German online news site Spiegel Online has more details on this: Heat plant in the car. It uses a high temperature (up to 550 Celsius) circuit using water and a low temperature one using ethanol (alcohol) (operating at 150 Celsius). Both are closed systems.
My constant conflict with open source software is that I feel its great for home users, the poor, schools and other individuals/organizations that can't afford 300 dollars for windows and 700 dollars for photoshop. The problem is how do I pay my bills?
But that problem is true with all stuff that comes out and is cheap and replaces another business. When photography became popular and usable by everyone (either by going to a photographer or later one buying your own camera), portrait painting was on the loss, and many painters went out of business. Does that mean that photography made easy for everyone was an error? Now digital photography takes over and makes chemical processing obsolete for most people, and only some specialized labs for certain types of photography will remain in the long run. Does that make digital photography a curse?
Information (as a structure, that contains non random patterns) has to be created only once, and then you only need copies. So there is only the initial cost of creating said patterns, and the cost of copy equals nearly zero. In economic terms it means: The fair price for the n+1 copy is the cost for the copy itself because all the other costs are already contained in the first n copies. If it weren't for the copyright law and the licenses a copy of a program would cost zero. That's the big advantage and the big risk of software (and other 'content' industries). On the one hand the costs to produce the next copy is zero, once the development is done. On the other hand you have to provide an incentive for paying nonzero for something that in economic terms has a fair price of zero.
If the incentive is void because there is a source of an equivalent product where the price is already zero, then the market price is zero. This is independent on the fact that a company or a programmer has to reimburse his costs. That's bad luck. But lets just imagine a world, where progress has slowed down so that for 95% of the people an old version of everything just does the job. Lets say in 2100 people are still using Win2000 for an OS. Win2000 by then is in the public domain, and free for everyone. What do you tell people to buy Win2100, if they can get their problems solved by Win2000 for free? Currently no software is that old that its copyright has expired. So the problem is other software that was put into public domain or is offered for free intentionally. But what when your older software versions start to bite you for free?
If they wouldn't pirate Photoshop, they would use GIMP, and complain about missing features, some of them would start scratching the itch and bringing GIMP on par with Photoshop, KPT would be recoded from scratch for GIMP, and finally even professional graphic shops would switch to GIMP and save the $700. So in the end no one would ever buy Photoshop, and the coders of Photoshop would be out of work and unpaid. Basicly piracy is the last thing that keeps Free Software from world domination.
For Europe there was always (since about at least 10 years) the HAFAS system, first coming on CD-ROM and covering only Germany's railway, later online and now including most public transports in Germany and some neighbouring countries. You just enter starting point and destination, time for travel (default: now) and get several schedules, stretching across different transport companies, if necessary, and you are done.
When I was in the Bay Area in 2000, I was trying to get a schedule with BART from Santa Clara, CA to the Napa Valley, and I failed utterly, because BART never managed to get a sensible online schedule put together. So Google is just bringing the U.S. region by region to the point where Europe was 10 years ago.
The definition used for the statistics (by an U.S. researcher btw.) was: 'armed conflict involving at least 1000 armed persons on each side'. And if you are talking about being kicked in the ass by the Germans: France was on the winning side in both World War I and World War II in the end. However they managed this may be subject to discussion, but they prevailed.
The connection to France becoming a losing party after turning republic is also flawed. In 1870 France was an imperium with Napoleon III being the emperor. Only after that France became a republic again after the conflict between the proletarian and the bourgeois side (the Commune of Paris 1870/71) was over. After that France did remarkably well until 1956 (battle of Dien Bien Pu).
A lot of wars France was losing in the last half of the 20th century were colonial wars. The U.S. never claimed any country formally as a colony, so the U.S. never had a chance to lose a similar war. The U.S. was instead losing a lot of supportive dictatorships during the 20th century (Cuba, Nicaragua, South Vietnam, Haiti, Iran...) or were having rebellious allegiancies (Angola, again Nicaragua, Venezuela...) which lost. This really improves the formal balance sheet for the U.S. because they could always claim not to be a party in the conflict.
Saxony is still a country. It just lost its state as Nation,:) being part of Germany since 1871. Nations that no longer exists would include Burgundy (now partly Belgium, Luxembourg and eastern France) and Venezia (Italy, Slovenia, southern Croatia, Albania, Greece).
No, it's just the central situation in western Europe, bordering to Great Britain, Spain, the Netherlands, Prussia and Austria, where all the other strong nation had only one or two strong neighbours.
PS: Austria no longer borders to France. It lost the Western Austria (Vorderoesterreich) in the aftermath of the Napoleon Wars. Same with the Netherlands. 1831 the southern provincies of the Netherlands segregated and formed the new country Belgium.
As a matter of fact: France fought the most wars in modern history (that's since the 15th century). It fought more than Great Britain, Prussia and Austria (the other large nations in Europe with many battles) together. As far as I remember, it was more than 2000 wars with french involvement, compared with 500-700 for each of the other nations. The U.S. come not even close to a 10th of the numbers of France. And if french troups have surrendered here and there, the quote is still smaller than that of other nations (otherwise France would not be here anymore).
Jokes about France surrendering sheet more light on uneducated jokers than on french national characteristics. If you really want a nation to make jokes about because of constant bad luck in battles, take Saxony: No major victory in battle since the second siege of Vienna. Frederick the Great of Prussia once joked: Saxony is like a sack of flour: You can beat it as often as you want, there is still something coming out (Yes, I am of saxon origin).
Normally the window that you scroll over gets a mouse event anyway, so if this is your window you'll be fine just to handle that event. To get ALL mouse events means that you are catching stuff that doesn't belong to your software. And here you need VERY good reason to have this.
Yeah... it has been that way for some time. 80GByte HDs are the cheapest you can get in reasonable numbers. 40 GByte HDs you don't get in larger amounts, only some remaining disk lying in the 'outdated' corner. And for 20 GByte you have to ask and accept a higher price.
Shut down for what? Developing yet another file transfer protocol? That's what BitTorrent is after all. And for the torrent links on BitTorrent.com, he probably sifts through them anyway and throws out those that look fishy to him. Basicly for BitTorrent the links are just a proof of concept: Here! That's what a torrent file looks like, and that's what happens if you click at it, and look, how many of them I can easily host on my rather small bandwith site. There is absolutely no point for him to have things there that could get him into trouble.
The same could be said about the Greeks. In about any town around the mediterran there were people understanding greek for about 2000 years (from 480 BC, battle at the Thermopyles, to 1543 AC, when Constantinople felt to the Turks). This doesn't make the mediterran a continious society.
And yes, the alphabet used by the ancient greek (and used today) was derived from the phoenicean, which in turn took them from the semitic people living in the eastern part of the mediterran (heck, even the name 'alphabet' is semitic: from 'aleph' [bull] and 'bet' [house], the first letters of the hebrew alphabet, which were adapted to the greek 'alpha' and 'beta'). So we have at least 3500 years of the same alphabet to denote the same phonems. And still we don't consider us being the same civilisation than the one of 1500 BC.
So to optimize code for HT we should make sure that different threads of the same process are running at the same small code and data fragments, so the code and data caches are not flushed due to thread switches. Process (and thus context) switches are bad enough with having to reload the whole cache. There are not many problems where this is the case (ray tracing comes to mind, where different rays run on the same vector set and use the same code to get evaluated).
If different threads have to run at different data or code fragments, then THOSE should be optimized in RAM in a way that they don't purge each other out of the cache. This can be achieved with sophisticated memory mapping, but I guess it's hell of an optimizing problem.
Normally cache lines are connected to physical memory pages. You can always tell from the address of the memory page in which cache line the data will be loaded if accessed. If you now have memory management that gives different threads sophisticatedly choosen memory pages, two threads never will trash each other's cache. In this case every thread will only see half of the cache, because its memory pages never get loaded in the other half of the cache, without the programmer needing to take care of this. Of course then the programmer has to take care that two concurrently running threads always use disjunct cache halves (basicly have 'even' and 'odd' threads and make sure, that never two of the same type are running in parallel).
Sony distributed copyrighted software without a proper license to do so. What stands in their contract with First 4 Internet doesn't really matter for the fact that this was infringing on copyright. You could only say that they acted in good faith, but it's still illegal. The road to Hell is paved with good intentions.
But there is one point where the GPL actually helps: The people who got the rootkit installed and running everytime they access the CD drive. If it weren't for the GPL they would also be in infringment on copyright, running software without a license to do so (the EULA is invalid, because it is attached to something Sony wasn't allowed to license to begin with). Luckily the GPL allows everyone to run derivative works as long as they don't distribute or modify them without further restrictions. But that's where it gets really messy. Think about the situation where a company starts to distribute something like this rootkit on all customer PCs, and this software wasn't ripped from something licensed under GPL, rather from someting licensed under a shareware license or something.
Now every of the half a million infected PCs isn't allowed to run a software, that is installed on it, can't be removed without risk to crash the computer, and will be started automaticly everytime the PC is turned on. Basicly this computer can never be turned on again (except for booting from another boot media, which in most cases means reinstall) without infringing copyright.
It basicly shows that you either have one big DRM scheme to rule them all, or just don't have one at all. Maybe this puts DRM on hold until the standardizing and licensing issues are sorted out. On the other hand: If there is a single DRM scheme for all, then a single weakness in the DRM scheme unlocks ALL content protected under this scheme.
According to both LGPL and GPL the one you get the software from is the distributor. He is the one responsible for adhering to the licenses. He can of course sue his own software provider later, but for now it's Sony that distributed the programs.
If Sony is providing the source code for the programs and restates that the software is unter GPL (thus giving you the right to modify and distribute your modification), then everything is fine between Sony and you though.
There have been several similar cases in Europe about this, and in every case the GPL has been found valid, and the violation of the license has been considered healed, if the final distributor was able to get hold of the source code and distribute this one too under GPL.
Check GPL v2.0 section 4: 4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.
For Sony this means: They lost the right to distribute the Program, and they will be in violation of the GPL until they start to comply with the GPL themselves (e.g. distributing the source and allowing modifications and redistribution under GPL).
Strange. Everyone I know who does Instant Messaging has an IRC nick. Seems to be connected to the fact that I hang out on IRC. (And no. I don't have an AIM or MSN name.)
Microsoft probably had to consent, otherwise they would have lost in court anyway. Microsoft Germany already lost a similar case in german court, where they tried to forbid a computer store (Snogard) to buy used licenses from people who don't need them (installing from a different source for instance) and bundle them with the shop's computer offerings.
Microsoft lost badly in court, mainly because of the First Sale doctrin (which in Germany is called Erschoepfungsgrundsatz).
Lindows got Microsoft's suit refused because 'Windows' is a generic english noun, and thus not trademarkable. So my analogy still holds. You can have a product in the software market with a name that is easily confusable with 'Windows' and still not infringe on Microsoft's trademark of 'Microsoft Windows'.
But Lindows was actually selling an operating system (heck, even a fully fledged software distribution!), and Microsoft in the end had to BUY the Lindows name to have the company changing the name to Linspire. So even in the (US-) software world 'Windows' is not trademarkable. (It's different in non english speaking countries, because there Windows is no generic noun.)
Slashdot Mirror
German online news site Spiegel Online has more details on this:
Heat plant in the car. It uses a high temperature (up to 550 Celsius) circuit using water and a low temperature one using ethanol (alcohol) (operating at 150 Celsius). Both are closed systems.
But that problem is true with all stuff that comes out and is cheap and replaces another business. When photography became popular and usable by everyone (either by going to a photographer or later one buying your own camera), portrait painting was on the loss, and many painters went out of business. Does that mean that photography made easy for everyone was an error? Now digital photography takes over and makes chemical processing obsolete for most people, and only some specialized labs for certain types of photography will remain in the long run. Does that make digital photography a curse?
Information (as a structure, that contains non random patterns) has to be created only once, and then you only need copies. So there is only the initial cost of creating said patterns, and the cost of copy equals nearly zero. In economic terms it means: The fair price for the n+1 copy is the cost for the copy itself because all the other costs are already contained in the first n copies. If it weren't for the copyright law and the licenses a copy of a program would cost zero. That's the big advantage and the big risk of software (and other 'content' industries). On the one hand the costs to produce the next copy is zero, once the development is done. On the other hand you have to provide an incentive for paying nonzero for something that in economic terms has a fair price of zero.
If the incentive is void because there is a source of an equivalent product where the price is already zero, then the market price is zero. This is independent on the fact that a company or a programmer has to reimburse his costs. That's bad luck. But lets just imagine a world, where progress has slowed down so that for 95% of the people an old version of everything just does the job. Lets say in 2100 people are still using Win2000 for an OS. Win2000 by then is in the public domain, and free for everyone. What do you tell people to buy Win2100, if they can get their problems solved by Win2000 for free? Currently no software is that old that its copyright has expired. So the problem is other software that was put into public domain or is offered for free intentionally. But what when your older software versions start to bite you for free?
If they wouldn't pirate Photoshop, they would use GIMP, and complain about missing features, some of them would start scratching the itch and bringing GIMP on par with Photoshop, KPT would be recoded from scratch for GIMP, and finally even professional graphic shops would switch to GIMP and save the $700. So in the end no one would ever buy Photoshop, and the coders of Photoshop would be out of work and unpaid. Basicly piracy is the last thing that keeps Free Software from world domination.
For Europe there was always (since about at least 10 years) the HAFAS system, first coming on CD-ROM and covering only Germany's railway, later online and now including most public transports in Germany and some neighbouring countries. You just enter starting point and destination, time for travel (default: now) and get several schedules, stretching across different transport companies, if necessary, and you are done.
When I was in the Bay Area in 2000, I was trying to get a schedule with BART from Santa Clara, CA to the Napa Valley, and I failed utterly, because BART never managed to get a sensible online schedule put together. So Google is just bringing the U.S. region by region to the point where Europe was 10 years ago.
The definition used for the statistics (by an U.S. researcher btw.) was: 'armed conflict involving at least 1000 armed persons on each side'. And if you are talking about being kicked in the ass by the Germans: France was on the winning side in both World War I and World War II in the end. However they managed this may be subject to discussion, but they prevailed.
The connection to France becoming a losing party after turning republic is also flawed. In 1870 France was an imperium with Napoleon III being the emperor. Only after that France became a republic again after the conflict between the proletarian and the bourgeois side (the Commune of Paris 1870/71) was over. After that France did remarkably well until 1956 (battle of Dien Bien Pu).
A lot of wars France was losing in the last half of the 20th century were colonial wars. The U.S. never claimed any country formally as a colony, so the U.S. never had a chance to lose a similar war. The U.S. was instead losing a lot of supportive dictatorships during the 20th century (Cuba, Nicaragua, South Vietnam, Haiti, Iran...) or were having rebellious allegiancies (Angola, again Nicaragua, Venezuela...) which lost. This really improves the formal balance sheet for the U.S. because they could always claim not to be a party in the conflict.
The member states of the Federal Republic of Germany are called 'countries'. Saxony is one of them: the Free State of Saxony.
My statement is similar to the statement, that Texas is a republic, but not a nation.
Saxony is still a country. It just lost its state as Nation, :) being part of Germany since 1871. Nations that no longer exists would include Burgundy (now partly Belgium, Luxembourg and eastern France) and Venezia (Italy, Slovenia, southern Croatia, Albania, Greece).
No, it's just the central situation in western Europe, bordering to Great Britain, Spain, the Netherlands, Prussia and Austria, where all the other strong nation had only one or two strong neighbours.
PS: Austria no longer borders to France. It lost the Western Austria (Vorderoesterreich) in the aftermath of the Napoleon Wars. Same with the Netherlands. 1831 the southern provincies of the Netherlands segregated and formed the new country Belgium.
As a matter of fact: France fought the most wars in modern history (that's since the 15th century). It fought more than Great Britain, Prussia and Austria (the other large nations in Europe with many battles) together. As far as I remember, it was more than 2000 wars with french involvement, compared with 500-700 for each of the other nations. The U.S. come not even close to a 10th of the numbers of France. And if french troups have surrendered here and there, the quote is still smaller than that of other nations (otherwise France would not be here anymore).
Jokes about France surrendering sheet more light on uneducated jokers than on french national characteristics. If you really want a nation to make jokes about because of constant bad luck in battles, take Saxony: No major victory in battle since the second siege of Vienna. Frederick the Great of Prussia once joked: Saxony is like a sack of flour: You can beat it as often as you want, there is still something coming out (Yes, I am of saxon origin).
Normally the window that you scroll over gets a mouse event anyway, so if this is your window you'll be fine just to handle that event. To get ALL mouse events means that you are catching stuff that doesn't belong to your software. And here you need VERY good reason to have this.
Yeah... it has been that way for some time. 80GByte HDs are the cheapest you can get in reasonable numbers. 40 GByte HDs you don't get in larger amounts, only some remaining disk lying in the 'outdated' corner. And for 20 GByte you have to ask and accept a higher price.
I guess iSupply has a pretty good idea how much rebate you get if you buy in bulk. :)
Shut down for what? Developing yet another file transfer protocol? That's what BitTorrent is after all. And for the torrent links on BitTorrent.com, he probably sifts through them anyway and throws out those that look fishy to him. Basicly for BitTorrent the links are just a proof of concept: Here! That's what a torrent file looks like, and that's what happens if you click at it, and look, how many of them I can easily host on my rather small bandwith site. There is absolutely no point for him to have things there that could get him into trouble.
The same could be said about the Greeks. In about any town around the mediterran there were people understanding greek for about 2000 years (from 480 BC, battle at the Thermopyles, to 1543 AC, when Constantinople felt to the Turks). This doesn't make the mediterran a continious society.
And yes, the alphabet used by the ancient greek (and used today) was derived from the phoenicean, which in turn took them from the semitic people living in the eastern part of the mediterran (heck, even the name 'alphabet' is semitic: from 'aleph' [bull] and 'bet' [house], the first letters of the hebrew alphabet, which were adapted to the greek 'alpha' and 'beta'). So we have at least 3500 years of the same alphabet to denote the same phonems. And still we don't consider us being the same civilisation than the one of 1500 BC.
So to optimize code for HT we should make sure that different threads of the same process are running at the same small code and data fragments, so the code and data caches are not flushed due to thread switches. Process (and thus context) switches are bad enough with having to reload the whole cache.
There are not many problems where this is the case (ray tracing comes to mind, where different rays run on the same vector set and use the same code to get evaluated).
If different threads have to run at different data or code fragments, then THOSE should be optimized in RAM in a way that they don't purge each other out of the cache. This can be achieved with sophisticated memory mapping, but I guess it's hell of an optimizing problem.
Normally cache lines are connected to physical memory pages. You can always tell from the address of the memory page in which cache line the data will be loaded if accessed. If you now have memory management that gives different threads sophisticatedly choosen memory pages, two threads never will trash each other's cache. In this case every thread will only see half of the cache, because its memory pages never get loaded in the other half of the cache, without the programmer needing to take care of this. Of course then the programmer has to take care that two concurrently running threads always use disjunct cache halves (basicly have 'even' and 'odd' threads and make sure, that never two of the same type are running in parallel).
Sony distributed copyrighted software without a proper license to do so. What stands in their contract with First 4 Internet doesn't really matter for the fact that this was infringing on copyright. You could only say that they acted in good faith, but it's still illegal. The road to Hell is paved with good intentions.
But there is one point where the GPL actually helps: The people who got the rootkit installed and running everytime they access the CD drive. If it weren't for the GPL they would also be in infringment on copyright, running software without a license to do so (the EULA is invalid, because it is attached to something Sony wasn't allowed to license to begin with). Luckily the GPL allows everyone to run derivative works as long as they don't distribute or modify them without further restrictions. But that's where it gets really messy. Think about the situation where a company starts to distribute something like this rootkit on all customer PCs, and this software wasn't ripped from something licensed under GPL, rather from someting licensed under a shareware license or something.
Now every of the half a million infected PCs isn't allowed to run a software, that is installed on it, can't be removed without risk to crash the computer, and will be started automaticly everytime the PC is turned on. Basicly this computer can never be turned on again (except for booting from another boot media, which in most cases means reinstall) without infringing copyright.
It basicly shows that you either have one big DRM scheme to rule them all, or just don't have one at all. Maybe this puts DRM on hold until the standardizing and licensing issues are sorted out. On the other hand: If there is a single DRM scheme for all, then a single weakness in the DRM scheme unlocks ALL content protected under this scheme.
According to both LGPL and GPL the one you get the software from is the distributor. He is the one responsible for adhering to the licenses. He can of course sue his own software provider later, but for now it's Sony that distributed the programs.
If Sony is providing the source code for the programs and restates that the software is unter GPL (thus giving you the right to modify and distribute your modification), then everything is fine between Sony and you though.
There have been several similar cases in Europe about this, and in every case the GPL has been found valid, and the violation of the license has been considered healed, if the final distributor was able to get hold of the source code and distribute this one too under GPL.
Check GPL v2.0 section 4:
4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.
For Sony this means: They lost the right to distribute the Program, and they will be in violation of the GPL until they start to comply with the GPL themselves (e.g. distributing the source and allowing modifications and redistribution under GPL).
Strange. Everyone I know who does Instant Messaging has an IRC nick. Seems to be connected to the fact that I hang out on IRC. (And no. I don't have an AIM or MSN name.)
Says who?
There is something called 'checking for facts' (like phoning the wife to make sure he is the one who he claims to be).
Microsoft probably had to consent, otherwise they would have lost in court anyway. Microsoft Germany already lost a similar case in german court, where they tried to forbid a computer store (Snogard) to buy used licenses from people who don't need them (installing from a different source for instance) and bundle them with the shop's computer offerings.
Microsoft lost badly in court, mainly because of the First Sale doctrin (which in Germany is called Erschoepfungsgrundsatz).
Lindows got Microsoft's suit refused because 'Windows' is a generic english noun, and thus not trademarkable. So my analogy still holds. You can have a product in the software market with a name that is easily confusable with 'Windows' and still not infringe on Microsoft's trademark of 'Microsoft Windows'.
But Lindows was actually selling an operating system (heck, even a fully fledged software distribution!), and Microsoft in the end had to BUY the Lindows name to have the company changing the name to Linspire.
So even in the (US-) software world 'Windows' is not trademarkable.
(It's different in non english speaking countries, because there Windows is no generic noun.)