On the other hand, if a company releases its code under the GPL and then accepts code from outside contributors under the GPL, the company no longer has the option of selling the software under a commercial license.
That's partially true, but it's a bit unfair. Without the GPL or some sort of free software license, you can't use those people's code at all (ie, it's theirs, you can only distribute it if they explicitly agree to license it to you.) From a company's point of view, the ideal situation is to use a license that's slightly less restrictive than the GPL, but then how do you get people to give back their source code changes when they distribute, as the GPL does? Some of them will still probably give you the source, but you pretty much just have to hope for it (big companies that are throwing lots of support behind a popular project can get away with this, but smaller companies probably can't.)
The GPL is non-exclusive, so if you could convince all of the contributors to additionally give you closed-source rights to their contributions (say, when they check their code into your tree), you could still release the project separately under a non-GPL license.
Again, this doesn't preclude anyone from also distributing the same code under the GPL.
will the government begin to make auto manufacturers integrate these in to every vehicle, making it so they can just mail you a ticket any time you exceed the speed limit?
There's been a certain amount of talk about adapting EZPass technology (automatic toll-paying tags in NY/NJ/DE) to catch speeders. Implementing this would be a snap; just measure the time a car takes between two tollbooths and mail out a ticket. Practically, it would be a political nightmare, and would very likely result in the speed limit rapidly being raised to 75MPH+.
I've driven the NJ Turnpike a few times (speed limit=55MPH), and I can assure you that only a very small minority of the state's drivers actually pay attention to the posted limits. If an automatic system like this were implemented, half the drivers in the state could rack up enough points for a license suspension, all inside a week. I'd hate to be the politician who authorized that particular course of action; you can guarantee there would be some changes made, and fast.
So your complaint isn't that Big Media is the cause of the current problem, but that they aren't using their market power to force the ad industry to change its practices. Well, that seems somewhat more plausible, but I'm still not sure I agree with it
Well, I'm not sure I do either:) It's just a theory, and while I believe there's probably some truth to it, it's fairly diabolical to imply that corporations are colluding to deliberately wipe out small online content companies. On the other hand, it's certainly not implausable; remember, when companies this big exist, such tactics are no longer "conspiracies", they're simply corporate policies.
Big companies aren't any happier when they lose money than small companies.
It bears mentioning that the advertising rates for the most trafficked sites (AOL, for instance) haven't plummeted the way the smaller sites' banner ad rates have. So these companies aren't losing quite as much money as their competitors. Also, think of how much money a company like Microsoft is willing to invest in something like Internet Explorer, simply to corner a related market a few years down the line. Big companies that focus on nothing but the short-term reward are the exception.
And it's not like the ad money that used to be spent on the web has shifted to print and broadcast media. It has simply dried up as companies have cut back or gone out of business.
Well, that's certainly partially true. But the traditional media companies are still getting the lion's share of the (admittedly smaller) pie. It's not particularly their fault that the economy's slowed down along with advertising spending.
It involves the intent of the programmer and the purpose of the program.
To my mind, it involves the intent of the "speaker", not the programmer. No matter what purpose the programmer intended for the code, if I give it out with the explicit purpose of dissemination useful information (and what is expression, really?), I should be protected. The constitution doesn't say anything about "functional" speech, mind you, and such classification should be reserved only for cases where it can be shown beyond a shadow of a doubt that the purpose of the speech was functional, and not-- as 2600 clearly intended-- to disseminate information. It would probably be so difficult to make this determination that speech would only rarely be classified as purely functional. And I'm alright with that; the constitution intended to protect speech and expression and we should always err to the safe side of things.
Again, I say this because Big Media has an enormous amount of power over the ad industry. At present, a handful of corporations controls advertisers' access to a vast spectrum of media. AOL/TW alone could force changes in the advertising structure on the net if they wanted to. At the moment, however, they're making enough money on their traditional outlets that they're willing to let the industry play possum. In other cases, large conglomerates own both traditional media outlets and ad agencies.
This is why I blame Big Media. They're not exactly taking an active role in reducing online ad revenue, they're simply promoting the crap out of their print and broadcast advertising to the detriment of the online equivalent. These companies have a vested interest in reducing competition; they can afford to lose millions of dollars maintaining essentially profitless sites, until they are the only major players (especially if they're just shuttling that revenue over to their offline publications.) Wait a couple of years and watch the renaissance in online advertising rates that occurs when AOL/TW and Viacom start throwing their weight around.
PS The above is one those paranoid anti-corporate predictions that almost always come true.
Do these security-educated beta testers spend most of their day trying to hack their OS, or do they just poke at it a bit like every other beta tester I've ever known?
The only way to seriously beat up an OS is to find people who are willing to spend their every waking minute trying to break it. These people are generally paid for their effort, or else they stand to gain something if they succeed. MS may be hiring such people, I'm just not convinced that they're hiring enough. And your average beta-tester-with-a-job doesn't cut it.
The one and only problem here is that THEY DON'T MAKE MONEY
This is certainly true, but doesn't go very deep. Why don't those sites make money? They've received millions and millions of visitors-- traffic was rarely a problem. One of the primary reasons some of these companies are failing is their inability to make decent rates off of online advertising. And who controls the rate structure?
The online advertising crash has often been blamed on low click-through rates. Perhaps that's just a silly standard to apply. Traditional media companies make tons of money slapping glossy ads onto dead trees with no percievable response, and why is that? Because the ad industry is comfortable with that way of doing things (or in some cases, the ad industry is owned by those media companies.)
There are certainly plenty of other reasons why these sites have had trouble-- bad management, over-investment, etc. But it's disingenuous to say that the old-fashioned big-media economy had nothing to do with it.
Typical beta testing will not uncover many of the really tricky security holes in a product. If you need proof of this, go look at the Critical Security Updates page for Windows NT 4.0. The full list of security patches since the original release reads like the first few chapters of the bible.
Unless MS is specifically recruiting thousands of beta-testers just to hack the security, they're not going to fix the important holes. I'm sure they're doing a certain amount of this sort of testing, but it clearly wasn't enough for their previous OSes. In any case, the sort of beta testing you suggest is generally not where you would expect to discover most of your security flaws. I really hope MS knows this.
Hmm. I'm thinking of the huge number of NT hacks that weren't preventable until MS threw together a patch. Many of these hacks couldn't be avoided simply by securing the box. I really hope Win2K doesn't have any of those, but how the hell do we know? It's a lot of code, and people are only now beginning to peck at it.
Well, duh. Closed source doesn't necessarily mean that nobody ever sees the code. In any case, I'm sure the NSA could get access to any source code they want, if they put enough pressure on. What closed source does mean is that they can't just make changes and hand them out. Everything they want to fix has to go back to Microsoft where it'll be slowly mixed in with the next set of 'enhancements'. The resulting upgrade, coming out a year or two down the line, might be just as vulnerable as the current version. The same thing happens with Linux, but a) it's possible to fork the code in order to avoid this problem and b) everyone can examine the source to see if anything broke.
First, I find it difficult to believe that there aren't a sufficient number of Java development jobs.
There's no shortage of Java work. However, it's not all in the most interesting areas. My company works with both Java and C++, depending on the particular area we're working on at the moment, and we're constantly turning away people with only Java experience. Some companies only work with Java, so they're ideal for those kinds of programmers. But larger companies with more diverse projects tend to look for people with multiple skills.
Again, if it were easy to go from Java to C++, this wouldn't be a problem. But for a programmer who has no experience with C++ or even C (ie, they aren't experienced with memory manipulation and the wonders of header files), making such a transfer is difficult, and most companies don't want a programmer learning these concepts in the middle of a project (can you say memory leak?) It's much more reasonable to expect a smooth transfer from Java/C++ to Perl or Python than from Java to C++.
Second, the point of a CS education should not be to learn the most popular language, but to learn the concepts of programming so that you can easily pick up new languages as required.
Absolutely. People have lots of time to pick up new languages after they've taken an intro class. I would withdraw all of the arguments I just made if I felt that Java were also the best introductory language. But I don't, and the reason again comes down to the fact that Java lacks certain basic concepts that you become accustomed to when you learn C or C++, concepts that you should understand long before you mess with OO. I also find that the Java intro classes I've seen have a hard time staying away from OO early on, and this confuses people when they're still trying to master subroutines.
The best course, as far as I'm concerned, is to begin with C because both Java and C++ more or less rip off C syntax. Learn the basic concepts, including pointers and work your way up to abstract data structures (which is more or less what OO really is.) Then go on to Java or C++ as you see fit.
What your portraying here is that the French gov made a simple resonable request of a corporation to comply with it's laws, and that corporation, for reasons unknown, deceided not to make an effort to do so. Although you may think the/. folks are a bit zeolous your portrayal of the French gov as making a meek request and the large Multinational corp. thumbing its nose at them doesn't quite seem plausable either...
Especially considering Yahoo went to the effort and expense of a) shutting down the auctions and b) dragging the whole thing back to court.
My bet is that the French gov't did not give them any real guarantee that they wouldn't be held liable at some later point, so rather than continuing with a cloud of uncertainty over their heads, they took the course of action that they did.
Short stints might not be the best way to teach programming. It takes a certain amount of time to become comfortable with programming structures; since most of those languages have essentially the same structures, you're forcing students to learn multiple syntaxes just when they're trying to make sense of program flow. Let students get comfortable, then later on let show them that they can do the same stuff in a handful of other languages.
I think multiple language experience is useful, but I'm afraid that what you describe might be similar to inserting short stints of Basque and Catalan into the first semesters of an intro Spanish course.
My college went from teaching C/C++ to teaching Java in their intro classes a few years after I'd finished them. Our experience with C and C++, however unpleasant those languages are to learn with, really helped people get jobs. Java is much less useful in this way. And while it's not too difficult to transfer from one to the other, I wouldn't trust a newly converted C++ programmer to write robust code farther than I could throw him/her. Moving from Java to C++ is a lot stickier than going the other way-- C++ to Java is almost effortless.
Java's like candy-- it takes care of so much of the annoying stuff that C++ programmers get used to dealing with. While C++ sucks in this way, if you're looking for a high-performance language you can't rely on a system with garbage collection and run-time linking. Not to mention that truly understanding what's going on with memory and pointers, while posing a steeper learning curve, really helps you understand how the machine deals with your programs.
The bandwidth of current cable infrastructures is often limited to the broadcast of about 25 - 40 video channels at the last mile
Actually, if you get rid of the analog channels and go exclusively to digital transmission, you can pack a considerable number of on-demand channels onto the coax. Each 6Mhz band can currently support 8-10 channels on standard digital cable networks (the lower channels are slightly more limited.) If the "live" channels are multicast, you can further reduce the bandwidth they require. Of course, this assumes that you're completely revise the format in which digital cable is transmitted; fortunately, these changes would mainly take place at the head end and the client, so they wouldn't require massive system-wide upgrades.
Some of the calculations I've seen give each household an average of between one and two on-demand connections apiece, if you assume a reasonably sized local loop (300-400 homes.) If your local loop has too many subscribers on it, as some networks tend to (think the former TCI empire), you're out of luck.
And don't forget that the "take rate" for these services is not going to be high in the beginning, so 30-40 channels per neighborhood might be acceptable.
So, even if you are using safeweb (don't know exactly what it does), if you are using a proxy, your requests will most likely be sent to the proxy in clear text.
The following is probably either obvious or completely wrong, but let me try it:
If the browser shows an icon for a secure web page, then you've got an SSL connection. SSL is end-to-end, correct? The Proxy can't mount what is effectively a man-in-the-middle attack without your browser knowing it (and therefore failing to display a secure icon.) The situation you mention above would be obvious to any user who knows enough to look in the corner of their browser window (or better yet, check the Security properties for the page), correct? And if the icon isn't there, they can't really claim to be using a secure connection (hopefully most porn-surfers know this much.)
Since the GET requests you make over an SSL connection are all encrypted, the only thing the proxy would see is which host you're trying to connect to (Safeweb over and over again), not the full request URL (eg/cgi-bin/redirect?url=www.stinkypanties.com... etc).
A better way to get around corporate monitoring of browsing habits is to convert the IP address of the site you want to visit into a decimal number. That should confuse your typical admin enough
I would imagine that a half-serious porn-monitoring operation would log the IP addresses and use a reverse lookup (or at least, look at a table of known porn sites.)
We don't come down against DeCSS just because it can be used for piracy; we should not come down on this because it can be used for snooping.
The difference is that we (the "Slashdot crowd") don't have the capability to declare things illegal. All we can do is make a lot of noise and hope the powers that be will consider our opposition before they head down a potentially abusive path.
If the MPAA had restricted its complaints about DeCSS to a few web pages and newspaper articles rather than heading to court to flex their newly purchased laws, I don't think we'd all be quite as annoyed with them.
For the same reason that we cannot allow terrorists to get what they want, we cannot tolerate a foreign power that strikes from behind cover of their civilian populace.
If you're talking about offensive weapons, or weapons of mass destruction, then I completely agree with you. Any nation that would hide such things among the civilian population must be stopped. On the other hand, when you're talking about a defensive system like a radar network (does a radar network "strike"?), you can't be quite as moralistic about it. There may be good reason to knock it out, but that doesn't justify the slaughter of many thousands of people.
If a mission cannot be accomplished without a massacre, the military should rethink the mission; if it's a life-or-death struggle, then I suppose there's no choice. If it's a "police action" or a political move, it may not justify killing many innocent people.
This is no different than the decision police officers make in a hostage situation. While there are situations in which the death of soe hostages is almost impossible to avoid, you do not invite those situations-- no matter how much it inconveniences you.
Or, more to the point, using the civilian population as a shield to hide weapons behind is a pretty low tactic.
A cellular network is not a weapon, and even if it is used as a radar system, there has to be a better way to take it out than by throwing bombs at the individual cell towers.
The only way to make sure that governments don't do this is to systematically destroy such installations, preferably with smart weapons to moderate (though not eliminate) civilian casualties
If the Serbians/Iraqis had located thousands of small, passive radar systems every couple of blocks, could we have possibly justified the kind of slaughter that would result from taking them out (even with smart bombs?) The US has changed the way it makes war since Vietnam (at least, the way it makes war in public.) I'm not saying that civilians don't already die in NATO bombing raids, but this would be a couple of orders of magnitude worse.
The military has learned that publicly slaughtering thousands of civilians is not acceptable in the sorts of engagements we've been having. And even if you do use Smart weapons, you're still going to have a hell of a time not killing many thousands of people. These "installations" you speak of are often attached to civilian buildings in heavily populated areas-- destroying them is not like targeting a factory and knowing there may be some collateral damage. It's like going after fleas with a shotgun (or at least, a rifle.)
So in essence, deploying a system like this gives your opponent carte blanche to destroy your civilian wireless communications network. This is a bad, BAD idea.
Cell towers are tiny things that are generally located in populated areas. The bombing pattern you suggest would make swiss cheese out of an undefended civilian center. Any government that would authorize that sort of attack is a monster, and your military should be doing everything possible to make sure that sort of attacker does not emerge victorious, no matter how many lives it costs. With such respect for life, do you really want to meet them in person?
In any case, portable cell towers are quite common-- I see them at concerts from time to time. I would imagine that after the enemy has wasted an enormous amount of time and smart ordinance blasting your fixed cell network to bits, you could simply turn a couple of hundred mobile transmitters on and start over again.
The bigger weakness is, how long do you think that cellular network is going to continue working once the USAF finds out it is vectoring in the interceptors?
I can only assume that the usefulness of this system would be limited to early-warning detection. Once an attacker starts throwing bombs at a country's infrastructure, they've given up surprise-- which has to make up 50% of the benefit afforded by Stealth technology.
if it was a case of price fixing by retailers there might be something worth investigating... but I doubt thats the case
I assume you remember this, but there was a recent court case where the RIAA had to defend itself against price fixing charges. Essentially, the recording industry offers "promotional" money to record stores in exchange for inflated retail prices on all CDs. If a store won't comply, it loses the money.
To me this just seems like politicians playing nice with consumer groups while keeping everything the same.
Region coding is a major thorn in the side of the RIAA, considering they're going to have to defend the practice in court case after court case for the next few years. They certainly don't need anyone drawing attention to it.
I believe this whole problem could be solved with compulsory licensing for distribution. Record/movie companies would still make money, but they wouldn't be able to use their copyright in such an anti-competitive way.
Note that these machines contain only unclassified data. While that information might certainly be useful if you aggregate enough of it and cross-reference, overwriting the data (one time or hopefully many) is probably enough to reduce that threat-- it's unlikely that "the enemy" will expend the resources necessary to acquire and probe thousands of machines for trace magnetic information without being noticed.
Slashdot Mirror
That's partially true, but it's a bit unfair. Without the GPL or some sort of free software license, you can't use those people's code at all (ie, it's theirs, you can only distribute it if they explicitly agree to license it to you.) From a company's point of view, the ideal situation is to use a license that's slightly less restrictive than the GPL, but then how do you get people to give back their source code changes when they distribute, as the GPL does? Some of them will still probably give you the source, but you pretty much just have to hope for it (big companies that are throwing lots of support behind a popular project can get away with this, but smaller companies probably can't.)
The GPL is non-exclusive, so if you could convince all of the contributors to additionally give you closed-source rights to their contributions (say, when they check their code into your tree), you could still release the project separately under a non-GPL license. Again, this doesn't preclude anyone from also distributing the same code under the GPL.
There's been a certain amount of talk about adapting EZPass technology (automatic toll-paying tags in NY/NJ/DE) to catch speeders. Implementing this would be a snap; just measure the time a car takes between two tollbooths and mail out a ticket. Practically, it would be a political nightmare, and would very likely result in the speed limit rapidly being raised to 75MPH+.
I've driven the NJ Turnpike a few times (speed limit=55MPH), and I can assure you that only a very small minority of the state's drivers actually pay attention to the posted limits. If an automatic system like this were implemented, half the drivers in the state could rack up enough points for a license suspension, all inside a week. I'd hate to be the politician who authorized that particular course of action; you can guarantee there would be some changes made, and fast.
Well, I'm not sure I do either :) It's just a theory, and while I believe there's probably some truth to it, it's fairly diabolical to imply that corporations are colluding to deliberately wipe out small online content companies. On the other hand, it's certainly not implausable; remember, when companies this big exist, such tactics are no longer "conspiracies", they're simply corporate policies.
Big companies aren't any happier when they lose money than small companies.
It bears mentioning that the advertising rates for the most trafficked sites (AOL, for instance) haven't plummeted the way the smaller sites' banner ad rates have. So these companies aren't losing quite as much money as their competitors. Also, think of how much money a company like Microsoft is willing to invest in something like Internet Explorer, simply to corner a related market a few years down the line. Big companies that focus on nothing but the short-term reward are the exception.
And it's not like the ad money that used to be spent on the web has shifted to print and broadcast media. It has simply dried up as companies have cut back or gone out of business.
Well, that's certainly partially true. But the traditional media companies are still getting the lion's share of the (admittedly smaller) pie. It's not particularly their fault that the economy's slowed down along with advertising spending.
To my mind, it involves the intent of the "speaker", not the programmer. No matter what purpose the programmer intended for the code, if I give it out with the explicit purpose of dissemination useful information (and what is expression, really?), I should be protected. The constitution doesn't say anything about "functional" speech, mind you, and such classification should be reserved only for cases where it can be shown beyond a shadow of a doubt that the purpose of the speech was functional, and not-- as 2600 clearly intended-- to disseminate information. It would probably be so difficult to make this determination that speech would only rarely be classified as purely functional. And I'm alright with that; the constitution intended to protect speech and expression and we should always err to the safe side of things.
This is why I blame Big Media. They're not exactly taking an active role in reducing online ad revenue, they're simply promoting the crap out of their print and broadcast advertising to the detriment of the online equivalent. These companies have a vested interest in reducing competition; they can afford to lose millions of dollars maintaining essentially profitless sites, until they are the only major players (especially if they're just shuttling that revenue over to their offline publications.) Wait a couple of years and watch the renaissance in online advertising rates that occurs when AOL/TW and Viacom start throwing their weight around.
PS The above is one those paranoid anti-corporate predictions that almost always come true.
The only way to seriously beat up an OS is to find people who are willing to spend their every waking minute trying to break it. These people are generally paid for their effort, or else they stand to gain something if they succeed. MS may be hiring such people, I'm just not convinced that they're hiring enough. And your average beta-tester-with-a-job doesn't cut it.
This is certainly true, but doesn't go very deep. Why don't those sites make money? They've received millions and millions of visitors-- traffic was rarely a problem. One of the primary reasons some of these companies are failing is their inability to make decent rates off of online advertising. And who controls the rate structure?
The online advertising crash has often been blamed on low click-through rates. Perhaps that's just a silly standard to apply. Traditional media companies make tons of money slapping glossy ads onto dead trees with no percievable response, and why is that? Because the ad industry is comfortable with that way of doing things (or in some cases, the ad industry is owned by those media companies.)
There are certainly plenty of other reasons why these sites have had trouble-- bad management, over-investment, etc. But it's disingenuous to say that the old-fashioned big-media economy had nothing to do with it.
Unless MS is specifically recruiting thousands of beta-testers just to hack the security, they're not going to fix the important holes. I'm sure they're doing a certain amount of this sort of testing, but it clearly wasn't enough for their previous OSes. In any case, the sort of beta testing you suggest is generally not where you would expect to discover most of your security flaws. I really hope MS knows this.
Hmm. I'm thinking of the huge number of NT hacks that weren't preventable until MS threw together a patch. Many of these hacks couldn't be avoided simply by securing the box. I really hope Win2K doesn't have any of those, but how the hell do we know? It's a lot of code, and people are only now beginning to peck at it.
Well, duh. Closed source doesn't necessarily mean that nobody ever sees the code. In any case, I'm sure the NSA could get access to any source code they want, if they put enough pressure on. What closed source does mean is that they can't just make changes and hand them out. Everything they want to fix has to go back to Microsoft where it'll be slowly mixed in with the next set of 'enhancements'. The resulting upgrade, coming out a year or two down the line, might be just as vulnerable as the current version. The same thing happens with Linux, but a) it's possible to fork the code in order to avoid this problem and b) everyone can examine the source to see if anything broke.
There's no shortage of Java work. However, it's not all in the most interesting areas. My company works with both Java and C++, depending on the particular area we're working on at the moment, and we're constantly turning away people with only Java experience. Some companies only work with Java, so they're ideal for those kinds of programmers. But larger companies with more diverse projects tend to look for people with multiple skills.
Again, if it were easy to go from Java to C++, this wouldn't be a problem. But for a programmer who has no experience with C++ or even C (ie, they aren't experienced with memory manipulation and the wonders of header files), making such a transfer is difficult, and most companies don't want a programmer learning these concepts in the middle of a project (can you say memory leak?) It's much more reasonable to expect a smooth transfer from Java/C++ to Perl or Python than from Java to C++.
Second, the point of a CS education should not be to learn the most popular language, but to learn the concepts of programming so that you can easily pick up new languages as required.
Absolutely. People have lots of time to pick up new languages after they've taken an intro class. I would withdraw all of the arguments I just made if I felt that Java were also the best introductory language. But I don't, and the reason again comes down to the fact that Java lacks certain basic concepts that you become accustomed to when you learn C or C++, concepts that you should understand long before you mess with OO. I also find that the Java intro classes I've seen have a hard time staying away from OO early on, and this confuses people when they're still trying to master subroutines.
The best course, as far as I'm concerned, is to begin with C because both Java and C++ more or less rip off C syntax. Learn the basic concepts, including pointers and work your way up to abstract data structures (which is more or less what OO really is.) Then go on to Java or C++ as you see fit.
Especially considering Yahoo went to the effort and expense of a) shutting down the auctions and b) dragging the whole thing back to court.
My bet is that the French gov't did not give them any real guarantee that they wouldn't be held liable at some later point, so rather than continuing with a cloud of uncertainty over their heads, they took the course of action that they did.
I think multiple language experience is useful, but I'm afraid that what you describe might be similar to inserting short stints of Basque and Catalan into the first semesters of an intro Spanish course.
Java's like candy-- it takes care of so much of the annoying stuff that C++ programmers get used to dealing with. While C++ sucks in this way, if you're looking for a high-performance language you can't rely on a system with garbage collection and run-time linking. Not to mention that truly understanding what's going on with memory and pointers, while posing a steeper learning curve, really helps you understand how the machine deals with your programs.
Actually, if you get rid of the analog channels and go exclusively to digital transmission, you can pack a considerable number of on-demand channels onto the coax. Each 6Mhz band can currently support 8-10 channels on standard digital cable networks (the lower channels are slightly more limited.) If the "live" channels are multicast, you can further reduce the bandwidth they require. Of course, this assumes that you're completely revise the format in which digital cable is transmitted; fortunately, these changes would mainly take place at the head end and the client, so they wouldn't require massive system-wide upgrades.
Some of the calculations I've seen give each household an average of between one and two on-demand connections apiece, if you assume a reasonably sized local loop (300-400 homes.) If your local loop has too many subscribers on it, as some networks tend to (think the former TCI empire), you're out of luck.
And don't forget that the "take rate" for these services is not going to be high in the beginning, so 30-40 channels per neighborhood might be acceptable.
The following is probably either obvious or completely wrong, but let me try it:
If the browser shows an icon for a secure web page, then you've got an SSL connection. SSL is end-to-end, correct? The Proxy can't mount what is effectively a man-in-the-middle attack without your browser knowing it (and therefore failing to display a secure icon.) The situation you mention above would be obvious to any user who knows enough to look in the corner of their browser window (or better yet, check the Security properties for the page), correct? And if the icon isn't there, they can't really claim to be using a secure connection (hopefully most porn-surfers know this much.)
Since the GET requests you make over an SSL connection are all encrypted, the only thing the proxy would see is which host you're trying to connect to (Safeweb over and over again), not the full request URL (eg /cgi-bin/redirect?url=www.stinkypanties.com... etc).
A better way to get around corporate monitoring of browsing habits is to convert the IP address of the site you want to visit into a decimal number. That should confuse your typical admin enough
I would imagine that a half-serious porn-monitoring operation would log the IP addresses and use a reverse lookup (or at least, look at a table of known porn sites.)
The difference is that we (the "Slashdot crowd") don't have the capability to declare things illegal. All we can do is make a lot of noise and hope the powers that be will consider our opposition before they head down a potentially abusive path.
If the MPAA had restricted its complaints about DeCSS to a few web pages and newspaper articles rather than heading to court to flex their newly purchased laws, I don't think we'd all be quite as annoyed with them.
Yes, and another nice independent thinker like Thomas would be great too.
If you're talking about offensive weapons, or weapons of mass destruction, then I completely agree with you. Any nation that would hide such things among the civilian population must be stopped. On the other hand, when you're talking about a defensive system like a radar network (does a radar network "strike"?), you can't be quite as moralistic about it. There may be good reason to knock it out, but that doesn't justify the slaughter of many thousands of people.
If a mission cannot be accomplished without a massacre, the military should rethink the mission; if it's a life-or-death struggle, then I suppose there's no choice. If it's a "police action" or a political move, it may not justify killing many innocent people.
This is no different than the decision police officers make in a hostage situation. While there are situations in which the death of soe hostages is almost impossible to avoid, you do not invite those situations-- no matter how much it inconveniences you.
A cellular network is not a weapon, and even if it is used as a radar system, there has to be a better way to take it out than by throwing bombs at the individual cell towers.
The only way to make sure that governments don't do this is to systematically destroy such installations, preferably with smart weapons to moderate (though not eliminate) civilian casualties
If the Serbians/Iraqis had located thousands of small, passive radar systems every couple of blocks, could we have possibly justified the kind of slaughter that would result from taking them out (even with smart bombs?) The US has changed the way it makes war since Vietnam (at least, the way it makes war in public.) I'm not saying that civilians don't already die in NATO bombing raids, but this would be a couple of orders of magnitude worse.
The military has learned that publicly slaughtering thousands of civilians is not acceptable in the sorts of engagements we've been having. And even if you do use Smart weapons, you're still going to have a hell of a time not killing many thousands of people. These "installations" you speak of are often attached to civilian buildings in heavily populated areas-- destroying them is not like targeting a factory and knowing there may be some collateral damage. It's like going after fleas with a shotgun (or at least, a rifle.)
Cell towers are tiny things that are generally located in populated areas. The bombing pattern you suggest would make swiss cheese out of an undefended civilian center. Any government that would authorize that sort of attack is a monster, and your military should be doing everything possible to make sure that sort of attacker does not emerge victorious, no matter how many lives it costs. With such respect for life, do you really want to meet them in person?
In any case, portable cell towers are quite common-- I see them at concerts from time to time. I would imagine that after the enemy has wasted an enormous amount of time and smart ordinance blasting your fixed cell network to bits, you could simply turn a couple of hundred mobile transmitters on and start over again.
I can only assume that the usefulness of this system would be limited to early-warning detection. Once an attacker starts throwing bombs at a country's infrastructure, they've given up surprise-- which has to make up 50% of the benefit afforded by Stealth technology.
Not if the Chinese bootlegs are ripped from a US DVD.
I assume you remember this, but there was a recent court case where the RIAA had to defend itself against price fixing charges. Essentially, the recording industry offers "promotional" money to record stores in exchange for inflated retail prices on all CDs. If a store won't comply, it loses the money.
To me this just seems like politicians playing nice with consumer groups while keeping everything the same.
Region coding is a major thorn in the side of the RIAA, considering they're going to have to defend the practice in court case after court case for the next few years. They certainly don't need anyone drawing attention to it.
I believe this whole problem could be solved with compulsory licensing for distribution. Record/movie companies would still make money, but they wouldn't be able to use their copyright in such an anti-competitive way.
Note that these machines contain only unclassified data. While that information might certainly be useful if you aggregate enough of it and cross-reference, overwriting the data (one time or hopefully many) is probably enough to reduce that threat-- it's unlikely that "the enemy" will expend the resources necessary to acquire and probe thousands of machines for trace magnetic information without being noticed.