says yes. No matter what the variance, any additional unbiased data improves your estimate. It may not improve it a lot, but it will improve it. In the case where there aren't enough sattelites of any one system visible though, it could mean the difference between working rather than failing.
I don't think that Galileo (which has an accuracy of 0.1 meters afaik) can be enhanced by some GPS satellites (which has an accuracy of 15 meters). They are way too old, the GPS satellites (at least, most of them). Well, it's hard to be newer that satellites which have yet to be launched... Let's revisit the accuracy when there is a Galileo constellation to speak of. Right now, the plans are for the <1m accuracy service to be encrypted, and require payment to access. To get the 0.1m accuracy, a base station (ala differential GPS) is required, so you're doing an apples to oranges comparison. The open/free service is 4m(x,y),8m(z).
Now, with the EU taking control, maybe the <1m service will now be free, but let's talk about how well it works when it actually exists.
And commenting on the article, if Galileo and GPS don't sync up thier clocks directly, I don't see how it will work. Of course you can combine them; It's just a question of how much additional information you can get. Worst case is that you have to treat them separately until the position is calculated, and you then combine the two independent readings, which should about halve the variance. That's nothing to sneeze at, and wouldn't require any information at all about relative clock skew. In the best case however, a device could track the long-term clock skew between the two systems (which should stay nearly fixed) by filtering on the time skew that brings separate readings into most agreement. The skew would take a long time to estimate, but once you have it, you should be able to mix and match satellites.
So, how did you store your personal data such that it depends on the internal module ABI? Remember that the userland/syscall ABI is stable, so you're referring to the driver/module interface. Maybe you should try a different way of storing important data than embedding it in driver code...
I think that is unfair. They are experimenting with a _different_ paradigm. And the article finishes with the call for further research in adapting to other terrains and the like. The article reads like an article on the MIT leg lab's "spring chicken" robot from almost 10 years ago. Just about every thing described for this robot was already done on that planar walker (on a larger more realistic scale, I might add). So, rather than comparing it with Asimo, they should compare it with the leg lab robots and tell us what is different from them. Don't build a new car and tell me how it is different from trucks, compare it to other cars. As I said though, this might not be the researchers' fault, since we have the percolated article to read rather than links to their original research.
These "baby steps" are precisely the concise research AI needs. It looks like they have made progress in how people _really_ walk. OK, so building on that foundation let's continue work on the interfaces and control units to sense, recognize and adapt to changing conditions. Rod Brooks has been pushing that kind of approach since the late 80s, and most of the AI community has long since been won over to a bottom-up approach. The kind of stodgy logic-only robotics, that you seem to want to contrast this work with, fell out of favor a long time ago. Bottom up design with statistics and learning are pretty standard in research circles now. So, please don't lecture the AI/robotics community about something we already know; For the most part you are preaching to the choir, and those isolated research groups that aren't sold on the approach will continue to resist anyway, and are best left alone (who knows, maybe they will turn out right in the end, so let them try their theories too).
If there is any generalization about real world AI that holds true, it is in how much the difficulty is under-appreciated. Exactly, which is why we shouldn't oversell it. The title should be "planar walking robot learns using multi-level feedback loops", since this isn't 3d (and most planar research hasn't transferred well), nor is it the first walker using passive dynamics, learning, or to operate on uneven terrain. If you don't tell it like it is (a) other researchers are annoyed you didn't reference them, and (b) the press gets confused later when you actually try to do something they thought you already did. If you want public support for a 3d walker later, you'd better be clear you are only doing 2d now.
If there's something the world probably didn't need, it's another planar walker. Of course, the researchers are probably quite honest about the limitations when applying this to full 3d walking, but all that is lost in the translation to an article and then a slashdot blurb.
No kidding. If Google censored the healthcare websites or put their rank artificially low, would that not be more evil? Why does "evil" on Slashdot now mean "anyone who doesn't agree with me". That must be "evil" as in "axis of evil", I guess... if you don't like the administration, don't act like them.
This case is like offering to sell a car to someone who is boycotting a bus. Not taking sides, just offering a service for a price. That's how the free market works.
And I would agree with both statements. I don't see how anyone can rage against Rush or Fox news, and then use a Moore film as "evidence". They all use the same approaches. Moore (and the others) believe that its ok to twist the facts, as long as your message is a good one. Of course, with distorted evidence, we can only trust that they are right, or watch multiple films as a same point-counterpoint. Spin is very common at all levels of journalism, but that doesn't mean it should be the ideal one strives toward. Instead I say that one ought to set an example above those you are complaining about -- don't just stoop to their level. But what do you expect when Che Guevara shirts are many times more popular than MLK ones?
Here's a timeline. So what should we have done, grounded planes for the month of August? I'm sure the hijackers could wait. Also, how many other reports and threats do you think we get? Should we act on all of those too? If that happened, the terrorists have already won. The loss of some freedom in our country is something they can already notch as a victory.
The general problem is that when you look back in hindsight, the right information was there, and some expert was saying exactly the right thing and had a perfect prediction. The problem is, at the time, you have tons of information and hundreds of experts saying conflicting things. Was the administration incompetent to the threat? Maybe, but the fact that they missed "signs" isn't going to tell us that.
The security expert Bruce Schneier has a lot of excellent commentary on the prevention of terrorism, which is quite worth reading. Trying to cope with any one threat is generally a waste of time, yet that's what the police do. It is better to think like a security researcher and keep thinking of the weakest attack vector and trying to deal with it.
My point was that Clinton got Yousef behind bars. Where is Bin Laden, the guy who attacked us on 9/11 and killed 3,000 people? First, there's a bit of difference between the money, power, and connection those leaders have -- like the difference between are car theif and a mobster. Second, getting leaders like that is largely symbolic. Bin Laden is no longer in direct control of followers, and if we kill him it won't slow down Al Qaeda much. I'm not saying we shouldn't pursue him, but I actually agree with the administration on this one; We are better off spending major resources elsewhere.
It's claimed that the target in Sudan was a chemical weapons lab. If it was destroyed, wouldn't that mean less chemical weapons in the hands of terrorist? After Iraq, how much do you trust that intelligence?
That's a good point. Clinton didn't have the political capital to launch an all-out war against terrorism. The American public didn't perceive the threat on that scale. However, I disagree that Clinton should have done absolutely nothing, as you claim, because he couldn't do *everything*. Well, maybe we could try to better inform the public of the risk, and make a public warning to our enemies. Then, when a time comes when we need to go "all out", we do so quickly. I'm not a big fan of limited war. It's not that you have to do everything, but you have to do enough, which usually means "a lot". Doing "a little" isn't usually worth it. Another good example are the generals who wanted 250k troops for Iraq, rather than the more limited 100k or so what we used. They were probably right, and winning the peace might have been a lot easier with enough troops in place; Do it fully or don't do it at all. That said, again hindsight is 20/20.
Clinton did foil the millennium bombing plot. True, and there are many we'll never hear about. Unfortunately, it usually seems like other nations are handing us the critical intelligence any time we have a decent success story.
You think he should have let that happen because he couldn't start an all-out war against Bin Laden? Meanwhile, Bush did in fact do absolutely nothing, and 9/11 happened. The deep question is this: Was seriously provoking Bin Laden a wise thing to do? In retrospect (20/20 again), I feel it may have been a bad idea. Terrorists feed off of being the victim, that is how they get their recruits.
It would have done a lot more good if the Republicans hadn't been screaming that Clinton was "wagging the dog" while they were trying to invent a reason to impeach him. True, but if he hadn't been screwing around that wouldn't have been a problem, now would it? The timing of the attack was terrible.
It also would have helped if Bush had negotiated with the Taliban when they offered to hand over Bin Laddin. And what, exactly, would we do with him? I can see the news stories now: "Government improperly holding Bin Laden without charges", "US negotiates with impressive regime, for what?" It probably wouldn't stop Al Qaeda either, just as stopping the leader of the insurgency in Iraq didn't do much. It may have prevented the 9/11 attacks, but again, hindsight is 20/20.
Sleeper cells! Oh, the drama! If you want to act that way, I guess I have to spell it out for you. Once the planning is done, and all the the people are in place, there is very little you can do to stop an attack. If they aren't communicating or planning anymore, what are you going to intercept? Maybe you advocate racial profiling?
Problem: all the successes we've had fighting terrorism have come from the boring, "pre 911 mindset", Constitution-abiding law enforcement. On the other hand, the Administration has it's Patriot Act, waterboarding, NSA wiretapping, kidnappings, endless detention w/o trial, And yet all that success was and is not enough to prevent multiple attacks from occurring. There are countries who spend a whole lot more time doing invasive counter-terrorism work, such as Israel, but they still can't stop everything. If you demand perfection, you're going to be disappointed. That was my point.
and yet is too fucking incompetent to know when an Arab company is about to take over administration of the largest ports in the U.S. Ok, I guess you really do advocate racial profiling. Hint: Most Arabs are not terrorists. Even if you want to profile in your example, what does the owner of a port matter, when nearly all of your longshoremen are US citizens?
That would only make any sense if other countries had open borders, and didn't mind giving jobs to immigrants. I can't think of any developed countries that do not try to protect internal jobs somewhat, and many EU members are very protective. France had riots over this very issue, both by jobless immigrants, and the French youth who wanted highly protected jobs to be maitained.
That said, what this law firm is doing is legal, though perhaps not very ethical. However, taking every advantage of loopholes is what you would expect from a lawyer, or for that matter a tax-man or politician. Politicians regularly skirt every law and regulation about campaign funding, yet that seems to be ok with them. The only talk is of reforming the campaign laws to change the boundaries. So, if you want this law firm to act differently, try writing better laws.
Clinton pursued and convicted Ramzi Yousef, the mastermind behind the first WTC bombing. Note that he didn't prevent the attack, he reacted to an attack that had already happened. Personally, that's all I would expect of a government, but people seem to think 9/11 was easily preventable. In that case, the original WTC attack should have been preventable as well.
Clinton launched cruise missile attacks against terrorist training camps in Sudan and Afghanistan A lot of good that did. Do you think that made us safer? Within one year of those ineffectual attacks, Bin Laden had raised the Bojinka plot from the dead, and recruited enough people to carry it out. What event do you think he used for motivation? Just like the Iraq war is not making us safer, those attacks ultimately put us in more danger, as it just made people angrier. You don't poke a bear with a stick; You either do enough to stop it completely, or you leave it alone. Of course, at the time there was no way of knowing how far they might take it, and congress wouldn't have let Clinton go on a major offensive against Bin Laden anyway. Hindsight is 20/20.
So he was attacking Osama Bin Laden and Al Qaida. Yes, but the sleeper cells came to the US under his watch, and the only easy time to stop someone like that is when they enter the country. After that, tracking someone's movements becomes quite difficult (and rightfully so, in any free society). If you insist on blaming the government, there's plenty of blame to go around. Or you can be honest, and admit that a government that respects people's privacy is not going to be able to prevent every possible attack. Clinton knew this, and Bush may know this although he doesn't act like it, although he's probably just giving in (or taking advantage of) the impossible task that the public is demanding.
Sorry I indeed misread. I thought you were claiming "Making modular software reliable and documented is hard" rather than "Making reliable, documented, modular software is hard". Since your "it" jumped across posts, you can see how I might make that mistake.
Yes, but over-engineering can happen if you take modularity (loose term, btw) too far and that can be just as devastating to a design as spaghetti code Of course, but for things as large as searching and indexing subsystems, it would be surprising to not have it as a module. Microsoft doesn't have a great track record either, with their attempts to make IE as integrated as possible, even when it really was modular. They were actively aiming for monolithic (or plain lying, take your pick), which is sad.
because now you're stuck supporting scenarios for which you never intended nor wanted your API to be used for. Well, that never stopped MS before on any of their not-completely documented interfaces (or OSS for that matter). They are good for the things they publish (well, up until Vista at least), but real programs often end up going beyond that into gray areas that end up changing. I don't think there would be anything wrong with saying an API is not guaranteed to be supported, although given the antitrust issue I guess they would have to show that they didn't purposely aim for incompatibility with a change (like with DR DOS).
Try working in a large software engineering house for several years where you have to make tradeoffs like this all the time and see what tough decisions you have to make sometimes. That isn't the only place where complex software is written. Try entering nine international robotics competitions with fixed deadlines, testing that can't be automated since it requires the physical world to operate, the uncertainty of pitting your system against another software/hardware system that you've never seen before, and the necessity of changing it during an ongoing competition. Now do all that with a team where member skills vary from college underclassmen to PhDs.
The software world isn't perfect and acting like you are is just you talking out your ass. I never claimed to be perfect, just that modularity should come first. Look at some of the posts about the architecture of spotlight; Despite a similar setup and a shorter development time, it has several places where an ISV can hook into the system.
No problem. This is also the source for the November date I gave you. It goes on to make some very good points about why search isn't 'middleware', and the fact that search was never included in the 2000 judgement is a very good argument as to why Microsoft are in the clear. Wow, that article is a piece of work. It pretty much ignores all the concrete information provided in the NYT article, thus it doesn't really rebut any of the arguments with meat on them: "Bias, what bias, they are both companies! Never mind any of those pesky details like possible conflicts of interest!" What do you expect from a guy who wrote "The Google Problem" with choice quotes such as "Microsoft is hugely and increasingly transparent, and uncharacteristically so among large, successful companies." and "Who better then to judge Google as a problem than Microsoft? If Microsoft is concerned, why aren't you?" Someone who thinks Microsoft is a "hugely transparent company" is not living in the same world as the rest of us.
Not to mention that Google have been (and still are) free to put their own front end on Vista's indexer in the same way they do with Spotlight on the Mac. Except that given Microsoft's latest proposed remedy, one still can't override search in explorer or several other places, just the one on the start menu. Spotlight, as has been pointed out elsewhere in the comments, is actually quite modular and all the bits can be replaced or overridden. That's all Google is asking for from Windows.
Of course I wouldn't. Read this and then tell me why that argument is even coherent. Well, let's see if I can make this clear. A DOJ official with a potential conflict of interest is acting unusual and as though he actually has a conflict of interest. In the whitehouse, officials with a potential conflict of interest did unusual things as though they had a conflict of interest. You claim that the DOJ official's past has "zero" to do with how he acts now. That same unsubstantiated claim can be applied to our friends in the whitehouse. If you want to substantiate it with a "because", then I might believe you, but you simply state that he has no conflict as a universal fact.
it still does not give basis to Google's ridiculous claim that somehow Microsoft are acting anti-competitively by having a way of searching for files in their operating system. That's a complete red herring. Google isn't asking for Microsoft to refrain from having a search feature, they are just asking for a way to replace it, if the user so chooses. Of course you know that, and are just trying to exaggerate the issue to make it absurd. It's true that Google could be wrong: If search can indeed be disabled and replaced, then there's no problem for Microsoft to remedy. In that case, Microsoft should have no problem documenting how and providing this information Googla and the DOJ, and everyone can go away satisfied. So far that hasn't happened.
TiVo should be fine unless the Linux kernel is relicensed, which is not going to happen any time soon, if ever. Even then, they could always just fork off at the last GPL2 kernel release, as its not like they desperately need most updates (are they even on 2.6 yet?). Any other libraries/interfaces they use outside of the kernel should be LGPL or MIT/BSD, so they should be fine unless some critical program/daemon they need to run becomes GPL3 only. I guess if that happens its time to adopt a BSD userspace.
Yes, they filed back in November - 4 days before Vista's release to business. They had clear knowledge of the existence of the indexing system since 2003, and a year of beta testing. What took them so long? Perhaps they were told before that the the product was in beta, and the the final release would be different. Others have claimed this, although there is little public information, so we don't know for sure. If you have evidence that Nov 4 is the first time they ever complained, I would like to see it. Anything else is speculation.
The rest of your post seems to argue with itself. You could read the article I linked. It's not that long.
You cite a memo that came from someone with an apparent conflict of interests - despite the fact that working for a company that once worked for Microsoft has absolutely zero bearing on how he does his job So, would you like to go on record that the Bush/Cheney ties with oil companies and Halliburton don't affect how they do their current jobs? After all, they cut all ties with their former employers, right?
and then carry on to say that the memo didn't in fact exist. Can you decide whether it was sent or not, because if it wasn't your outrage kind of falls flat. Um, what it means is that the state lawmakers were suprised to get a memo requesting something that no other head had ever requested. The existence of the memo is not in question, which would be obvious if you read the NY Times article. As phrased though, perhaps it is a little unclear, so here's another sentence of context:
Mr. Barnett's memo dismissing Google's claims, sent to state attorneys general around the nation, alarmed many of them, they and other lawyers from five states said. Some state officials said they believed that Google's complaint had merit. They also said that they could not recall receiving a request by any head of the Justice Department's antitrust division to drop any inquiry.
If you want to argue with facts presented in the NY Times, perhaps you could at least cite some sources.
Guess what, Google complained a long time ago, but someone at the DOJ tried to get it dismissed. It turns out that DOJ official used to be a lawyer at a law firm representing Microsoft. The only thing that kept the complaint alive was action by state attorneys, who disagreed with the DOJ official's memo, and were concerned by the fact that the memo even existed:
Some state officials said they believed that Google's complaint had merit. They also said that they could not recall receiving a request by any head of the Justice Department's antitrust division to drop any inquiry. If you were a cynic before, you must really be cynical now...
All of these *can*, very theoretically, be done. Actually, it's not merely theoretical, because Debian can do that right now. It's amazing how far you can get with a good modular design.
Slashdot Mirror
Now, with the EU taking control, maybe the <1m service will now be free, but let's talk about how well it works when it actually exists.
That's a great argument for seven years ago, but selective availability is ancient history now.
sorry, couldn't help myself. As far as price instability, I don't think there's an OSEC yet.
So, how did you store your personal data such that it depends on the internal module ABI? Remember that the userland/syscall ABI is stable, so you're referring to the driver/module interface. Maybe you should try a different way of storing important data than embedding it in driver code...
If there's something the world probably didn't need, it's another planar walker. Of course, the researchers are probably quite honest about the limitations when applying this to full 3d walking, but all that is lost in the translation to an article and then a slashdot blurb.
No kidding. If Google censored the healthcare websites or put their rank artificially low, would that not be more evil? Why does "evil" on Slashdot now mean "anyone who doesn't agree with me". That must be "evil" as in "axis of evil", I guess... if you don't like the administration, don't act like them.
This case is like offering to sell a car to someone who is boycotting a bus. Not taking sides, just offering a service for a price. That's how the free market works.
And I would agree with both statements. I don't see how anyone can rage against Rush or Fox news, and then use a Moore film as "evidence". They all use the same approaches. Moore (and the others) believe that its ok to twist the facts, as long as your message is a good one. Of course, with distorted evidence, we can only trust that they are right, or watch multiple films as a same point-counterpoint. Spin is very common at all levels of journalism, but that doesn't mean it should be the ideal one strives toward. Instead I say that one ought to set an example above those you are complaining about -- don't just stoop to their level. But what do you expect when Che Guevara shirts are many times more popular than MLK ones?
Note to Ms Gould: There's a difference between a tag you wear at work, and something semi-permanently implanted in your body.
Here's a timeline. So what should we have done, grounded planes for the month of August? I'm sure the hijackers could wait. Also, how many other reports and threats do you think we get? Should we act on all of those too? If that happened, the terrorists have already won. The loss of some freedom in our country is something they can already notch as a victory.
The security expert Bruce Schneier has a lot of excellent commentary on the prevention of terrorism, which is quite worth reading. Trying to cope with any one threat is generally a waste of time, yet that's what the police do. It is better to think like a security researcher and keep thinking of the weakest attack vector and trying to deal with it. My point was that Clinton got Yousef behind bars. Where is Bin Laden, the guy who attacked us on 9/11 and killed 3,000 people? First, there's a bit of difference between the money, power, and connection those leaders have -- like the difference between are car theif and a mobster. Second, getting leaders like that is largely symbolic. Bin Laden is no longer in direct control of followers, and if we kill him it won't slow down Al Qaeda much. I'm not saying we shouldn't pursue him, but I actually agree with the administration on this one; We are better off spending major resources elsewhere. It's claimed that the target in Sudan was a chemical weapons lab. If it was destroyed, wouldn't that mean less chemical weapons in the hands of terrorist? After Iraq, how much do you trust that intelligence? That's a good point. Clinton didn't have the political capital to launch an all-out war against terrorism. The American public didn't perceive the threat on that scale. However, I disagree that Clinton should have done absolutely nothing, as you claim, because he couldn't do *everything*. Well, maybe we could try to better inform the public of the risk, and make a public warning to our enemies. Then, when a time comes when we need to go "all out", we do so quickly. I'm not a big fan of limited war. It's not that you have to do everything, but you have to do enough, which usually means "a lot". Doing "a little" isn't usually worth it. Another good example are the generals who wanted 250k troops for Iraq, rather than the more limited 100k or so what we used. They were probably right, and winning the peace might have been a lot easier with enough troops in place; Do it fully or don't do it at all. That said, again hindsight is 20/20. Clinton did foil the millennium bombing plot. True, and there are many we'll never hear about. Unfortunately, it usually seems like other nations are handing us the critical intelligence any time we have a decent success story. You think he should have let that happen because he couldn't start an all-out war against Bin Laden? Meanwhile, Bush did in fact do absolutely nothing, and 9/11 happened. The deep question is this: Was seriously provoking Bin Laden a wise thing to do? In retrospect (20/20 again), I feel it may have been a bad idea. Terrorists feed off of being the victim, that is how they get their recruits.
That would only make any sense if other countries had open borders, and didn't mind giving jobs to immigrants. I can't think of any developed countries that do not try to protect internal jobs somewhat, and many EU members are very protective. France had riots over this very issue, both by jobless immigrants, and the French youth who wanted highly protected jobs to be maitained.
That said, what this law firm is doing is legal, though perhaps not very ethical. However, taking every advantage of loopholes is what you would expect from a lawyer, or for that matter a tax-man or politician. Politicians regularly skirt every law and regulation about campaign funding, yet that seems to be ok with them. The only talk is of reforming the campaign laws to change the boundaries. So, if you want this law firm to act differently, try writing better laws.
Amazing, isn't it? I guess John C Dvorak was right after all...
TiVo should be fine unless the Linux kernel is relicensed, which is not going to happen any time soon, if ever. Even then, they could always just fork off at the last GPL2 kernel release, as its not like they desperately need most updates (are they even on 2.6 yet?). Any other libraries/interfaces they use outside of the kernel should be LGPL or MIT/BSD, so they should be fine unless some critical program/daemon they need to run becomes GPL3 only. I guess if that happens its time to adopt a BSD userspace.
Mr. Barnett's memo dismissing Google's claims, sent to state attorneys general around the nation, alarmed many of them, they and other lawyers from five states said. Some state officials said they believed that Google's complaint had merit. They also said that they could not recall receiving a request by any head of the Justice Department's antitrust division to drop any inquiry.
If you want to argue with facts presented in the NY Times, perhaps you could at least cite some sources.
Hey, where are all the replies? Did iluvcapra's facts get in the way of your argument?