You're right that a game taking 30 seconds to start is not such a performance hit that people will submit to built-in, Windows based Trusted Computing rather than use an open source version if one ever exists. But I see little evidence that a Linux version of the tools does or will exist. Remember that the Trusted Computing can be built into the DVD and mapped into the vendor's specific game or DVD player. Unless you can successfully use the vendor's application in your open source world, and make Trusted Computing work inside that open source basis, you're screwed.
And if opening and saving encrypted documents takes 10 seconds rather than 2, and if playing the authorized content of a DVD cannot be done in real time or in real-time and high resolution with open source tools because they don't have the built-in decryption available, we'll see quite a lot of pressure to use the closed source tools.
I agree with your analysis. Protecting boot loaders from viruses and preventing bootable CD's from being able to boot a secured system are very useful applications for something like Trusted Computing.
However, take a good look at all those click-through tools for setting a new boot loader or manipulating the TPM module that you've described. They're Windows based, closed source, and Trusted Computing authenticated. That means that to add a new bootloader, you have to use the Windows controlled to load, guess what? A Windows bootloader!
As near as I can tell, Trusted Computing cannot be legally reverse engineered due to the DMCA. So creating Linux tools for accessing the boot loader controls or other DRM managed hardware becomes extremely difficult.
Please separate your ideas so they can be addressed one at a time, instead of all jumbled up.
The performance hit should be very slight with this integrated DRM. That's fundamental to the whole design of Trusted Computing: if consumers see a real performance hit, they will avoid the DRM controlled media or software.
There are serious trade-offs in having integrated encryption vs. ease of recovering documents. Trusted Computing has decided on, and implemented, a system where all keys are either centrally managed or can be created only with a centrally managed signature. This eases recovering lost keys, but it means a huge level of trust must exist in the central key authority: they can, and history shows that they will, turn over the keys for their own political or business reasons. And there really is no need to have such a central authority if your goal is user protection: the use and spread of PGP has shown how to do that.
No, the centralized key management is clearly focused on DRM. And the movie and software companies have repeatedly demonstrated their willingness to restrict you from legal use of their material. If they weren't, they'd have created or permitted a Linux DVD player years ago and avoided the whole libdvdcss mess where they sued people and got a portable DVD decryption library outlawed in numerous countries, a library that has been expressed as 2 lines of Perl code as a demonstration of how easy it is to do.
It's a very good question: I know that the Trusted Computing system is able to check boot loaders for signatures, and prevent or permit them from being used. But I bet that VMware is looking very hard at exactly how to manage this.
And as such usage grows, I expect to see a growing market for gelatin: the old Gummi Finger experiments, described in detail here on Slashdot, explain in clear language how to make a fake fingerprint that will work more than 85% of the time.
There are several benefits of Trusted Computing for average users, benefits that are possible by other means but have never gotten the support to be implemented widely and become a standard practice. These include the ability to really authenticate patches for your operating system and software, so when you download your patches you haven't just automatically clicked on some flakey SSL key somewhere and accepted a Windows patch from Rootkits-R-Us.
Second, it does provide a robust and hopefully well integrated set of tools for encrypting and authenticating email and other documents. This is very useful for business contracts or sending personally encrypted email, and will wind up built right into Microsoft's email clients.
Third, getting real encryption into broad use has been a nightmare for Windows and other largely deployed OS's because of US laws about selling encryption tools overseas: Since the US government has bought into this project, they won't be blocking its sale overseas or otherwise interfering in its development. Those regulations have repeatedly been a huge burden for software vendors.
And you seem to have missed that while Microsoft has sworn on a stack of bibles that right now, they have no intention to mandate the use of Trusted Computing in their operating system or in Microsoft Office, it's only a matter of time before they provide features that only work with Trusted Computing, and then mandate it for new products. Expect it to be almost mandatory in all CPU's within five years, and for Microsoft OS's and core Windows software to require it within 10 years.
Oh, that's a different matter. The offering up your thumbprint to use, say, Photoshop, is for commercial artwork. Stealing movies before they get released and selling cheap DVD's of them in China is a huge deal to Hollywood, and this is aimed at forcing the artists to really authenticate themselves. Getting them to actually use passwords and use them properly has proven quite difficult.
You're missing a lot of details about this software. It's closed source, and a violation of the DMCA to reverse engineer it. That means writing an open source version of the encryption/decryption tools is going to be a nightmare.
Second, running it at the OS level instead of the hardware level of the built-in features of the Intel CPU's is going to really slow it down: that will probably hurt performance a lot of open source versions of the Trusted Computing tools, even if they're legally created.
Third, the next logical stage of Trusted Computing is hardware locking: motherboards that won't load unsigned boot loaders, or won't access DVD drives or hard drives without being authenticated with Trusted Computing licenses to be held by OS distributions or DVD drive and software vendors. This can be used to block open source operating systems from even booting, or to prevent Trusted Computing managed DVD drives from being able to read DVD's that have Trusted Computing signed DVD's in them without a Trusted Computing signed media player.
It's very nasty, and it's at the core of why Microsoft and Hollywood are collaborating so well in this project.
The designers of Trusted Computing have thought very hard about how to transfer and manage semi-private keys: I've seen some good presentations on it, and it has a public/private key handling capability, much like PGP. One problem is that to load new keys, those new keys have to be signed by an existing key owner. And to do that, you basically need to buy them from Microsoft, who can and will put in back doors.
Second, Microsoft seems to be keeping the master keys, and you can be quite certain that the US federal government is keeping additional keys or a copy of the master keys. And the US government has demonstrated repeatedly that there's always some idiot willing to violate other people's privacy for political or criminal goals: sometimes that idiot is the president, such as Nixon in the Watergate scandal or Bush with the recent discovery of warrantless wire taps. The deliberate difficulty of creating your own private keys on the fly, for private users, means that both Microsoft and the US government will always have access to your most private materials.
It's certainly possible to create authentication and encryption systems where there need be no law enforcement access, or master keys. PGPphone did this many years ago, and PGP does it now. But it seems a requirement of all corporate grade systems now that the manufacturer cooperate with federal demands for backdoor access: notice how the Clipper Chip was pulled from the market when it was discovered how to use it with your own private keys.
This has limits. Spammers, at least criminal spammers, are currently using a lot of "zombied" machines. They'll happily rent time on those machines to each other, and your home machines will be used to send everyone else's spam. So this micro-payment system, and the "first 1000 free" proposals, simply make spam slightly harder. The workarounds are already in place and broadly used.
Actually convicting phishers of wire fraud, however, would be a big step towards eliminating at least that particular spam. The phishers operate for weeks, even months at a time with fake addresses and fake websites because their ISP's refuse to act against obviously criminal behavior, and law enforcement can't be bothered to get the subpoenas, check the logs and the billing records, or check the bank records to follow the money. It's a related problem to spamming, and it's much easier to convict people for.
These ways of affecting a host are not very complex. But there was a fascinating SF story years ago, about a behavior-affecting prion that encouraged people to give blood. This created a vector for transmission, but because people felt this need to give blood they seemed to become better for it and feel they were doing something good in their lives, so they felt somewhat morally better and became better people for it.
This isn't always why layoffs happen. I've seen companies developing and releasing a good product, project market growth over the next few years that justify bringing some new staff, then have to cut back wildly as the market evaporates. It happened to a lot of good companies in the dotcom crash, and in the wake of 9/11, even if the companies were well run and actually doing good work.
When people won't buy even good products, or stop traveling out of justified fear, you have to scale back your staff. The trick is to dump, without a big golden parachute, those "big idea" people who thought that all that expansion was a good idea and spent all the investor's money on office chairs for executives.
This is exactly right. The problem of hardware that can run only signed binaries is at the core of the Trusted Computing initiative, formerly called Palladium and confusingly renamed. Trusted Computing is clearly designed to apply "digital rights management" at the core of new hardware and operating systems, to create DVD's, CD's, files, file systems, and boot records that can only be run by the approved software and hardware combination, with whatever built-in limitations that setup contains.
The new GPL tries to address these risks somewhat, by preventing GPL licensed tools from being used to cut off other software from access to the information or to the system. This is especially true for boot loaders and device drivers: the Trusted Computing tools can easily be further integrated into the BIOS and the CPU itself to prevent non-signed boot loaders from being able to even boot a computer with a non-Trusted-Computing signed boot loader, or to access a hard drive or DVD drive to boot from without a Trusted-Computing signed OS.
Your claim that time didn't exist "outside the universe" only applies if the matter in the Big Bang was the entire universe. There are plenty of theories that it was not: there are other theories that involve multiple simultaneously existing universes, the "multiverse", and the then there's the various "Cyclic" universe models that say at some future point, if the universe has enough matter in it, the universie will collapse back to a single object and again explode.
It's a fascinating subject for discussion and analysis: please don't throw out the language of "before the Big Bang" because it doesn't apply to a specific version of the theory.
I've seen the claims you mention. The claim that the rate of expansion of the universe is increasing is interesting, but frankly impossible to tell directly. We don't have a long enough timeline to directly measure it, and the precise distance and age and nature of objects billions of light years away is very difficult to determine. I don't think we can trust the claim that the rate of expansion is increasing yet: it does lend some credence to the Steady State theory of the universe, but anyone who's done high school laboratory science can show you how very, very small changes in the original measurements of anything subtle can lead you to wild differences in conclusions.
The results certainly do need to be understood, but I wouldn't call them conclusive disproof of anything yet.
And I have friends without degrees who get their resumes discarded even before job interviews if they make it clear on their transcript. They've learned to fudge: their qualifications are excellent, but they wound up doing other things their final year (such as having kids, founding start-ups, or building houses for the poor).
But not having that degree often gets the HR person to take you off the list of candidates: it's a real problem for some very skilled people.
There's a lot of compelling evidence for some sort of Big Bang. The universe is clearly expanding: further galaxies show a distinct "red shift", a change in the light coming from that can be measured and shows how fast away they're going. Their distance is estimated by looking for bright, measurably bright objects like nova or supernova and extrapolating their distance from the brightness.
The further away they are, the faster they seem to be going. That hints at some sort of event, roughly 10 billion years ago, that forced them all away and in fact created these objects. That's coupled with a background microwave radiation we'd expect from a universe at about 3 degrees Kelvin, as if the matter that spread out has cooled down to about that average temperature.
Other theories, such as the "Cyclic" theory assume that the universe keeps exploding and contracting, but it's hard to detect enough matter in the universe to allow it to re-contract from gravitation. Or the "Steady-State" theory assumes that the matter, the universe itself somehow keeps regenerating itself over time: some weird quantum ideas describe universes where matter forms from vacuum, but those theories don't predict the actual measurements very well.
So there are 3 common theories: the Big Bang explains the existing evidence well, but leaves people wondering "what happened before" and "what will happen later". Like gravity or light, the basic facts seem well explained, but there are weird details that do require more work to really understand.
Your metaphor is confusing. I thought the mastodon was an arctic creature, in the cold wastes of the world, who benefitted and increased their range of influence from having the Sun's influence lessened.
I spoke with Mormons and former Mormons. It's hard to separate out the facts of inbreeding from the racial slurs, I admit. The isolation of the original Mormon group for some generations in Utah, founded by a cult movement driven across America by local communities unwilling to accept them in each State as they gathered, and their very small racial diversity (only white members (with basically no black, Asian, or native American members whatsoever), and the sexual policies of multiple wives (which meant only their local leaders need join) meant they have serious inbreeding issues.
Their intense interest in genealogies is a big help in reducing the problem, by helping avoid first cousins marrying, and it's very helpful indeed for genealogical research. And they believe in big families, so the group has spread out quite a bit from their relatively small origins, but if you look at those family trees, their family trees all seem be tightly interwoven for those first 3 generations. You're going to have common inherited diseases in a group like that.
They do other world-class research in problems common among their heavily inbred population: some blood disorders and oddly enough, deafness get quite a lot of funding and research coming out of Utah.
It's not that the web sites are necessarily unscrupulous. Many of them are having spyware and adware installers put in due to poor security of the webserver, and others are due to unscrupulous web-site consultants slipping in their favorite spyware to serve purposes that have nothing to do with the site they are setting up. And the broad vulnerabilities of a poorly set up webhosting site can load literally thousands of websites with spyware.
NASA does have enough to deal with, but political and even religious involvement in scientific research is built into human history. Since early religious interpretation of the stars, used to predict when to plant and when to reap crops, they've been interwoven. When religious leaders make statements about their deity bringing famine to the ungodly, and some smart aleck points out that overgrazing by the local baron's goats have changed the watercourses that supply the farmers, religion and fincance and politics have been interwoven. Expecting them not to interact is like expecting light to be a wave: it looks that way in certain conditions, but if you examine it too closely it can surprise you.
Now, we can expect a group like NASA who get funded to do research and exploration to keep the science as rigorous and well-grounded as possible: when the politics gets in the way of that so blatantly and so deliberately aimed as to cause fraud, the politics should lose. But you can certainly expect the politicians to react to this by cutting funding where possible: that's part of the price of discovering an unpleasant truth.
Please tell me you're kidding. The spike trips are very dangerous at high speed, and require you to get in front of the fleeing car.
No, my concern for this technology is its ubiquitous use to tag innocent people. If all a cop has to say to explain why they're tracking you is "oh, we missed this other guy and hit your car with that tracer", then we're just plain going to see abuses of the technology. And given current federal policy of secret illegal warrantless wiretapping and secret prisons in foreign countries where torture is allowed, there's really no place to go with complaints about such abuses unless you have the resources to get it pursued by your state police, or to get Congress interested.
Hey, I prefer to eat vegan. Sauteed, maybe with a cream sauce. It's easy to catch them if lay some alfalfa and wheat germ out on a sample tray at your local health food co-op.
First, forget round robin DNS. It simply doesn't work. You cannot rely on the DNS clients to select the round-robin IP addresses in order except under very rare circumstances.
Second, if it absolutely has to work, you can rent services from Akamai, who have a huge and under-used network for exactly this service. Let them deal with distributing the load from one feed to thousands of servers worldwide: they think it will make them lots of money. They will play more clever DNS games than merely round-robining.
Third, forget streaming live if you can. Set up a download site for a Bittorrent feed. If lots of people want to see the download, let them distribute the load among themselves.
Fourth, forget streaming Windows Media. Use Quicktime or Real if you have to stream: they both put much less load on your servers and your server can handle a lot more clients.
Slashdot Mirror
You're right that a game taking 30 seconds to start is not such a performance hit that people will submit to built-in, Windows based Trusted Computing rather than use an open source version if one ever exists. But I see little evidence that a Linux version of the tools does or will exist. Remember that the Trusted Computing can be built into the DVD and mapped into the vendor's specific game or DVD player. Unless you can successfully use the vendor's application in your open source world, and make Trusted Computing work inside that open source basis, you're screwed.
And if opening and saving encrypted documents takes 10 seconds rather than 2, and if playing the authorized content of a DVD cannot be done in real time or in real-time and high resolution with open source tools because they don't have the built-in decryption available, we'll see quite a lot of pressure to use the closed source tools.
I agree with your analysis. Protecting boot loaders from viruses and preventing bootable CD's from being able to boot a secured system are very useful applications for something like Trusted Computing. However, take a good look at all those click-through tools for setting a new boot loader or manipulating the TPM module that you've described. They're Windows based, closed source, and Trusted Computing authenticated. That means that to add a new bootloader, you have to use the Windows controlled to load, guess what? A Windows bootloader! As near as I can tell, Trusted Computing cannot be legally reverse engineered due to the DMCA. So creating Linux tools for accessing the boot loader controls or other DRM managed hardware becomes extremely difficult.
Please separate your ideas so they can be addressed one at a time, instead of all jumbled up.
The performance hit should be very slight with this integrated DRM. That's fundamental to the whole design of Trusted Computing: if consumers see a real performance hit, they will avoid the DRM controlled media or software.
There are serious trade-offs in having integrated encryption vs. ease of recovering documents. Trusted Computing has decided on, and implemented, a system where all keys are either centrally managed or can be created only with a centrally managed signature. This eases recovering lost keys, but it means a huge level of trust must exist in the central key authority: they can, and history shows that they will, turn over the keys for their own political or business reasons. And there really is no need to have such a central authority if your goal is user protection: the use and spread of PGP has shown how to do that.
No, the centralized key management is clearly focused on DRM. And the movie and software companies have repeatedly demonstrated their willingness to restrict you from legal use of their material. If they weren't, they'd have created or permitted a Linux DVD player years ago and avoided the whole libdvdcss mess where they sued people and got a portable DVD decryption library outlawed in numerous countries, a library that has been expressed as 2 lines of Perl code as a demonstration of how easy it is to do.
It's a very good question: I know that the Trusted Computing system is able to check boot loaders for signatures, and prevent or permit them from being used. But I bet that VMware is looking very hard at exactly how to manage this.
And as such usage grows, I expect to see a growing market for gelatin: the old Gummi Finger experiments, described in detail here on Slashdot, explain in clear language how to make a fake fingerprint that will work more than 85% of the time.
There are several benefits of Trusted Computing for average users, benefits that are possible by other means but have never gotten the support to be implemented widely and become a standard practice. These include the ability to really authenticate patches for your operating system and software, so when you download your patches you haven't just automatically clicked on some flakey SSL key somewhere and accepted a Windows patch from Rootkits-R-Us.
Second, it does provide a robust and hopefully well integrated set of tools for encrypting and authenticating email and other documents. This is very useful for business contracts or sending personally encrypted email, and will wind up built right into Microsoft's email clients.
Third, getting real encryption into broad use has been a nightmare for Windows and other largely deployed OS's because of US laws about selling encryption tools overseas: Since the US government has bought into this project, they won't be blocking its sale overseas or otherwise interfering in its development. Those regulations have repeatedly been a huge burden for software vendors.
And you seem to have missed that while Microsoft has sworn on a stack of bibles that right now, they have no intention to mandate the use of Trusted Computing in their operating system or in Microsoft Office, it's only a matter of time before they provide features that only work with Trusted Computing, and then mandate it for new products. Expect it to be almost mandatory in all CPU's within five years, and for Microsoft OS's and core Windows software to require it within 10 years.
Oh, that's a different matter. The offering up your thumbprint to use, say, Photoshop, is for commercial artwork. Stealing movies before they get released and selling cheap DVD's of them in China is a huge deal to Hollywood, and this is aimed at forcing the artists to really authenticate themselves. Getting them to actually use passwords and use them properly has proven quite difficult.
You're missing a lot of details about this software. It's closed source, and a violation of the DMCA to reverse engineer it. That means writing an open source version of the encryption/decryption tools is going to be a nightmare.
Second, running it at the OS level instead of the hardware level of the built-in features of the Intel CPU's is going to really slow it down: that will probably hurt performance a lot of open source versions of the Trusted Computing tools, even if they're legally created.
Third, the next logical stage of Trusted Computing is hardware locking: motherboards that won't load unsigned boot loaders, or won't access DVD drives or hard drives without being authenticated with Trusted Computing licenses to be held by OS distributions or DVD drive and software vendors. This can be used to block open source operating systems from even booting, or to prevent Trusted Computing managed DVD drives from being able to read DVD's that have Trusted Computing signed DVD's in them without a Trusted Computing signed media player.
It's very nasty, and it's at the core of why Microsoft and Hollywood are collaborating so well in this project.
The designers of Trusted Computing have thought very hard about how to transfer and manage semi-private keys: I've seen some good presentations on it, and it has a public/private key handling capability, much like PGP. One problem is that to load new keys, those new keys have to be signed by an existing key owner. And to do that, you basically need to buy them from Microsoft, who can and will put in back doors.
Second, Microsoft seems to be keeping the master keys, and you can be quite certain that the US federal government is keeping additional keys or a copy of the master keys. And the US government has demonstrated repeatedly that there's always some idiot willing to violate other people's privacy for political or criminal goals: sometimes that idiot is the president, such as Nixon in the Watergate scandal or Bush with the recent discovery of warrantless wire taps. The deliberate difficulty of creating your own private keys on the fly, for private users, means that both Microsoft and the US government will always have access to your most private materials.
It's certainly possible to create authentication and encryption systems where there need be no law enforcement access, or master keys. PGPphone did this many years ago, and PGP does it now. But it seems a requirement of all corporate grade systems now that the manufacturer cooperate with federal demands for backdoor access: notice how the Clipper Chip was pulled from the market when it was discovered how to use it with your own private keys.
This has limits. Spammers, at least criminal spammers, are currently using a lot of "zombied" machines. They'll happily rent time on those machines to each other, and your home machines will be used to send everyone else's spam. So this micro-payment system, and the "first 1000 free" proposals, simply make spam slightly harder. The workarounds are already in place and broadly used.
Actually convicting phishers of wire fraud, however, would be a big step towards eliminating at least that particular spam. The phishers operate for weeks, even months at a time with fake addresses and fake websites because their ISP's refuse to act against obviously criminal behavior, and law enforcement can't be bothered to get the subpoenas, check the logs and the billing records, or check the bank records to follow the money. It's a related problem to spamming, and it's much easier to convict people for.
These ways of affecting a host are not very complex. But there was a fascinating SF story years ago, about a behavior-affecting prion that encouraged people to give blood. This created a vector for transmission, but because people felt this need to give blood they seemed to become better for it and feel they were doing something good in their lives, so they felt somewhat morally better and became better people for it.
It was fascinating reading.
This isn't always why layoffs happen. I've seen companies developing and releasing a good product, project market growth over the next few years that justify bringing some new staff, then have to cut back wildly as the market evaporates. It happened to a lot of good companies in the dotcom crash, and in the wake of 9/11, even if the companies were well run and actually doing good work.
When people won't buy even good products, or stop traveling out of justified fear, you have to scale back your staff. The trick is to dump, without a big golden parachute, those "big idea" people who thought that all that expansion was a good idea and spent all the investor's money on office chairs for executives.
This is exactly right. The problem of hardware that can run only signed binaries is at the core of the Trusted Computing initiative, formerly called Palladium and confusingly renamed. Trusted Computing is clearly designed to apply "digital rights management" at the core of new hardware and operating systems, to create DVD's, CD's, files, file systems, and boot records that can only be run by the approved software and hardware combination, with whatever built-in limitations that setup contains.
The new GPL tries to address these risks somewhat, by preventing GPL licensed tools from being used to cut off other software from access to the information or to the system. This is especially true for boot loaders and device drivers: the Trusted Computing tools can easily be further integrated into the BIOS and the CPU itself to prevent non-signed boot loaders from being able to even boot a computer with a non-Trusted-Computing signed boot loader, or to access a hard drive or DVD drive to boot from without a Trusted-Computing signed OS.
Your claim that time didn't exist "outside the universe" only applies if the matter in the Big Bang was the entire universe. There are plenty of theories that it was not: there are other theories that involve multiple simultaneously existing universes, the "multiverse", and the then there's the various "Cyclic" universe models that say at some future point, if the universe has enough matter in it, the universie will collapse back to a single object and again explode.
It's a fascinating subject for discussion and analysis: please don't throw out the language of "before the Big Bang" because it doesn't apply to a specific version of the theory.
I've seen the claims you mention. The claim that the rate of expansion of the universe is increasing is interesting, but frankly impossible to tell directly. We don't have a long enough timeline to directly measure it, and the precise distance and age and nature of objects billions of light years away is very difficult to determine. I don't think we can trust the claim that the rate of expansion is increasing yet: it does lend some credence to the Steady State theory of the universe, but anyone who's done high school laboratory science can show you how very, very small changes in the original measurements of anything subtle can lead you to wild differences in conclusions.
The results certainly do need to be understood, but I wouldn't call them conclusive disproof of anything yet.
And I have friends without degrees who get their resumes discarded even before job interviews if they make it clear on their transcript. They've learned to fudge: their qualifications are excellent, but they wound up doing other things their final year (such as having kids, founding start-ups, or building houses for the poor).
But not having that degree often gets the HR person to take you off the list of candidates: it's a real problem for some very skilled people.
There's a lot of compelling evidence for some sort of Big Bang. The universe is clearly expanding: further galaxies show a distinct "red shift", a change in the light coming from that can be measured and shows how fast away they're going. Their distance is estimated by looking for bright, measurably bright objects like nova or supernova and extrapolating their distance from the brightness.
The further away they are, the faster they seem to be going. That hints at some sort of event, roughly 10 billion years ago, that forced them all away and in fact created these objects. That's coupled with a background microwave radiation we'd expect from a universe at about 3 degrees Kelvin, as if the matter that spread out has cooled down to about that average temperature.
Other theories, such as the "Cyclic" theory assume that the universe keeps exploding and contracting, but it's hard to detect enough matter in the universe to allow it to re-contract from gravitation. Or the "Steady-State" theory assumes that the matter, the universe itself somehow keeps regenerating itself over time: some weird quantum ideas describe universes where matter forms from vacuum, but those theories don't predict the actual measurements very well.
So there are 3 common theories: the Big Bang explains the existing evidence well, but leaves people wondering "what happened before" and "what will happen later". Like gravity or light, the basic facts seem well explained, but there are weird details that do require more work to really understand.
Your metaphor is confusing. I thought the mastodon was an arctic creature, in the cold wastes of the world, who benefitted and increased their range of influence from having the Sun's influence lessened.
I spoke with Mormons and former Mormons. It's hard to separate out the facts of inbreeding from the racial slurs, I admit. The isolation of the original Mormon group for some generations in Utah, founded by a cult movement driven across America by local communities unwilling to accept them in each State as they gathered, and their very small racial diversity (only white members (with basically no black, Asian, or native American members whatsoever), and the sexual policies of multiple wives (which meant only their local leaders need join) meant they have serious inbreeding issues.
Their intense interest in genealogies is a big help in reducing the problem, by helping avoid first cousins marrying, and it's very helpful indeed for genealogical research. And they believe in big families, so the group has spread out quite a bit from their relatively small origins, but if you look at those family trees, their family trees all seem be tightly interwoven for those first 3 generations. You're going to have common inherited diseases in a group like that.
They do other world-class research in problems common among their heavily inbred population: some blood disorders and oddly enough, deafness get quite a lot of funding and research coming out of Utah.
It's not that the web sites are necessarily unscrupulous. Many of them are having spyware and adware installers put in due to poor security of the webserver, and others are due to unscrupulous web-site consultants slipping in their favorite spyware to serve purposes that have nothing to do with the site they are setting up. And the broad vulnerabilities of a poorly set up webhosting site can load literally thousands of websites with spyware.
NASA does have enough to deal with, but political and even religious involvement in scientific research is built into human history. Since early religious interpretation of the stars, used to predict when to plant and when to reap crops, they've been interwoven. When religious leaders make statements about their deity bringing famine to the ungodly, and some smart aleck points out that overgrazing by the local baron's goats have changed the watercourses that supply the farmers, religion and fincance and politics have been interwoven. Expecting them not to interact is like expecting light to be a wave: it looks that way in certain conditions, but if you examine it too closely it can surprise you.
Now, we can expect a group like NASA who get funded to do research and exploration to keep the science as rigorous and well-grounded as possible: when the politics gets in the way of that so blatantly and so deliberately aimed as to cause fraud, the politics should lose. But you can certainly expect the politicians to react to this by cutting funding where possible: that's part of the price of discovering an unpleasant truth.
Please tell me you're kidding. The spike trips are very dangerous at high speed, and require you to get in front of the fleeing car.
No, my concern for this technology is its ubiquitous use to tag innocent people. If all a cop has to say to explain why they're tracking you is "oh, we missed this other guy and hit your car with that tracer", then we're just plain going to see abuses of the technology. And given current federal policy of secret illegal warrantless wiretapping and secret prisons in foreign countries where torture is allowed, there's really no place to go with complaints about such abuses unless you have the resources to get it pursued by your state police, or to get Congress interested.
Hey, I prefer to eat vegan. Sauteed, maybe with a cream sauce. It's easy to catch them if lay some alfalfa and wheat germ out on a sample tray at your local health food co-op.
First, forget round robin DNS. It simply doesn't work. You cannot rely on the DNS clients to select the round-robin IP addresses in order except under very rare circumstances.
Second, if it absolutely has to work, you can rent services from Akamai, who have a huge and under-used network for exactly this service. Let them deal with distributing the load from one feed to thousands of servers worldwide: they think it will make them lots of money. They will play more clever DNS games than merely round-robining.
Third, forget streaming live if you can. Set up a download site for a Bittorrent feed. If lots of people want to see the download, let them distribute the load among themselves.
Fourth, forget streaming Windows Media. Use Quicktime or Real if you have to stream: they both put much less load on your servers and your server can handle a lot more clients.