Because they don't have any successful tech companies of their own to boss around? Europe virtually ignored the tech industry for decades, and now that they've realized the ship has sailed and they're not on it, they're "getting back" at the successful companies by annoying them with pointless fines.
A little of both I guess. It sounds like you want to clone gcc so that the community can have a non-GPL-licensed compiler suite, but I don't understand why you think that is necessary.
and I really do not see how you can say they predate themselves!
Then you don't know the difference between OSS and Free Software.
RMS started the Free Software movement. Others (notably Eric Raymond) decided to start a competing movement called Open Source Software. These are not the same thing at all. Though technically a Free Software license like the GPL is also an Open Source license (OSS is a superset of Free Software), the purposes of the two movements are entirely different.
Personally as I see it the developer chooses a licence to avoid the hassle of working out what rights they hand out - but it is THEIR work and does not belong to whatever faction has started playing games in the FSF recently.
The copyrights to the gcc toolchain belong to the FSF -- they ARE the owners of the work! It has long been a condition to work on the official fork: if you want your patches to go everywhere, you assign copyright. Developers that don't like that are free to make their own forks (as with Emacs vs. XEmacs), but FSF has had enough developers who are OK with it to now have the definitive version of gcc.
And if you think GPLv3 is a recent "game" from a "faction" in the FSF, you haven't been paying attention for about 20 years. FSF has ALWAYS been about copyleft. They predate the OSS movement by a decade and Usenet is littered with the ashes of long flamewars about the GPL license.
Why am I starting to get the feeling that outside of the FSF no one is going to adopt v3?
It's far more than FSF, it is everyone who has seriously considered the GPL and found that it is appropriate for their needs. For me, I chose GPL because I want to retain the option to get paid for closed-source forks of my work. (No takers yet.:) )
All it takes is ONE GPL "V3 or later" userland program to be essential and the whole house of cards falls down. C++ is gaining new language features soon. Samba will need to be updated to the next release of Windows networking. Wine will need to support Vista features. I'm sure with a few minutes I could make this list run to several dozen critical apps whose developers chose (L)GPL and will move to V3.
The Two licenses are incompatible and will creat some expensive legal challenges.
The two licenses are incompatible, but that will not create problems for any projects except the Linux kernel. Most GPL code out there is licensed under "V2 or later", meaning with each additional version of the GPL released by the FSF the code can automatically be released under the new license.
When GPLv3 is ready, the FSF is going to fork all of the code they have copyright to and make the fork "V3 or later". Since "V3 or later" is included within the set "V2 or later", this is OK. However, people who want the "V2 or later" version are going to have to use a separate fork. Since the FSF has no plans to maintain both forks, this will force people wanting to do "V2 or later" to do their own work. Also, they cannot take "V3 or later" code and bring back to their "V2 or later" branch, since that would violate the copyright terms on the "V3 or later" fork.
The Linux kernel is a special case. Parts of the code are "V2 only" and parts are "V2 or later". This combination is OK. However, it would essentially be impossible to move the Linux kernel to "V3 only" or "V3 and later" because EVERY contributor with "V2 only" code in the kernel must agree to this change, and several of them have died and their estates own the copyright on their contributions.
I'm a Debian user. I want a fully-configurable, easy to install/use, desktop AND server oriented distro. In fact, I have Debian installed on 8 servers in a Beowulf-style configuration, 2 workstations, and 1 laptop.
No. It shouldn't. The potential damage that could occur with random Joe Q Public having access to the entire methodology behind the storage of people's most private data, without even the legal protection of an NDA is just... astronomical.
What kind of damage could Joe Q Public do if they had access to database schema and application code? Medical records are already protected (or so we think) by elaborate security measures, without the proper passwords just having the codebase poses no risk to the data itself. You sound like you are arguing from the same vein that OpenBSD is much more vulnerable because the source is available when history shows the opposite.
I think the poster of the root of this particular thread is just another of those anti-copyright zealots who think that every single thing developed should be public domain.
And I think you're the kind of idiot who over-generalizes someone's entire position from a single question. But then I've got evidence for that opinion.
I know very well the difference between public domain, BSD, GPL, and various closed-source licenses and the reasons people might choose each; I've got my own ~45,000 LOC project out there on Sourceforge licensed under GPL; I've written ~100,000 LOC over the last 5 years for both private employers and government entities; I'm reasonably good about avoiding both video/music and software piracy.
Yet I still think most government grant-funded software should be public domain ala BSD. My research area is currently at the mercy of a dozen or so proprietary programs developed at significant public expense. I want those programs we have all already collectively paid for to be available to the rest of us like they used to be, and I would even be thrilled if several private companies developed competing forks. As it is, we have to tread very carefully to use the programs our colleagues have developed over the last 30 years lest we incur their wrath and lose the ability to publish our results.
See the web site http://www.bannedbygaussian.org/ for an example of this behavior. Ironically enough, one of the people responsible for BannedByGaussian has their own program, NWChem, only available to PIs at http://www.emsl.pnl.gov/docs/nwchem/download.html . Two competing programs, one originally funded by government money and now closed source, and the other still government funded but only available to established scientists. Both used in publications, both costing hundreds of thousands of dollars of public money in development, yet neither available for everyone to use as they wish.
could you give us any insight into why you think it would be in the public domain?
The law regarding software and publicly-funded inventions has not always been as it is now. It used to be the case that most significant publicly-funded software HAD to be in the public domain, which AFAIK is why we have the BSD license today. Also witness early versions of Gaussian (quantum chemistry).
These days lots of 100% publicly-funded software is not automatically released to the public domain but instead held ransom by the author or university with a separate license permitting unlimited government use. This directly affects me: essentially ALL of the current quantum chemistry code that produces publishable results is no longer free for everyone to use. Though most programs come with source (the have to for some of the systems we need to run it on), their license restrictions are very onerous for developers: only the PI can register to download it, or it costs 5000 euros per seat, or it cannot be ported to other platforms, etc. One program even revokes licenses from academics who use competing software in the same domain! And this almost ALL software written by tenured professors and their graduate students funded from government grants.
I think we all did much better with the old formula. University-developed code should be available for everyone to use, even if that means someone can later come along and compete with a closed-source version.
I'm curious if the Swedish system more closely resembles the current USA system or the old USA system.
Ah yes, I've encountered that before. find | xargs is normally how I handle those situations -- find emits the filenames, and xargs can group a bunch of them together for a command.
Now, you could say, the open-sourced firmware was never proprietary to begin with somehow, but that's just semantics clearly, Linksys thought of it as proprietary and weren't planning to release the sources until the outside pressure made them do it.
ALL code licenses are "just semantics", because all laws are "just semantics". Many companies have taken F/OSS code and pretended that they had the right to make it their own and were stopped by the legal copyright owners from doing so. They were in fact making derivative works of someone else's work.
Now try to do what the GP actually said: find a company that first developed a piece of proprietary software, added a piece of F/OSS code to it, and was then forced to open up their original work as a result.
This was NOT an emulation environment, but a seperate Win16 subsystem that runs on the desktop side by side Win32.
Are you saying that Apple emulated the entire processor, memory, etc., of a computer ala VMWare and Win16 does not, or are you saying something else? What is the distinction between "emulation" and "subsystem" ?
and old Windows games automatically scale across them because Vista handles this, don't be surprised when the rest of the industry cries foul because other OSes don't have the ability to take advantage of the multicore GPUs without application being written specifically for them.
Why would Windows be able to provide multi-core GPU performance for old games presumably not written specifically to take advantage of it, but the other OSes would be unable to do the same for their apps once they had rewritten their graphics subsystems?
By definition, science MUST look to natural explanations of phenomena. If the supernatural can be measured and tested, then it is no longer supernatural.
Personally, I think that people should stop conflating the feeling of "sacred and/or sublime" with the word "supernatural". Scientific atheists have no trouble with feeling or honoring the sacred; it is religious people who assume that the entire domain of the Platonic Good belongs only to them.
As of 1994, in Texas it was legal for a personal phone call to be recorded if one person on the end was aware of the fact. So you could easily get a Radio Shack recorder and tape your own calls to your heart's content. However, it was a felony charge to record someone else's conversation without their knowledge.
Slashdot Mirror
Because they don't have any successful tech companies of their own to boss around? Europe virtually ignored the tech industry for decades, and now that they've realized the ship has sailed and they're not on it, they're "getting back" at the successful companies by annoying them with pointless fines.
Care to back any of that up?
A little of both I guess. It sounds like you want to clone gcc so that the community can have a non-GPL-licensed compiler suite, but I don't understand why you think that is necessary.
and I really do not see how you can say they predate themselves!
Then you don't know the difference between OSS and Free Software.
RMS started the Free Software movement. Others (notably Eric Raymond) decided to start a competing movement called Open Source Software. These are not the same thing at all. Though technically a Free Software license like the GPL is also an Open Source license (OSS is a superset of Free Software), the purposes of the two movements are entirely different.
We *need* an alternative to GCC.
Why?
Personally as I see it the developer chooses a licence to avoid the hassle of working out what rights they hand out - but it is THEIR work and does not belong to whatever faction has started playing games in the FSF recently.
The copyrights to the gcc toolchain belong to the FSF -- they ARE the owners of the work! It has long been a condition to work on the official fork: if you want your patches to go everywhere, you assign copyright. Developers that don't like that are free to make their own forks (as with Emacs vs. XEmacs), but FSF has had enough developers who are OK with it to now have the definitive version of gcc.
And if you think GPLv3 is a recent "game" from a "faction" in the FSF, you haven't been paying attention for about 20 years. FSF has ALWAYS been about copyleft. They predate the OSS movement by a decade and Usenet is littered with the ashes of long flamewars about the GPL license.
In Orwellian terms: Indemnification, good. Covenant not to sue, bad. Ridiculous.
Doesn't seem so ridiculous to me.
Indemnification: "If the bully tries to beat my friends up I will come to their defense."
Covenant: "If you become my friend the bully promises not to beat you up."
Why am I starting to get the feeling that outside of the FSF no one is going to adopt v3?
:) )
It's far more than FSF, it is everyone who has seriously considered the GPL and found that it is appropriate for their needs. For me, I chose GPL because I want to retain the option to get paid for closed-source forks of my work. (No takers yet.
All it takes is ONE GPL "V3 or later" userland program to be essential and the whole house of cards falls down. C++ is gaining new language features soon. Samba will need to be updated to the next release of Windows networking. Wine will need to support Vista features. I'm sure with a few minutes I could make this list run to several dozen critical apps whose developers chose (L)GPL and will move to V3.
The Two licenses are incompatible and will creat some expensive legal challenges.
The two licenses are incompatible, but that will not create problems for any projects except the Linux kernel. Most GPL code out there is licensed under "V2 or later", meaning with each additional version of the GPL released by the FSF the code can automatically be released under the new license.
When GPLv3 is ready, the FSF is going to fork all of the code they have copyright to and make the fork "V3 or later". Since "V3 or later" is included within the set "V2 or later", this is OK. However, people who want the "V2 or later" version are going to have to use a separate fork. Since the FSF has no plans to maintain both forks, this will force people wanting to do "V2 or later" to do their own work. Also, they cannot take "V3 or later" code and bring back to their "V2 or later" branch, since that would violate the copyright terms on the "V3 or later" fork.
The Linux kernel is a special case. Parts of the code are "V2 only" and parts are "V2 or later". This combination is OK. However, it would essentially be impossible to move the Linux kernel to "V3 only" or "V3 and later" because EVERY contributor with "V2 only" code in the kernel must agree to this change, and several of them have died and their estates own the copyright on their contributions.
Get it now?
Government doesn't have any business trying to run companies.
Since all businesses legally exist only by government charter, I'd say it does.
I'm a Debian user. I want a fully-configurable, easy to install/use, desktop AND server oriented distro. In fact, I have Debian installed on 8 servers in a Beowulf-style configuration, 2 workstations, and 1 laptop.
Do I not count?
No. It shouldn't. The potential damage that could occur with random Joe Q Public having access to the entire methodology behind the storage of people's most private data, without even the legal protection of an NDA is just... astronomical.
What kind of damage could Joe Q Public do if they had access to database schema and application code? Medical records are already protected (or so we think) by elaborate security measures, without the proper passwords just having the codebase poses no risk to the data itself. You sound like you are arguing from the same vein that OpenBSD is much more vulnerable because the source is available when history shows the opposite.
I think the poster of the root of this particular thread is just another of those anti-copyright zealots who think that every single thing developed should be public domain.
And I think you're the kind of idiot who over-generalizes someone's entire position from a single question. But then I've got evidence for that opinion.
I know very well the difference between public domain, BSD, GPL, and various closed-source licenses and the reasons people might choose each; I've got my own ~45,000 LOC project out there on Sourceforge licensed under GPL; I've written ~100,000 LOC over the last 5 years for both private employers and government entities; I'm reasonably good about avoiding both video/music and software piracy.
Yet I still think most government grant-funded software should be public domain ala BSD. My research area is currently at the mercy of a dozen or so proprietary programs developed at significant public expense. I want those programs we have all already collectively paid for to be available to the rest of us like they used to be, and I would even be thrilled if several private companies developed competing forks. As it is, we have to tread very carefully to use the programs our colleagues have developed over the last 30 years lest we incur their wrath and lose the ability to publish our results.
See the web site http://www.bannedbygaussian.org/ for an example of this behavior. Ironically enough, one of the people responsible for BannedByGaussian has their own program, NWChem, only available to PIs at http://www.emsl.pnl.gov/docs/nwchem/download.html . Two competing programs, one originally funded by government money and now closed source, and the other still government funded but only available to established scientists. Both used in publications, both costing hundreds of thousands of dollars of public money in development, yet neither available for everyone to use as they wish.
could you give us any insight into why you think it would be in the public domain?
The law regarding software and publicly-funded inventions has not always been as it is now. It used to be the case that most significant publicly-funded software HAD to be in the public domain, which AFAIK is why we have the BSD license today. Also witness early versions of Gaussian (quantum chemistry).
These days lots of 100% publicly-funded software is not automatically released to the public domain but instead held ransom by the author or university with a separate license permitting unlimited government use. This directly affects me: essentially ALL of the current quantum chemistry code that produces publishable results is no longer free for everyone to use. Though most programs come with source (the have to for some of the systems we need to run it on), their license restrictions are very onerous for developers: only the PI can register to download it, or it costs 5000 euros per seat, or it cannot be ported to other platforms, etc. One program even revokes licenses from academics who use competing software in the same domain! And this almost ALL software written by tenured professors and their graduate students funded from government grants.
I think we all did much better with the old formula. University-developed code should be available for everyone to use, even if that means someone can later come along and compete with a closed-source version.
I'm curious if the Swedish system more closely resembles the current USA system or the old USA system.
...why isn't it already in the public domain?
Ah yes, I've encountered that before. find | xargs is normally how I handle those situations -- find emits the filenames, and xargs can group a bunch of them together for a command.
Example: One of our file-transfer bash scripts used ls to gather the file list. This has the obvious file count issue.
Could you explain the "obvious file count issue"?
Yes, because the thread would top 500 posts of pissed off people who want to rant, increasing the ad views.
That's why bad news is more popular than good news.
Now, you could say, the open-sourced firmware was never proprietary to begin with somehow, but that's just semantics clearly, Linksys thought of it as proprietary and weren't planning to release the sources until the outside pressure made them do it.
ALL code licenses are "just semantics", because all laws are "just semantics". Many companies have taken F/OSS code and pretended that they had the right to make it their own and were stopped by the legal copyright owners from doing so. They were in fact making derivative works of someone else's work.
Now try to do what the GP actually said: find a company that first developed a piece of proprietary software, added a piece of F/OSS code to it, and was then forced to open up their original work as a result.
So what do YOU do when a single meeting spans two timezones?
This was NOT an emulation environment, but a seperate Win16 subsystem that runs on the desktop side by side Win32.
Are you saying that Apple emulated the entire processor, memory, etc., of a computer ala VMWare and Win16 does not, or are you saying something else? What is the distinction between "emulation" and "subsystem" ?
and old Windows games automatically scale across them because Vista handles this, don't be surprised when the rest of the industry cries foul because other OSes don't have the ability to take advantage of the multicore GPUs without application being written specifically for them.
Why would Windows be able to provide multi-core GPU performance for old games presumably not written specifically to take advantage of it, but the other OSes would be unable to do the same for their apps once they had rewritten their graphics subsystems?
By definition, science MUST look to natural explanations of phenomena. If the supernatural can be measured and tested, then it is no longer supernatural.
Personally, I think that people should stop conflating the feeling of "sacred and/or sublime" with the word "supernatural". Scientific atheists have no trouble with feeling or honoring the sacred; it is religious people who assume that the entire domain of the Platonic Good belongs only to them.
Not a bad start, but I think that last point could be clarified to explicitly include the social sciences.
Thanks so much for providing our daily dose of "Poor People Suck" to offset all the empathy around here lately.
As of 1994, in Texas it was legal for a personal phone call to be recorded if one person on the end was aware of the fact. So you could easily get a Radio Shack recorder and tape your own calls to your heart's content. However, it was a felony charge to record someone else's conversation without their knowledge.
its got some inherent security problems (like not being able to set a dir to forbid execution without making it a NOEXEC partition)
/path/to/executable" ?
Can't you always just do "/bin/ld.so