It does if the drivers have to pass a standardized method of certification and testing on both 32 and 64-bit kernels.
Food for thought:
Before Vista was released, 64-bit drivers were hard to find for Windows. They were frequently buggy, rarely supported every feature, and as a result, Windows XP 64-bit is barely acknowledges as a legitimate OS release, and has received little fanfare. With Vista, and the introduction of 64-bit as a consumer OS feature widely available, and closer alignment to the server platform that would soon (2008 R2) become 64-bit only, this has changed. Why? Not because of anything I said, actually. It's because WHQL and Windows Logo certification didn't necessarily require 64-bit drivers to be released. When Vista was under development, it was decided the only thing that would drag the hardware vendors into the 21st century was to require 64-bit drivers under a number of categories (I'm unfamiliar if there are exceptions for certification.) This means that, for example, every shipping driver in Windows Vista that received WHQL certification has a 32 and a 64 bit version.
This requirement from Microsoft killed Vista's reputation, as we all know. Why is Windows 7 being reviewed so much better? In many ways, it's a similar OS, with a similar kernel. There are performance improvements, some very nice new features, etc. I could talk about those at length. But the reason it's getting fantastic reviews is that since three years ago, anyone who was anyone who wanted to release hardware that worked with Windows and get it certified had to make sure a 32-bit and a 64-bit driver was released.
I only mention all of this because this is one of the primary reason why certain things are supported on certain SKUs of Windows. I outlined differences between Windows XP and Windows Vista in terms of driver certification, but that's just the beginning. For the most part, consumer drivers should work with PAE, but may not, and as far as I know, this is an issue with the requirements Microsoft laid out. For consumer OSes and server OSes, it was decided that the driver requirements differed and support for certain features was an unnecessary burden on hardware manufacturers. It is those logo requirements that drive many vendors to add features.
You'd be surprised by the list of logo requirements and how they benefit end users. For example, Windows Vista logo testing required cameras to support Media Transfer Protocol, for speeding file transfer. It requires printers to support the Windows Color System, for accurate color space representation. It requires DVI connectors on logo bearing monitors. It requires QoS and topology discovery in routers, switches, and network hardware that is logo tested. These are all things that have real benefits, no? Companies are not required to get logo testing, but it's a win for consumers, even those who never use Microsoft's OS.
So yes, signing does have a thing or two to do with quality, if the right agency is doing the signing and certification.
No, I don't work for Microsoft, I just despise the people who think that every corporation can do no good. Does it right Microsoft's wrongs? No. Don't pretend I think it does!
(a.) It's obvious to an expert in the craft. People have been executing this patent's claims for years. (b.) It's a software patent. (c.) There is tremendous prior art.
Honestly, why are we having this discussion at all?
Like me, I'm guessing you hate software patents. They make your job annoying, they make my job annoying. They tend to produce no net worth, yet create huge risks and liabilities for young companies and startups. Software patents have become such a joke that companies like IBM and Microsoft submit several, perhaps dozens of patents a day to make sure every possible process is covered. Why? So that when push comes to shove, they can show the other team's lawyers their patent warchest and cross license.
Even if it's a unique "technique," this is something obvious to the craft.
It's amazing what flies come out of the woodwork when Microsoft gets sued. No, sorry, I meant stupid. It's stupid what I'm seeing on Slashdot. Everyone who hates software patents and has any grasp of current technology would realize i4i just another company building a patent warchest, and somehow it crossed their mind to sue Microsoft.
Or they worked with i4i long enough to realize their patent was pretty thoroughly covered by prior art, considering their "invention" is a method of separating data from metadata describing the formatting, styling, etc, of the data.
It looks like if this patent holds up, they could hold the online world at gunpoint and demand everyone who uses CSS to pay up. I'm sure OpenOffice applies too, because it supports XML based styling and the like.
The really, really funny thing about this case is it has brought out more trolls than I can ever remember though. I mean, truly vociferous, loud and obnoxious trolls yelling, "THEY'RE NOT A PATENT TROLL!"
I have to wonder, is i4i astroturfing or is the community really so blind as to not realize this is yet another harmful case of software patent litigation? Who cares if it's Microsoft, it's a software patent, likely one with prior art, and frankly, sounds obvious to anyone who has worked with computers.
It seems like it would cover Adobe Dreamweaver, or any GUI editor that edits an XML-based data and styling application, where the data and the metacode defining the styles are separately stored but selected through some procedure.
For example, Notepad++, an open source notepad implementation, stores style information in a document separate from the storage of the text, which always remains plaintext and without any markup except that which is defined in a separate metadata file. In fact, Notepad++ goes a bit further and defines the method of mapping external to the document itself, allowing entirely plaintext documents, batch files, and dozens of programming languages to be syntax highlighted. That's just one example, and I believe Notepad++ is based off a root engine, Scintilla?
Regardless, I feel like I can continue to think of applications that violate the patent all day. The only reason I4i went after Microsoft in an East Texas courtroom is because they knew they could get either a lot of money, or a lot of publicity, or absurdly, they actually think they deserve the sole right to implement separately stored metadata + data in XML formats.
Taking drafting for a semester in high school was the best thing that ever happened to my ability to print block letters on forms and things. And it shows, because when I write in that form it's vastly more legible for myself and anyone who has to read it.
Who knew, that professions based on precision would create the most workable and long-lasting script, while professions based on creativity and nuance would create the one that eventually falls by the wayside for being too complex and annoying to discern? How ever could that have happened?
All the important operations tend to be random. For a file server, you may have twenty people accessing files simultaneously. Or a hundred, or a thousand. For a webserver, it'll be hitting dozens or hundreds of static pages and, if you have database backend, that's almost entirely random as well.
For people consolidating physical servers to virtual servers, you now have two, three, ten or twenty VMs running on one machine. If every one of those VMs tries to do a "sequential" IO, it gets interlaced by the hypervisor into all the other sequential IOs. No hypervisor would dare tell all the other VMs to sit back and wait so that every IO is sequential. That delay could be seconds or minutes or hours.
Now imagine all that, and take into account that the latest Intel SSD gets around 6600 IOPS read and write. A good, fast hard drive gets 200. So you could put thirty three hard drives in RAID 0 and have the same number of IOPS, and your latency would still be worse. All the RAID0 really does for you is give you a nice big queue pipeline, like in a CPU. Your IO doesn't really get done faster, but you can have many more running simultaneously.
Given that SSDs are easily three to four times faster on sequential IO and an order of magnitude faster on random IO, I don't think it's that implausible to believe that the industry isn't ready.
This is different from AI, and is coming from someone whose expertise on the subject is demonstrable. He's not talking about AI, he's talking about simulating all of the tissue in a human brain and providing it with stimuli to determine reactions.
He's not saying it'll necessarily be a good ol' buddy ol' pal right off the bat. Probably not. Probably won't even be capable of simple arithmetic for years. On the other hand, we could simulate things like lesions effecting far away parts of the brain, various known "paths" that signals travel in the brain and ways to alter those paths or correct flaws, etc.
As well, we could simulate the effect of various drugs on large-scale phenomena in the brain to help try and understand (a.) what a drug will do before it undergoes testing, and (b.) why exactly it is that makes these drugs work so well. Both questions are currently unanswerable. We know what a drug does, but rarely do we understand the full extent of why a particular drug helps certain conditions.
The one that worked immediately created a blackhole and ended humanity in that universe. By the anthropic principle, we will never make the LHC work, and will instead see one failure after another and wonder why fate has cursed us so.
From what I've read on the issue, such as Feynman's books and other novels targeted toward those of us who do not have a complete grasp of quantum mechanics, you are wrong.
Caveat emptor, this is merely what I've read:
Classical mechanics as explained by Feynman were the result of the sum of all possible histories, among other interpretations. Regardless of one's interpretation, Feynman and others found that as you crunch the math for larger and larger quantities of particles, the results closer and closer approximate what we think of as classical physics. As a result, classical physics is an approximation of quantum mechanics, which is a theory of how the universe really works.
Leaking? Jesus, the beta and the RC were both widely and easily available, and you can still download the RC.
The leaks are just that, they really would rather you not play around with the unofficial builds which have a lot of other debug functionality turned on a lot of the time.
Actually, it depends on the reliability. 95% reliability becomes 90.25% reliability. 50% reliability becomes 25% reliability. 1% reliability becomes 0.01% reliability.
So if your drives are very reliable, it's very slightly less than twice the failure rate. If your drives are not reliable, then it asymptotically approaches an infinitely greater risk of failure.
Then you're going to have to try very hard to invalidate 80 years of science that has culminated in microprocessing technology that relies on quantum effects, thus allowing you to post on Slashdot.
I'm genuinely interested in your alternative theory for all things quantum. I don't think you have a theory, but I won't let that stop you.
Right now, we have very successful theories that explain how LEDs work, how your computer's CPU works, etc. What I don't see is an alternative theory to explain all of those things rigorously.
I'm a layperson on the subject but have read extensively, and my understanding is yes, there is a "quantum superposition" and you can prove it by running exactly the same experiment over and over again, and getting different results each time. You take the average of the results and that's the answer to your problem.
Hence, a quantum algorithm only has a probability of arriving at the correct answer. Executing a quantum algorithm several times gives you increasingly better odds in polynomial time, and because there are quantum algorithms for non-polynomial problems (see: Shor's Algorithm), you can simply repeat the algorithm and refine your result and get a right answer much faster.
While you can never guarantee your answer is 100% correct, I believe, you can get arbitrarily close. And for problems that are intractable for modern computers, problems that would require enormous supercomputers, the correct answer can be derived with, theoretically, much smaller and much faster quantum computers if you are willing to accept some small probability of error.
In addition, the solutions to these problems can be checked rather quickly, for example, if in attempting to factor the number 15, you get the number 6, a classical computer could check this very easily. Even when you scale up the numbers immensely, the verification step is not much more difficult. So Shor's Algorithm on a quantum computer, combined with a fast bignum library on a classical computer, could determine factors much more rapidly than any classical computer could on its own given current knowledge.
Pictures shot in digital or transferred from digital intermediate tend to come out blurry when put on film no matter what, and are a pain to focus. That said, the quality is somewhere between 2K (around 1080P) and 4K (four times the pixels of 1080P, an image 4000 pixels wide by ~2000 pixels tall.)
More like, when he stops stealing TVs, you stop charging him with theft.
They're charging Microsoft with crimes that they haven't yet committed, and every indication seems to be Microsoft is willing to meet some rather absurd demands.
Slashdot Mirror
A $75 video card has maybe 1/4 the performance of a $150 video card, from what I've seen.
On the other hand, going from $150 to $300 often means "using two cards simultaneously," so the performance is maybe 90% better, not 300% better.
And the difference between two $150 cards and two $200 cards is smaller yet.
It does if the drivers have to pass a standardized method of certification and testing on both 32 and 64-bit kernels.
Food for thought:
Before Vista was released, 64-bit drivers were hard to find for Windows. They were frequently buggy, rarely supported every feature, and as a result, Windows XP 64-bit is barely acknowledges as a legitimate OS release, and has received little fanfare. With Vista, and the introduction of 64-bit as a consumer OS feature widely available, and closer alignment to the server platform that would soon (2008 R2) become 64-bit only, this has changed. Why? Not because of anything I said, actually. It's because WHQL and Windows Logo certification didn't necessarily require 64-bit drivers to be released. When Vista was under development, it was decided the only thing that would drag the hardware vendors into the 21st century was to require 64-bit drivers under a number of categories (I'm unfamiliar if there are exceptions for certification.) This means that, for example, every shipping driver in Windows Vista that received WHQL certification has a 32 and a 64 bit version.
This requirement from Microsoft killed Vista's reputation, as we all know. Why is Windows 7 being reviewed so much better? In many ways, it's a similar OS, with a similar kernel. There are performance improvements, some very nice new features, etc. I could talk about those at length. But the reason it's getting fantastic reviews is that since three years ago, anyone who was anyone who wanted to release hardware that worked with Windows and get it certified had to make sure a 32-bit and a 64-bit driver was released.
I only mention all of this because this is one of the primary reason why certain things are supported on certain SKUs of Windows. I outlined differences between Windows XP and Windows Vista in terms of driver certification, but that's just the beginning. For the most part, consumer drivers should work with PAE, but may not, and as far as I know, this is an issue with the requirements Microsoft laid out. For consumer OSes and server OSes, it was decided that the driver requirements differed and support for certain features was an unnecessary burden on hardware manufacturers. It is those logo requirements that drive many vendors to add features.
You'd be surprised by the list of logo requirements and how they benefit end users. For example, Windows Vista logo testing required cameras to support Media Transfer Protocol, for speeding file transfer. It requires printers to support the Windows Color System, for accurate color space representation. It requires DVI connectors on logo bearing monitors. It requires QoS and topology discovery in routers, switches, and network hardware that is logo tested. These are all things that have real benefits, no? Companies are not required to get logo testing, but it's a win for consumers, even those who never use Microsoft's OS.
So yes, signing does have a thing or two to do with quality, if the right agency is doing the signing and certification.
No, I don't work for Microsoft, I just despise the people who think that every corporation can do no good. Does it right Microsoft's wrongs? No. Don't pretend I think it does!
They are a patent troll because:
(a.) It's obvious to an expert in the craft. People have been executing this patent's claims for years.
(b.) It's a software patent.
(c.) There is tremendous prior art.
Honestly, why are we having this discussion at all?
Like me, I'm guessing you hate software patents. They make your job annoying, they make my job annoying. They tend to produce no net worth, yet create huge risks and liabilities for young companies and startups. Software patents have become such a joke that companies like IBM and Microsoft submit several, perhaps dozens of patents a day to make sure every possible process is covered. Why? So that when push comes to shove, they can show the other team's lawyers their patent warchest and cross license.
Even if it's a unique "technique," this is something obvious to the craft.
It's amazing what flies come out of the woodwork when Microsoft gets sued. No, sorry, I meant stupid. It's stupid what I'm seeing on Slashdot. Everyone who hates software patents and has any grasp of current technology would realize i4i just another company building a patent warchest, and somehow it crossed their mind to sue Microsoft.
Or they worked with i4i long enough to realize their patent was pretty thoroughly covered by prior art, considering their "invention" is a method of separating data from metadata describing the formatting, styling, etc, of the data.
It looks like if this patent holds up, they could hold the online world at gunpoint and demand everyone who uses CSS to pay up. I'm sure OpenOffice applies too, because it supports XML based styling and the like.
The really, really funny thing about this case is it has brought out more trolls than I can ever remember though. I mean, truly vociferous, loud and obnoxious trolls yelling, "THEY'RE NOT A PATENT TROLL!"
I have to wonder, is i4i astroturfing or is the community really so blind as to not realize this is yet another harmful case of software patent litigation? Who cares if it's Microsoft, it's a software patent, likely one with prior art, and frankly, sounds obvious to anyone who has worked with computers.
On the contrary, it does hint XML documents. It also does HTML, and it can do a variety of languages.
So in the case of hinting an XML document, it has XML data, with XML styling and XML definitions of how to lay out styles.
Have you read the patent?
It seems like it would cover Adobe Dreamweaver, or any GUI editor that edits an XML-based data and styling application, where the data and the metacode defining the styles are separately stored but selected through some procedure.
For example, Notepad++, an open source notepad implementation, stores style information in a document separate from the storage of the text, which always remains plaintext and without any markup except that which is defined in a separate metadata file. In fact, Notepad++ goes a bit further and defines the method of mapping external to the document itself, allowing entirely plaintext documents, batch files, and dozens of programming languages to be syntax highlighted. That's just one example, and I believe Notepad++ is based off a root engine, Scintilla?
Regardless, I feel like I can continue to think of applications that violate the patent all day. The only reason I4i went after Microsoft in an East Texas courtroom is because they knew they could get either a lot of money, or a lot of publicity, or absurdly, they actually think they deserve the sole right to implement separately stored metadata + data in XML formats.
Why would it be any different?
He's using the RMS definition of free, which... *checks news headlines* this week has more restrictions than it did last week.
Taking drafting for a semester in high school was the best thing that ever happened to my ability to print block letters on forms and things. And it shows, because when I write in that form it's vastly more legible for myself and anyone who has to read it.
Who knew, that professions based on precision would create the most workable and long-lasting script, while professions based on creativity and nuance would create the one that eventually falls by the wayside for being too complex and annoying to discern? How ever could that have happened?
All the important operations tend to be random. For a file server, you may have twenty people accessing files simultaneously. Or a hundred, or a thousand. For a webserver, it'll be hitting dozens or hundreds of static pages and, if you have database backend, that's almost entirely random as well.
For people consolidating physical servers to virtual servers, you now have two, three, ten or twenty VMs running on one machine. If every one of those VMs tries to do a "sequential" IO, it gets interlaced by the hypervisor into all the other sequential IOs. No hypervisor would dare tell all the other VMs to sit back and wait so that every IO is sequential. That delay could be seconds or minutes or hours.
Now imagine all that, and take into account that the latest Intel SSD gets around 6600 IOPS read and write. A good, fast hard drive gets 200. So you could put thirty three hard drives in RAID 0 and have the same number of IOPS, and your latency would still be worse. All the RAID0 really does for you is give you a nice big queue pipeline, like in a CPU. Your IO doesn't really get done faster, but you can have many more running simultaneously.
Given that SSDs are easily three to four times faster on sequential IO and an order of magnitude faster on random IO, I don't think it's that implausible to believe that the industry isn't ready.
This is different from AI, and is coming from someone whose expertise on the subject is demonstrable. He's not talking about AI, he's talking about simulating all of the tissue in a human brain and providing it with stimuli to determine reactions.
He's not saying it'll necessarily be a good ol' buddy ol' pal right off the bat. Probably not. Probably won't even be capable of simple arithmetic for years. On the other hand, we could simulate things like lesions effecting far away parts of the brain, various known "paths" that signals travel in the brain and ways to alter those paths or correct flaws, etc.
As well, we could simulate the effect of various drugs on large-scale phenomena in the brain to help try and understand (a.) what a drug will do before it undergoes testing, and (b.) why exactly it is that makes these drugs work so well. Both questions are currently unanswerable. We know what a drug does, but rarely do we understand the full extent of why a particular drug helps certain conditions.
It's likely Dreamspark will get Windows Server 2008 R2, not Windows 7.
The one that worked immediately created a blackhole and ended humanity in that universe. By the anthropic principle, we will never make the LHC work, and will instead see one failure after another and wonder why fate has cursed us so.
You mean the most expensive scientific machine ever made makes... duct tape?
From what I've read on the issue, such as Feynman's books and other novels targeted toward those of us who do not have a complete grasp of quantum mechanics, you are wrong.
Caveat emptor, this is merely what I've read:
Classical mechanics as explained by Feynman were the result of the sum of all possible histories, among other interpretations. Regardless of one's interpretation, Feynman and others found that as you crunch the math for larger and larger quantities of particles, the results closer and closer approximate what we think of as classical physics. As a result, classical physics is an approximation of quantum mechanics, which is a theory of how the universe really works.
I save mine in Arial, Times New Roman and Wingdings.
Is this sufficient?
Leaking? Jesus, the beta and the RC were both widely and easily available, and you can still download the RC.
The leaks are just that, they really would rather you not play around with the unofficial builds which have a lot of other debug functionality turned on a lot of the time.
Actually, it depends on the reliability. 95% reliability becomes 90.25% reliability. 50% reliability becomes 25% reliability. 1% reliability becomes 0.01% reliability.
So if your drives are very reliable, it's very slightly less than twice the failure rate. If your drives are not reliable, then it asymptotically approaches an infinitely greater risk of failure.
Statistically speaking. :)
Then you're going to have to try very hard to invalidate 80 years of science that has culminated in microprocessing technology that relies on quantum effects, thus allowing you to post on Slashdot.
I'm genuinely interested in your alternative theory for all things quantum. I don't think you have a theory, but I won't let that stop you.
Right now, we have very successful theories that explain how LEDs work, how your computer's CPU works, etc. What I don't see is an alternative theory to explain all of those things rigorously.
I'm a layperson on the subject but have read extensively, and my understanding is yes, there is a "quantum superposition" and you can prove it by running exactly the same experiment over and over again, and getting different results each time. You take the average of the results and that's the answer to your problem.
Hence, a quantum algorithm only has a probability of arriving at the correct answer. Executing a quantum algorithm several times gives you increasingly better odds in polynomial time, and because there are quantum algorithms for non-polynomial problems (see: Shor's Algorithm), you can simply repeat the algorithm and refine your result and get a right answer much faster.
While you can never guarantee your answer is 100% correct, I believe, you can get arbitrarily close. And for problems that are intractable for modern computers, problems that would require enormous supercomputers, the correct answer can be derived with, theoretically, much smaller and much faster quantum computers if you are willing to accept some small probability of error.
In addition, the solutions to these problems can be checked rather quickly, for example, if in attempting to factor the number 15, you get the number 6, a classical computer could check this very easily. Even when you scale up the numbers immensely, the verification step is not much more difficult. So Shor's Algorithm on a quantum computer, combined with a fast bignum library on a classical computer, could determine factors much more rapidly than any classical computer could on its own given current knowledge.
In short, you're wrong.
On the other hand, I have dealt with GPL programs that ask me to agree to the GPL before I download.
I have Windows 7 installed on my x60 and it went fine with a USB CD/DVD drive.
Pictures shot in digital or transferred from digital intermediate tend to come out blurry when put on film no matter what, and are a pain to focus. That said, the quality is somewhere between 2K (around 1080P) and 4K (four times the pixels of 1080P, an image 4000 pixels wide by ~2000 pixels tall.)
Projectionist, lol.
More like, when he stops stealing TVs, you stop charging him with theft.
They're charging Microsoft with crimes that they haven't yet committed, and every indication seems to be Microsoft is willing to meet some rather absurd demands.