This is already in development. It's the Trusted Computing initiative, formerly known as Palladium, and it's a very big security effort. The benefit of its features are high: on-board high-speed encryption and authentication, easily available to users. The danger is also high: this is supposed to integrate with Dirital Rights Management and provide hardware level control of access to DVD writers, hard drives, boot loaders, system kernels, and secure operations called from withing software. That means that unless you can get the autohrization and the money to buy a highly authorized key from, say, the Microsoft key provisioning service, you will have difficulty writing and especially publishing open source tools that access those features.
He may be cautious about his non-disclosure agreements, or not want his former very political co-worker to play political games on his own career. Cowardly, but completely understandable: helping protect sources from reprisals is one of the reasons for anonymous posting on Slashdot.
They built EL4 RPM's? Excellent. That means they'll run on CentOS (at www.centos.org) for those sites who are too cheap to buy RHEL or who pay an internal Linux support person instead of paying support fees to RedHat.
That's very odd. I went into a field I enjoyed, was underpaid for 20 years, and suddenly my field took off and money became very easy to earn.
Nevery underestimate the fiscal benefit of going into something you love and getting very, very, very good at it, as opposed to starting out with a better pay scale but loathing the work and thus never achieving greatness, or even quality.
Copyrighted and illegal to download are two very different things. What the EULA says and what copyright law actually says are also two wildly, wildly, wildly differnent things.
I'm not a lawyer, but since duplicating a CD that you've purchased for backup reasons has been defended in court successfully several times, and since a user still requires an authorization key to install the software and usually actively agrees to the license agreement, I think that the copyright violation of copying such a CD for a user who bought the software is small, if it exists at all. For those users, it's also pretty harmless. They're not stealing, they're not reselling, they're not costing Microsoft revenue unless the "shipping fees" for replacing CD's are in fact a profit item for them. (I suspect they are!)
But the point is that such a destructive payload would be a nasty cracker kiddie's delight, and the the risk of it needs to be taken seriously by BitTorrent authors and users. That isn't the merely annoying risk of failing to download a pirated music file. That's capable of trashing your hardware, or the hardware of someone innocent who gets handed that destructive CD not knowing its source.
Have you ever filled out Microsoft's registration forms or tried to get them to send you media for a product you legitimately own and have the license key decal on the hardware? And have you had to yank out the hardware to read the terrible font on those decals and try to type in those numbers, where the B looks like an 8?
That waste of my time all by itself alone is worth $50, at consulting rates, and not something I can afford to do in the field all the time. And if I haven't found the time to burn a CD image of the original CD, or my friend I'm visiting with is short of time, I'm delighted to verify that they have a legitimate license and point them to the very fast BitTorrent or Kazaa download sites.
This week, it's only heavily copyrighted files that are real piracy targets. Next week, though, it could be the Windows XP installation CD image, which people download because they lost their CD but still have the license key on the side of their computer, and it will overwrite the first few blocks of the disk with random digits 1,000,000 times to try and ruin their hard drives while displaying the Windows installation images.
All the shared network tools are vulnerable to this sof abuse.
Oh, it's worse than this. As soon as a moderator starts moderating, vetting content, it's a public admission that the moderator deliberately downloaded and reviewed the content. That gives RIAA a place and a name, or at least a hosting address, to send their attack lawyers. You know, the ones that threaten single moms with a year in court if they don't pony up $10,000 in a "settlement" because they downloaded a DVD movie that they already happen to own, but their copy is damaged because of their kids handling it with peanut butter smeared fingers?
I was reading about this in the Ney York Times a few weeks ago. It was a fascinating case among many other badly aimed lawsuits. I understand the judge finally threw out the case, but the mom could hardly afford the legal fees and time off of work to deal with the mess in court. Adn the attack lawyers absolutely could not be reasoned with.
Why, yes. A young person could learn what I've learned. It would take 5 years of dedicated, training-only workttime to learn what I've picked up twice that amount of hands-on and development, especially the lessons of "don't buy from these guys because they lie", or "how to make friends and contribute patches to important software projects so that your requests for fixes are taken seriously".
They would of course be unable to do anything else while trying to learn it at such an accelerated rate, and at the end of that time, they'd be as much of a problem for political middle managers because they'd be worth one heck of a lot of salary and tell their managers why some things can't be done, and why others will cost more than they budgeted, and why the unpatched laptop that the president uses to telecommute is a big security hole. Tjem they'd leave, because they think they can hop elsewhere as a new person with that incredible set of skills and qualifications.
Wait! I already do all that work! Why don't we save the 5 years of doing nothing for the new guy who will flee when he gets married or decides to become a nun or make other life changes, and let me do my job?
You've obviously never tried to recover data from paper tape, 5 year old floppies, or 15 year-old half-ing magtape. People forget the weird formats they kept things, especially proprietary backup formats. When you couple that with companies that wrote the software closing their doors and never releasing their source code, and with the physical fragility of these media, and the unwillingness to spend the time and money to transfer them to the latest readable format, you face a tremendous loss of original scientific and historical data.
Even old SCSI drives can be a nightmare to recover data from if they've been sitting on a shelf for 5 years.
GPG is also used to sign the tarball, at least with the source tarball sites I've been working with lately. Those sites involve security tools, so the authors are a bit pickier than most.
You're absolutely correct to be concerned. Some bloggers do lie, just as some lobbyists and journalists lie, and it's up to us as readers and up to editors of blog-recommendation sites as editors to be careful.
But the public revelations of US torture in Abu Ghraib are the result of websites and bloggers. What has been the core of US federal response to them? Forbidding digital cameras amoung prison guards....
RPM's are a bit more interesting. The content of the RPM is MD5 signed, but almost all RPM packages are also GPG signed as a package. This means that downloading things from redhat.com or mandrake.com, they have a GPG signature that is checked by most installers as a default and that the package is signed with, swearing that it was compiled by the people you love and trust.
If you're not checking the GPG signatures, you have no idea what is in the package anyway: it could have been built by the 3l33t cracker kiddy of the week, and it's why the FTP or HTTP repositories for public RPM's need to be secure and managed. and why publicly distributed packages from random sources need to be built from source.
Now, getting the MD5 signatures to match on a fake source tarball: that's clearly a bit of a risk to us downloaders, although that's been a risk for a while. It's nowhere near the risk of a badly secured site having the tarballs *AND* MD5 sums replaced. That's why really cautious sites publish MD5 sums and GPG or PGP sign those.
You mean someone actually investigates Ebay fraud cases? *HAH*. Only when the victim's name shows up in the paper or it's many hundreds of thousands of dollars, or the Ebay phishing spammers would have been out of business 2 years ago.
Way too many people get ripped off via Ebay, especially via credit card fraud. The credit card companies often write it off as a loss and make it good for the legitimate customers ripped off, but it's still massive amounts of fraud, and they simply don't investigate modest thefts.
PGPPhone had this high level of end-to-end security almost 20 years ago. It used on RSA, which still had a valid patent, but the PGP web of trust is pretty good and you can always generate your own new PGP keys and publish only the public part.
A modest re-write to operate on TCP instead of modems should be quite straightforward.
The keys are not held by the user: the keys are held by Skype, and are thus perfectly amenable to a Skype controlled man-in-the-middle monitoring. By opening their capabilities to monitoring by US law enforcement, and by getting US Department of Commerce approval for its use and export to non-restricted countries, I'm sure that the relevant federal agencies are falling over themselves to make Skype or another similarly tappable system the de facto standard.
Remember, unless you're the only one who owns the keys, your communications are not secure from anyone who can steal or borrow or liberate with a foolishly granted warrant the keys to your communications. And federal handling of telephone privacy has been horrible, as demonstrated by the FBI history if mis-handled wiretaps and political monitoring.
It's not as bad as countries where all foreign phone calls are automatically monitored by a secret policeman, but with computer technologies similar to the Carnivore email monitoring system, it's a big problem for privacy.
Few folks have the equipment and desire to set up local 2-day backup power for their VOIP setups: at the building level, it's usually much cheaper and safer to keep a pair of analog phone lines for emergencies, and rely on the Telco backup power systems (which tend to be massively over-supplied, for lots of very good reasons).
The flaws could be predicted, and they were predicted. Auto-downloading included icons and attachments while supporting things like ActiveX to install them, along with complete mishandling of URL publication in the Address field, coupled with Microsoft email and web tools tendencies to hide actual contents of clickable links and display only part of the URL are the absolute opposite of good design.
It doesn't matter how many pieces of duct tape you put on the pile of jello they call the source code for Internet Explorer, until they're willing to abandon some of the amazingly bad ideas of their public demos and its listed features, it will remain vulnerable to the most trivial forms of attack.
The desktop isn't the problem, and it's most certainly not 95% Microsoft at the server level.
For small scale, less than 50 users, MS is very common. For large scale, more than 50, it doesn't scale well. The interdependencies of what are functionally very different parts of the system (user authentication, file services, DNS, email, calendar, etc.) just don't work well in a large shop based on Microsoft. And the licensing costs are prohibitive at the medium scale, unless you've got an educational discount or someone willing to buy an MSDN license and violate it to install at least 3 distinct MS servers for general use.
Even Microsoft can't use it at the large scale of over 1000 users: hotmail and MSN were running on Suns, the last time I looked.
No, it's the difference between doing things in 5 minutes vs. 15 minutes (such as installing OpenOffice), or the difference between doing something 500,000 times that takes only 0.01 seconds and 0.0337 seconds (parsing a genetic database for matching strings, which is vastly faster in efficiently written C or even C++ because it has to actually read from and write to local files and just doesn't need the overhead of Java to do that.)
Java is very useful when having good layers of abstraction between your programmers and their specific platform lets you handwave a rather large set of those problems into the realm of the Java machine. But when you actually want lightning performance and to share data among the layers of abstraction in order to detect errors or to skip steps that don't need to be done, forget it.
The fact that they wrote it is interesting. But Java has not lived up to its promise of "write once, run anywhere", and it's proven to be horribly inefficient for things that actually write to disk. Layers of abstraction, but there are so many in the typical large Java application that it will run at a fraction of the speed and with many times the local memory and disk burden of similar applications written in C.
It would be absolutely wonderful if we could get them to simply use LinuxBIOS, but I'm afraid that compatibility with whatever cockamamie Windows management tools exist will preclude that on desktops for a while, But check out http://www.linuxbios.org/index.php/Main_Page for details. This stuff is very, very interesting for replacing the closed source, API violating, debris cluttered, bug-ridden, and debris cluttered cruft most BIOS's use.
Now, I know you're kidding. The amount of time my friends waste with dumb Windows help requests would more than cover the time they'd spend with dumb Linux requests.
Dell has offered RedHat before, on servers especially. And there are smaller computer shops all around the country that offer RedHat on custom machines. I wonder what got Dell to use Mandrive? And whether they will offer it on laptops, which has always been the hardest to integrate Linux onto?
No, it means no. Please re-read the post asking if this data applies to Stephen Hawking's theories about black hole evaporation. This data does not apply to Hawking's fascinating models because the black holes theoretically involved in quasars are much, much, much too large for the trickle of energy radiated by what Hawking described as the "temperature" of a black hole to be remotely noticeable or have any effect within the expected lifetime of our ability to observe it.
Slashdot Mirror
This is already in development. It's the Trusted Computing initiative, formerly known as Palladium, and it's a very big security effort. The benefit of its features are high: on-board high-speed encryption and authentication, easily available to users. The danger is also high: this is supposed to integrate with Dirital Rights Management and provide hardware level control of access to DVD writers, hard drives, boot loaders, system kernels, and secure operations called from withing software. That means that unless you can get the autohrization and the money to buy a highly authorized key from, say, the Microsoft key provisioning service, you will have difficulty writing and especially publishing open source tools that access those features.
He may be cautious about his non-disclosure agreements, or not want his former very political co-worker to play political games on his own career. Cowardly, but completely understandable: helping protect sources from reprisals is one of the reasons for anonymous posting on Slashdot.
The site seems to be slash-dotted.
They built EL4 RPM's? Excellent. That means they'll run on CentOS (at www.centos.org) for those sites who are too cheap to buy RHEL or who pay an internal Linux support person instead of paying support fees to RedHat.
That's very odd. I went into a field I enjoyed, was underpaid for 20 years, and suddenly my field took off and money became very easy to earn.
Nevery underestimate the fiscal benefit of going into something you love and getting very, very, very good at it, as opposed to starting out with a better pay scale but loathing the work and thus never achieving greatness, or even quality.
I'm looking, I'm looking.
A bit of an overbite, but fabulous legs. I'll give you two camels and a goat for her.
OK, two goats.
Copyrighted and illegal to download are two very different things. What the EULA says and what copyright law actually says are also two wildly, wildly, wildly differnent things.
I'm not a lawyer, but since duplicating a CD that you've purchased for backup reasons has been defended in court successfully several times, and since a user still requires an authorization key to install the software and usually actively agrees to the license agreement, I think that the copyright violation of copying such a CD for a user who bought the software is small, if it exists at all. For those users, it's also pretty harmless. They're not stealing, they're not reselling, they're not costing Microsoft revenue unless the "shipping fees" for replacing CD's are in fact a profit item for them. (I suspect they are!)
But the point is that such a destructive payload would be a nasty cracker kiddie's delight, and the the risk of it needs to be taken seriously by BitTorrent authors and users. That isn't the merely annoying risk of failing to download a pirated music file. That's capable of trashing your hardware, or the hardware of someone innocent who gets handed that destructive CD not knowing its source.
Have you ever filled out Microsoft's registration forms or tried to get them to send you media for a product you legitimately own and have the license key decal on the hardware? And have you had to yank out the hardware to read the terrible font on those decals and try to type in those numbers, where the B looks like an 8?
That waste of my time all by itself alone is worth $50, at consulting rates, and not something I can afford to do in the field all the time. And if I haven't found the time to burn a CD image of the original CD, or my friend I'm visiting with is short of time, I'm delighted to verify that they have a legitimate license and point them to the very fast BitTorrent or Kazaa download sites.
This week, it's only heavily copyrighted files that are real piracy targets. Next week, though, it could be the Windows XP installation CD image, which people download because they lost their CD but still have the license key on the side of their computer, and it will overwrite the first few blocks of the disk with random digits 1,000,000 times to try and ruin their hard drives while displaying the Windows installation images.
All the shared network tools are vulnerable to this sof abuse.
Oh, it's worse than this. As soon as a moderator starts moderating, vetting content, it's a public admission that the moderator deliberately downloaded and reviewed the content. That gives RIAA a place and a name, or at least a hosting address, to send their attack lawyers. You know, the ones that threaten single moms with a year in court if they don't pony up $10,000 in a "settlement" because they downloaded a DVD movie that they already happen to own, but their copy is damaged because of their kids handling it with peanut butter smeared fingers?
I was reading about this in the Ney York Times a few weeks ago. It was a fascinating case among many other badly aimed lawsuits. I understand the judge finally threw out the case, but the mom could hardly afford the legal fees and time off of work to deal with the mess in court. Adn the attack lawyers absolutely could not be reasoned with.
Why, yes. A young person could learn what I've learned. It would take 5 years of dedicated, training-only workttime to learn what I've picked up twice that amount of hands-on and development, especially the lessons of "don't buy from these guys because they lie", or "how to make friends and contribute patches to important software projects so that your requests for fixes are taken seriously".
They would of course be unable to do anything else while trying to learn it at such an accelerated rate, and at the end of that time, they'd be as much of a problem for political middle managers because they'd be worth one heck of a lot of salary and tell their managers why some things can't be done, and why others will cost more than they budgeted, and why the unpatched laptop that the president uses to telecommute is a big security hole. Tjem they'd leave, because they think they can hop elsewhere as a new person with that incredible set of skills and qualifications.
Wait! I already do all that work! Why don't we save the 5 years of doing nothing for the new guy who will flee when he gets married or decides to become a nun or make other life changes, and let me do my job?
You've obviously never tried to recover data from paper tape, 5 year old floppies, or 15 year-old half-ing magtape. People forget the weird formats they kept things, especially proprietary backup formats. When you couple that with companies that wrote the software closing their doors and never releasing their source code, and with the physical fragility of these media, and the unwillingness to spend the time and money to transfer them to the latest readable format, you face a tremendous loss of original scientific and historical data.
Even old SCSI drives can be a nightmare to recover data from if they've been sitting on a shelf for 5 years.
GPG is also used to sign the tarball, at least with the source tarball sites I've been working with lately. Those sites involve security tools, so the authors are a bit pickier than most.
You're absolutely correct to be concerned. Some bloggers do lie, just as some lobbyists and journalists lie, and it's up to us as readers and up to editors of blog-recommendation sites as editors to be careful.
But the public revelations of US torture in Abu Ghraib are the result of websites and bloggers. What has been the core of US federal response to them? Forbidding digital cameras amoung prison guards....
RPM's are a bit more interesting. The content of the RPM is MD5 signed, but almost all RPM packages are also GPG signed as a package. This means that downloading things from redhat.com or mandrake.com, they have a GPG signature that is checked by most installers as a default and that the package is signed with, swearing that it was compiled by the people you love and trust.
If you're not checking the GPG signatures, you have no idea what is in the package anyway: it could have been built by the 3l33t cracker kiddy of the week, and it's why the FTP or HTTP repositories for public RPM's need to be secure and managed. and why publicly distributed packages from random sources need to be built from source.
Now, getting the MD5 signatures to match on a fake source tarball: that's clearly a bit of a risk to us downloaders, although that's been a risk for a while. It's nowhere near the risk of a badly secured site having the tarballs *AND* MD5 sums replaced. That's why really cautious sites publish MD5 sums and GPG or PGP sign those.
You mean someone actually investigates Ebay fraud cases? *HAH*. Only when the victim's name shows up in the paper or it's many hundreds of thousands of dollars, or the Ebay phishing spammers would have been out of business 2 years ago.
Way too many people get ripped off via Ebay, especially via credit card fraud. The credit card companies often write it off as a loss and make it good for the legitimate customers ripped off, but it's still massive amounts of fraud, and they simply don't investigate modest thefts.
PGPPhone had this high level of end-to-end security almost 20 years ago. It used on RSA, which still had a valid patent, but the PGP web of trust is pretty good and you can always generate your own new PGP keys and publish only the public part.
A modest re-write to operate on TCP instead of modems should be quite straightforward.
The keys are not held by the user: the keys are held by Skype, and are thus perfectly amenable to a Skype controlled man-in-the-middle monitoring. By opening their capabilities to monitoring by US law enforcement, and by getting US Department of Commerce approval for its use and export to non-restricted countries, I'm sure that the relevant federal agencies are falling over themselves to make Skype or another similarly tappable system the de facto standard.
Remember, unless you're the only one who owns the keys, your communications are not secure from anyone who can steal or borrow or liberate with a foolishly granted warrant the keys to your communications. And federal handling of telephone privacy has been horrible, as demonstrated by the FBI history if mis-handled wiretaps and political monitoring.
It's not as bad as countries where all foreign phone calls are automatically monitored by a secret policeman, but with computer technologies similar to the Carnivore email monitoring system, it's a big problem for privacy.
Few folks have the equipment and desire to set up local 2-day backup power for their VOIP setups: at the building level, it's usually much cheaper and safer to keep a pair of analog phone lines for emergencies, and rely on the Telco backup power systems (which tend to be massively over-supplied, for lots of very good reasons).
The flaws could be predicted, and they were predicted. Auto-downloading included icons and attachments while supporting things like ActiveX to install them, along with complete mishandling of URL publication in the Address field, coupled with Microsoft email and web tools tendencies to hide actual contents of clickable links and display only part of the URL are the absolute opposite of good design.
It doesn't matter how many pieces of duct tape you put on the pile of jello they call the source code for Internet Explorer, until they're willing to abandon some of the amazingly bad ideas of their public demos and its listed features, it will remain vulnerable to the most trivial forms of attack.
The desktop isn't the problem, and it's most certainly not 95% Microsoft at the server level.
For small scale, less than 50 users, MS is very common. For large scale, more than 50, it doesn't scale well. The interdependencies of what are functionally very different parts of the system (user authentication, file services, DNS, email, calendar, etc.) just don't work well in a large shop based on Microsoft. And the licensing costs are prohibitive at the medium scale, unless you've got an educational discount or someone willing to buy an MSDN license and violate it to install at least 3 distinct MS servers for general use.
Even Microsoft can't use it at the large scale of over 1000 users: hotmail and MSN were running on Suns, the last time I looked.
No, it's the difference between doing things in 5 minutes vs. 15 minutes (such as installing OpenOffice), or the difference between doing something 500,000 times that takes only 0.01 seconds and 0.0337 seconds (parsing a genetic database for matching strings, which is vastly faster in efficiently written C or even C++ because it has to actually read from and write to local files and just doesn't need the overhead of Java to do that.)
Java is very useful when having good layers of abstraction between your programmers and their specific platform lets you handwave a rather large set of those problems into the realm of the Java machine. But when you actually want lightning performance and to share data among the layers of abstraction in order to detect errors or to skip steps that don't need to be done, forget it.
The fact that they wrote it is interesting. But Java has not lived up to its promise of "write once, run anywhere", and it's proven to be horribly inefficient for things that actually write to disk. Layers of abstraction, but there are so many in the typical large Java application that it will run at a fraction of the speed and with many times the local memory and disk burden of similar applications written in C.
It would be absolutely wonderful if we could get them to simply use LinuxBIOS, but I'm afraid that compatibility with whatever cockamamie Windows management tools exist will preclude that on desktops for a while, But check out http://www.linuxbios.org/index.php/Main_Page for details. This stuff is very, very interesting for replacing the closed source, API violating, debris cluttered, bug-ridden, and debris cluttered cruft most BIOS's use.
Now, I know you're kidding. The amount of time my friends waste with dumb Windows help requests would more than cover the time they'd spend with dumb Linux requests. Dell has offered RedHat before, on servers especially. And there are smaller computer shops all around the country that offer RedHat on custom machines. I wonder what got Dell to use Mandrive? And whether they will offer it on laptops, which has always been the hardest to integrate Linux onto?
No, it means no. Please re-read the post asking if this data applies to Stephen Hawking's theories about black hole evaporation. This data does not apply to Hawking's fascinating models because the black holes theoretically involved in quasars are much, much, much too large for the trickle of energy radiated by what Hawking described as the "temperature" of a black hole to be remotely noticeable or have any effect within the expected lifetime of our ability to observe it.