I do not see any reason why your hypothetical dolt should ever have to recompile a kernel. If Win4Lin requires this, then W4L is outside the reach of your user. QEMU Accelerator and VMWare for Linux do not require this, and they have substantiatly similar hooks in the kernel.
As for the nVidia drivers, I've used them for years, and I have never had to recompile the kernel. In fact, all I have ever needed to do was rerun the driver update after upgrading the kernel.
Finally, converting from RH9 to FC3 is no less complicated than converted from Windows NT 4.0 to Windows XP. Your argument would have been in the realm of reality if you were describing package upgrades within FC3, but of course, that's such a simple matter, there would hardly be anything to say about it.
Two people involved in this mess have already "committed suicide" under very unusual circumstances. One of them was a Canopy employee who just received 3 million in options vested immediately. The other was Ray Noorda's daughter. She supposedly shot her self in the head after winning the Norda/Canopy court case. It has also been pointed out that both were practicing Mormons, a group with a very low rate of suicide as their admission to heaven is forfeit if they take their own life.
Whether Ms. Jones has anything to fear is immaterial. As I read it, the police themselves have requested that she publicly establish that she has no intention of commiting suicide. I can think of a lot of reasons for that. If someone was stalking Ms. Jones, then a faked suicide has just been eliminated from the playbook.
Tridge did not have a BitKeeper license and was not bound to it in any way. This is a case were a license between two parties was voided by the issuer for actions by an independent third-party. Anyone can see this was an unstable situation. Whether Tridge voided Torvald's agreement with McVoy intentionally or not is irrelevent. He was never a party to their agreement.
This guy writes like a car reviewer that has never seen under the hood of a car.
"Ah, still using four tires, I see. And there's a steering wheel, too. Still, the color is nice, and the radio looks expensive."
I prefer reviews that do more than comment on the styling and goo-gahs. I expect him to drive the damn thing, pop the hood and see what has changed. Offer an explanation of why and how these changes improve the user's day-to-day work experience.
I propose that the reason he's discounted the other 198 claimed features is that he has not the foggiest idea what they do or how they fit into the future of Mac OS X.
At least two of the uses you specified, antivirus repositories and Windows Update Service, are simply unwanted overhead for your existing Windows deployment. That Linux is not helping you there is not really a surprise is it? As for your existing inventory software, talk to the vendor. If they did not do anything excessively clever when they wrote it, it may even run under Wine.
Our company has off-loaded 100% of web-hosting, mail, and file/printer sharing to Linux. The entire company server network runs without any additional attention 9 out of 10 days. We have a single admin managing 50 machines in four states and we have excess capacity. Using Linux where it's strongest is the best way to get a feel for its capabilities.
No case law is necessary since the definition will suffice. A license is a grant, not a piece of paper. If you could resurrect a void license by making a photocopy, the legal system would be turned on its ear. No lawyer would attempt to argue that for fear of sanctions.
Althought there are per-copy licenses out there, the GPL in particular is not, even if you would like it to be so. Moreover, subsequent revisions of the same product will be covered by the same license, so obtaining a later version will not provide you with a new license, merely another copy of the old license which was violated.
No license is per-copy unless it explicitly says so.
-Hope
You Misunderstand When and How the GPL is Applied.
on
CherryOS Goes Open Source
·
· Score: 2, Informative
The only "due process" that exists is before a jury, and that is after they've been hauled into court for copyright violation. The first time the GPL goes into action is before the judge, and it goes like this (courtesy of Eben Moglen):
EM: Your Honor, these people are distributing our copyrighted software. Please make them stop.
GPLViolator: We invoke the GPL as a defense.
EM: Your Honor, according to clause 4, when they violated the GPL here here and here, it was revoked and void, therefore they cannot use the GPL as a defense.
GV: (I'm screwed...) Let's settle.
To date, no one has gone beyond the "let's settle" stage. The only way they could do that would be to deny they violated the GPL and request a jury to make a decision. Good luck convincing a judge of that.
GoDaddy's practices leave something to be desired - anyone running anything even remotely controversal, especially if adult/porn related, does best to avoid GoDaddy when registering the hosting domain(s).
And why is that? Do they provide bad service to these domains? Do they lose their DNS records? All I ask of my registrar is that the root DNS servers stay updated. What else can go wrong? I'm genuinely curious.
Domain owners are not asking for anonymity. They are asking that a reasonable buffer be put between them and the public. When I walk into a shopping mall, I am obviously in a public place. No one has the right to know my name. No one has the right to know my address. This does not mean that this information should be protected by force of law. It does mean that I should not be required to provide it to any and everyone simply because I am in public.
For what its worth, the license is not a per-copy ticket. If you redistribute the code without explicit permission of the author you are technically violating copyright law. The GPL is an affirmative defense against that charge. If you violate the GPL, it is void, and therefore not a defense; hence, you must stop distributing.
If they'd given due credit to PearPC this wouldn't have happened, but don't you think that if they DO bring their behaviour into line with the PearPC license (GPL), they should be given a chance?
That decision is solely up to the PearPC developers and CherryOS has already told them to talk to a lawyer. I think it is fair to say that CherryOS had its chance.
As for whether the user gains or loses in this transaction -- even if they gain a GUI, they lose the source. That alone is not a good transaction in my opinion. Anyone can write a GUI for PearPC, including CherryOS. What they cannot do is deprive the user of that source code.
My company is based in Chicago. I live in Arizona. If I get a second job in New York, will I soon have %300 of my base income taxed? This is ridiculous. I can see paying tax on wages earned while in the state. I can see paying taxes on wages earned from a specific company in a state. But %100 of my income as a basis for taxation? I don't think so.
Apple does not maintain Fink so there is no reason to assume that they will provide details about their proprietary codecs to the OSS community. What Fink users have been providing back is help with endian problems since code written for x86 will not necessarily run correctly on the PPC. This helps Linux on PPC as well.
As for GUI applications, most Fink X programs run as "second-class citizens" on the OSX desktop (keyboard shortcuts are often not transmitted to the correct (if any) application window for instance) so the incentive to work directly with those code-bases is lessened. More often, the code itself is ported directly to the native OSX GUI which opens the question whether those changes can be incorporated back into the original project. Certainly, that depends on the project itself.
Their site requires cookies for state management. I had them turned off for all but specific sites. Perhaps an assert(cookie_is_valid) would be in order before that session lookup.
Session Timeout
Your session has timed out. Please register again.
Perhaps posting to slashdot wasn't the best way to collect data, but it will give them something to think about with regards to server application reliability.
Actually, the grandparent post is valid. It takes a couple of days for technical news discussed on Slashdot and reported by the Register to trickle up to the less technically-savvy investors. On a whim, I took $1000 to just over $5000 in January and February of 2000 using this exact strategy. Unfortunately, with the bubble now popped, the tech industry is already depressed so it's not obvious what kind of response you'll see today. Still, panic is panic.
For what it's worth, I went on vacation using that money and when I came back all my other investments were tanked -- net loss all around. Long-term diversification strategies are a better bet. Day trading is still gambling after all.
Interestingly enough, I had a proper rebuttal, and in the process of looking at your original post for the quote necessary to secure it, I discovered that I misread the intent of your post. For that, I apologize.
Your statement that the price is determined by the buyer's perception of the value is absolutely correct. I interpreted your statement to imply that the value was in fact zero. That is not what you stated at all, even if you were implying it.
However, even if the buyer's bid price vastly exceeds the asking price, the discrepency is arbitrable and nominally resolves to the asking price, zero, since the "seller" is generally an automated agent, a computer on the internet. All other aspects of the package such as service, support, and warm-fuzzy feelings are separable since they can be obtained with or without the software.
Finally, discrepencies between theoretical values and actual values are the heart of research in both Economics and Physics, but you will have a difficult time convincing me that Ph.D.'s in Economics are successfully throwing out Supply and Demand in favor of something that conforms better to observation. I spent a decade working with Economic Ph.D.'s, candidates, and instructors as my previous work was enabling them to test their theories in actual markets. Only the standard Supply and Demand models ever came close, although initial iterations were more the subject of game theory.
This is probably going to go way over your head, but for all intents and purposes, your understanding of how a price is determined is severly flawed. The price is determined by one thing only -- the intersection of the supply curve and the demand curve. That's it.
In your simple painting example, the supply curve has one item to sell, so the price will be determined exclusively by demand if the sellers minimum price is met.
What you have with Linux is a complete reversal of the curve. For commodity linux, the supply curve is infinitely long at a zero price. Since the demand cannot possibly be infinite, the demand curve must intersect the supply curve at a price of zero. It has nothing whatsoever to do with the buyer's utility function, the quantity they wish to purchase, or the "perceived value" of the software. In fact, every single buyer could bid a million dollars and the price would still be zero.
If you do not understand that, you missed the fundamental theory behind supply and demand.
For what it's worth, few Japanese people will argue with your statement about Pearl Harbor, even if it would make them uncomfortable to discuss it. However, what you have stated is not fully correct in the sense that the Japanese did not simply bomb the United States one Sunday morning without provocation -- many, many political and military events preclude Pearl Harbor. In particular, the United States was blocking fuel shipments from the Phillipines to Japan at the time. This was one of many "acts of war" that the Japanese claimed as pretense for initiating all-out conflict in the Pacific. The political situation in China was also a major factor, but that is less direct.
History is written by the winners. Had the United State's aircraft carriers been in port, the outcome of the pacific war may have been less certain, in which case December 7th would have been marked in the Japanese history books as a decisive response to the oil embargo, or some such. As it is, major portions of the war and the time preceding it are hardly mentioned in Japanese history texts at all. Some call it revisionism; I call it shame. In the end, war is a very inefficient means to resolve conflicts; however, any time someone dictates another's actions by force, the situation ultimately devolves into warfare of one type or another. Neither the Japanese or the United States can absolve themselves of that responsibility.
In my experience listening to dozens of Japanese WWII vets through-out Japan talk about their experience, I've come to the conclusion that they are very similar to American war vets. The principal difference is that they lost and that they felt betrayed by their leaders for endangering their country. I suspect that the successful U.S. occupation helped with that opinion as many of the Japanese vets I talked to seemed grateful that the U.S. provided them a means to retain their pride. General MacArthur was spot on in that, and many other regards.
You brought up an interesting example and one that I use often. Someone recently pointed out a different order that apparently improves things a percent on some processors. Preload the cache by reading the first column. On a parallel pipeline processor, each read will stall waiting for main memory, but many of the reads will stall in parallel. Also, the whole cache line will be loaded which could be a substantial amount of the array depending. As the y/x loop finally begins, main memory will be arriving for future reads on upcoming rows.
Personally, I have not put this to practice, but it was an interesting thought anyway.
Microsoft is not enforcing what operating systems Microsoft Office will run on. It is enforcing what operating systems it will provide an automatic update service to.
One of the principal ideas behind consumer protect laws goes like this: a business may group customers into classes. Those classes may be defined by a number of legal means including inking of individual contracts and public or private pricing structures. It cannot be based on all the usual items like race, color, and sex. Some other non-obvious things that such classes cannot be based on: state of residence, height, last name, and most importantly, third-party business relationships. In short, customers have a legal expectation of equivalent service to all other customers in their class. If someone sues for breach of this law, they are stating that they did not receive equivalent service to members of their class. The company's defense must state that the service was in fact "equivalent" (to be decided by a jury) or by what means this customer is classified differently (also to be decided by a jury). If they state that downloading via a separate product using a separate website is equivalent, they lose. If they state "usage of third-party product" as a means of classification, they lose because that is not a legally justifiable means of classifying your customers. In fact, it's explicitly illegal. If they state, "failure to use our other product", then they lose because that is illegal for both anti-trust reasons and consumer protection reasons. In short, they lose. Eventually. After the lawyers are done.
Meanwhile, by publicly offering updates to their customers, Microsoft is establishing an open contract to customers of a given product class. Any intentional action by Microsoft that prevents a consumer from receive equivalent service to any other customer in his class based on the fact that he uses WINE or does not use Windows, is illegal.
This does not mean that Microsoft has to lift a finger to help WINE connect and get updates. It does not mean that it must provide updates to non-paying customers. It does mean that it must provide the exact same service to all customers who paid for Microsoft Office. In this regard, it has broken the law, since it is providing a degraded service with the consumer class partitioned illegally.
With regard to AIM, AOL provides the service to paying AOL customers. Gaim and similar users are not entitled to nor have any expectation to free service from AOL. What would be illegal is if AOL's software refused to operate on a computer that had a competitor's product also installed. That would be a violation of consumer rights law.
Informative, but I don't see anything that damns Microsoft outright.
Specifically, I was referring to this item: "The actionable wrong lies in the inducement to break the contract or to sever the relationship, not in the kind of contract or relationship so disrupted, whether it is written or oral, enforceable or not enforceable." Microsoft has induced a situation in which the customer must discontinue use of WINE, a third-party product, un-related the Office product, or forfeit access to equivalent product upgrades that the consumer is legally entitled to based on consumer protection laws. That covers "inducement to break contract as well as sever [business] relationship" between the consumer and companies selling WINE-based products. The contract is implied fitness and warranty. The relationship is anything a jury defines it as, but in particular, future sales.
There is a common misconception that you share which is that an object can be identified by a dictionary definition. This is not so, and let me explain why.
To identify an object, it is necessary to determine what it is not, rather than what it is. For example, if I define a car to be "a four wheeled passenger vehicle powered by a combustion engine" one would not be able to distinguish it from a van or a pickup truck or a school bus. In fact, no matter how much additional detail is added to the definition, there is a possibility that two dissimilar things may meet that description and not be the aforementioned object.
Here too with WINE, certainly it is similar to an emulator. It shares many of the same traits as an emulator. What an emulator is not however, is a native implementation of an API. That alone invalidates any possibility of WINE being an emulator.
Faulting the dictionary for not providing enough information to invalidate WINE as an emulator is to fail to understand what the dictionary is providing. It states: if you have an object, and that object is an "emulator", then one or more of the following definitions is true. It makes no other assurances, and thus, if you have an object and it is not an emulator, the dictionary can tell you absolutely nothing about it.
In summary, a dictionary can tell you what something means; it cannot tell you if something you have can be called by a particular name. That is determined largely be consensus which you have accurately pointed out to be muddled. However, the authority for establishing criteria for identifying an object rests with the people who created the criteria in the first place, the experts in the field, so it is their consensus alone that matters. Everyone else is simply incorrect.
Most languages have words in common usage that differ from their correct usage. The common usage does not become more correct as the number of people using it increases. This situation is no different.
What you have stated is not true. While Microsoft has no responsibility to uphold their quality guarantees if the product is used contrary to their expectations, they cannot deny their publicly advertised and generally available services to their own customers without violating consumer protection laws. Furthermore, the stated "system requirements" can recommend specific products, but it cannot mandate them. That would be product bundling, and that is illegal under anti-trust laws. Even if Microsoft does not put "or equivalent" after those operating systems, legally, that is how it is interpreted in a court of law.
If you would like an example, have you ever considered why the car industry cannot prevent you from putting a third-party stereo in your car or have a non-certified mechanic work on your engine, and yet you can still have your engine serviced by the dealership? If they denied you service on any similar basis, they would be breaking the law.
Slashdot Mirror
I do not see any reason why your hypothetical dolt should ever have to recompile a kernel. If Win4Lin requires this, then W4L is outside the reach of your user. QEMU Accelerator and VMWare for Linux do not require this, and they have substantiatly similar hooks in the kernel.
As for the nVidia drivers, I've used them for years, and I have never had to recompile the kernel. In fact, all I have ever needed to do was rerun the driver update after upgrading the kernel.
Finally, converting from RH9 to FC3 is no less complicated than converted from Windows NT 4.0 to Windows XP. Your argument would have been in the realm of reality if you were describing package upgrades within FC3, but of course, that's such a simple matter, there would hardly be anything to say about it.
-Hope
Two people involved in this mess have already "committed suicide" under very unusual circumstances. One of them was a Canopy employee who just received 3 million in options vested immediately. The other was Ray Noorda's daughter. She supposedly shot her self in the head after winning the Norda/Canopy court case. It has also been pointed out that both were practicing Mormons, a group with a very low rate of suicide as their admission to heaven is forfeit if they take their own life.
Whether Ms. Jones has anything to fear is immaterial. As I read it, the police themselves have requested that she publicly establish that she has no intention of commiting suicide. I can think of a lot of reasons for that. If someone was stalking Ms. Jones, then a faked suicide has just been eliminated from the playbook.
-Hope
Tridge did not have a BitKeeper license and was not bound to it in any way. This is a case were a license between two parties was voided by the issuer for actions by an independent third-party. Anyone can see this was an unstable situation. Whether Tridge voided Torvald's agreement with McVoy intentionally or not is irrelevent. He was never a party to their agreement.
-Hope
This guy writes like a car reviewer that has never seen under the hood of a car.
"Ah, still using four tires, I see. And there's a steering wheel, too. Still, the color is nice, and the radio looks expensive."
I prefer reviews that do more than comment on the styling and goo-gahs. I expect him to drive the damn thing, pop the hood and see what has changed. Offer an explanation of why and how these changes improve the user's day-to-day work experience.
I propose that the reason he's discounted the other 198 claimed features is that he has not the foggiest idea what they do or how they fit into the future of Mac OS X.
-Hope
At least two of the uses you specified, antivirus repositories and Windows Update Service, are simply unwanted overhead for your existing Windows deployment. That Linux is not helping you there is not really a surprise is it? As for your existing inventory software, talk to the vendor. If they did not do anything excessively clever when they wrote it, it may even run under Wine.
Our company has off-loaded 100% of web-hosting, mail, and file/printer sharing to Linux. The entire company server network runs without any additional attention 9 out of 10 days. We have a single admin managing 50 machines in four states and we have excess capacity. Using Linux where it's strongest is the best way to get a feel for its capabilities.
-Hope
And backslashdot would have Pro-Microsoft articles.
No case law is necessary since the definition will suffice. A license is a grant, not a piece of paper. If you could resurrect a void license by making a photocopy, the legal system would be turned on its ear. No lawyer would attempt to argue that for fear of sanctions.
-Hope
Althought there are per-copy licenses out there, the GPL in particular is not, even if you would like it to be so. Moreover, subsequent revisions of the same product will be covered by the same license, so obtaining a later version will not provide you with a new license, merely another copy of the old license which was violated.
No license is per-copy unless it explicitly says so.
-Hope
The only "due process" that exists is before a jury, and that is after they've been hauled into court for copyright violation. The first time the GPL goes into action is before the judge, and it goes like this (courtesy of Eben Moglen):
EM: Your Honor, these people are distributing our copyrighted software. Please make them stop.
GPLViolator: We invoke the GPL as a defense.
EM: Your Honor, according to clause 4, when they violated the GPL here here and here, it was revoked and void, therefore they cannot use the GPL as a defense.
GV: (I'm screwed...) Let's settle.
To date, no one has gone beyond the "let's settle" stage. The only way they could do that would be to deny they violated the GPL and request a jury to make a decision. Good luck convincing a judge of that.
-Hope
GoDaddy's practices leave something to be desired - anyone running anything even remotely controversal, especially if adult/porn related, does best to avoid GoDaddy when registering the hosting domain(s).
And why is that? Do they provide bad service to these domains? Do they lose their DNS records? All I ask of my registrar is that the root DNS servers stay updated. What else can go wrong? I'm genuinely curious.
-Hope
Domain owners are not asking for anonymity. They are asking that a reasonable buffer be put between them and the public. When I walk into a shopping mall, I am obviously in a public place. No one has the right to know my name. No one has the right to know my address. This does not mean that this information should be protected by force of law. It does mean that I should not be required to provide it to any and everyone simply because I am in public.
-Hope
For what its worth, the license is not a per-copy ticket. If you redistribute the code without explicit permission of the author you are technically violating copyright law. The GPL is an affirmative defense against that charge. If you violate the GPL, it is void, and therefore not a defense; hence, you must stop distributing.
-HopeOS
If they'd given due credit to PearPC this wouldn't have happened, but don't you think that if they DO bring their behaviour into line with the PearPC license (GPL), they should be given a chance?
That decision is solely up to the PearPC developers and CherryOS has already told them to talk to a lawyer. I think it is fair to say that CherryOS had its chance.
As for whether the user gains or loses in this transaction -- even if they gain a GUI, they lose the source. That alone is not a good transaction in my opinion. Anyone can write a GUI for PearPC, including CherryOS. What they cannot do is deprive the user of that source code.
-Hope
My company is based in Chicago. I live in Arizona. If I get a second job in New York, will I soon have %300 of my base income taxed? This is ridiculous. I can see paying tax on wages earned while in the state. I can see paying taxes on wages earned from a specific company in a state. But %100 of my income as a basis for taxation? I don't think so.
-Hope
Apple does not maintain Fink so there is no reason to assume that they will provide details about their proprietary codecs to the OSS community. What Fink users have been providing back is help with endian problems since code written for x86 will not necessarily run correctly on the PPC. This helps Linux on PPC as well.
As for GUI applications, most Fink X programs run as "second-class citizens" on the OSX desktop (keyboard shortcuts are often not transmitted to the correct (if any) application window for instance) so the incentive to work directly with those code-bases is lessened. More often, the code itself is ported directly to the native OSX GUI which opens the question whether those changes can be incorporated back into the original project. Certainly, that depends on the project itself.
-Hope
Their site requires cookies for state management. I had them turned off for all but specific sites. Perhaps an assert(cookie_is_valid) would be in order before that session lookup.
-Hope
Session Timeout
Your session has timed out. Please register again.
Perhaps posting to slashdot wasn't the best way to collect data, but it will give them something to think about with regards to server application reliability.
-Hope
Actually, the grandparent post is valid. It takes a couple of days for technical news discussed on Slashdot and reported by the Register to trickle up to the less technically-savvy investors. On a whim, I took $1000 to just over $5000 in January and February of 2000 using this exact strategy. Unfortunately, with the bubble now popped, the tech industry is already depressed so it's not obvious what kind of response you'll see today. Still, panic is panic.
For what it's worth, I went on vacation using that money and when I came back all my other investments were tanked -- net loss all around. Long-term diversification strategies are a better bet. Day trading is still gambling after all.
-Hope
Interestingly enough, I had a proper rebuttal, and in the process of looking at your original post for the quote necessary to secure it, I discovered that I misread the intent of your post. For that, I apologize.
Your statement that the price is determined by the buyer's perception of the value is absolutely correct. I interpreted your statement to imply that the value was in fact zero. That is not what you stated at all, even if you were implying it.
However, even if the buyer's bid price vastly exceeds the asking price, the discrepency is arbitrable and nominally resolves to the asking price, zero, since the "seller" is generally an automated agent, a computer on the internet. All other aspects of the package such as service, support, and warm-fuzzy feelings are separable since they can be obtained with or without the software.
Finally, discrepencies between theoretical values and actual values are the heart of research in both Economics and Physics, but you will have a difficult time convincing me that Ph.D.'s in Economics are successfully throwing out Supply and Demand in favor of something that conforms better to observation. I spent a decade working with Economic Ph.D.'s, candidates, and instructors as my previous work was enabling them to test their theories in actual markets. Only the standard Supply and Demand models ever came close, although initial iterations were more the subject of game theory.
-Hope
This is probably going to go way over your head, but for all intents and purposes, your understanding of how a price is determined is severly flawed. The price is determined by one thing only -- the intersection of the supply curve and the demand curve. That's it.
In your simple painting example, the supply curve has one item to sell, so the price will be determined exclusively by demand if the sellers minimum price is met.
What you have with Linux is a complete reversal of the curve. For commodity linux, the supply curve is infinitely long at a zero price. Since the demand cannot possibly be infinite, the demand curve must intersect the supply curve at a price of zero. It has nothing whatsoever to do with the buyer's utility function, the quantity they wish to purchase, or the "perceived value" of the software. In fact, every single buyer could bid a million dollars and the price would still be zero.
If you do not understand that, you missed the fundamental theory behind supply and demand.
-Hope
For what it's worth, few Japanese people will argue with your statement about Pearl Harbor, even if it would make them uncomfortable to discuss it. However, what you have stated is not fully correct in the sense that the Japanese did not simply bomb the United States one Sunday morning without provocation -- many, many political and military events preclude Pearl Harbor. In particular, the United States was blocking fuel shipments from the Phillipines to Japan at the time. This was one of many "acts of war" that the Japanese claimed as pretense for initiating all-out conflict in the Pacific. The political situation in China was also a major factor, but that is less direct.
History is written by the winners. Had the United State's aircraft carriers been in port, the outcome of the pacific war may have been less certain, in which case December 7th would have been marked in the Japanese history books as a decisive response to the oil embargo, or some such. As it is, major portions of the war and the time preceding it are hardly mentioned in Japanese history texts at all. Some call it revisionism; I call it shame. In the end, war is a very inefficient means to resolve conflicts; however, any time someone dictates another's actions by force, the situation ultimately devolves into warfare of one type or another. Neither the Japanese or the United States can absolve themselves of that responsibility.
In my experience listening to dozens of Japanese WWII vets through-out Japan talk about their experience, I've come to the conclusion that they are very similar to American war vets. The principal difference is that they lost and that they felt betrayed by their leaders for endangering their country. I suspect that the successful U.S. occupation helped with that opinion as many of the Japanese vets I talked to seemed grateful that the U.S. provided them a means to retain their pride. General MacArthur was spot on in that, and many other regards.
-Hope
You brought up an interesting example and one that I use often. Someone recently pointed out a different order that apparently improves things a percent on some processors. Preload the cache by reading the first column. On a parallel pipeline processor, each read will stall waiting for main memory, but many of the reads will stall in parallel. Also, the whole cache line will be loaded which could be a substantial amount of the array depending. As the y/x loop finally begins, main memory will be arriving for future reads on upcoming rows.
Personally, I have not put this to practice, but it was an interesting thought anyway.
-Hope
Microsoft is not enforcing what operating systems Microsoft Office will run on. It is enforcing what operating systems it will provide an automatic update service to.
One of the principal ideas behind consumer protect laws goes like this: a business may group customers into classes. Those classes may be defined by a number of legal means including inking of individual contracts and public or private pricing structures. It cannot be based on all the usual items like race, color, and sex. Some other non-obvious things that such classes cannot be based on: state of residence, height, last name, and most importantly, third-party business relationships. In short, customers have a legal expectation of equivalent service to all other customers in their class. If someone sues for breach of this law, they are stating that they did not receive equivalent service to members of their class. The company's defense must state that the service was in fact "equivalent" (to be decided by a jury) or by what means this customer is classified differently (also to be decided by a jury). If they state that downloading via a separate product using a separate website is equivalent, they lose. If they state "usage of third-party product" as a means of classification, they lose because that is not a legally justifiable means of classifying your customers. In fact, it's explicitly illegal. If they state, "failure to use our other product", then they lose because that is illegal for both anti-trust reasons and consumer protection reasons. In short, they lose. Eventually. After the lawyers are done.
Meanwhile, by publicly offering updates to their customers, Microsoft is establishing an open contract to customers of a given product class. Any intentional action by Microsoft that prevents a consumer from receive equivalent service to any other customer in his class based on the fact that he uses WINE or does not use Windows, is illegal.
This does not mean that Microsoft has to lift a finger to help WINE connect and get updates. It does not mean that it must provide updates to non-paying customers. It does mean that it must provide the exact same service to all customers who paid for Microsoft Office. In this regard, it has broken the law, since it is providing a degraded service with the consumer class partitioned illegally.
With regard to AIM, AOL provides the service to paying AOL customers. Gaim and similar users are not entitled to nor have any expectation to free service from AOL. What would be illegal is if AOL's software refused to operate on a computer that had a competitor's product also installed. That would be a violation of consumer rights law.
Informative, but I don't see anything that damns Microsoft outright.
Specifically, I was referring to this item: "The actionable wrong lies in the inducement to break the contract or to sever the relationship, not in the kind of contract or relationship so disrupted, whether it is written or oral, enforceable or not enforceable." Microsoft has induced a situation in which the customer must discontinue use of WINE, a third-party product, un-related the Office product, or forfeit access to equivalent product upgrades that the consumer is legally entitled to based on consumer protection laws. That covers "inducement to break contract as well as sever [business] relationship" between the consumer and companies selling WINE-based products. The contract is implied fitness and warranty. The relationship is anything a jury defines it as, but in particular, future sales.
-Hope
There is a common misconception that you share which is that an object can be identified by a dictionary definition. This is not so, and let me explain why.
To identify an object, it is necessary to determine what it is not, rather than what it is. For example, if I define a car to be "a four wheeled passenger vehicle powered by a combustion engine" one would not be able to distinguish it from a van or a pickup truck or a school bus. In fact, no matter how much additional detail is added to the definition, there is a possibility that two dissimilar things may meet that description and not be the aforementioned object.
Here too with WINE, certainly it is similar to an emulator. It shares many of the same traits as an emulator. What an emulator is not however, is a native implementation of an API. That alone invalidates any possibility of WINE being an emulator.
Faulting the dictionary for not providing enough information to invalidate WINE as an emulator is to fail to understand what the dictionary is providing. It states: if you have an object, and that object is an "emulator", then one or more of the following definitions is true. It makes no other assurances, and thus, if you have an object and it is not an emulator, the dictionary can tell you absolutely nothing about it.
In summary, a dictionary can tell you what something means; it cannot tell you if something you have can be called by a particular name. That is determined largely be consensus which you have accurately pointed out to be muddled. However, the authority for establishing criteria for identifying an object rests with the people who created the criteria in the first place, the experts in the field, so it is their consensus alone that matters. Everyone else is simply incorrect.
Most languages have words in common usage that differ from their correct usage. The common usage does not become more correct as the number of people using it increases. This situation is no different.
-Hope
What you have stated is not true. While Microsoft has no responsibility to uphold their quality guarantees if the product is used contrary to their expectations, they cannot deny their publicly advertised and generally available services to their own customers without violating consumer protection laws. Furthermore, the stated "system requirements" can recommend specific products, but it cannot mandate them. That would be product bundling, and that is illegal under anti-trust laws. Even if Microsoft does not put "or equivalent" after those operating systems, legally, that is how it is interpreted in a court of law.
If you would like an example, have you ever considered why the car industry cannot prevent you from putting a third-party stereo in your car or have a non-certified mechanic work on your engine, and yet you can still have your engine serviced by the dealership? If they denied you service on any similar basis, they would be breaking the law.
-Hope