There needs to be some way on Slashdot to moderate the occasional, ultra-exceptional post up to about +10, give the post it's own front-page article, something to distinguish it from the hundreds of "ordinary" +5, Insightfuls that get rated every day.
I can't give you any more points, so I'll just reply with this pointlessly wordy "me too" - this is such a post.
Ultimately hardware options are not a solution pirates can use, since watermarking could easily identify which person freed some content from DRM.
Unless, of course, that content is ever sold for cash, like every video on the racks at every store from Best Buy to Walmart. Will popular videos all start going "direct-to-internet" any time soon? I doubt it. Publishers may talk a big talk about how much money they estimate they lose from piracy, but when they have to put that estimate up against the amount they know they'll lose from keeping their products out of every big retail outlet in the world, I doubt it'll seem so huge.
As soon as you need actual hardware to pirate the signal, copying movies becomes a restricted occupation again, just like selling free cable boxes.
No, copying movies the first time becomes a restricted occupation. Once a single unencrypted copy exists, then making a million more is no more difficult than it is today.
Whip out your favorite P2P client, and search for some copyrighted video. Do you see a hundred different rips made by each of the hundred different people sharing a copy? No, you see one or two of the best rips, each with hundreds of identical copies shared, in part because the swarming download protocols and hashing algorithms fundamentally encourage that behavior.
So what difference will in-monitor DRM make? Instead of having a few zealous groups using software to rip tons of movies that are then shared by millions of people, we'll have a few zealous groups using hardware to rip tons of movies that are then shared by hundreds of millions of people.
Wait - why will there be more people sharing these rips? Because most people will own some of the billions of non-DRM-capable monitors in existance, and the moron DRM-using publishers will have thus made it impossible for them to play a full-quality copy of these videos unless they have an illegal copy. Publishers couldn't do anything more stupid if they put a "Download free movies on P2P! It's the best!" advertisement at the start of every show!
You sound as if you think there's a finite list of terrorists out there, and as soon as we scratch every name off the list then the terrorism problem will be solved! Not a chance - those people were made into terrorists, they weren't born that way. It doesn't matter how many of this generation's terrorists die for their crimes, if there's another equally large generation coming right after them.
I'm not agreeing with the idea that the way to stop the creation of new terrorists from religious zealots is to "treat them better" or "stop offending them" - for all I know it may be just the opposite. But we do need to understand these people, desperately, because it's only understanding or dumb luck that's going to allow us to stop the terrorist meme, and I'm not feeling very lucky.
I admit vengeance sounds pretty nice, but I'd gladly trade it for a more scientific understanding of the sociology of violence. The question of how we make more dead terrorists isn't nearly as important as the question of how we protect more live innocents. If capital punishment for mass murderers is part of that, then fine, but don't lose sight of the goal just because one step along the way is more emotionally compelling.
Isn't this sort of thing calculable farther in advance? There shouldn't be a whole lot of angular momentum being added or subtracted from the Earth's rotation.
Do I need a new glibc? Or any other POSIX library, for that matter? If this is a new announcement then presumably every implementation of mktime(), localtime(), gmtime(), etc. needs to be updated.
This isn't just about parallel computing - in fact if you'll read the article you'll see that they're using MPI for handling parallelism! Grid computing isn't about reinventing inter-node communications - it's more about inventing inter-node scheduling.
Your cluster - is it so fast that you're never stuck waiting for jobs to finish? If not, then you could probably benefit from being able to borrow time on someone's larger system. Is your cluster so well-utilized that the load's always around 1? If not then you've probably got spare capacity that someone else could benefit from. The fact that both you and those others are using MPI is necessary but insufficient to allow you to cooperate.
Three foreigners: a businessman, physicist, and mathematician, are talking about the country they're all visiting for the first time.
Suddenly, the businessman points out the window in surprise. "Look at that! The sheep in Scotland are black!"
Amused at how readily his new friend jumps to conclusions, the physicist corrects him: "No, all we can be certain of is that some of the sheep in Scotland are black."
The mathematician looks out the window himself, and corrects the both of them: "We know there exists a sheep in Scotland which is black on at least one side."
Yeah, I know, we all hate making slide shows, but if you can develop a stage presence and get gigs with large enough software companies then you'll really have an opportunity to make your physical activity and body language work for you.
Here, try a simple exercise to see if this might be the right career track for you. Stand up, and start stomping your feet and clapping your hands, while screaming "Developers! Developers! Developers! Developers!"
I fail to see why you concluded that any reference frame used for FTL must be backwards in time compared to our own, as light does not travel instantaneously, therefore it is hypothetically possible for something to move faster than light without traveling backwards in time.
See my other comment. It is possible to travel faster than light without moving backwards in time in a given reference frame; it is not possible to travel faster than light in one reference frame without moving backwards in time in some other frame.
Actually, I believe some laboratories have managed to accellerate light to faster than its normal speed, though I can't be bothered to dig up any articles on it.
I assume you're talking about the experiments here. Those showed a "group velocity" faster than light, but not faster than light transmission of mass/energy or information and so no problems with causality (and no utility for FTL travel).
And yes, FTL is often used for ambiance, as, quite frankly, it's rather difficult to have a multi-world society without FTL, as mere communications would take decades at the least, and interstellar trade would be pretty much nonexistant.
"Difficult" is just a dismissive way of saying "would make a fascinating story". See the first books of the "Rissa and Tregare" series by FM Busby for my favorite examples.
Besides, how many writers out there are really capable of describing a civilization fully utilizing a hundred billion stars? Not nearly as many as the number of writers who think they're up to it and are mistaken. Look at the universe-shrinking cameos in the Star Wars prequels: from watching the movies, you'd think Chewbacca was the only Wookie in the galaxy, not one out of a billion.
Of course, the easiest way to have a multi-world society without putting those worlds decades and centuries apart is to put them around the same star. It's ironic that people who think we need FTL to make space small enough to write about don't realize just how big space really is. There is enough free-falling raw material and solar power to support quintillions of people in this solar system alone - surely that's enough for a few interesting stories here too.;-)
The biggest problem in my mind is that paths which are faster than light in one reference frame are all backwards in time in some other reference frame.
Can you please provide an example of this?
Any space-like interval (two points in space-time whose separation distance exceeds the speed of light multiplied by their separation time, with both measured in the same reference frame) is an example, because for any two such points there is another reference frame in which they are simultaneous. If AB is such an interval, then even if in your reference frame event A occurred before event B, then in the reference frame of someone traveling fast enough (relative to your velocity) in the direction from A to B they were simultaneous. And, in the reference frame of someone traveling even faster in the AB direction, event B occurred first.
If you were at event B, event A would be in your past in the original reference frame. If you then accelerate to a very high speed away from A, then in your new reference frame A can be in your future. An FTL drive which works independently of reference frame (the way all physical laws work as far as we know) could then take you to A's location in space but prior to A in time. To finish the loop, decelerate back to the original reference frame (which now puts B in your future), then FTL back to B's location in space but at an earlier time than you left.
I'm moderately sure he does understand the problems. I do, yet I don't think FTL is impossible. The only thing I know to likely be impossible is to accellerate a mass to beyond the speed of light in normal spacetime.
That's the first problem with FTL that comes to mind for many people, but I don't think it's the most important. The biggest problem in my mind is that paths which are faster than light in one reference frame are all backwards in time in some other reference frame.
Does this mean that FTL is so ridiculous that it should be off-limits to science fiction? Well, no (certainly even time travel shouldn't be off-limits, however unlikely!), but I suspect that many stories which use FTL travel are doing so because they want it for ambiance, not because they also want time travel in their story or because they're intentionally abandoning relativity.
I seem to remember that one Arthur C. Clark has been officialy recognized as the "inventor" of the satelite concept...
Not quite. AFAIK Clarke was the first person to publish the idea of geosynchronous communications satellites, but the idea of artificial satellites in general is much older.
Why would a story set in an interstellar spaceship suddenly become too mundane if that spaceship is limited to light speed? Would there be too much of the "present day" in a story about the lives of some of the quintillions of people an average solar system could support in orbital cities? Are nanomachines too boring when authors are careful not to turn them into thermodynamics-defying magic dust?
Nobody wants science fiction stripped of the fiction, some people just don't want it all stripped of the science. Science fantasy can still be entertaining, but it shouldn't be allowed to slip into otherwise consistent science fiction any more than traditional fantasy should corrupt traditional fiction. I suspect most of the Slashdot readers currently whining about how "why does everything have to be based on real facts" would turn the TV off in disgust if the next episode of "24" featured a nuclear bomb stolen by leprechauns or if "CSI" started occasionally solving mysteries with magic spells.
I mainly practice safe networking with a Linksys router/firewall at work and an OpenBSD gateway at home.
Does your firewall block outgoing HTTP connections and incoming email? If not, then it's not going to help against attacks like this PNG bug which are propagated through user-pulled data rather than attacker-pushed port connections. Such attacks exist for Linux, too. There is no such thing as "safe networking", and the only way to come close is to keep every connected computer up to date. I think Fedora still comes with up2date searching for updates in the background and displaying the results on a panel icon. Unless you use something else for security updates you ought to be clicking on that every time it finds something new.
After all, Japan always had to import 100% of their petroleum needs, hence the reason why they've always emphasized high energy efficiency.
Not always. At one point, they emphasized using a powerful military to protect overseas oil assets, even going so far as to preemptively attack the largest threat to that empire. The result ended up including over a million Japanese casualties - it's not surprising they've become much less enamored of that strategy than the USA.
50fps (at 1024x768) in Half Life 2 according to Tom's Hardware, and $64 (including shipping) according to Pricewatch. There are probably better values out there, this is just the first model I thought to check.
Yeah, I know, if you want 1600x1200 + 8xAA + 8xAF you'll have to pay through the nose for it, but if you just want the image quality that a console gives you, you can get it on the PC cheaply.
I can't see how that round shape is going to be able to clean the corners of my kitchen tile. If a vacuum misses the corners of the carpet, that's not a big problem - they don't get much foot traffic anyway. But in the kitchen? The corners are where every little spill accumulates!
No, I'm going to stick with my original mopping plan: waiting until one of the stains attains sentience, then negotiating with it to defend me against the others.
It's hell to read a 100% width article on a large monitor even after blowing up the fontsize.
By "it's hell", do you mean "I have to shrink my browser window horizontally"? Those of us who like reading 100% of our screen width can't widen fixed-width pages, but a page that respects the reader's browser preferences can be as narrow as you want it to be.
Unless the torrent you're downloading contains commercials, including those from your local market, you are paying for fuck-all. You actually believe the $0.20 per month SCIFI gets from you entitles you to their entire lineup, commercial-free?
On my computer, I get Sci-Fi's entire lineup, commercial free, by pressing a "skip" button whenever a commercial starts and jumping immediately to the resumed show. I know that function's either hidden or nonexistant on commercial PVRs, but it's really only an incremental improvement on "mute" and "fast forward" anyway. Even permanently cutting out commercials on programs I want to archive is something that's always been possible for anyone with two VCRs and too much time on their hands.
So is what I'm doing unethical? Morally wrong but allowed via legal loophole? Illegal?
I hope not. If TV channel owners are expecting me to watch those commercials, they probably ought to have me sign something to that effect. On the other hand, if the Sci-Fi channel gets 20% of my viewing time but 0.4% of my cable bill, perhaps I'm not the one with whom they should be renegotiating a contract.
I've seen case law (wish I remember where) saying APIs aren't copyrightable - you can't necessarily just #include "whatever.h" into your program, but you can rewrite a "my_whatever.h" to declare the same symbols and binary compatible structs. If that's true, if a GPL program has an existing binary plugin loading interface, and if you're willing to do a bit of redundant work, you can compile a module that meets that interface and isn't legally a derivative work.
Can you redistribute that plugin without source code along with the GPL program? I have no idea. But I'll bet you can redistribute the plugin alone, never distribute the GPL program, and be in the clear. Remember, you can't sue someone for "GPL violations" - nobody has to agree to the GPL. GPL violations are only important because they remove a defense against copyright infringement lawsuits. So, if you're never going to redistribute copyrighted material in the first place, why worry about whether you would be licensed to do so or not?
That "abstract_api1" better be calling system(" proprietary_prog args") if you want to be GPL-compatible. What's more, you'd better be sure that proprietary_prog doesn't even include more GPL header files than fair use allows.
What you're describing is LGPL-compatible: with the LGPL you only have to release code to a library and modifications to that library. With the GPL, you have to release modifications to the entire derived work.
Now, you might be able to convince a court that sufficiently abstractly linked code isn't a derived work. I'd at least make sure nothing is statically linked in that case, but I know RMS considers even dynamically linked binary-only code to be a GPL violation - who knows what a court would say. I know Nvidia's walking on thin ice with this one - Linus considers binary modules that don't require core kernel changes to be GPL-compatible, but IIRC there are other kernel developers who disagree and just aren't angry enough to sue about it.
And since energy = power * time and distance = velocity * time, that means fuel consumption = power per velocity. In other words, if power goes up as V^3, fuel consumption goes up as V^2.
You're right that real mileage is much more complicated; I was just making a quick joke reply to someone who didn't seem to remember the exact definition of a drag coefficient. I wasn't expecting a half dozen replies making the same physics error...
Slashdot Mirror
There needs to be some way on Slashdot to moderate the occasional, ultra-exceptional post up to about +10, give the post it's own front-page article, something to distinguish it from the hundreds of "ordinary" +5, Insightfuls that get rated every day.
I can't give you any more points, so I'll just reply with this pointlessly wordy "me too" - this is such a post.
Ultimately hardware options are not a solution pirates can use, since watermarking could easily identify which person freed some content from DRM.
Unless, of course, that content is ever sold for cash, like every video on the racks at every store from Best Buy to Walmart. Will popular videos all start going "direct-to-internet" any time soon? I doubt it. Publishers may talk a big talk about how much money they estimate they lose from piracy, but when they have to put that estimate up against the amount they know they'll lose from keeping their products out of every big retail outlet in the world, I doubt it'll seem so huge.
As soon as you need actual hardware to pirate the signal, copying movies becomes a restricted occupation again, just like selling free cable boxes.
No, copying movies the first time becomes a restricted occupation. Once a single unencrypted copy exists, then making a million more is no more difficult than it is today.
Whip out your favorite P2P client, and search for some copyrighted video. Do you see a hundred different rips made by each of the hundred different people sharing a copy? No, you see one or two of the best rips, each with hundreds of identical copies shared, in part because the swarming download protocols and hashing algorithms fundamentally encourage that behavior.
So what difference will in-monitor DRM make? Instead of having a few zealous groups using software to rip tons of movies that are then shared by millions of people, we'll have a few zealous groups using hardware to rip tons of movies that are then shared by hundreds of millions of people.
Wait - why will there be more people sharing these rips? Because most people will own some of the billions of non-DRM-capable monitors in existance, and the moron DRM-using publishers will have thus made it impossible for them to play a full-quality copy of these videos unless they have an illegal copy. Publishers couldn't do anything more stupid if they put a "Download free movies on P2P! It's the best!" advertisement at the start of every show!
Nothing will stop them until they are dead.
You sound as if you think there's a finite list of terrorists out there, and as soon as we scratch every name off the list then the terrorism problem will be solved! Not a chance - those people were made into terrorists, they weren't born that way. It doesn't matter how many of this generation's terrorists die for their crimes, if there's another equally large generation coming right after them.
I'm not agreeing with the idea that the way to stop the creation of new terrorists from religious zealots is to "treat them better" or "stop offending them" - for all I know it may be just the opposite. But we do need to understand these people, desperately, because it's only understanding or dumb luck that's going to allow us to stop the terrorist meme, and I'm not feeling very lucky.
I admit vengeance sounds pretty nice, but I'd gladly trade it for a more scientific understanding of the sociology of violence. The question of how we make more dead terrorists isn't nearly as important as the question of how we protect more live innocents. If capital punishment for mass murderers is part of that, then fine, but don't lose sight of the goal just because one step along the way is more emotionally compelling.
Isn't this sort of thing calculable farther in advance? There shouldn't be a whole lot of angular momentum being added or subtracted from the Earth's rotation.
Do I need a new glibc? Or any other POSIX library, for that matter? If this is a new announcement then presumably every implementation of mktime(), localtime(), gmtime(), etc. needs to be updated.
This isn't just about parallel computing - in fact if you'll read the article you'll see that they're using MPI for handling parallelism! Grid computing isn't about reinventing inter-node communications - it's more about inventing inter-node scheduling.
Your cluster - is it so fast that you're never stuck waiting for jobs to finish? If not, then you could probably benefit from being able to borrow time on someone's larger system. Is your cluster so well-utilized that the load's always around 1? If not then you've probably got spare capacity that someone else could benefit from. The fact that both you and those others are using MPI is necessary but insufficient to allow you to cooperate.
I just can't check for traffic jams on MoPac yet.
i ndex.html
http://www.statesman.com/traffic/content/traffic/
Three foreigners: a businessman, physicist, and mathematician, are talking about the country they're all visiting for the first time.
Suddenly, the businessman points out the window in surprise. "Look at that! The sheep in Scotland are black!"
Amused at how readily his new friend jumps to conclusions, the physicist corrects him: "No, all we can be certain of is that some of the sheep in Scotland are black."
The mathematician looks out the window himself, and corrects the both of them: "We know there exists a sheep in Scotland which is black on at least one side."
Yeah, I know, we all hate making slide shows, but if you can develop a stage presence and get gigs with large enough software companies then you'll really have an opportunity to make your physical activity and body language work for you.
Here, try a simple exercise to see if this might be the right career track for you. Stand up, and start stomping your feet and clapping your hands, while screaming "Developers! Developers! Developers! Developers!"
I fail to see why you concluded that any reference frame used for FTL must be backwards in time compared to our own, as light does not travel instantaneously, therefore it is hypothetically possible for something to move faster than light without traveling backwards in time.
;-)
See my other comment. It is possible to travel faster than light without moving backwards in time in a given reference frame; it is not possible to travel faster than light in one reference frame without moving backwards in time in some other frame.
Actually, I believe some laboratories have managed to accellerate light to faster than its normal speed, though I can't be bothered to dig up any articles on it.
I assume you're talking about the experiments here. Those showed a "group velocity" faster than light, but not faster than light transmission of mass/energy or information and so no problems with causality (and no utility for FTL travel).
And yes, FTL is often used for ambiance, as, quite frankly, it's rather difficult to have a multi-world society without FTL, as mere communications would take decades at the least, and interstellar trade would be pretty much nonexistant.
"Difficult" is just a dismissive way of saying "would make a fascinating story". See the first books of the "Rissa and Tregare" series by FM Busby for my favorite examples.
Besides, how many writers out there are really capable of describing a civilization fully utilizing a hundred billion stars? Not nearly as many as the number of writers who think they're up to it and are mistaken. Look at the universe-shrinking cameos in the Star Wars prequels: from watching the movies, you'd think Chewbacca was the only Wookie in the galaxy, not one out of a billion.
Of course, the easiest way to have a multi-world society without putting those worlds decades and centuries apart is to put them around the same star. It's ironic that people who think we need FTL to make space small enough to write about don't realize just how big space really is. There is enough free-falling raw material and solar power to support quintillions of people in this solar system alone - surely that's enough for a few interesting stories here too.
Can you please provide an example of this?
Any space-like interval (two points in space-time whose separation distance exceeds the speed of light multiplied by their separation time, with both measured in the same reference frame) is an example, because for any two such points there is another reference frame in which they are simultaneous. If AB is such an interval, then even if in your reference frame event A occurred before event B, then in the reference frame of someone traveling fast enough (relative to your velocity) in the direction from A to B they were simultaneous. And, in the reference frame of someone traveling even faster in the AB direction, event B occurred first.
If you were at event B, event A would be in your past in the original reference frame. If you then accelerate to a very high speed away from A, then in your new reference frame A can be in your future. An FTL drive which works independently of reference frame (the way all physical laws work as far as we know) could then take you to A's location in space but prior to A in time. To finish the loop, decelerate back to the original reference frame (which now puts B in your future), then FTL back to B's location in space but at an earlier time than you left.
I'm moderately sure he does understand the problems. I do, yet I don't think FTL is impossible. The only thing I know to likely be impossible is to accellerate a mass to beyond the speed of light in normal spacetime.
That's the first problem with FTL that comes to mind for many people, but I don't think it's the most important. The biggest problem in my mind is that paths which are faster than light in one reference frame are all backwards in time in some other reference frame.
Does this mean that FTL is so ridiculous that it should be off-limits to science fiction? Well, no (certainly even time travel shouldn't be off-limits, however unlikely!), but I suspect that many stories which use FTL travel are doing so because they want it for ambiance, not because they also want time travel in their story or because they're intentionally abandoning relativity.
I seem to remember that one Arthur C. Clark has been officialy recognized as the "inventor" of the satelite concept...
Not quite. AFAIK Clarke was the first person to publish the idea of geosynchronous communications satellites, but the idea of artificial satellites in general is much older.
Why would a story set in an interstellar spaceship suddenly become too mundane if that spaceship is limited to light speed? Would there be too much of the "present day" in a story about the lives of some of the quintillions of people an average solar system could support in orbital cities? Are nanomachines too boring when authors are careful not to turn them into thermodynamics-defying magic dust?
Nobody wants science fiction stripped of the fiction, some people just don't want it all stripped of the science. Science fantasy can still be entertaining, but it shouldn't be allowed to slip into otherwise consistent science fiction any more than traditional fantasy should corrupt traditional fiction. I suspect most of the Slashdot readers currently whining about how "why does everything have to be based on real facts" would turn the TV off in disgust if the next episode of "24" featured a nuclear bomb stolen by leprechauns or if "CSI" started occasionally solving mysteries with magic spells.
I mainly practice safe networking with a Linksys router/firewall at work and an OpenBSD gateway at home.
Does your firewall block outgoing HTTP connections and incoming email? If not, then it's not going to help against attacks like this PNG bug which are propagated through user-pulled data rather than attacker-pushed port connections. Such attacks exist for Linux, too. There is no such thing as "safe networking", and the only way to come close is to keep every connected computer up to date. I think Fedora still comes with up2date searching for updates in the background and displaying the results on a panel icon. Unless you use something else for security updates you ought to be clicking on that every time it finds something new.
Redhat still trying to figure out how to lure the opensource community back.
Are you talking about the open source community that includes people like Alan Cox, Ingo Molnar, Havoc Pennington, and Owen Taylor? It never left.
Are you part of a new, anti-RedHat OSS community? What have you written?
After all, Japan always had to import 100% of their petroleum needs, hence the reason why they've always emphasized high energy efficiency.
Not always. At one point, they emphasized using a powerful military to protect overseas oil assets, even going so far as to preemptively attack the largest threat to that empire. The result ended up including over a million Japanese casualties - it's not surprising they've become much less enamored of that strategy than the USA.
50fps (at 1024x768) in Half Life 2 according to Tom's Hardware, and $64 (including shipping) according to Pricewatch. There are probably better values out there, this is just the first model I thought to check.
Yeah, I know, if you want 1600x1200 + 8xAA + 8xAF you'll have to pay through the nose for it, but if you just want the image quality that a console gives you, you can get it on the PC cheaply.
I can't see how that round shape is going to be able to clean the corners of my kitchen tile. If a vacuum misses the corners of the carpet, that's not a big problem - they don't get much foot traffic anyway. But in the kitchen? The corners are where every little spill accumulates!
No, I'm going to stick with my original mopping plan: waiting until one of the stains attains sentience, then negotiating with it to defend me against the others.
It's hell to read a 100% width article on a large monitor even after blowing up the fontsize.
By "it's hell", do you mean "I have to shrink my browser window horizontally"? Those of us who like reading 100% of our screen width can't widen fixed-width pages, but a page that respects the reader's browser preferences can be as narrow as you want it to be.
Unless the torrent you're downloading contains commercials, including those from your local market, you are paying for fuck-all. You actually believe the $0.20 per month SCIFI gets from you entitles you to their entire lineup, commercial-free?
On my computer, I get Sci-Fi's entire lineup, commercial free, by pressing a "skip" button whenever a commercial starts and jumping immediately to the resumed show. I know that function's either hidden or nonexistant on commercial PVRs, but it's really only an incremental improvement on "mute" and "fast forward" anyway. Even permanently cutting out commercials on programs I want to archive is something that's always been possible for anyone with two VCRs and too much time on their hands.
So is what I'm doing unethical? Morally wrong but allowed via legal loophole? Illegal?
I hope not. If TV channel owners are expecting me to watch those commercials, they probably ought to have me sign something to that effect. On the other hand, if the Sci-Fi channel gets 20% of my viewing time but 0.4% of my cable bill, perhaps I'm not the one with whom they should be renegotiating a contract.
I've seen case law (wish I remember where) saying APIs aren't copyrightable - you can't necessarily just #include "whatever.h" into your program, but you can rewrite a "my_whatever.h" to declare the same symbols and binary compatible structs. If that's true, if a GPL program has an existing binary plugin loading interface, and if you're willing to do a bit of redundant work, you can compile a module that meets that interface and isn't legally a derivative work.
Can you redistribute that plugin without source code along with the GPL program? I have no idea. But I'll bet you can redistribute the plugin alone, never distribute the GPL program, and be in the clear. Remember, you can't sue someone for "GPL violations" - nobody has to agree to the GPL. GPL violations are only important because they remove a defense against copyright infringement lawsuits. So, if you're never going to redistribute copyrighted material in the first place, why worry about whether you would be licensed to do so or not?
The books are much better. They are not solely for kids.
Big Harry Potter fan?
That "abstract_api1" better be calling system(" proprietary_prog args") if you want to be GPL-compatible. What's more, you'd better be sure that proprietary_prog doesn't even include more GPL header files than fair use allows.
What you're describing is LGPL-compatible: with the LGPL you only have to release code to a library and modifications to that library. With the GPL, you have to release modifications to the entire derived work.
Now, you might be able to convince a court that sufficiently abstractly linked code isn't a derived work. I'd at least make sure nothing is statically linked in that case, but I know RMS considers even dynamically linked binary-only code to be a GPL violation - who knows what a court would say. I know Nvidia's walking on thin ice with this one - Linus considers binary modules that don't require core kernel changes to be GPL-compatible, but IIRC there are other kernel developers who disagree and just aren't angry enough to sue about it.
And since energy = power * time and distance = velocity * time, that means fuel consumption = power per velocity. In other words, if power goes up as V^3, fuel consumption goes up as V^2.
...
You're right that real mileage is much more complicated; I was just making a quick joke reply to someone who didn't seem to remember the exact definition of a drag coefficient. I wasn't expecting a half dozen replies making the same physics error