Because you may well be mistaken about what is incorrect and what is just a different notation. Take a look at a random decent reference for a second and scroll down to momentum.
You will see that if you call the mass of the object its rest mass, then you need a relativistic correction term. If you call the mass of the object its total mass including the mass of its kinetic energy, Newton's formulation still holds.
It is entirely a question of definition. Here is a random argument for defining the mass one way. OTOH my background is in math. I learned physics from older texts, and from the point of view of a mathematician the concept of relativistic mass is a lot cleaner than carrying around the correction term.
Now when I am explicit about saying that I am using the word mass to mean mass with the relativistic correction (ie with the mass of the kinetic energy counted as mass), what I said becomes both unambiguous and clearly correct.
I can also point out that from the point of view of a purist it is cleaner to think that way. For instance a fraction of your mass is the mass of the heat of your body, which is just kinetic energy. Likewise a fraction of the mass of the Solar System is from the motion of the planets around the Sun. To me the mass of a system of objects should just be the sum of their masses. Using relativistic masses that is true. Using rest masses it doesn't quite add up.
Someone I know writes device drivers for a living. One Halloween he put together a costume out of free NT installation CD's he had been given that he was never going to use.
I nearly died laughing when I saw it.:-)
I am not sure that he would have resold them. Something about it being an, "abuse of computer rights"...:-) (Actually he probably would have.)
What I said is absolutely correct in special relativity. The momentum of something is its mass times its velocity. You have to include the mass of its kinetic energy as part of its mass of course, but why should kinetic mass be treated any differently than any other kind of mass? As an extreme example, photons have a mass, but they have have no rest mass.
As for conservation of momentum and energy, the old sci.physics FAQ had an neat section on that pointing out that even stating global conservation laws in general relativity is problematical. And yes, I have heard of the 3K background radiation. I have even observed informed discussions on whether there is a non-conservation of energy going on in that cooling process.
The sail temperature issue is non-trivial. If the mass of the energy making up a photon is m, the energy is m*c*c, and the momentum is m*c. So no, the energy is not 0.5mv^2, it is off by a full factor of 2. Big whoop. The ratio between energy and momentum scales approximately linearly with speed (varying by a full factor of 2), and that is what matters here. Unless you have a very reflective surface, what you can deliver is limited by when your sail will melt.
Incidentally the wavelength is irrelevant to how efficient light sails are. Assuming that you get a good reflection, they are all equally inefficient because of the ratio between energy and momentum that is a basic consequence of SR.
Do you have any other "corrections" for me?
Ben
PS Trivia note. Newton's statement of his three laws is actually completely correct in SR. The (slightly) incorrect f=ma formulation came after.
And decided that the moderators would likely misunderstand the response I had in mind.
I was thinking something like, "You are no Einstein" and pointing him at E=MC^2. Then following up with the particle-wave duality and mentioning the experimental issues with diffracting a herd of elephants.
All true, but given the level of science around here, some moderator would be sure to misunderstand.:-/
There are considerable difficulties with the laser. For an interesting and detailed discussion, try this paper. (For people who don't know, the Dr. Forward they refer to is the same Robert L. Forward that writes sci-fi occasionally.)
1) Lasers have existed for some time and have no problems with focus.
2) That is the potential catch. Even if you reflect as much as you can, the remainder still adds up to a lot of heat. They claim it can work.
3) That kind of technical detail is less important than the proof of concept.
4) For all intents and purposes, the range of a laser is unlimited.
5) "Crisp!" is more like it. This beam would have a reasonable momentum and one heck of a lot of energy. Light is very inefficient in that way. (The only win is that you don't have to carry your energy around.)
6) In answer to another person's question, tears are a smaller deal than you would think. You want the sail itself to be light-weight, and therefore patches of sail would be connected to lines that hold the tension. Tears would degrade the sail over time, but unless you lose major cables, you are OK.
The interesting thing about/. is how little basic science a lot of the people who think they are qualified to answer your question have.
Anyways, here are the answers:
1. I have no idea how long until they try this in space.
2. Yes, by Newton's third law the source is pushed back. As long as that is a planet, you won't care much.
3. a) Think ping-pong balls. When you hit them you can change the direction of motion without changing their speed. That is a transfer of momentum.
b) Momentum is mass*velocity. The velocity of light in a vacuum is constant. The mass is dependent on the frequency. So light coming from a gravity well, or from something receeding from us loses energy, and the photon is "red-shifted".
c) They would want to avoid absorbing photons since that would burn up the sail very fast. Energy scales as velocity squared, and light moves pretty fast, so it delivers a *lot* of heat for the momentum.
Err, no they don't. The light is absorbed, not reflected. If it was reflected then they'd have to slow down to transfer momentum, which is obviously impossible.
Wrong. If you redirect the light, then by good old conservation of momentum you have to have a transfer of momentum. No slowdown needed. The light is reflected for several reasons:
1) If you absorb it your sail would quickly burn up.
2) By reflecting it the light has a bigger change of momentum, giving more push.
3) By choosing the direction of reflection you get some steering capability.
4) There is a theoretical double-sail configuration for braking, but it only works if you are using reflection. (One of the potential interstellar craft on the drawing board uses this design, see the book "Indistinguishable From Magic" for details.)
Robert L. Forward's book of that title has an interesting section on virtually anything to do with space that geeks could be interested in.
Space sails in particular were more interesting than I thought. Yes, there is a braking system designed using them. And the general idea is that we would set up a laser on Mercury, possibly solar powered, and use that to propel an interstellar craft. The win? The craft does not need to carry its own fuel. That is not a small win.
Anyways if you are interested in space, and are interested in what physicists who do blue sky research for NASA think about, read this book. You won't regret it.
I hate people talking how 2.4 will fix everything, 2.2 surely didn't.
Where did I say that 2.4 will fix everything?
I said that there is a specific problem, known in 2.2 that has turned up before, that is a potential explanation for this bad result.
There are other known (and fixed) scheduler problems.
Encountering any combination of these in 2.2 benchmarks is to be expected. Don't make these out to be more or less than indications that 2.2 had some obvious room for improvement.
I am sure that 2.4 will have more problems. However many problems that turned up in benchmarking 2.2 have been fixed (because they turned up in benchmarking 2.2), and preliminary benchmarks of 2.4 (eg the recent SpecWeb result where it nearly tripled Windows 2000 on a similar 4 CPU box) indicate this.
Now will 2.4 be ready for the enterprise, as they like to say? Not really. First of all until it has been through a few point releases, I would expect some significant bugs. (To be expected in any software.) Aside from that issue, it lacks many managability tools, a volume manager, more work needs to be done on failover, journaling filesystems are needed, etc. I have been convinced by Larry McVoy's argument that further work on SMP is not needed, NUMA (done through clustering and virtual operating systems) is.
These are known problems. Work is being done on them. However there will be room for complaint about Linux vs more mature systems for some time to come. However problems are getting solved, and Linux is moving up the food chain, fast.
The "thundering herd" problem that was identified in Mindcraft and fixed in 2.4, isn't that still present in RedHat 6.1? (BTW calling it "Linux 6.1" really irritated me.) That could explain a sudden drop-off. It is not a problem, not a problem, then suddenly becomes a problem and as soon as you get a slow-down, you get a real traffic jam.
When I see attempts to integrate Microsoft products and push them through their paces, I see bugs popping up. YMMV, but that seems to be a pretty common experience.
Now looking at your user info you have a career that involves your status as a pundit on Microsoft products. Watching those products wane in influence may be pretty personally threatening. But that doesn't mean that Microsoft doesn't fully deserve what is happening to it. And it doesn't mean that things are better today than they would be if Microsoft had not set out to abuse its monopoly position.
One of the outcomes of the Mindcraft saga was this wonderful set of benchmarks by C't.
One of the things that they did is force tests that stressed various parts of the OS. For me one of the more telling ones was the selection against many files, where the ability to serve off of disk (as opposed to out of RAM) was being pushed.
Linux won, of course. But I wonder whether Win2K is better at this than NT was...
Don't allow a post to be posted when it matches an existing post. (Check for matches in a 1-way hash-value first. Otherwise the effort of this matching will be too hard on the database.)
This reduces the problem.
Then put in a scan for "known signatures" of specific spammers. (For instance that mug of beer.) To find those add a "spam" option for moderators. If a user starts being moded for spam, investigate the pattern...
One of the things that software blurs is the distinction between software distribution and a public performance. Indeed in today's world an increasingly common use of GPLed software is public performance. (Think of slash!)
Unfortunately the GPL nowhere explicitly tackles the issues of public performance. From my reading, term 0 explicitly disclaims any relevance to public performance. Others may disagree. But I think that with the direction that this industry is heading it would be extremely useful to have a version 3 produced that explicitly addressed the question of public performance.
Cheers, Ben
PS I would also like to see section 3 of the GPL replace "operating system" with the more general "basic operating environment". Why? Well can you GPL a Java applet? Look at the definition of "source code for a work". The standard Java libraries are often not part of the OS. But they are probably distributed with your JVM, and it makes no sense to make you worry about it!
Have you ever tried to maintain a Perl system? (Hint, try "perldoc Test".)
Have you ever compared how much nicer it is to have an automated test-suite than it is to not have one?
Trust me, any large distributed project that is used across multiple platforms can benefit from a test-suite. Making it easy to build one is a good thing. Being able to drive an interactive gui program that you may not have source for from a command line is in and of itself darned useful. (The lack of an integrated ability to be so driven during batch processing is a major drawback of most of them, and is one of the things that I dislike about having to work with stuff that was thrown together in VB.)
OK, now that I have addressed usefulness, let me address Debian. Debian's problems would be totally resolved by this. The program would be distributed with dynamically linked against a program with no license problems. The fact that most users would actually link it to Qt after you distribute it is just fine. The GPL doesn't prevent you from doing that, it only prevents you from distributing it that way.
Get permission from Troll to use their interface files (literally).
Put in useless stubs for everything you don't want.
Turn it into a scriptable toolkit, usable for writing automated tests against. Basically do for Qt-based applications what Expect does for interactive command-line applications.
Start working on an automated test suite, similar to the one that (for instance) Perl uses.
This isn't a hard task, read the license for yourself.
It is a far more subtle document than one would think just reading it, and sometimes it has surprising consequences. For instance here is a summary of Debian's issues with Qt. Can these be addressed? Yes. Are they somehow claiming that Qt is a derivative work of KDE? Nope. They are claiming that KDE is a work that cannot be distributed by the terms of its own license.
But I think that there is a way to solve this problem for once and for all which is not discussed. Produce a GPL-friendly interface to Qt with a GPL-friendly implementation behind it. Have it support the full programming interface, but don't worry too much about it "looking nice" or usability for the end-user. Ship KDE with this and with some easy way to switch from this (bad) default to Qt. Then KDE can ship in compliance with the GPL completely separately from any license issues that Qt may have.
Looking at that, section 406 (c) makes it utterly clear that a claim of "as is" or something similar to that (eg what the GPL does) is sufficient to disclaim all warranties except that of section 401. Section 401 only applies to information, not programs. (ie Software is in the clear, geographica databases are not.) And that can be disclaimed with words to the effect of, "There is no warranty against interference with your enjoyment of the information or against infringement."
A lot of proprietary software cannot be legally run at all.
Hey...:-)
IANAL but I would strongly doubt that your interpretation would stand up. After all you are supposed to use common sense. Actually running a program would probably be covered by something like fair use.
The public performance issue I raised earlier I still have questions about (though Karsten outlined a possible solution) but your explanation of the warranty issue sounds valid to me.
The key is not in the wording of the GPL, but rather in carefully following the instructions for how to apply the GPL to your program.
However when it comes to warranties, I am somewhat concerned. Term 5 explicitly says that you can use the software without accepting the license. Terms 11 and 12 are the disclaimer. What happens if Joe random user uses Linux, loses data because of a kernel bug, and then sues Linus saying that he had used the software but never agreed to the GPL as is his right (see term 5) and therefore had never accepted that Linus is not liable for any loss of data and business on his part!
When you do that in the background the OS will have to start paging in the encrypted file. It will stay in memory until that memory is needed for something else. But the encrypted file is a worst case scenario for the compression algorithm.
If someone has an encrypted filesystem this will actually be an extremely common case!
Really, your "Oh we will never hit that in the real world" is exactly how programmers f*ck up time after time again. You may not see how it will happen, but it will happen eventually and people will get hosed.
Deal with your corner cases before turning your code lose on the world, please.
Slashdot Mirror
Because you may well be mistaken about what is incorrect and what is just a different notation. Take a look at a random decent reference for a second and scroll down to momentum.
You will see that if you call the mass of the object its rest mass, then you need a relativistic correction term. If you call the mass of the object its total mass including the mass of its kinetic energy, Newton's formulation still holds.
It is entirely a question of definition. Here is a random argument for defining the mass one way. OTOH my background is in math. I learned physics from older texts, and from the point of view of a mathematician the concept of relativistic mass is a lot cleaner than carrying around the correction term.
Now when I am explicit about saying that I am using the word mass to mean mass with the relativistic correction (ie with the mass of the kinetic energy counted as mass), what I said becomes both unambiguous and clearly correct.
I can also point out that from the point of view of a purist it is cleaner to think that way. For instance a fraction of your mass is the mass of the heat of your body, which is just kinetic energy. Likewise a fraction of the mass of the Solar System is from the motion of the planets around the Sun. To me the mass of a system of objects should just be the sum of their masses. Using relativistic masses that is true. Using rest masses it doesn't quite add up.
Regards,
Ben
Someone I know writes device drivers for a living. One Halloween he put together a costume out of free NT installation CD's he had been given that he was never going to use.
:-)
:-) (Actually he probably would have.)
I nearly died laughing when I saw it.
I am not sure that he would have resold them. Something about it being an, "abuse of computer rights"...
Cheers,
Ben
What I said is absolutely correct in special relativity. The momentum of something is its mass times its velocity. You have to include the mass of its kinetic energy as part of its mass of course, but why should kinetic mass be treated any differently than any other kind of mass? As an extreme example, photons have a mass, but they have have no rest mass.
As for conservation of momentum and energy, the old sci.physics FAQ had an neat section on that pointing out that even stating global conservation laws in general relativity is problematical. And yes, I have heard of the 3K background radiation. I have even observed informed discussions on whether there is a non-conservation of energy going on in that cooling process.
The sail temperature issue is non-trivial. If the mass of the energy making up a photon is m, the energy is m*c*c, and the momentum is m*c. So no, the energy is not 0.5mv^2, it is off by a full factor of 2. Big whoop. The ratio between energy and momentum scales approximately linearly with speed (varying by a full factor of 2), and that is what matters here. Unless you have a very reflective surface, what you can deliver is limited by when your sail will melt.
Incidentally the wavelength is irrelevant to how efficient light sails are. Assuming that you get a good reflection, they are all equally inefficient because of the ratio between energy and momentum that is a basic consequence of SR.
Do you have any other "corrections" for me?
Ben
PS Trivia note. Newton's statement of his three laws is actually completely correct in SR. The (slightly) incorrect f=ma formulation came after.
And decided that the moderators would likely misunderstand the response I had in mind.
:-/
I was thinking something like, "You are no Einstein" and pointing him at E=MC^2. Then following up with the particle-wave duality and mentioning the experimental issues with diffracting a herd of elephants.
All true, but given the level of science around here, some moderator would be sure to misunderstand.
Cheers,
Ben
There are considerable difficulties with the laser. For an interesting and detailed discussion, try this paper. (For people who don't know, the Dr. Forward they refer to is the same Robert L. Forward that writes sci-fi occasionally.)
Cheers,
Ben
1) Lasers have existed for some time and have no problems with focus.
2) That is the potential catch. Even if you reflect as much as you can, the remainder still adds up to a lot of heat. They claim it can work.
3) That kind of technical detail is less important than the proof of concept.
4) For all intents and purposes, the range of a laser is unlimited.
5) "Crisp!" is more like it. This beam would have a reasonable momentum and one heck of a lot of energy. Light is very inefficient in that way. (The only win is that you don't have to carry your energy around.)
6) In answer to another person's question, tears are a smaller deal than you would think. You want the sail itself to be light-weight, and therefore patches of sail would be connected to lines that hold the tension. Tears would degrade the sail over time, but unless you lose major cables, you are OK.
Cheers,
Ben
The interesting thing about /. is how little basic science a lot of the people who think they are qualified to answer your question have.
Anyways, here are the answers:
1. I have no idea how long until they try this in space.
2. Yes, by Newton's third law the source is pushed back. As long as that is a planet, you won't care much.
3.
a) Think ping-pong balls. When you hit them you can change the direction of motion without changing their speed. That is a transfer of momentum.
b) Momentum is mass*velocity. The velocity of light in a vacuum is constant. The mass is dependent on the frequency. So light coming from a gravity well, or from something receeding from us loses energy, and the photon is "red-shifted".
c) They would want to avoid absorbing photons since that would burn up the sail very fast. Energy scales as velocity squared, and light moves pretty fast, so it delivers a *lot* of heat for the momentum.
Cheers,
Ben
Err, no they don't. The light is absorbed, not reflected. If it was reflected then they'd have to slow down to transfer momentum, which is obviously impossible.
Wrong. If you redirect the light, then by good old conservation of momentum you have to have a transfer of momentum. No slowdown needed. The light is reflected for several reasons:
1) If you absorb it your sail would quickly burn up.
2) By reflecting it the light has a bigger change of momentum, giving more push.
3) By choosing the direction of reflection you get some steering capability.
4) There is a theoretical double-sail configuration for braking, but it only works if you are using reflection. (One of the potential interstellar craft on the drawing board uses this design, see the book "Indistinguishable From Magic" for details.)
Sorry,
Ben
Robert L. Forward's book of that title has an interesting section on virtually anything to do with space that geeks could be interested in.
Space sails in particular were more interesting than I thought. Yes, there is a braking system designed using them. And the general idea is that we would set up a laser on Mercury, possibly solar powered, and use that to propel an interstellar craft. The win? The craft does not need to carry its own fuel. That is not a small win.
Anyways if you are interested in space, and are interested in what physicists who do blue sky research for NASA think about, read this book. You won't regret it.
Cheers,
Ben
I hate people talking how 2.4 will fix everything, 2.2 surely didn't.
Where did I say that 2.4 will fix everything?
I said that there is a specific problem, known in 2.2 that has turned up before, that is a potential explanation for this bad result.
There are other known (and fixed) scheduler problems.
Encountering any combination of these in 2.2 benchmarks is to be expected. Don't make these out to be more or less than indications that 2.2 had some obvious room for improvement.
I am sure that 2.4 will have more problems. However many problems that turned up in benchmarking 2.2 have been fixed (because they turned up in benchmarking 2.2), and preliminary benchmarks of 2.4 (eg the recent SpecWeb result where it nearly tripled Windows 2000 on a similar 4 CPU box) indicate this.
Now will 2.4 be ready for the enterprise, as they like to say? Not really. First of all until it has been through a few point releases, I would expect some significant bugs. (To be expected in any software.) Aside from that issue, it lacks many managability tools, a volume manager, more work needs to be done on failover, journaling filesystems are needed, etc. I have been convinced by Larry McVoy's argument that further work on SMP is not needed, NUMA (done through clustering and virtual operating systems) is.
These are known problems. Work is being done on them. However there will be room for complaint about Linux vs more mature systems for some time to come. However problems are getting solved, and Linux is moving up the food chain, fast.
Regards,
Ben
An immediate thought.
The "thundering herd" problem that was identified in Mindcraft and fixed in 2.4, isn't that still present in RedHat 6.1? (BTW calling it "Linux 6.1" really irritated me.) That could explain a sudden drop-off. It is not a problem, not a problem, then suddenly becomes a problem and as soon as you get a slow-down, you get a real traffic jam.
Just guessing...
Ben
When I see attempts to integrate Microsoft products and push them through their paces, I see bugs popping up. YMMV, but that seems to be a pretty common experience.
Now looking at your user info you have a career that involves your status as a pundit on Microsoft products. Watching those products wane in influence may be pretty personally threatening. But that doesn't mean that Microsoft doesn't fully deserve what is happening to it. And it doesn't mean that things are better today than they would be if Microsoft had not set out to abuse its monopoly position.
Regards,
Ben
I can personally vouch that gcc is widely used, for instance on Wall St.
Perhaps online trading systems are not complex enough for you?
And if you want something useful, look at Perl. Built with gcc. Running on FreeBSD. Sounds like kid's play? Sounds like Yahoo to me!
Regards,
Ben
Microsoft heard that people were setting up clusters so that they could get fail-over.
Therefore Microsoft outdid themselves and made it a standard OS feature that it would fail over and over and over and...
Cheers,
Ben
One of the outcomes of the Mindcraft saga was this wonderful set of benchmarks by C't.
One of the things that they did is force tests that stressed various parts of the OS. For me one of the more telling ones was the selection against many files, where the ability to serve off of disk (as opposed to out of RAM) was being pushed.
Linux won, of course. But I wonder whether Win2K is better at this than NT was...
Cheers,
Ben
Don't allow a post to be posted when it matches an existing post. (Check for matches in a 1-way hash-value first. Otherwise the effort of this matching will be too hard on the database.)
This reduces the problem.
Then put in a scan for "known signatures" of specific spammers. (For instance that mug of beer.) To find those add a "spam" option for moderators. If a user starts being moded for spam, investigate the pattern...
Imperfect, but it would really help.
Cheers,
Ben
IANAL but here are some previous thoughts of mine on the GPL.
One of the things that software blurs is the distinction between software distribution and a public performance. Indeed in today's world an increasingly common use of GPLed software is public performance. (Think of slash!)
Unfortunately the GPL nowhere explicitly tackles the issues of public performance. From my reading, term 0 explicitly disclaims any relevance to public performance. Others may disagree. But I think that with the direction that this industry is heading it would be extremely useful to have a version 3 produced that explicitly addressed the question of public performance.
Cheers,
Ben
PS I would also like to see section 3 of the GPL replace "operating system" with the more general "basic operating environment". Why? Well can you GPL a Java applet? Look at the definition of "source code for a work". The standard Java libraries are often not part of the OS. But they are probably distributed with your JVM, and it makes no sense to make you worry about it!
Have you ever tried to maintain a Perl system? (Hint, try "perldoc Test".)
Have you ever compared how much nicer it is to have an automated test-suite than it is to not have one?
Trust me, any large distributed project that is used across multiple platforms can benefit from a test-suite. Making it easy to build one is a good thing. Being able to drive an interactive gui program that you may not have source for from a command line is in and of itself darned useful. (The lack of an integrated ability to be so driven during batch processing is a major drawback of most of them, and is one of the things that I dislike about having to work with stuff that was thrown together in VB.)
OK, now that I have addressed usefulness, let me address Debian. Debian's problems would be totally resolved by this. The program would be distributed with dynamically linked against a program with no license problems. The fact that most users would actually link it to Qt after you distribute it is just fine. The GPL doesn't prevent you from doing that, it only prevents you from distributing it that way.
Cheers,
Ben
Get permission from Troll to use their interface files (literally).
Put in useless stubs for everything you don't want.
Turn it into a scriptable toolkit, usable for writing automated tests against. Basically do for Qt-based applications what Expect does for interactive command-line applications.
Start working on an automated test suite, similar to the one that (for instance) Perl uses.
Cheers,
Ben
This isn't a hard task, read the license for yourself.
It is a far more subtle document than one would think just reading it, and sometimes it has surprising consequences. For instance here is a summary of Debian's issues with Qt. Can these be addressed? Yes. Are they somehow claiming that Qt is a derivative work of KDE? Nope. They are claiming that KDE is a work that cannot be distributed by the terms of its own license.
But I think that there is a way to solve this problem for once and for all which is not discussed. Produce a GPL-friendly interface to Qt with a GPL-friendly implementation behind it. Have it support the full programming interface, but don't worry too much about it "looking nice" or usability for the end-user. Ship KDE with this and with some easy way to switch from this (bad) default to Qt. Then KDE can ship in compliance with the GPL completely separately from any license issues that Qt may have.
Cheers,
Ben
Someone else gave a link to the text of UCITA.
Looking at that, section 406 (c) makes it utterly clear that a claim of "as is" or something similar to that (eg what the GPL does) is sufficient to disclaim all warranties except that of section 401. Section 401 only applies to information, not programs. (ie Software is in the clear, geographica databases are not.) And that can be disclaimed with words to the effect of, "There is no warranty against interference with your enjoyment of the information or against infringement."
Cheers,
Ben
A lot of proprietary software cannot be legally run at all.
:-)
Hey...
IANAL but I would strongly doubt that your interpretation would stand up. After all you are supposed to use common sense. Actually running a program would probably be covered by something like fair use.
Cheers,
Ben
The public performance issue I raised earlier I still have questions about (though Karsten outlined a possible solution) but your explanation of the warranty issue sounds valid to me.
The key is not in the wording of the GPL, but rather in carefully following the instructions for how to apply the GPL to your program.
Cheers,
Ben
Here is the license and from a previous discussion, some of my analysis.
However when it comes to warranties, I am somewhat concerned. Term 5 explicitly says that you can use the software without accepting the license. Terms 11 and 12 are the disclaimer. What happens if Joe random user uses Linux, loses data because of a kernel bug, and then sues Linus saying that he had used the software but never agreed to the GPL as is his right (see term 5) and therefore had never accepted that Linus is not liable for any loss of data and business on his part!
What next?
Regards,
Ben
mmap() a large encrypted file. Start decrypting.
When you do that in the background the OS will have to start paging in the encrypted file. It will stay in memory until that memory is needed for something else. But the encrypted file is a worst case scenario for the compression algorithm.
If someone has an encrypted filesystem this will actually be an extremely common case!
Really, your "Oh we will never hit that in the real world" is exactly how programmers f*ck up time after time again. You may not see how it will happen, but it will happen eventually and people will get hosed.
Deal with your corner cases before turning your code lose on the world, please.
Regards,
Ben