I greatly admire Canada. I frequently vacation there and have been to BC, Alberta, Ontario, and Quebec a number of times. I wish that the US were more like it. It is a much more humane country than the US and more pleasant to live in in so many ways.
However, it is completely bogus for a Canadian to attack the US on environmental issues, as this is one area where Canada lags the US. Canada has no equivalent to the US Endangered Species Act; Canada burns way too much coal (Ontario coal-burning is a major contributor to acid rain in the Northeast US and the Maritimes); Canada is destroying its (temperate) rain forest at a rate comparable to Brazil's, and British Columbia has already enacted policies that the Bush anti-enviro forces can only dream about (ending the establishment of any new wilderness areas or any limits on exploiting lands that haven't yet been exploited). The focus on resource extraction is natural for a country with so much, but continuation of such policies will ruin much of what makes Canada such a great place. It will just take longer because of the lower population density.
Patents don't make a bit of difference. If you want to claim that GPL'ed code violates a patent, you can sue. Or you can grant the public the right to use the patent in GPLed code (as IBM has done, in the case of some register allocation patents they have, so that GCC can use them). But you can't have it both ways. If you distribute GPLed code, you must grant the world a right to redistribute the code under the GPL. Otherwise, the GPL text itself says that you forfeit your rights under the GPL. This forfeit would mean that SCO can no longer distribute the Linux kernel.
Actually, Debian does not avoid this problem. Because even Debian unstable tends to lag the rest of the world, it's often quite difficult to try out a very new Gnome or KDE app, if that app depends on newer versions of shared libraries than the ones packaged in Debian.
Hundreds of people have slipped in copyrighted code to the Linux kernel. The whole thing is copyrighted. But anyway...
I have on several occasions seen junior programmers at proprietary software companies take code off the net or out of a textbook, and try to build it into a proprietary product (e.g. they need to do an FFT, so they get one out of "Numerical Algorithms in C"). This is because they have no clue about the difference between "the source code is published" and "it's public domain". Such codes get ripped out when found, or licenses are negotiated when it's too hard (at least at ethical companies),but it does happen.
The point is, it does happen that proprietary code contains illegally copied software, so SCO had better be careful what they allege.
They have said that they found common code. One possibility is that someone illegally copied SCO-owned code to Linux. Another possibility is that both kernels got the code from the same source: from BSD, for example, or by implementing an algorithm out of a textbook. A third possibility, one mentioned by Bruce Perens, is that SCO stole the code from Linux.
Wrong: the National Museum drew scholars from all around the world, and in a free society, would be a major tourist attraction. All that money coming in feeds people.
Studies have shown, for example, that New York's art museums contribute far more to New York's economy than all its sports teams combined.
What you're missing is that much of what is valuable in Unix was produced by the contributors to BSD, at Berkeley and elsewhere, and this extra code was always licensed under the loosest terms that could be achieved given that, to be useful, a Unix source license was needed. This work was done collaboratively by a community. The community was "gated", true, but in the early days any university CS department could get the Unix source license for a song. So, in that sense, the spirit of the free software / open source community is as old as Unix.
Certainly Microsoft has introduced some innovations, but their real expertise is in borrowing or acquiring ideas from others, and then (after a number of releases where the customers shake out the bugs) coming up with a very clean implementation. And they keep on making incremental improvements when others rest on their laurels: the Apple fans correctly said "Windows 95 = Apple 84", but what they forgot to say was "Apple 95 = Apple 84" (Apple got off their butts since then and have made real progress).
The NT kernel is a clone of the DEC VAX/VMS operating system, perpetuating many ideas that first appeared in DEC's RSX-11/M OS. This is no accident; Microsoft hired Dave Cutler, the main architect of VMS, to do NT for them. VMS's virtual memory system predates the first virtual memory Unix (which was done at Berkeley, not by Bell Labs). DOS and CP/M can also trace their roots back to DEC operating systems (particularly to RT-11).
So Microsoft's stuff is quite old as well. They did have good taste as to what they copied, but because they copied from so many sources, they have a ton of cruft.
I am a developer who makes a living from my work. I use the GNU tools to develop proprietary software. I've also contributed to some open source/free software. My company has greatly benefited by having these tools available for free. How have I been discriminated against? Because I can't add an enhancement to GCC and sell it?
True, Stallman would like to go beyond the GPL and have only free software in the world. But the GPL itself does not achieve this.
People who need more than 4Gb on one machine usually want to be able to address it all in one process (in my field, electronic design automation, this is becoming increasingly common for laying out or analyzing a big chip), so they'll want 64 bits. I think Opteron's going to be the winner here if AMD can reliably deliver.
If the previous owners of Unix couldn't win against BSD, which was written by people who had full access to the Bell Labs source code and gradually replaced it bit by bit, what gives you the idea that they could possibly win against Linux, which was developed completely independently?
The market thinks the case is bogus. They are asking for $1 billion. Their current market capitalization is $37 million. If, as an analyst, I gave SCO a 5% chance of prevailing, that would be $50 million right there, so we can conclude that stock analysts don't even give SCO a 5% chance of collecting $1 billion from IBM (OK, to be fair, deduct expenses and payments to lawyers, and it's not even a 10% chance).
No, they are not alleging that "some IBM guys copied source code from System V to Linux": you will not find such words anywhere in the complaint. They are apparently claiming trade secret violations
SCO doesn't have a very strong tool: the previous owners of Unix already sued the free software world (specifically the University of California, over BSD) and lost (or rather, accepted a settlement that amounted to defeat).
I think that SCO wants IBM to buy them just to end the nuisance suit. That's probably what they are going for.
IBM has no power to "put it there", they don't control Linux. Also, any such effort would have been caught already; there are vast numbers of people out there who have been Unix source code licensees; someone would have noticed.
Second, SCO continues to distribute Linux, even after making these allegations. That would appear, under the GPL, to kill their claims: continuing to distribute means that everyone can use the code under GPL terms. Trying to add restrictions to GPLed code violates GPL terms, resulting in the loss of the license to copy and modify. So SCO can sue Linux or distribute Linux, but not both.
I understand that the way around that one is to use large pages, to decrease the overhead for each page.
Pages can be of variable sizes.
I'm not a kernel expert, but there was a
talk on this topic at the last Ottawa Linux Symposium.
I'll tell you what the US did: it had a copyright and patent regime that refused to recognize foreign copyrights and patents, and proceeded to "steal" British "intellectual property" to create an American industry. It passed high tariffs, to keep out foreign competition. In other words, it did the very things that the US (through its organs, the IMF, World Bank, and WTO) forbids developing countries to do.
No country ever got rich by opening itself for exploitation by foreign companies and by practicing "free trade".
I had actually used the CERN line-mode www interface before Mosaic came out, just to check out the ravings of this pompous Brit I heard about
(by the name of Tim Berners-Lee) who was raving that this thing called the World Wide Web was intended to contain the sum of all human knowlege. But Mosaic was a huge leap forward.
However, when Mosaic first came out, a lot of folks in my department were using it as a better interface to Gopher, since in 1993 there was far more interesting stuff available via Gopher than via HTTP. Of course that didn't last long.
Brett, we all know that you don't like the GPL, but your old troll is irrelevant to this argument. The Debian folks are completely happy with the GPL; they object to the GFDL.
The GPL explicitly meets the Debian definition of "free", the Debian Free Software Guidelines say so. Go read it.
When you assign code to the FSF, you sign a contract. The contract grants back to the contributor the right to continue to use and distribute the code, under whatever terms desired. So, if Walter Bright contributed code to the FSF, he could continue to use it in his own, proprietary software.
Because of the contracts signed with contributors, the FSF is required to maintain the software as free, forever, and the lawyers have legal standing to defend the copyright. This is a good thing.
North American problem only
on
Dial-A-Cam
·
· Score: 3, Interesting
The need for new area codes only affects countries like the US and Canada, that use fixed-length telephone numbers. Europe uses a variable-length approach that lets frequently dialed numbers be short, while numbers for devices can be as long as you want. You can hand a prefix to an organization and they can populate it with phone numbers however they want, much like DNS. You never have to change phone numbers.
Slashdot Mirror
I greatly admire Canada. I frequently vacation there and have been to BC, Alberta, Ontario, and Quebec a number of times. I wish that the US were more like it. It is a much more humane country than the US and more pleasant to live in in so many ways.
However, it is completely bogus for a Canadian to attack the US on environmental issues, as this is one area where Canada lags the US. Canada has no equivalent to the US Endangered Species Act; Canada burns way too much coal (Ontario coal-burning is a major contributor to acid rain in the Northeast US and the Maritimes); Canada is destroying its (temperate) rain forest at a rate comparable to Brazil's, and British Columbia has already enacted policies that the Bush anti-enviro forces can only dream about (ending the establishment of any new wilderness areas or any limits on exploiting lands that haven't yet been exploited). The focus on resource extraction is natural for a country with so much, but continuation of such policies will ruin much of what makes Canada such a great place. It will just take longer because of the lower population density.
Patents don't make a bit of difference. If you want to claim that GPL'ed code violates a patent, you can sue. Or you can grant the public the right to use the patent in GPLed code (as IBM has done, in the case of some register allocation patents they have, so that GCC can use them). But you can't have it both ways. If you distribute GPLed code, you must grant the world a right to redistribute the code under the GPL. Otherwise, the GPL text itself says that you forfeit your rights under the GPL. This forfeit would mean that SCO can no longer distribute the Linux kernel.
Actually, Debian does not avoid this problem. Because even Debian unstable tends to lag the rest of the world, it's often quite difficult to try out a very new Gnome or KDE app, if that app depends on newer versions of shared libraries than the ones packaged in Debian.
The GPL places absolutely no restrictions on a user: a user is someone who runs the program and gets a result.
The GPL only restricts someone who wishes to modify or distribute the program, which is not something that a user normally does.
Hundreds of people have slipped in copyrighted code to the Linux kernel. The whole thing is copyrighted. But anyway ...
I have on several occasions seen junior programmers at proprietary software companies take code off the net or out of a textbook, and try to build it into a proprietary product (e.g. they need to do an FFT, so they get one out of "Numerical Algorithms in C"). This is because they have no clue about the difference between "the source code is published" and "it's public domain". Such codes get ripped out when found, or licenses are negotiated when it's too hard (at least at ethical companies),but it does happen.
The point is, it does happen that proprietary code contains illegally copied software, so SCO had better be careful what they allege.
They have said that they found common code. One possibility is that someone illegally copied SCO-owned code to Linux. Another possibility is that both kernels got the code from the same source: from BSD, for example, or by implementing an algorithm out of a textbook. A third possibility, one mentioned by Bruce Perens, is that SCO stole the code from Linux.
Wrong: the National Museum drew scholars from all around the world, and in a free society, would be a major tourist attraction. All that money coming in feeds people.
Studies have shown, for example, that New York's art museums contribute far more to New York's economy than all its sports teams combined.
What you're missing is that much of what is valuable in Unix was produced by the contributors to BSD, at Berkeley and elsewhere, and this extra code was always licensed under the loosest terms that could be achieved given that, to be useful, a Unix source license was needed. This work was done collaboratively by a community. The community was "gated", true, but in the early days any university CS department could get the Unix source license for a song. So, in that sense, the spirit of the free software / open source community is as old as Unix.
Certainly Microsoft has introduced some innovations, but their real expertise is in borrowing or acquiring ideas from others, and then (after a number of releases where the customers shake out the bugs) coming up with a very clean implementation. And they keep on making incremental improvements when others rest on their laurels: the Apple fans correctly said "Windows 95 = Apple 84", but what they forgot to say was "Apple 95 = Apple 84" (Apple got off their butts since then and have made real progress).
The NT kernel is a clone of the DEC VAX/VMS operating system, perpetuating many ideas that first appeared in DEC's RSX-11/M OS. This is no accident; Microsoft hired Dave Cutler, the main architect of VMS, to do NT for them. VMS's virtual memory system predates the first virtual memory Unix (which was done at Berkeley, not by Bell Labs). DOS and CP/M can also trace their roots back to DEC operating systems (particularly to RT-11).
So Microsoft's stuff is quite old as well. They did have good taste as to what they copied, but because they copied from so many sources, they have a ton of cruft.
Read the DFSG, section 10. I quote:
You cannot argue with this verbatim quote.
I am a developer who makes a living from my work. I use the GNU tools to develop proprietary software. I've also contributed to some open source/free software. My company has greatly benefited by having these tools available for free. How have I been discriminated against? Because I can't add an enhancement to GCC and sell it?
True, Stallman would like to go beyond the GPL and have only free software in the world. But the GPL itself does not achieve this.
People who need more than 4Gb on one machine usually want to be able to address it all in one process (in my field, electronic design automation, this is becoming increasingly common for laying out or analyzing a big chip), so they'll want 64 bits. I think Opteron's going to be the winner here if AMD can reliably deliver.
The freeness of a society is inversely proportional to the number of things you can write about that will get you in trouble.
If the previous owners of Unix couldn't win against BSD, which was written by people who had full access to the Bell Labs source code and gradually replaced it bit by bit, what gives you the idea that they could possibly win against Linux, which was developed completely independently?
The market thinks the case is bogus. They are asking for $1 billion. Their current market capitalization is $37 million. If, as an analyst, I gave SCO a 5% chance of prevailing, that would be $50 million right there, so we can conclude that stock analysts don't even give SCO a 5% chance of collecting $1 billion from IBM (OK, to be fair, deduct expenses and payments to lawyers, and it's not even a 10% chance).
No, they are not alleging that "some IBM guys copied source code from System V to Linux": you will not find such words anywhere in the complaint. They are apparently claiming trade secret violations
SCO doesn't have a very strong tool: the previous owners of Unix already sued the free software world (specifically the University of California, over BSD) and lost (or rather, accepted a settlement that amounted to defeat).
I think that SCO wants IBM to buy them just to end the nuisance suit. That's probably what they are going for.
IBM has no power to "put it there", they don't control Linux. Also, any such effort would have been caught already; there are vast numbers of people out there who have been Unix source code licensees; someone would have noticed.
Second, SCO continues to distribute Linux, even after making these allegations. That would appear, under the GPL, to kill their claims: continuing to distribute means that everyone can use the code under GPL terms. Trying to add restrictions to GPLed code violates GPL terms, resulting in the loss of the license to copy and modify. So SCO can sue Linux or distribute Linux, but not both.
OK, you're hired!
I understand that the way around that one is to use large pages, to decrease the overhead for each page. Pages can be of variable sizes. I'm not a kernel expert, but there was a talk on this topic at the last Ottawa Linux Symposium.
I'll tell you what the US did: it had a copyright and patent regime that refused to recognize foreign copyrights and patents, and proceeded to "steal" British "intellectual property" to create an American industry. It passed high tariffs, to keep out foreign competition. In other words, it did the very things that the US (through its organs, the IMF, World Bank, and WTO) forbids developing countries to do.
No country ever got rich by opening itself for exploitation by foreign companies and by practicing "free trade".
I had actually used the CERN line-mode www interface before Mosaic came out, just to check out the ravings of this pompous Brit I heard about (by the name of Tim Berners-Lee) who was raving that this thing called the World Wide Web was intended to contain the sum of all human knowlege. But Mosaic was a huge leap forward.
However, when Mosaic first came out, a lot of folks in my department were using it as a better interface to Gopher, since in 1993 there was far more interesting stuff available via Gopher than via HTTP. Of course that didn't last long.
Brett, we all know that you don't like the GPL, but your old troll is irrelevant to this argument. The Debian folks are completely happy with the GPL; they object to the GFDL.
The GPL explicitly meets the Debian definition of "free", the Debian Free Software Guidelines say so. Go read it.
When you assign code to the FSF, you sign a contract. The contract grants back to the contributor the right to continue to use and distribute the code, under whatever terms desired. So, if Walter Bright contributed code to the FSF, he could continue to use it in his own, proprietary software.
Because of the contracts signed with contributors, the FSF is required to maintain the software as free, forever, and the lawyers have legal standing to defend the copyright. This is a good thing.
The need for new area codes only affects countries like the US and Canada, that use fixed-length telephone numbers. Europe uses a variable-length approach that lets frequently dialed numbers be short, while numbers for devices can be as long as you want. You can hand a prefix to an organization and they can populate it with phone numbers however they want, much like DNS. You never have to change phone numbers.
Surely you know that the EU is in the process of passing DMCA-like laws.