Have you actually seen how an undersea cable is constructed? I find it highly dubious that someone would attempt tapping an undersea cable. Sure it sounds great on paper, but in practice doing it would require specialized custom made equipment to do it. And even then you would still have the possibility of the tap being discovered when the cable is having maintenance performed on it.
I'm sorry that you don't but it, but it is true. Special cases are possible to prove, but your talking about a program running in a general purpose system, not a specialized embedded hardware platform that is rigidly controlled. But the complexity of the problem has very little to do with the Halting problem, however now that you've brought it up if you think about it your reasoning is simply wrong.
If a Turing machine encounters an undefined state transition that is the definition of an logically inconsistent situation. However you can't know if your TM is going to halt or not because its possible for your TM to run forever. But this has nothing to do with what I was talking about. If you try to mathematically prove that a program is correct you simply cannot do it beyond a very simple level of complexity. Special cases are not sufficient enough proof for the completeness and correctness of a program.
This would be akin to your compiler being able to tell you that a program you have written has exactly X number of bugs and then listing them out to you. It doesnt exactly work this way, as while you can catch syntatic errors early on and some kinds of logical problems there are set of more complicated deep problems that there exist no NP polynomial time algorithm or way to compute the results of.
"8 year old stuff still compiles mostly, its fluid."
Uhm, so is Linux the bedrock of computing or is it the agile warrior able to adapt to its changing foes? I'm a bit confused.
I don't know what 8 year old code you think would still compile against todays Linux. Between major changes from the pre 2.0 kernel days to now I can think of plenty of code that would break.
And then you've got your personal best friend in the world, a new version of glibc just around the corner to break things once in a while, but thats not Linux per se since Linux is just a kernel. But its all of the FOSS/FSF software that makes a Linux DISTRO.
Now show me a piece of 8 year old code that will compile on a current distro without barfing or having its./configure script changed and I might begin to see your point. But I doubt your argument holds true for enough pieces of FOSS software to be truly relevant.
The bigger cost doesnt come from porting the software to Linux.
The big cost comes when its time for the users to be retrained to use the software under Linux.
This typically means their IT has to either be outsourced to someone who can support a Linux workstation environment, or they need to develop that skillset in the inhouse IT staff. That is not something that all organizations can do sometimes. I believe in using the best tool possible, but if at the end of the day that tool breaks and no one knows how to fix it youve just created an even bigger problem.
"My personal feeling is that the use of unstable software (no matter who it is by) should be outlawed from mission-critical systems where lives really are at stake."
Uh, has anyone told you that engineering software with zero bugs is not possible for any system of software being even the most basic of functionality? Software has bugs. Period.
You can do many things to try and minimize the number of bugs, but you can not eliminate all of them all of the time.
Beyond a certain level of complexity it becomes mathematically infeasible to prove that a software program is 100% "correct".
While I think Peter Gutmann is a capable fellow on the tech side I fail to see one piece of actual FACT presented in his paper that purports to support his rashly concluded suppositions.
He doesnt have access to the Vista source code
He doesnt have access into any special insight(s) to the hardware manufacturers
His entire argument is based off of speculation and nothing factually provable
Nothing in his paper actually examines the "cost" of the Vista DRM
No medical imaging system in use today or anytime in the future is EVER going to use HDCP or its like because thats something that the MPAA wants for *its* content, it could care less if anyone else used it
Elimination of Unified drivers is a complete and total fabrication on his part. The reason NVIDA's Forceware drivers fragmented recently was because of the new G80 chip architecture being a total departure from previous chips. Not because of bloody Vista DRM demands.
I could go on, but this entire paper is rife with statements he never backs up with any kind of rational assertion. Just for amusement here are things he says that he never shows evidence for:
"...in order to work, Vista's content protection must be able to violate the laws of physics, something that's unlikely to happen no matter how much the content industry wishes it were possible."
---Riiiiiight.
"In order to prevent the creation of hardware emulators of protected output devices, Vista requires a Hardware Functionality Scan (HFS) that can be used to uniquely fingerprint a hardware device to ensure that it's (probably) genuine."
---This existed in XP of course, after all XP had to create a table of hashes for your hardware. But how he describes it simply doesnt work that way.
"...this leads to a problem: It's no longer possible to tell if a graphics chip is situated on a plug-in card or attached to the motherboard, since as far as the system is concerned they're both just devices sitting on the AGP/PCIe bus."
---But wait! If the HFS is so damned smart and intelligent in knowing if hardware will obey the Almighty DRM Commands or not, then it should be trivial for HFS to tell if your bleeding video card sits in an AGP, PCI-E slot or is onboard the motherboard. So which is it? Smart AND dumb? Let me guess, you'd just tout that MS is capable of writing that kind of software to fulfill your little bit of paranoia.
"Once a weakness is found in a particular driver or device, that driver will have its signature revoked by Microsoft"
---Riiiiight. Microsoft will issue a Windows Update to MILLIONS of users, QA, test, and engineer a single patch or piece of code just to update Vista to DISABLE A DRIVER FOR ONE VIDEO CARD? What does Occam's Razor tell you is more likely? Device driver signatures arent even NECESSARY! Half the drivers you install right now dont even come with a proper signature to begin with. You can install unsigned drivers on Vista 32-bit (I havent tried it with Vista 64-bit) and it works.
"Cannot go to market until it works to specification... potentially more respins of hardware" -- ATI.
---Why is this quote even here? Graphics card makers already operate under this modus operandi, otherwise they would be shipping BROKEN HARDWARE.
"The high-end graphics and audio market are dominated entirely by gamers, who will do anything to gain the tiniest bit of extra performance, like buying Bigfoot Networks' $250 "Killer NIC" ethernet card in the hope that it'll help reduce their network latency by a few milliseconds."
---Oh brother, please, freaking spare me what you think of "gamers" it pains me. I cant think of a single gamer I know of that thinks the bloody Killer NIC is anything but a giant waste of money.
I could go on here, but what is the point. By Gutmann's own admission in the acknowledgments section of the "paper" came from other people "involved" in Vista or other aspects of the ecosystem, but wishing to remain anonymous. So I cant tell what in this pape
Physical access to the disk while mounted is pretty much the only way to defated this encryption. All of the other methods mentioned are workable, but rely on brute force attacks that would take longer then the lifetime of the universe to crack. Unless you get lucky and can find evidence of the master password in the slack space on the disk
Yes but the buffering in Quicktime actually works! Whereas in any product from Real all too often you'd see that damn "Buffering..." message forever with no video starting up or worse it would start for a few seconds and then go right back to "Buffering"!
I can just imagine folks the world over will be beating on their door to license such wonderfully working software!
Or people could just do MPEG-4 or Quicktime streaming and never have to deal with the unending stream of "BUffering...." seen in almost any Real Networks product.
So lets see here, its Slashot...oooooh shiny...uh what was I doing? Oh yeah lets postpostpost about the Intel Macs. Post the story first and read the articles linked in the story later....if ever.
Seriously this story is stupid.
1. NO ONE HAS A WORKING ACCELERATED VIDEO DRIVER WORKING ON XP UNDER MAC
2. The person who submitted this story THINKS that their MIGHT be a driver because they saw those two videos on YouTube of all places of someone with a Mac Mini that showed Half-Life 2 running as well as on a MacBook Pro.
3. If you read the forums on onmac.net and osx86project NO ONE has working accelerated video.
4. Blanka and narf are NOT working on the video drivers
Stop accepting every stupid story submission about Wikipedia.
It's BORING.
Seriously, its like opening up Popular Science to see an article about how Scientific American discovered there were some factual discrepancies in Encyclopedia Britannica Vol 24 45th Edition entry on Underwater Basketweaving.
Discussions at Slashdot are relevant? That's the first time I've ever heard that. Must be due to the Great Hot-Grits Famine of 2003-2005. The fact of the matter is, Slashdot's story submission system blows. On Digg it's much easier to find esoteric stories that the stuffy story submission soup-nazis here would have rejected because it wasn't the right shade of purple, or the content wasn't just pink enough.
I'm sorry but I am a big fan of open source! I've used Linux all the way back from 94 with Slackware on a zillion floppies. However these days I use OpenBSD for servers and at home I have a PowerBook now to go along with my Windows desktop. That being said however, if there is one thing I have learned about Linux drivers over the years is that they will bite you in the ass sometimes. FOSS device drivers are not always "better", in fact sometimes they can perfectly decent hardware into utter crap on the platform. Just look at the whole mess that happened with SATA support on Linux! End result is a narror range of SATA controllers that are fully supported well and a broader range of controllers that work okay, but if they blow up the develop list is just going to say "the chipset is crap". Yet those same chipset manage to work under Windows fine? So to sum up:
FOSS device driver stability/quality != Apple device driver stability/quality
Slashdot Mirror
Look at a Lenovo W540. I've got an older W530 model but its basically a workhorse and seems to have good cooling.
Have you actually seen how an undersea cable is constructed? I find it highly dubious that someone would attempt tapping an undersea cable. Sure it sounds great on paper, but in practice doing it would require specialized custom made equipment to do it. And even then you would still have the possibility of the tap being discovered when the cable is having maintenance performed on it.
I find it hilarious my post has been modded as flamebait since that whats Gutmanns paper reads like
I'm sorry that you don't but it, but it is true. Special cases are possible to prove, but your talking about a program running in a general purpose system, not a specialized embedded hardware platform that is rigidly controlled. But the complexity of the problem has very little to do with the Halting problem, however now that you've brought it up if you think about it your reasoning is simply wrong.
If a Turing machine encounters an undefined state transition that is the definition of an logically inconsistent situation. However you can't know if your TM is going to halt or not because its possible for your TM to run forever. But this has nothing to do with what I was talking about. If you try to mathematically prove that a program is correct you simply cannot do it beyond a very simple level of complexity. Special cases are not sufficient enough proof for the completeness and correctness of a program.
This would be akin to your compiler being able to tell you that a program you have written has exactly X number of bugs and then listing them out to you. It doesnt exactly work this way, as while you can catch syntatic errors early on and some kinds of logical problems there are set of more complicated deep problems that there exist no NP polynomial time algorithm or way to compute the results of.
Haha, very funny.
How about something, you know, useful?
"8 year old stuff still compiles mostly, its fluid."
./configure script changed and I might begin to see your point. But I doubt your argument holds true for enough pieces of FOSS software to be truly relevant.
Uhm, so is Linux the bedrock of computing or is it the agile warrior able to adapt to its changing foes? I'm a bit confused.
I don't know what 8 year old code you think would still compile against todays Linux. Between major changes from the pre 2.0 kernel days to now I can think of plenty of code that would break.
And then you've got your personal best friend in the world, a new version of glibc just around the corner to break things once in a while, but thats not Linux per se since Linux is just a kernel. But its all of the FOSS/FSF software that makes a Linux DISTRO.
Now show me a piece of 8 year old code that will compile on a current distro without barfing or having its
The bigger cost doesnt come from porting the software to Linux.
The big cost comes when its time for the users to be retrained to use the software under Linux.
This typically means their IT has to either be outsourced to someone who can support a Linux workstation environment, or they need to develop that skillset in the inhouse IT staff. That is not something that all organizations can do sometimes. I believe in using the best tool possible, but if at the end of the day that tool breaks and no one knows how to fix it youve just created an even bigger problem.
Its hard to balance sometimes.
"My personal feeling is that the use of unstable software (no matter who it is by) should be outlawed from mission-critical systems where lives really are at stake."
Uh, has anyone told you that engineering software with zero bugs is not possible for any system of software being even the most basic of functionality? Software has bugs. Period.
You can do many things to try and minimize the number of bugs, but you can not eliminate all of them all of the time.
Beyond a certain level of complexity it becomes mathematically infeasible to prove that a software program is 100% "correct".
While I think Peter Gutmann is a capable fellow on the tech side I fail to see one piece of actual FACT presented in his paper that purports to support his rashly concluded suppositions.
He doesnt have access to the Vista source code
He doesnt have access into any special insight(s) to the hardware manufacturers
His entire argument is based off of speculation and nothing factually provable
Nothing in his paper actually examines the "cost" of the Vista DRM
No medical imaging system in use today or anytime in the future is EVER going to use HDCP or its like because thats something that the MPAA wants for *its* content, it could care less if anyone else used it
Elimination of Unified drivers is a complete and total fabrication on his part. The reason NVIDA's Forceware drivers fragmented recently was because of the new G80 chip architecture being a total departure from previous chips. Not because of bloody Vista DRM demands.
I could go on, but this entire paper is rife with statements he never backs up with any kind of rational assertion. Just for amusement here are things he says that he never shows evidence for:
"...in order to work, Vista's content protection must be able to violate the laws of physics, something that's unlikely to happen no matter how much the content industry wishes it were possible."
---Riiiiiight.
"In order to prevent the creation of hardware emulators of protected output devices, Vista requires a Hardware Functionality Scan (HFS) that can be used to uniquely fingerprint a hardware device to ensure that it's (probably) genuine."
---This existed in XP of course, after all XP had to create a table of hashes for your hardware. But how he describes it simply doesnt work that way.
"...this leads to a problem: It's no longer possible to tell if a graphics chip is situated on a plug-in card or attached to the motherboard, since as far as the system is concerned they're both just devices sitting on the AGP/PCIe bus."
---But wait! If the HFS is so damned smart and intelligent in knowing if hardware will obey the Almighty DRM Commands or not, then it should be trivial for HFS to tell if your bleeding video card sits in an AGP, PCI-E slot or is onboard the motherboard. So which is it? Smart AND dumb? Let me guess, you'd just tout that MS is capable of writing that kind of software to fulfill your little bit of paranoia.
"Once a weakness is found in a particular driver or device, that driver will have its signature revoked by Microsoft"
---Riiiiight. Microsoft will issue a Windows Update to MILLIONS of users, QA, test, and engineer a single patch or piece of code just to update Vista to DISABLE A DRIVER FOR ONE VIDEO CARD? What does Occam's Razor tell you is more likely? Device driver signatures arent even NECESSARY! Half the drivers you install right now dont even come with a proper signature to begin with. You can install unsigned drivers on Vista 32-bit (I havent tried it with Vista 64-bit) and it works.
"Cannot go to market until it works to specification... potentially more respins of hardware" -- ATI.
---Why is this quote even here? Graphics card makers already operate under this modus operandi, otherwise they would be shipping BROKEN HARDWARE.
"The high-end graphics and audio market are dominated entirely by gamers, who will do anything to gain the tiniest bit of extra performance, like buying Bigfoot Networks' $250 "Killer NIC" ethernet card in the hope that it'll help reduce their network latency by a few milliseconds."
---Oh brother, please, freaking spare me what you think of "gamers" it pains me. I cant think of a single gamer I know of that thinks the bloody Killer NIC is anything but a giant waste of money.
I could go on here, but what is the point. By Gutmann's own admission in the acknowledgments section of the "paper" came from other people "involved" in Vista or other aspects of the ecosystem, but wishing to remain anonymous. So I cant tell what in this pape
Physical access to the disk while mounted is pretty much the only way to defated this encryption. All of the other methods mentioned are workable, but rely on brute force attacks that would take longer then the lifetime of the universe to crack. Unless you get lucky and can find evidence of the master password in the slack space on the disk
Ken Kutaragi probably also thinks that gold-plating his nuts for a few grand is "probably too cheap" for the "value for your money".
Yes but the buffering in Quicktime actually works! Whereas in any product from Real all too often you'd see that damn "Buffering..." message forever with no video starting up or worse it would start for a few seconds and then go right back to "Buffering"!
Now RealNetwork's owns a patent on the "Buffering...." nonsense.
I can just imagine folks the world over will be beating on their door to license such wonderfully working software!
Or people could just do MPEG-4 or Quicktime streaming and never have to deal with the unending stream of "BUffering...." seen in almost any Real Networks product.
Would someone PLEASE submit Slashdot to the Web Pages That Suck people ASAP?
Thanks!
Uhm, am the only one who sees the obvious issue of we can't ALL be managers?
And isn't one of the problems of business organizations one of being too management heavy?
If I'm wrong, then correct me, please.
Drama drama drama blog drama drama Amazon drama blog smug smug drama blog Amazon smung drama blah.
Why is this news we give a flying leap about?
We don't care.
So lets see here, its Slashot...oooooh shiny...uh what was I doing? Oh yeah lets postpostpost about the Intel Macs. Post the story first and read the articles linked in the story later....if ever.
Seriously this story is stupid.
1. NO ONE HAS A WORKING ACCELERATED VIDEO DRIVER WORKING ON XP UNDER MAC
2. The person who submitted this story THINKS that their MIGHT be a driver because they saw those two videos on YouTube of all places of someone with a Mac Mini that showed Half-Life 2 running as well as on a MacBook Pro.
3. If you read the forums on onmac.net and osx86project NO ONE has working accelerated video.
4. Blanka and narf are NOT working on the video drivers
5. No pooftas!
6. See #1. Repeat.
Sony is waiting for Toy Story to finish rendering on the box.
I'll make this dead simple.
What you said is utter bullshit.
Go learn how real worms work bub.
So advanced that now can you no longer look at the souce but you can't even load the nessus.org website anymore. Take that GPL!
Stop accepting every stupid story submission about Wikipedia.
It's BORING.
Seriously, its like opening up Popular Science to see an article about how Scientific American discovered there were some factual discrepancies in Encyclopedia Britannica Vol 24 45th Edition entry on Underwater Basketweaving.
Discussions at Slashdot are relevant? That's the first time I've ever heard that. Must be due to the Great Hot-Grits Famine of 2003-2005. The fact of the matter is, Slashdot's story submission system blows. On Digg it's much easier to find esoteric stories that the stuffy story submission soup-nazis here would have rejected because it wasn't the right shade of purple, or the content wasn't just pink enough.
I'm sorry but I am a big fan of open source! I've used Linux all the way back from 94 with Slackware on a zillion floppies. However these days I use OpenBSD for servers and at home I have a PowerBook now to go along with my Windows desktop. That being said however, if there is one thing I have learned about Linux drivers over the years is that they will bite you in the ass sometimes. FOSS device drivers are not always "better", in fact sometimes they can perfectly decent hardware into utter crap on the platform. Just look at the whole mess that happened with SATA support on Linux! End result is a narror range of SATA controllers that are fully supported well and a broader range of controllers that work okay, but if they blow up the develop list is just going to say "the chipset is crap". Yet those same chipset manage to work under Windows fine? So to sum up:
FOSS device driver stability/quality != Apple device driver stability/quality
We better start working on a new definition for "intelligent" and "design".
God damn religous whackjobs.