I think I see what you're saying - but the thing is that it's exactly as you say - except that it takes years to know if the cure worked. If it did, then the body will fix itself etc. If it didn't, then it's not good.
One more aspect of "cure" vs "treatment" (again, IMHO) is an ability to deterministically state "cured" based on a solid metric, not merely "Well, its been a while, it looks like it is gone."
o it really limited my top speed (or would make the PID run away if the target was higher than 127).
Yea, I believe I have since added PID overflow detection in the code. I haven't worked on it in quite a while. Jobs, wife, kids, etc.
The important part of the project, for me, was to do the PID algorithm based on measured time on a standard i.e. non real-time kernel. I was pretty happy with the results.
I think it is time, however, to refurbish the project with a nifty dual/quad core CPU and solid state hard disk. I'll probably have to design my own DC->DC power supply because all the ones I can find, at a reasonable price, are like 60W~75W.
I'd love to know what the distinction between a "cure" and "removal or killing of cancerous cells" is. If you get all the cells, which is often not as hard as you'd think, that seems to me like a cure.
This is a sensitive subject for some, so I'd like to say at the beginning that this is an observation. I lost my father to lung cancer, my niece had/has thyroid cancer. I did a little thinking about it all.
A "cure" (to me) means that you fix the initial problem and the body can then finish the job and heal with no chance of spontaneous relapse. Cancer treatment consists of killing every cell that will die under chemotherapy and/or radiation and hoping that it was enough and hoping that the body can recover.
There is a difference but it is difficult to articulate clearly.
Mostly just "stable enough", I am still working on tuning it since I switched to the Arduino for motor control. The event interface seems to provide the data quickly enough to fit into my loop but I don't know enough about what is going on underneath to know if the mouse is sending events as it gets them. "Close enough for Government work" is the phrase that comes to mind.
The problem with the USB mouse interface is that there is no polling mechanism. You get the events when the mouse thinks you want them. You can't control the period and you can't be sure the time-frame in which all the clicks happened.
Older mice with USB and PS/2 connectors (US witha PS/2 adapter) exhibit this behavior on USB yet work fine using PS/2.
Because of the lack of determinism in the USB mouse protocol it isn't well suited. In a practical sense, and in keeping with the $500 "close enough" philosophy, it can probably work. It will have trouble in low speed precision movement, but will work well enough on cumulative "cross the room" motion.
The biggest issue you will have is "dead reckoning" because on the diametrically opposed motor design, the relative motion of the two wheels has to be pretty accurate. Then again, inconsistent surfaces are probably a greater source of error.
I also wrote an encoder driver to use the Linux event interface rather than the ps2 interface so I could use a USB mouse encoder.
With USB, I could not get the mouse to send events unless and until it wanted too. The PS/2 interface allowed a fairly stable polling system from which I could calculate the interval for PID. We you able to get a stable PID system or, like most of the project, "stable enough" for actual work.
At Denning we had a mobile robot security guard. It could roam a factory or warehouse looking for intruders. it had sonar, radar, and other things.
Notifying people of appointments, delivering small objects, and serving drinks is not only possible, it is probably the easiest set of tasks that you can do.
I have a project on-line that allows you to build a basic robot for $500. It has PWM motor control and basic tips on building the base. It uses a PS/2 mouse to do wheel encoders. (cheap) and using a USB A-D/D-A board to control stuff. (I won't give the URL for fear of slashdotting my server.)
So, my two points: 1) It is possible they are doing what they say they can do. 2) Its fairly trivial if you have the time to waste.
Bittorrent already has something much better than SSL (for this particular purpose) called Message Stream Encryption [azureuswiki.com] that hides the fact that it is even conforming to any particular protocol -- it is, for all outside purposes, a wall of random data from the moment the connection is opened.
It isn't "better" than SSL, it is a different approach. It obscures the nature of the protocol but does not protect the security of the data.
The benefits of SSL is the fact that it isn't obscured. Automated wave shapers that can't identify a protocol don't automatically limit or block it. SSL is a standard, you can't block it. All secure web sites use it. If you block SSL, you block internet usability for e-commerce.
So, they HAVE to let SSL packets go because there is no defensible position that it is some sort of pirate application. That being said, they can only see the SSL protocol, NOT the application protocol it is protecting.
Its just going to push every last communication on to SSL.
At the network level, it is almost impossible to determine the nature of the communication without inspecting the packets. They can't eliminate peer to peer packets because things like instant messaging, vonage, skype, ssh, etc. will fail as well and these are legitimate non-RIAA objectionable services. So, once ALL the p2p systems start using SSL, even the ISPs will be powerless to stop P2P without making their system largely unusable.
Besides, ISPs have more to gain by adding a "premium" service (at a price premium) that allows better P2P performance and user priority for gaming.
A wise man manages to find happiness whatever the circumstances.
"What do you take me for, an idiot? [when asked if he was happy]" Charles De Gaulle
I don't think a wise man "finds" happiness. I think it is more likely that a wise man CREATES his happiness. A wise man knows that happiness comes from within.
You'll find wise men in slums, who will say they are content with their life.
Do not confuse wisdom with contentment. They are not the same thing.
You'll find fools who try to change the world because they are unhappy of the way their US cellphone fails to connect in Japan.
Perhaps, but the desire to make change is not a foolish impulse.
Thomas Jefferson and James Madison are most responsible for the Constitution and the Bill of Rights, two documents which represent HUGE change in the history of western civilization. (Along with the Magna Carta of course). Would you consider Thomas Jefferson and James Madison to be fools?
No fools don't make changes, they are incapable of the seriousness and thought that change requires.
"The Wise adapts himself to the world. The Fool adapts the world to himself. Therefore, all progress depends on the Fool."
I dispute the basic premise that a wise man adapts himself to the world.
A wise man knows what things are possible to change and in a realistic time frame. A fool does not understand these things and thus fails to accomplish anything.
Seriously, I never thought it possible, but the universe has bent in such a way that the extreme insipid evil that are xmax carols inflicted on poor helpless sharks is funny. Really funny. Sad... but really funny!
Maybe we could find the actual sharks who have attacked humans and force them to listen to xmax music, ala "Clockwork Orange" to teach them a lesson. Let them free so that they will tell other sharks about the horrors.
Of course, then we will see sharks beaching themselves like dolphins just to get the stupid tunes out of their tiny little brains.
That they would put this on a war ship indicates that politics was put in front of technology. Sorry, say what you want Microsoft and its apologists, Windows is neither reliable nor secure. That is a fact, not an opinion.
This isn't flame bait and I won't respond to anyone who wants to debate the facts because they are, in fact, fact.
The last 8 years has made it virtually impossible to state something conclusively without some "independent" group disputing the facts with junk science or out right lies. This tactic has paralyzed efforts from anti-smoking to global warming.
It allows people with a vested interest to force a debate of facts at the "policy" level by people who can easily be confused by carefully crafted BS. This is the tactic: they can't win the debate amongst experts, so they create a bogus dispute of facts with FUD or astro-turf, and then force people who don't know the background to decide between facts and factiness. (apologies to Stephen Colbert)
I know that everyone is saying Steve Jobs is fine, but there are some things to think about:
(1) Steve Jobs is an egotist. (2) The only thing Steve Jobs loves more than Apple is the sound of his own voice talking about Apple. (3) There is no "cure" for cancer. There are only methods for the removal or killing of cancerous cells.
I don't care what the "business" mumbo jumbo given is, for Steve Jobs to miss out on the chance to be on-stage and suck up the mac fanboi adulation, he's got to be very sick.
I am an Atheist, and I reject that programming languages have any connection with Religion. There is no actual belief system.
I know it is a joke, but on a philosophical level, it feels wrong.
I have loved computers since 1977 when I read Byte magazine on how to build an RCA 1802 Elf computer. I saved my money and built it. It was exciting! It was triumph of intellect over inexperience. I built it, fixed it, and got it to work. I read books, I bought tools. I conquered the subject.
Even viewing the act of programming as a religion is an insult to the pure intellectual pursuit that it is. As anyone who has done anything interesting will tell you, it is often the pure force of intellect that allows you to accomplish.
net Neutrality is like election finance reform. The people trying to gain access are all for it, but once access is gained, the urgency seems to fade away.
Google needs net neutrality where it is weak, but exploits sweet heart deals where it is strong. The ISPs should be careful, in this economy, the infrastructure that they depend on can be bought by Google or Microsoft. More over, if Google or Microsoft could buy or build a few major backbones, they'll be screaming bloody murder FOR net neutrality.
I think Google has done the numbers, though. They are banking on semi-truck sized compact portable data centers and using existing the existing backbone as merely the pipeline for cache coherency. So when you run google apps, you are getting your applications only a few hops away without sprint in the way.
I will paraphrase an old expression, never under estimate the data bandwidth of a semi-truck sized data center driving two days across country. Think about the number of raw terabytes that can be shipped vs transfered over the backbone.
I saw my post judged as flamebait, it has now moved on to something more appropriate, IMHO.
In a Quixotic fashion, I have to wonder why claiming that the XP downgrade can only happen in a monopolistic environment is flamebait.
Here is my reasoning: In a competitive environment, vendors don't have the luxury of customers asking for a prior release. Customer's say F*^&K YOU and buy your competitor. They don't, in this case, give you MORE money for an OLDER product. Seriously? Really? Can I have that business model, pleeesseee?
Microsoft through its defacto and proprietary standards hold your data and interoperability hostage. If you bought into Microsoft's product portfolio, you basically sealed your fate. Any purchase of Microsoft "now" will have payments later. You can pay a little on a regular basis to maintain, or you can suffer the balloon payment and cash out of the Microsoft tread-mill by transfering all your legacy data a more standards based system.
This opinion is supported by all the various anti-trust actions against Microsoft. If stating conclusions supported by multiple courts of law in the U.S.A. and the EU is flamebait, then I think the definition of flamebait has changed in the last 10 years.
This is something that can only happen when there is a monopoly involved. If there were a real and competitive environment, a vendor would not be able to do this to their customers without them choosing the competition.
It is so bad, that they aren't choosing competition, they are choosing to keep their previous product. Its pathetic.
It's a fair assumption. You can still build tube circuits. Some people, like audiophiles and guitar purists, will use nothing else.
I don't know how old you are, but my being 45 years old, I think I have some insight that may help you.
Time was, long ago, you could go to radio shack and buy tubes. There were hundreds of different types. For all purposes that transistors do now. There was even a tube-tester in each radio shack so you could find a "bad" tube and buy a replacement.
Fast forward today. There may be some audio-phool luddites who only use tubes, but that does not mean it is in any way a viable technology or that a tube manufactured 40 years ago is usable today. The socket type, the voltages, the gain, and the plate configuration could be anything.
Now, take this analogy to microchip proms. Just 25 years ago most chips were DIP types with a 0.01 pin spacing. These days it is all surface mount or PGA.
Would you call music on an 8-track tape usable today? How about Betamax? VHS is dying as we speak. Floppies? QIC-02 tapes? DEC-RKO drives? Have an old MFM-hard disk around?
It just simply can not happen. Space needs grow too fast for any standard to last and be practical for any long period of time. Just look at hard disks. 10 years ago, 15G was a fairly large hard disk. These days disks are 40 times that size.
Your best strategy is to just keep copying it on to new medium as time goes on.
It is just not possible to create a digital storage system that will be archive quality and guarantee it will be accessible in a time frame measured in decades.
Burnable CDs and DVDs degrade, the foil comes off and the plastic corrodes. Magnetic tape may be a good idea, but will a computer, 10~20 years from now, support the storage media?
I have stuff on 360K 5 1/4 inch floppies. If I could FIND a drive that worked, could I get the data? maybe. What about my CP/M floppies?
A 10 year time-frame may be doable. 20 year time is problematic. 30 years? Not a chance.
Then cite the GPLv2 where it makes it clear what is "mere aggregation" and what isn't.
In contract law, where there are complex terms and concepts in the contract, supporting documents are used to interpret their definitions. The FAQ you posted as a link is a pretty clear discussion about what is meant by "mere aggregation."
Then your simplistic advice which started this thread isn't so clear anymore, now is it?
Actually it is very clear. Everything created by people is imperfect, and anyone can create a discussion and argument where black looks like white for the sake of an argument. This is exactly the FUD you are employing. FUD -- "fear uncertainty & doubt"
My original posted rules:
(1) if its LGPL, link to it, but don't modify it. If you need to modify it, make the modification in the form of a generic API extension and call it from the application. Make your extensions public.
(2) If it is GPL, make it a service and call it through a socket.
(3) If it is a kernel module, there seems to be some wiggle room there, otherwise make a public mini-driver and a proprietary user space app.
These have not been refuted. All you have done is raise a border condition that in all likelihood can't be accomplished without violating standard GPL copyright to begin with.
Specifically, your concern is this part, I assume: But if the semantics of the communication are intimate enough, exchanging complex internal data structures, that too could be a basis to consider the two parts as combined into a larger program.
This section of the FAQ is legal conjecture about what the FSF thinks might be an issue. And while I can see their point, in practice it would be difficult to keep with the separate process space guideline and "exchanging complex internal data structures" guideline without red flags going up as it would almost assuredly mean you are using common structure and data definitions during the build process which would be source of the GPL violation.
You are citing authorities I have no way to verify.
Yup. anyone could claim anything. One must evaluate based on what is said and independent corroborating facts instead of reputation. Its harder, but better.
Furthermore, lawyers can be wrong, just like anybody else, and case law is what counts.
Case law is important *only* if there is serious ambiguity and a hostile party. One can always ASK the FSF if they have an issue BEFORE it becomes an issue. The WANT to help!!
Furthermore, the FUD thing is old. Courts only LIMIT rights based on law, they do not grant rights. The GPL defines what rights you have to GPL code. The only thing the court can do is limit the FSF's rights, and thus far the GPL has been held up as valid in all tests.
I have a number of friends who are lawyers, and through them I have gained an appreciation for the law. Law suits are a tricky thing, who pre-judges the suit's validity or merit before it goes to court?
You can't have a lesser authority as gate keeper because that just creates an expensive layer of bureaucracy. It would only allow well funded people through the system while over burdening poorer people.
So, the only fair way to judge a suit on merit is in court, any other way is just makes the situation worse.
The system is not perfect, any system devised and operated by humans is imperfect. How do you fix it without harming the people the system is intended to protect?
Slashdot Mirror
I think I see what you're saying - but the thing is that it's exactly as you say - except that it takes years to know if the cure worked. If it did, then the body will fix itself etc. If it didn't, then it's not good.
One more aspect of "cure" vs "treatment" (again, IMHO) is an ability to deterministically state "cured" based on a solid metric, not merely "Well, its been a while, it looks like it is gone."
o it really limited my top speed (or would make the PID run away if the target was higher than 127).
Yea, I believe I have since added PID overflow detection in the code. I haven't worked on it in quite a while. Jobs, wife, kids, etc.
The important part of the project, for me, was to do the PID algorithm based on measured time on a standard i.e. non real-time kernel. I was pretty happy with the results.
I think it is time, however, to refurbish the project with a nifty dual/quad core CPU and solid state hard disk. I'll probably have to design my own DC->DC power supply because all the ones I can find, at a reasonable price, are like 60W~75W.
I'd love to know what the distinction between a "cure" and "removal or killing of cancerous cells" is. If you get all the cells, which is often not as hard as you'd think, that seems to me like a cure.
This is a sensitive subject for some, so I'd like to say at the beginning that this is an observation. I lost my father to lung cancer, my niece had/has thyroid cancer. I did a little thinking about it all.
A "cure" (to me) means that you fix the initial problem and the body can then finish the job and heal with no chance of spontaneous relapse. Cancer treatment consists of killing every cell that will die under chemotherapy and/or radiation and hoping that it was enough and hoping that the body can recover.
There is a difference but it is difficult to articulate clearly.
Mostly just "stable enough", I am still working on tuning it since I switched to the Arduino for motor control. The event interface seems to provide the data quickly enough to fit into my loop but I don't know enough about what is going on underneath to know if the mouse is sending events as it gets them. "Close enough for Government work" is the phrase that comes to mind.
The problem with the USB mouse interface is that there is no polling mechanism. You get the events when the mouse thinks you want them. You can't control the period and you can't be sure the time-frame in which all the clicks happened.
Older mice with USB and PS/2 connectors (US witha PS/2 adapter) exhibit this behavior on USB yet work fine using PS/2.
Because of the lack of determinism in the USB mouse protocol it isn't well suited. In a practical sense, and in keeping with the $500 "close enough" philosophy, it can probably work. It will have trouble in low speed precision movement, but will work well enough on cumulative "cross the room" motion.
The biggest issue you will have is "dead reckoning" because on the diametrically opposed motor design, the relative motion of the two wheels has to be pretty accurate. Then again, inconsistent surfaces are probably a greater source of error.
I also wrote an encoder driver to use the Linux event interface rather than the ps2 interface so I could use a USB mouse encoder.
With USB, I could not get the mouse to send events unless and until it wanted too. The PS/2 interface allowed a fairly stable polling system from which I could calculate the interval for PID. We you able to get a stable PID system or, like most of the project, "stable enough" for actual work.
At Denning we had a mobile robot security guard. It could roam a factory or warehouse looking for intruders. it had sonar, radar, and other things.
Notifying people of appointments, delivering small objects, and serving drinks is not only possible, it is probably the easiest set of tasks that you can do.
I have a project on-line that allows you to build a basic robot for $500. It has PWM motor control and basic tips on building the base. It uses a PS/2 mouse to do wheel encoders. (cheap) and using a USB A-D/D-A board to control stuff. (I won't give the URL for fear of slashdotting my server.)
So, my two points: 1) It is possible they are doing what they say they can do. 2) Its fairly trivial if you have the time to waste.
Bittorrent already has something much better than SSL (for this particular purpose) called Message Stream Encryption [azureuswiki.com] that hides the fact that it is even conforming to any particular protocol -- it is, for all outside purposes, a wall of random data from the moment the connection is opened.
It isn't "better" than SSL, it is a different approach. It obscures the nature of the protocol but does not protect the security of the data.
The benefits of SSL is the fact that it isn't obscured. Automated wave shapers that can't identify a protocol don't automatically limit or block it. SSL is a standard, you can't block it. All secure web sites use it. If you block SSL, you block internet usability for e-commerce.
So, they HAVE to let SSL packets go because there is no defensible position that it is some sort of pirate application. That being said, they can only see the SSL protocol, NOT the application protocol it is protecting.
Its just going to push every last communication on to SSL.
At the network level, it is almost impossible to determine the nature of the communication without inspecting the packets. They can't eliminate peer to peer packets because things like instant messaging, vonage, skype, ssh, etc. will fail as well and these are legitimate non-RIAA objectionable services. So, once ALL the p2p systems start using SSL, even the ISPs will be powerless to stop P2P without making their system largely unusable.
Besides, ISPs have more to gain by adding a "premium" service (at a price premium) that allows better P2P performance and user priority for gaming.
A wise man manages to find happiness whatever the circumstances.
"What do you take me for, an idiot? [when asked if he was happy]"
Charles De Gaulle
I don't think a wise man "finds" happiness. I think it is more likely that a wise man CREATES his happiness. A wise man knows that happiness comes from within.
You'll find wise men in slums, who will say they are content with their life.
Do not confuse wisdom with contentment. They are not the same thing.
You'll find fools who try to change the world because they are unhappy of the way their US cellphone fails to connect in Japan.
Perhaps, but the desire to make change is not a foolish impulse.
Thomas Jefferson and James Madison are most responsible for the Constitution and the Bill of Rights, two documents which represent HUGE change in the history of western civilization. (Along with the Magna Carta of course). Would you consider Thomas Jefferson and James Madison to be fools?
No fools don't make changes, they are incapable of the seriousness and thought that change requires.
"The Wise adapts himself to the world. The Fool adapts the world to himself. Therefore, all progress depends on the Fool."
I dispute the basic premise that a wise man adapts himself to the world.
A wise man knows what things are possible to change and in a realistic time frame. A fool does not understand these things and thus fails to accomplish anything.
Seriously, I never thought it possible, but the universe has bent in such a way that the extreme insipid evil that are xmax carols inflicted on poor helpless sharks is funny. Really funny. Sad... but really funny!
Maybe we could find the actual sharks who have attacked humans and force them to listen to xmax music, ala "Clockwork Orange" to teach them a lesson. Let them free so that they will tell other sharks about the horrors.
Of course, then we will see sharks beaching themselves like dolphins just to get the stupid tunes out of their tiny little brains.
That they would put this on a war ship indicates that politics was put in front of technology. Sorry, say what you want Microsoft and its apologists, Windows is neither reliable nor secure. That is a fact, not an opinion.
This isn't flame bait and I won't respond to anyone who wants to debate the facts because they are, in fact, fact.
The last 8 years has made it virtually impossible to state something conclusively without some "independent" group disputing the facts with junk science or out right lies. This tactic has paralyzed efforts from anti-smoking to global warming.
It allows people with a vested interest to force a debate of facts at the "policy" level by people who can easily be confused by carefully crafted BS. This is the tactic: they can't win the debate amongst experts, so they create a bogus dispute of facts with FUD or astro-turf, and then force people who don't know the background to decide between facts and factiness. (apologies to Stephen Colbert)
I know that everyone is saying Steve Jobs is fine, but there are some things to think about:
(1) Steve Jobs is an egotist.
(2) The only thing Steve Jobs loves more than Apple is the sound of his own voice talking about Apple.
(3) There is no "cure" for cancer. There are only methods for the removal or killing of cancerous cells.
I don't care what the "business" mumbo jumbo given is, for Steve Jobs to miss out on the chance to be on-stage and suck up the mac fanboi adulation, he's got to be very sick.
I am an Atheist, and I reject that programming languages have any connection with Religion. There is no actual belief system.
I know it is a joke, but on a philosophical level, it feels wrong.
I have loved computers since 1977 when I read Byte magazine on how to build an RCA 1802 Elf computer. I saved my money and built it. It was exciting! It was triumph of intellect over inexperience. I built it, fixed it, and got it to work. I read books, I bought tools. I conquered the subject.
Even viewing the act of programming as a religion is an insult to the pure intellectual pursuit that it is. As anyone who has done anything interesting will tell you, it is often the pure force of intellect that allows you to accomplish.
net Neutrality is like election finance reform. The people trying to gain access are all for it, but once access is gained, the urgency seems to fade away.
Google needs net neutrality where it is weak, but exploits sweet heart deals where it is strong. The ISPs should be careful, in this economy, the infrastructure that they depend on can be bought by Google or Microsoft. More over, if Google or Microsoft could buy or build a few major backbones, they'll be screaming bloody murder FOR net neutrality.
I think Google has done the numbers, though. They are banking on semi-truck sized compact portable data centers and using existing the existing backbone as merely the pipeline for cache coherency. So when you run google apps, you are getting your applications only a few hops away without sprint in the way.
I will paraphrase an old expression, never under estimate the data bandwidth of a semi-truck sized data center driving two days across country. Think about the number of raw terabytes that can be shipped vs transfered over the backbone.
I saw my post judged as flamebait, it has now moved on to something more appropriate, IMHO.
In a Quixotic fashion, I have to wonder why claiming that the XP downgrade can only happen in a monopolistic environment is flamebait.
Here is my reasoning: In a competitive environment, vendors don't have the luxury of customers asking for a prior release. Customer's say F*^&K YOU and buy your competitor. They don't, in this case, give you MORE money for an OLDER product. Seriously? Really? Can I have that business model, pleeesseee?
Microsoft through its defacto and proprietary standards hold your data and interoperability hostage. If you bought into Microsoft's product portfolio, you basically sealed your fate. Any purchase of Microsoft "now" will have payments later. You can pay a little on a regular basis to maintain, or you can suffer the balloon payment and cash out of the Microsoft tread-mill by transfering all your legacy data a more standards based system.
This opinion is supported by all the various anti-trust actions against Microsoft. If stating conclusions supported by multiple courts of law in the U.S.A. and the EU is flamebait, then I think the definition of flamebait has changed in the last 10 years.
Blame it on Linux's lack of marketing.
Perhaps, that and Microsoft using its monopoly defacto standards and formats to keep the customers hostage.
This is something that can only happen when there is a monopoly involved. If there were a real and competitive environment, a vendor would not be able to do this to their customers without them choosing the competition.
It is so bad, that they aren't choosing competition, they are choosing to keep their previous product. Its pathetic.
It's a fair assumption. You can still build tube circuits. Some people, like audiophiles and guitar purists, will use nothing else.
I don't know how old you are, but my being 45 years old, I think I have some insight that may help you.
Time was, long ago, you could go to radio shack and buy tubes. There were hundreds of different types. For all purposes that transistors do now. There was even a tube-tester in each radio shack so you could find a "bad" tube and buy a replacement.
Fast forward today. There may be some audio-phool luddites who only use tubes, but that does not mean it is in any way a viable technology or that a tube manufactured 40 years ago is usable today. The socket type, the voltages, the gain, and the plate configuration could be anything.
Now, take this analogy to microchip proms. Just 25 years ago most chips were DIP types with a 0.01 pin spacing. These days it is all surface mount or PGA.
Would you call music on an 8-track tape usable today? How about Betamax? VHS is dying as we speak. Floppies? QIC-02 tapes? DEC-RKO drives? Have an old MFM-hard disk around?
It just simply can not happen. Space needs grow too fast for any standard to last and be practical for any long period of time. Just look at hard disks. 10 years ago, 15G was a fairly large hard disk. These days disks are 40 times that size.
Your best strategy is to just keep copying it on to new medium as time goes on.
Wouldn't a PROM chip last upwards of 50 years?
You are assuming a PROM chip of today will be usable in 50 years? When was the last time you SAW a vacuum tube?
There is a part you are missing about copyright itself, and that is the term "based" from "based on the Program" which is analogous to "derived from."
This is not mere aggregation.
It is just not possible to create a digital storage system that will be archive quality and guarantee it will be accessible in a time frame measured in decades.
Burnable CDs and DVDs degrade, the foil comes off and the plastic corrodes. Magnetic tape may be a good idea, but will a computer, 10~20 years from now, support the storage media?
I have stuff on 360K 5 1/4 inch floppies. If I could FIND a drive that worked, could I get the data? maybe. What about my CP/M floppies?
A 10 year time-frame may be doable. 20 year time is problematic. 30 years? Not a chance.
Then cite the GPLv2 where it makes it clear what is "mere aggregation" and what isn't.
In contract law, where there are complex terms and concepts in the contract, supporting documents are used to interpret their definitions. The FAQ you posted as a link is a pretty clear discussion about what is meant by "mere aggregation."
Then your simplistic advice which started this thread isn't so clear anymore, now is it?
Actually it is very clear. Everything created by people is imperfect, and anyone can create a discussion and argument where black looks like white for the sake of an argument. This is exactly the FUD you are employing. FUD -- "fear uncertainty & doubt"
My original posted rules:
(1) if its LGPL, link to it, but don't modify it. If you need to modify it, make the modification in the form of a generic API extension and call it from the application. Make your extensions public.
(2) If it is GPL, make it a service and call it through a socket.
(3) If it is a kernel module, there seems to be some wiggle room there, otherwise make a public mini-driver and a proprietary user space app.
These have not been refuted. All you have done is raise a border condition that in all likelihood can't be accomplished without violating standard GPL copyright to begin with.
If you can, cite a case law that covers something like using sockets to avoid GPLing your program.
You are looking as courts as the one's granting the rights, not the GPL and the FSF, this is the wrong approach.
It's ambiguous because the license itself is ambiguous on this issue.
again, FUD.
Reposting the FSF link:
http://www.fsf.org/licensing/licenses/gpl-faq.html#MereAggregation [fsf.org]
Specifically, your concern is this part, I assume:
But if the semantics of the communication are intimate enough, exchanging complex internal data structures, that too could be a basis to consider the two parts as combined into a larger program.
This section of the FAQ is legal conjecture about what the FSF thinks might be an issue. And while I can see their point, in practice it would be difficult to keep with the separate process space guideline and "exchanging complex internal data structures" guideline without red flags going up as it would almost assuredly mean you are using common structure and data definitions during the build process which would be source of the GPL violation.
You are citing authorities I have no way to verify.
Yup. anyone could claim anything. One must evaluate based on what is said and independent corroborating facts instead of reputation. Its harder, but better.
Furthermore, lawyers can be wrong, just like anybody else, and case law is what counts.
Case law is important *only* if there is serious ambiguity and a hostile party. One can always ASK the FSF if they have an issue BEFORE it becomes an issue. The WANT to help!!
Furthermore, the FUD thing is old. Courts only LIMIT rights based on law, they do not grant rights. The GPL defines what rights you have to GPL code. The only thing the court can do is limit the FSF's rights, and thus far the GPL has been held up as valid in all tests.
I have a number of friends who are lawyers, and through them I have gained an appreciation for the law. Law suits are a tricky thing, who pre-judges the suit's validity or merit before it goes to court?
You can't have a lesser authority as gate keeper because that just creates an expensive layer of bureaucracy. It would only allow well funded people through the system while over burdening poorer people.
So, the only fair way to judge a suit on merit is in court, any other way is just makes the situation worse.
The system is not perfect, any system devised and operated by humans is imperfect. How do you fix it without harming the people the system is intended to protect?