The fact that filesharers still bother downloading and sharing those files reflects that there is actual demand for that content
Just remember that demand is highly affected by price, and when the price of a product with even an infinitesimal perceived value reaches zero, the demand will increase sharply. Thus, you cannot reasonably compare demand indicated by download counts at a pirate site with demand in the marketplace where the traded items would actually cost money. Then everyone starts making cost/benefit calculations, and the demand will be very different.
And someone is willing to steal it and not pay you anything; quite a few people, perhaps.
What the GP might have meant is that his works cannot easily be stolen. I work for a company writing software that provides services to users (mostly other companies). We do not distribute any software, so the only thing that you can steal is our web pages. Sure, they can substitute me for another programmer, but they still have to pay someone to write the software that they need to continue to provide and improve the company's services.
Only a minority of the paid software developers work in companies that sell packaged software to customers. Many more work developing software for internal use only, either to provide services to external entities, or to aid internal business processes. Pirating such software is much harder than pirating packaged software available to the entire world.
Maybe this freudian slip actually shows the issues with GPL. Although atomically the GPL is clear by itself, using the GPL with other licenses basically forces the whole project to be GPL.
But how is this relevant to the discussion at hand? Microsoft doesn't seem to be contributing a lot to BSD projects nor GPL projects alike. So using this example to point out so-called "issues" in the GPL gives the logical conclusion that those same "issues" plague the BSD license too.
Libraries that clone existing non-GPL libraries should use the LGPL instead of the GPL, since there is no point in using the GPL, as users will only use the other alternatives.
Libraries that provide some unique functionality on the other hand, could very well be licensed under the GPL, to encourage wider use of the GPL by possible clients of the library.
Remember that licenses must be written in Legalese. Legalese is a very verbose language. What the GP wrote will be much larger when translated to Legalese.
Like proprietary licenses, MIT/BSD licenses (heavily/exclusively favouring the user) get little debate because they are unambiguous in both their intents and their effects.
I don't really think that the many heated debates over the GPL is due to any ambiguities in the license, but rather because of very different opinions on copyleft and other aspects of the GPL.
The fact that you link your program with a GPLd library doesn't magically change the copyright on it. Your program would have to be licensed under the GPL or a GPL-compatible license, but it's still YOUR program with YOUR copyright.
Exactly. And I also find this similar to the anti-tivoization clauses in the GPLv3. Detractors claim that those clauses try to extend the copyright of a program to the hardware it runs on. By an analogous argument to yours, this can be shown to be completely off base.
This makes it pretty unrealistic to build a business plan around selling software derived from GPL software commercially.
IIRC, there is a company that sells GPL-licensed router firmwares, providing the source only to their customers. If one of their customers leak the source to third parties, and the company finds out about it, the company voids any right to firmware updates for that customer.
I think that the FSF grudgingly conceded that this may be allowed by the letter of the GPLv2, but very far from its intent.
AFAIK, only the copyright holder can sue for infringement, not just any random guy. You might already know this, but I felt that it might be good to point this out anyway.
This is exactly why the LGPL was created. Or sometimes you will have a GPL lib with the linking or classpath exception.
Can anyone explain the difference between the LGPL and GPL with linking/classpath exception? They seem pretty similar, and I really never completely understood why both of these exist.
What he shouldn't do is claim that it's open source
So now the BSD fanatics try to appropriate the term "open source" and claim that something is only open source if it falls within their (your) own definition.
The argument is that Red Hat, IBM, Novell, and other companies wouldn't have contributed to Linux if they weren't forced to by the GPL.
Not necessarily. Another possibility is that those companies would not contribute if the GPL did not prevent competitors from taking their contributions proprietary.
That may be so for most operating systems, but since Vista, Windows apps can assign memory priorities to their allocated memory. This was introduced to counter the after lunch syndrome where a memory intensive process (...) has evicted all your procceses memory while you were eating. Processes can also register to be notified when their memory is evicted and may choose to just drop the memory instead of having it swapped to disk.
Interesting. That might be useful for some applications that implement their own caching. But most applications that I know of do not have large amounts of memory that they can do without on a short notice.
(e.g. a search indexer that springs into action on idle conditions
Clever implementations of search indexers doesn't need to run on idle conditions. It runs once when the OS is installed, and any changes after that are intercepted by subscribing to file system events, such as inotify in Linux or FSEvents in Mac OS X.
A process such as Windows Defender or the search indexer can now allocate memory with a lower-than-normal priority. This will not cause higher prioritzed memory to be swapped out and this memory will be the first to be evicted when higher-priority memory is allocated.
But what does that do to backwards compatibility? Do applications that use those features become Vista only, or can they run even if the OS does not support those features? Of course in that case, there would be no memory prioritization, but that is probably less important than being able to run the applications at all.
Windows Superfetch is an example of a user-mode process which leverage this memory infrastructure to allocate memory in a speculative manner. If memory is allocated with a higher priority, Superfetch readily gives up the memory with no additional disk access.
Sounds like a normal block/page cache, except for the speculative allocation. The prioritization may be implicit in other systems, but it certainly exists, as the block cache is never prioritized over memory allocation requests.
So while most operating systems indeed do not have a standardized way to tell applications to give up memory, this is exactly what Windows Vista, Server 2008 and later feature.
It's too bad such details won't make it into mainstream marketing. I'm not enough into Windows to read their developer news (I'm more of a *nix guy), where such information probably would show up. Such features you describe are certainly more interesting to me than animated wallpapers and translucent window frames.
I certainly do, but you, on the other hand, do not seem to know what you are talking about.
Unused RAM is RAM that is not being used. By anything. You open up task manager and see 500mb of RAM used out of 3gb, well, 2.5gb is not being used. That is unused RAM
Certain operating system tools do not make a clear distinction between really unused RAM and RAM that is not used by applications. In the latter case, the so-called "unused" RAM is not unused at all, but rather used for the page cache.
Where the hell did you come up with this disk cache BS? The opposite happens - the OS pulls data out of RAM and caches it onto disk when the data has been loaded but is not expected to be used for a while, or other programs need the RAM. This is called a PageFile in windows, and a Swap partition in Linux.
The disk cache (or page cache) is certainly not BS. Both the page cache and swap file/partition are used, but for different purposes. The disk cache is used to cache frequently used disk blocks in the memory unused by applications, while the swap file/partition is used to temporarily free up physical memory when applications request memory but most or all is already used.
I don't think you have much programming background at all
My programming background is pretty extensive.
or you'd know that directly pointing to addresses hasn't been done in a long, long time. Some languages, like C and others still allow it, but 99.99% of the time it is a very bad idea. It's the kind of shit you do with assembly language programing.
Here you are mixing up two concepts: Virtual memory and managed code. Using physical addresses in applications hasn't been used in a long time, since they use virtual memory addresses instead. The translation is then done by the MMU (Memory Management Unit) together with the operating system, using the page tables that list which virtual pages are mapped into which physical pages (or swap areas). Even applications written in C or assembly language can not use physical addresses in any modern operating system, as they are completely managed by the kernel and MMU.
Managed code is a totally different matter. Languages such as Java, C#, and pretty much all scripting languages use managed code. Pointers are not used by application programmers in those languages, instead, programmers use another layer of indirection to access objects, such as references. These references are then mapped to (virtual) memory addresses by the virtual machine or interpreter that the applications run in.
The only thing that is -somewhat- like your imaginary "disk caching" is Vista's ReadyBoost, which moves frequently used data to a FLASH drive in order to improve responsiveness.
Disk caching as I explained it exists on all modern operating systems, including Windows, Linux and Mac OS X. The Vista ReadyBoost is an improvement on this, since it actually preloads the flash drive with frequently used application data, enabling quick use even the first time after system boot.
Now, what you might be talking about, in a confused illiterate monkey sort of way, is the process by which the OS moves data out of physical RAM and onto disk cache (the swap or pagefile).
No, that is not what I'm talking about.
In no way, shape, or form does this mean the OS takes commonly used data and loads it into RAM.
No, the OS doesn't do that by itself, but I also never said that. What happens is the following: Each time a block is read from the hard drive (as requested by an application), it is loaded into the disk cache. This cache is made up of all RAM that is not used by the OS or applications. When applications require more heap space, the disk cache size is automatically reduced, and data in the cache
If the whole RAM isn't used, you're just wasting it.
That's hardly the case. Unused RAM is used as a disk cache, so that frequently read disk blocks reside in RAM instead of on disk. This makes reading them extremely fast. If applications allocate memory willy-nilly just because it's there, there won't be any memory left for the disk cache, and your system might become very slow. And if even more memory is allocated, the system will start paging stuff in and out of memory, slowing stuff down even more.
Applications should use the amount of memory they need, preferably leaving caching to the operating system. Applications generally cannot know when other processes need the RAM better, and AFAIK cannot be told to release memory by other applications in any standardized way. Not to mention that they are not told when the disk cache becomes too small, or when paging starts to occur, since the operating system reasonably assumes that memory allocated is required by the application, not just held for the fun of it.
The doctor said that nicotine treatment "would offer significant advancement over current treatments".
There is a difference between nicotine treatment (your words) and nicotine-based treatment (the doctor's words). The latter can imply that you are not treated with nicotine itself, but with some derivative of nicotine, e.g. with certain properties removed or minimized.
AFAIK, nicotine is a vasoconstrictor, which means it constricts the blood vessels in the body. This puts increased strain on your heart, and may prevent the extremities (arms and legs) from getting enough blood. If you shook hands with a smoker, you may have noticed that they often have cold hands, which is due to this low blood supply.
So oppening it up to public bidding? What the point when there is only one vendor who provides the product you're looking to buy?
Where I live (Sweden), that's not legal. You have to specify what the product should do, not which product you want to buy. For example, a couple of years ago, a county opened a bidding process for home PCs for employees, and specified that to be accepted, the computers offered must be equipped with an Intel Pentium 4 processor. AMD felt excluded by this requirement, brought the county to court, and won.
Of course, there are ways around that, such as specifying a list of features that only one product can match. But still, even in that case, the bidding process allows you to choose the cheapest vendor that carries that product.
Slashdot Mirror
The fact that filesharers still bother downloading and sharing those files reflects that there is actual demand for that content
Just remember that demand is highly affected by price, and when the price of a product with even an infinitesimal perceived value reaches zero, the demand will increase sharply. Thus, you cannot reasonably compare demand indicated by download counts at a pirate site with demand in the marketplace where the traded items would actually cost money. Then everyone starts making cost/benefit calculations, and the demand will be very different.
And someone is willing to steal it and not pay you anything; quite a few people, perhaps.
What the GP might have meant is that his works cannot easily be stolen. I work for a company writing software that provides services to users (mostly other companies). We do not distribute any software, so the only thing that you can steal is our web pages. Sure, they can substitute me for another programmer, but they still have to pay someone to write the software that they need to continue to provide and improve the company's services.
Only a minority of the paid software developers work in companies that sell packaged software to customers. Many more work developing software for internal use only, either to provide services to external entities, or to aid internal business processes. Pirating such software is much harder than pirating packaged software available to the entire world.
Maybe this freudian slip actually shows the issues with GPL. Although atomically the GPL is clear by itself, using the GPL with other licenses basically forces the whole project to be GPL.
But how is this relevant to the discussion at hand? Microsoft doesn't seem to be contributing a lot to BSD projects nor GPL projects alike. So using this example to point out so-called "issues" in the GPL gives the logical conclusion that those same "issues" plague the BSD license too.
Libraries shouldn't use the normal GPL.
Libraries that clone existing non-GPL libraries should use the LGPL instead of the GPL, since there is no point in using the GPL, as users will only use the other alternatives.
Libraries that provide some unique functionality on the other hand, could very well be licensed under the GPL, to encourage wider use of the GPL by possible clients of the library.
Remember that licenses must be written in Legalese. Legalese is a very verbose language. What the GP wrote will be much larger when translated to Legalese.
Only if you're in the business of selling copies of said program.
Like proprietary licenses, MIT/BSD licenses (heavily/exclusively favouring the user) get little debate because they are unambiguous in both their intents and their effects.
I don't really think that the many heated debates over the GPL is due to any ambiguities in the license, but rather because of very different opinions on copyleft and other aspects of the GPL.
The fact that you link your program with a GPLd library doesn't magically change the copyright on it. Your program would have to be licensed under the GPL or a GPL-compatible license, but it's still YOUR program with YOUR copyright.
Exactly. And I also find this similar to the anti-tivoization clauses in the GPLv3. Detractors claim that those clauses try to extend the copyright of a program to the hardware it runs on. By an analogous argument to yours, this can be shown to be completely off base.
This makes it pretty unrealistic to build a business plan around selling software derived from GPL software commercially.
IIRC, there is a company that sells GPL-licensed router firmwares, providing the source only to their customers. If one of their customers leak the source to third parties, and the company finds out about it, the company voids any right to firmware updates for that customer.
I think that the FSF grudgingly conceded that this may be allowed by the letter of the GPLv2, but very far from its intent.
where some GPL zealot can potentially sue.
AFAIK, only the copyright holder can sue for infringement, not just any random guy. You might already know this, but I felt that it might be good to point this out anyway.
This is exactly why the LGPL was created. Or sometimes you will have a GPL lib with the linking or classpath exception.
Can anyone explain the difference between the LGPL and GPL with linking/classpath exception? They seem pretty similar, and I really never completely understood why both of these exist.
I have something against people claiming to be open when doing so.
That's your problem. Deal with it. You BSD fanatics do not have a monopoly on the term "open source".
What he shouldn't do is claim that it's open source
So now the BSD fanatics try to appropriate the term "open source" and claim that something is only open source if it falls within their (your) own definition.
Would you pay for patent invalidation litigation if $BIGCORP accuses your OSS project of patent infringement? Litigation can be very expensive.
The GPL just makes it too much of a hassle in most cases
The GPL is not a problem at all for companies that do not distribute software.
The argument is that Red Hat, IBM, Novell, and other companies wouldn't have contributed to Linux if they weren't forced to by the GPL.
Not necessarily. Another possibility is that those companies would not contribute if the GPL did not prevent competitors from taking their contributions proprietary.
This is a common myth around the GPL, that it enforces a reciprocation agreement. It doesn't.
No, but in practice, that is usually what happens in the end.
Really? Every company I've ever worked at has used and contributed to BSD and LGPL projects, but not touched GPL with a barge pole.
Could it possibly be because you spread some FUD about the GPL in these companies? :P
That may be so for most operating systems, but since Vista, Windows apps can assign memory priorities to their allocated memory. This was introduced to counter the after lunch syndrome where a memory intensive process (...) has evicted all your procceses memory while you were eating. Processes can also register to be notified when their memory is evicted and may choose to just drop the memory instead of having it swapped to disk.
Interesting. That might be useful for some applications that implement their own caching. But most applications that I know of do not have large amounts of memory that they can do without on a short notice.
(e.g. a search indexer that springs into action on idle conditions
Clever implementations of search indexers doesn't need to run on idle conditions. It runs once when the OS is installed, and any changes after that are intercepted by subscribing to file system events, such as inotify in Linux or FSEvents in Mac OS X.
A process such as Windows Defender or the search indexer can now allocate memory with a lower-than-normal priority. This will not cause higher prioritzed memory to be swapped out and this memory will be the first to be evicted when higher-priority memory is allocated.
But what does that do to backwards compatibility? Do applications that use those features become Vista only, or can they run even if the OS does not support those features? Of course in that case, there would be no memory prioritization, but that is probably less important than being able to run the applications at all.
Windows Superfetch is an example of a user-mode process which leverage this memory infrastructure to allocate memory in a speculative manner. If memory is allocated with a higher priority, Superfetch readily gives up the memory with no additional disk access.
Sounds like a normal block/page cache, except for the speculative allocation. The prioritization may be implicit in other systems, but it certainly exists, as the block cache is never prioritized over memory allocation requests.
So while most operating systems indeed do not have a standardized way to tell applications to give up memory, this is exactly what Windows Vista, Server 2008 and later feature.
It's too bad such details won't make it into mainstream marketing. I'm not enough into Windows to read their developer news (I'm more of a *nix guy), where such information probably would show up. Such features you describe are certainly more interesting to me than animated wallpapers and translucent window frames.
Do you have any idea what you are talking about?
I certainly do, but you, on the other hand, do not seem to know what you are talking about.
Unused RAM is RAM that is not being used. By anything. You open up task manager and see 500mb of RAM used out of 3gb, well, 2.5gb is not being used. That is unused RAM
Certain operating system tools do not make a clear distinction between really unused RAM and RAM that is not used by applications. In the latter case, the so-called "unused" RAM is not unused at all, but rather used for the page cache.
Where the hell did you come up with this disk cache BS? The opposite happens - the OS pulls data out of RAM and caches it onto disk when the data has been loaded but is not expected to be used for a while, or other programs need the RAM. This is called a PageFile in windows, and a Swap partition in Linux.
The disk cache (or page cache) is certainly not BS. Both the page cache and swap file/partition are used, but for different purposes. The disk cache is used to cache frequently used disk blocks in the memory unused by applications, while the swap file/partition is used to temporarily free up physical memory when applications request memory but most or all is already used.
I don't think you have much programming background at all
My programming background is pretty extensive.
or you'd know that directly pointing to addresses hasn't been done in a long, long time. Some languages, like C and others still allow it, but 99.99% of the time it is a very bad idea. It's the kind of shit you do with assembly language programing.
Here you are mixing up two concepts: Virtual memory and managed code. Using physical addresses in applications hasn't been used in a long time, since they use virtual memory addresses instead. The translation is then done by the MMU (Memory Management Unit) together with the operating system, using the page tables that list which virtual pages are mapped into which physical pages (or swap areas). Even applications written in C or assembly language can not use physical addresses in any modern operating system, as they are completely managed by the kernel and MMU.
Managed code is a totally different matter. Languages such as Java, C#, and pretty much all scripting languages use managed code. Pointers are not used by application programmers in those languages, instead, programmers use another layer of indirection to access objects, such as references. These references are then mapped to (virtual) memory addresses by the virtual machine or interpreter that the applications run in.
The only thing that is -somewhat- like your imaginary "disk caching" is Vista's ReadyBoost, which moves frequently used data to a FLASH drive in order to improve responsiveness.
Disk caching as I explained it exists on all modern operating systems, including Windows, Linux and Mac OS X. The Vista ReadyBoost is an improvement on this, since it actually preloads the flash drive with frequently used application data, enabling quick use even the first time after system boot.
Now, what you might be talking about, in a confused illiterate monkey sort of way, is the process by which the OS moves data out of physical RAM and onto disk cache (the swap or pagefile).
No, that is not what I'm talking about.
In no way, shape, or form does this mean the OS takes commonly used data and loads it into RAM.
No, the OS doesn't do that by itself, but I also never said that. What happens is the following: Each time a block is read from the hard drive (as requested by an application), it is loaded into the disk cache. This cache is made up of all RAM that is not used by the OS or applications. When applications require more heap space, the disk cache size is automatically reduced, and data in the cache
If the whole RAM isn't used, you're just wasting it.
That's hardly the case. Unused RAM is used as a disk cache, so that frequently read disk blocks reside in RAM instead of on disk. This makes reading them extremely fast. If applications allocate memory willy-nilly just because it's there, there won't be any memory left for the disk cache, and your system might become very slow. And if even more memory is allocated, the system will start paging stuff in and out of memory, slowing stuff down even more.
Applications should use the amount of memory they need, preferably leaving caching to the operating system. Applications generally cannot know when other processes need the RAM better, and AFAIK cannot be told to release memory by other applications in any standardized way. Not to mention that they are not told when the disk cache becomes too small, or when paging starts to occur, since the operating system reasonably assumes that memory allocated is required by the application, not just held for the fun of it.
The doctor said that nicotine treatment "would offer significant advancement over current treatments".
There is a difference between nicotine treatment (your words) and nicotine-based treatment (the doctor's words). The latter can imply that you are not treated with nicotine itself, but with some derivative of nicotine, e.g. with certain properties removed or minimized.
Does nicotine alone have long-term health effects
AFAIK, nicotine is a vasoconstrictor, which means it constricts the blood vessels in the body. This puts increased strain on your heart, and may prevent the extremities (arms and legs) from getting enough blood. If you shook hands with a smoker, you may have noticed that they often have cold hands, which is due to this low blood supply.
So oppening it up to public bidding? What the point when there is only one vendor who provides the product you're looking to buy?
Where I live (Sweden), that's not legal. You have to specify what the product should do, not which product you want to buy. For example, a couple of years ago, a county opened a bidding process for home PCs for employees, and specified that to be accepted, the computers offered must be equipped with an Intel Pentium 4 processor. AMD felt excluded by this requirement, brought the county to court, and won.
Of course, there are ways around that, such as specifying a list of features that only one product can match. But still, even in that case, the bidding process allows you to choose the cheapest vendor that carries that product.
This is required by EU rules
Switzerland is not a member of the EU.