Not to mention that a nuclear rocket can be fueled by just about any gas.
Could you explain this? I would expect a nuclear rocket to be fueled by radioactive materials. (I presume you are not talking about deuterium fusion.) Are you referring to reaction mass? I would imagine a wide variety of gasses would server adequately as reaction mass.
It has been proposed that a lander or a robotic factory could use solar energy to crack water to use as fuel for more conventional rocket motors. This depends on finding a reliable source of large amounts of water, which appears not to be a trivial task on Mars.
Has a nuclear powered rocket ever started from solid ground and achieved Martian escape velocity? If not, development of this technology could be expensive.
Only one thing got me. On the final review-your-vote screen, it took a minute to find the scrollbar because it was on the left side of the screen, while I expect to see them on the right (except for xterm and emacs). I think someone with limited or no experience with computer interfaces would have had trouble reviewing his vote. These things may be easy to learn, but they are not intuitive.
A receipt is not enough: It has to be collected and retained like an old fashioned paper ballot. In which case, electronic voting offers nothing.
Not true. The receipts can be collected but ignored unless a recount is necessary. Also, votes recorded on a printed receipt will never be ambiguous; any ballot on which the voter marks her vote could have partial, smeared, or multiple marks, but a piece of paper with "Vote for President: George Washington" printed on it cannot be misunderstood.
"Speaking the X protocol is key to interoperability from Unix to Unix."
How about "key to interoperability between X client and X server". Remember that X was implemented on VMS as well as on Unix, not to mention the version in X terminals and various emulators for MSWindows and Mac.
Finally, an irrefutable argument for using Perl instead of shell scripts.
Almost.
But you could equally well conclude that the solution is to install GNU bash on all these machines. Scripts will be just as portable between machines running GNU bash as they would be in Perl.
"But then I wouldn't be able to hear the clicky sound of the keys when I'm typing."
My XT clone (I think that's what it was) did in fact make little clicky sounds with the speaker when keys were pressed. (It was possible to turn it off.) It was a useful sort of feedback for people used to loud, clunky keyboards or typewriters.
"wherever the offices are in the world these things will be provided by these companies or such like."
You caught TF's glaring mistake. A Compaq computer bought for a tech-support desk in India does no more for the American economy than the same computer bought for a tech-support desk in the US.
The fallacy that whatever makes more money for a US company must be good for US workers and consumers is subtle in comparison. (It may be true in many or most cases, but it is not a priori true.)
By googling I found a lot of discussions about whether the "must retain... this list of conditions" clause of the BSDL conflicts with the GPL, including a few on lkml. Nobody seems to have any answer except "It doesn't conflict because the FSF says it doesn't," (not an actual quote) which isn't any answer at all.
In fact, I came across several licenses which were almost exactly new-BSDL, but with the additional clause
Alternatively, this product may be distributed under the terms of the GNU General Public License (GPL), in which case the provisions of the GNU GPL are required instead of the above restrictions. (This clause is necessary due to a potential bad interaction between the GNU GPL and the restrictions contained in a BSD-style copyright.)
(a direct quote). This means some people think that some BSDL-like requirements might conflict with the GPL.
(Note that the lkml thread you pointed to in the other message is about an old-style BSDL with advertising clause.)
I think it may be the clause in the GPL that says "publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty" that people use to get around the compatibility issue. If the BSDL list of conditions is considered part of the copyright/warranty information under the GPL, then the "must retain... this list of conditions" is not inconsistent; in fact retaining the list would be required under the GPL. That is a pretty twisted way of reading it, though. I would hate to see someone have to argue in court what counts as an additional restriction under the GPL and what does not.
It still leaves the problem of what to do with a file that mixes GPL and BSDL code. I guess you need to say "Portions of this code are copyright 1991 The Regents of the University of California, and available under the following restrictions... Portions of the code are copyright whoever released under the GPL." I don't think anyone ever does that, though.
"If you seriously think that somehow what I said (that you can't change the license on the code, only use that code as allowed by the license) contradicts that, you're going to have to explain yourself, because I certainly don't see how that follows."
You said "If I put in a notice that the entire thing is now GPL, I'm simply wrong". I asked if this meant illegal, to which you replied, "I don't know, you'd have to ask a lawyer." So, possibly yes.
But the GPL says that you can't require anything other than what is required by the GPL. If a mix of GPL and BSDL code needs a notice saying that part of the code is available under the BSDL, then this extra requirement means it cannot be released under the GPL.
Also, the GPL states that derivative works can only be distributed under the GPL. Say Alice writes program A and releases it under the BSDL, and Bob takes part of A and puts it in his program B released under the GPL (copyright Alice and Bob). If Carol receives B under the GPL, she can only distribute a derivative work C under the GPL. That is what the GPL says. It doesn't say "unless that code was released under a different license by someone else." She has to get the code from that someone else under that license.
Am I reading the GPL too literally? Isn't that the way legal documents are meant to be taken?
This is a serious claim. Can you substantiate it? To the best of my knowledge, all the quotes I have attributed to you have been done by a direct cut-and-paste from your posts. If I included a specific quote, in quotation marks, and attributed to you incorrectly then I certainly owe you a apology.
I believe that I understand your point of view. I am simply trying to use a concrete example and the GPL to show that your point of view, if correct, has serious consequences regarding the mixing of BSDL and GPL code. If you won't address the concrete example I proposed or comment on the portion of the GPL that I quoted, no communication is possible.
I am amazed at how you dance around trying to avoid the direct quote of your previous message. It's like you want to take it back, but can't admit it was a mistake. Quite fascinating. Briefly.
"And for the second time I'm going to have to correct you for putting words in my mouth. I did not say that, or anything vaguely resembling it, you're saying it, and you're wrong."
You said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong." I am trying to understand what you meant when you said this. Do you think it is legal to distribute code with incorrect licensing information? I would think no. If a program has a single licensing notice that says it is GPLed, even though part of it came from a BSDL program, is the licensing notice incorrect? You said it is. Has the Linux kernel ever contained code taken from a BSDL source? I think it has. Has the documentation with the Linux kernel always claimed the whole thing was GPLed? I think so.
When you said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong," did you mean something equivalent to, "If I put a notice that no piece of the thing is available to the public except under the GPL, I'm simply wrong"? That is clearly true, but is not the most obvious way to interpret your statement.
"No I'm not. It's perfectly legal for them to contain that code - it's allowed by the license the copyright holders chose in this case. But it's not under the GPL - it's still under the BSDL."
You said that program containing some bits originally licensed under the BSDL cannot claim to be entirely GPLed. The kernel has at some point included some BSD code but has always claimed to be released entirely under the GPL. According to you, distributing these kernels was illegal because they claimed that some BSDL code was GPL code.
You said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong." Clearly it is not legal to distribute code with inaccurate licensing information.
"It's either available under the GPL, or a GPL-compatible license."
The GPL says, "Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions." Every work that is a derivative work of a GPLed program is itself under the GPL. If I modify a BSDL program and release under the GPL, anything you take from it and distribute has to be released under the GPL. If you want to release it under the BSDL, you have to take it from source which is BSDL (or compatible).
You may feel that this makes the GPL an illegal license. But that is what it says.
If I put in a notice that the entire thing is now GPL, I'm simply wrong
You are saying that Linux kernels that contained some BSD code were distributed illegally, because they were licensed only under the GPL. Under your philosophy, work licensed in any fashion other than the GPL can never be incorporated into GPLed code, because extra licensing information would be required, but extra requirements are not allowed under the GPL.
A license is permission to do something. Saying code is under the GPL means that people have permission to do certain things with it that they otherwise would not have under copyright law. If I release a program under the GPL, that says you have all the permissions listed in the GPL. It does not say that those are the only permissions you have. It does not say that no portions of the code are available under other licenses. The fact that some piece of the program is also available from someone else under a BSD license does not change the fact that I am distributing it under the GPL.
All the code in a GPLed program is, by definition, under the GPL. This has nothing to do with the fact that the Regents of the University of California may hold the copyright to some bit of it, and they allow you to do lots more stuff with their part than the GPL says.
"The Apache License can take away your right to use the software"
You are right that there could be a narrow region of incompatibility. Say that there is a GPLed code that violates my patent, i.e. it uses a patented technique without permission. I can legally run that code, but no one else can without my permission. However, there is no legal way for anyone to copy, modify, or distribute the code.
If the code were instead under the Apache license, I could still allow third parties to use my patented techniques (for a fee), but I could not myself run Apache, but only if Apache contained material under someone else's patent.
This seems like such a trivial case that it is not worth worrying about. If am suing everyone in the world who has a server running Apache, whether or not I can run Apache myself is not too important. They should be able to resolve this.
"Why do all these licenses need to exist if they're compatible with the GPL? Again, I don't expect an answer."
Too bad:-) "Compatible with the GPL" means that the license is less restrictive than the GPL and contains no incompatible provisions, so that derivative works can be licensed under the GPL. Many people want to release their code under a BSD license to allow different types of usage, one of which is inclusion in GPLed programs.
Some of the incompatible ones are really weird. Look at xinetd. Because of the wacko license, the current maintainer needed to get explicit permission from the original author to increment the version number! It also has "The author of the modifications... will be responsible for any wrong behavior of the modified software." I am sure the original author meant to say, "If someone else mucks with the code, don't come crying to me when it breaks." But this sounds a lot like saying authors of modifications will be financially and legally liable for any bugs in their code. This is not something we want even a hint of ambiguity about.
"If this is handled correctly, we'd ideally end up with something like 4 licenses"
Very good idea. If there are a few well-written licenses designed explicitly to cover the spectrum of what people want, then there will be much less of a tendency for people to roll their own.
"by what right may I take a piece of BSD software, and relicence it, given I am not the owner of the copyright? What legal right do I have to do that?
That would be the BSD license.
That's kinda the whole point of the thing. You can't reassign the copyright, but that is not the same thing.
This is the whole difference between the GPL and the BSDL. BSDL allows redistribution under a more restrictive license, and the GPL only allows redistribution under the GPL. Saying "you would equally be able to take Microsoft Word and relicence it as BSD" is nonsense because I have no license to distribute MSWord at all.
Read the BSD license. "Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:" It never says I can't distribute it with different conditions as long as my conditions include those in the original license.
"Because what the Apache license says isnt what is at issue. The Apache license could claim GPL compatiblity until it turns blue, but as long as the GPL isnt compatible with it you cant distribute GPL code together with, and linked to it.
I think the point the parent was making is that the license should explicitly say that it is compatible. He meant add a clause to the Apache license that says something like this: "Notwithstanding any other provision in this license, the SOFTWARE or any derivative work may be modified and distributed under the GNU General Public License (version x.y.z)."
Thanks for pointing out Moglen's response. The license was changed because of his concerns. For example, under the draft license, you would be unable to distribute Apache if you made any patent allegations against anyone who has contributed to Apache. (Under this license, if Caldera had ever given one line of code to Apache, IBM would now be prohibited from distributing it because of their counterclaims against SCOG.) It was rewritten to only apply to patent claims against the Apache code itself. This is a huge difference.
Legal questions are always hard, but it certainly sounds like the Apache people are trying hard to be GPL-compatible, and have reasonable arguments that they are.
"...released source code is horrible for security. Look at OpenBSD..."
Just maybe there is a difference between an open development process, like OpenBSD, where incremental changes are examined before becoming part of the production code and dumping on the web hundreds of meg of source of a finished product which has an installed base of millions. Open source OS's get security from having many people looking at code submissions and the opportunity to find and fix dangerous bugs before they are exploited. Making a bunch of Windows source code available on the net does neither of these things.
"Seriously though, I really think the new Firefox name is very clear, strong, easy-to-say, unique and the logo is really cool as well as red pandas being sooo cute."
Slashdot Mirror
Could you explain this? I would expect a nuclear rocket to be fueled by radioactive materials. (I presume you are not talking about deuterium fusion.) Are you referring to reaction mass? I would imagine a wide variety of gasses would server adequately as reaction mass.
It has been proposed that a lander or a robotic factory could use solar energy to crack water to use as fuel for more conventional rocket motors. This depends on finding a reliable source of large amounts of water, which appears not to be a trivial task on Mars.
Has a nuclear powered rocket ever started from solid ground and achieved Martian escape velocity? If not, development of this technology could be expensive.
Your argument would make sense if a navy aircraft carrier could fly.
Only one thing got me. On the final review-your-vote screen, it took a minute to find the scrollbar because it was on the left side of the screen, while I expect to see them on the right (except for xterm and emacs). I think someone with limited or no experience with computer interfaces would have had trouble reviewing his vote. These things may be easy to learn, but they are not intuitive.
Not true. The receipts can be collected but ignored unless a recount is necessary. Also, votes recorded on a printed receipt will never be ambiguous; any ballot on which the voter marks her vote could have partial, smeared, or multiple marks, but a piece of paper with "Vote for President: George Washington" printed on it cannot be misunderstood.
How about "key to interoperability between X client and X server". Remember that X was implemented on VMS as well as on Unix, not to mention the version in X terminals and various emulators for MSWindows and Mac.
Maybe someone would like to put out a version of firefox using a logo which is a picture of a firefox, not of a fox.
Just a thought.
Almost.
But you could equally well conclude that the solution is to install GNU bash on all these machines. Scripts will be just as portable between machines running GNU bash as they would be in Perl.
My XT clone (I think that's what it was) did in fact make little clicky sounds with the speaker when keys were pressed. (It was possible to turn it off.) It was a useful sort of feedback for people used to loud, clunky keyboards or typewriters.
For the last time, Linux is not a derived work of BSD or any other "Unix". You SCOG astroturfers make me ill.
You caught TF's glaring mistake. A Compaq computer bought for a tech-support desk in India does no more for the American economy than the same computer bought for a tech-support desk in the US.
The fallacy that whatever makes more money for a US company must be good for US workers and consumers is subtle in comparison. (It may be true in many or most cases, but it is not a priori true.)
In fact, I came across several licenses which were almost exactly new-BSDL, but with the additional clause
(a direct quote). This means some people think that some BSDL-like requirements might conflict with the GPL.(Note that the lkml thread you pointed to in the other message is about an old-style BSDL with advertising clause.)
I think it may be the clause in the GPL that says "publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty" that people use to get around the compatibility issue. If the BSDL list of conditions is considered part of the copyright/warranty information under the GPL, then the "must retain... this list of conditions" is not inconsistent; in fact retaining the list would be required under the GPL. That is a pretty twisted way of reading it, though. I would hate to see someone have to argue in court what counts as an additional restriction under the GPL and what does not.
It still leaves the problem of what to do with a file that mixes GPL and BSDL code. I guess you need to say "Portions of this code are copyright 1991 The Regents of the University of California, and available under the following restrictions... Portions of the code are copyright whoever released under the GPL." I don't think anyone ever does that, though.
You said "If I put in a notice that the entire thing is now GPL, I'm simply wrong". I asked if this meant illegal, to which you replied, "I don't know, you'd have to ask a lawyer." So, possibly yes.
But the GPL says that you can't require anything other than what is required by the GPL. If a mix of GPL and BSDL code needs a notice saying that part of the code is available under the BSDL, then this extra requirement means it cannot be released under the GPL.
Also, the GPL states that derivative works can only be distributed under the GPL. Say Alice writes program A and releases it under the BSDL, and Bob takes part of A and puts it in his program B released under the GPL (copyright Alice and Bob). If Carol receives B under the GPL, she can only distribute a derivative work C under the GPL. That is what the GPL says. It doesn't say "unless that code was released under a different license by someone else." She has to get the code from that someone else under that license.
Am I reading the GPL too literally? Isn't that the way legal documents are meant to be taken?
This is a serious claim. Can you substantiate it? To the best of my knowledge, all the quotes I have attributed to you have been done by a direct cut-and-paste from your posts. If I included a specific quote, in quotation marks, and attributed to you incorrectly then I certainly owe you a apology.
I believe that I understand your point of view. I am simply trying to use a concrete example and the GPL to show that your point of view, if correct, has serious consequences regarding the mixing of BSDL and GPL code. If you won't address the concrete example I proposed or comment on the portion of the GPL that I quoted, no communication is possible.
I am amazed at how you dance around trying to avoid the direct quote of your previous message. It's like you want to take it back, but can't admit it was a mistake. Quite fascinating. Briefly.
You said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong." I am trying to understand what you meant when you said this. Do you think it is legal to distribute code with incorrect licensing information? I would think no. If a program has a single licensing notice that says it is GPLed, even though part of it came from a BSDL program, is the licensing notice incorrect? You said it is. Has the Linux kernel ever contained code taken from a BSDL source? I think it has. Has the documentation with the Linux kernel always claimed the whole thing was GPLed? I think so.
When you said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong," did you mean something equivalent to, "If I put a notice that no piece of the thing is available to the public except under the GPL, I'm simply wrong"? That is clearly true, but is not the most obvious way to interpret your statement.
Could you explain the distinction?
Is it that "sublicense" means acting within the provisions of the original license, and "relicense" means replacing the original license?
You said that program containing some bits originally licensed under the BSDL cannot claim to be entirely GPLed. The kernel has at some point included some BSD code but has always claimed to be released entirely under the GPL. According to you, distributing these kernels was illegal because they claimed that some BSDL code was GPL code. You said, "If I put in a notice that the entire thing is now GPL, I'm simply wrong." Clearly it is not legal to distribute code with inaccurate licensing information.
"It's either available under the GPL, or a GPL-compatible license."
The GPL says, "Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions." Every work that is a derivative work of a GPLed program is itself under the GPL. If I modify a BSDL program and release under the GPL, anything you take from it and distribute has to be released under the GPL. If you want to release it under the BSDL, you have to take it from source which is BSDL (or compatible).
You may feel that this makes the GPL an illegal license. But that is what it says.
You are saying that Linux kernels that contained some BSD code were distributed illegally, because they were licensed only under the GPL. Under your philosophy, work licensed in any fashion other than the GPL can never be incorporated into GPLed code, because extra licensing information would be required, but extra requirements are not allowed under the GPL.
A license is permission to do something. Saying code is under the GPL means that people have permission to do certain things with it that they otherwise would not have under copyright law. If I release a program under the GPL, that says you have all the permissions listed in the GPL. It does not say that those are the only permissions you have. It does not say that no portions of the code are available under other licenses. The fact that some piece of the program is also available from someone else under a BSD license does not change the fact that I am distributing it under the GPL.
All the code in a GPLed program is, by definition, under the GPL. This has nothing to do with the fact that the Regents of the University of California may hold the copyright to some bit of it, and they allow you to do lots more stuff with their part than the GPL says.
You are right that there could be a narrow region of incompatibility. Say that there is a GPLed code that violates my patent, i.e. it uses a patented technique without permission. I can legally run that code, but no one else can without my permission. However, there is no legal way for anyone to copy, modify, or distribute the code.
If the code were instead under the Apache license, I could still allow third parties to use my patented techniques (for a fee), but I could not myself run Apache, but only if Apache contained material under someone else's patent.
This seems like such a trivial case that it is not worth worrying about. If am suing everyone in the world who has a server running Apache, whether or not I can run Apache myself is not too important. They should be able to resolve this.
Too bad :-) "Compatible with the GPL" means that the license is less restrictive than the GPL and contains no incompatible provisions, so that derivative works can be licensed under the GPL. Many people want to release their code under a BSD license to allow different types of usage, one of which is inclusion in GPLed programs.
Some of the incompatible ones are really weird. Look at xinetd. Because of the wacko license, the current maintainer needed to get explicit permission from the original author to increment the version number! It also has "The author of the modifications... will be responsible for any wrong behavior of the modified software." I am sure the original author meant to say, "If someone else mucks with the code, don't come crying to me when it breaks." But this sounds a lot like saying authors of modifications will be financially and legally liable for any bugs in their code. This is not something we want even a hint of ambiguity about.
"If this is handled correctly, we'd ideally end up with something like 4 licenses"
Very good idea. If there are a few well-written licenses designed explicitly to cover the spectrum of what people want, then there will be much less of a tendency for people to roll their own.
That would be the BSD license.
That's kinda the whole point of the thing. You can't reassign the copyright, but that is not the same thing.
This is the whole difference between the GPL and the BSDL. BSDL allows redistribution under a more restrictive license, and the GPL only allows redistribution under the GPL. Saying "you would equally be able to take Microsoft Word and relicence it as BSD" is nonsense because I have no license to distribute MSWord at all.
Read the BSD license. "Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:" It never says I can't distribute it with different conditions as long as my conditions include those in the original license.
I think the point the parent was making is that the license should explicitly say that it is compatible. He meant add a clause to the Apache license that says something like this: "Notwithstanding any other provision in this license, the SOFTWARE or any derivative work may be modified and distributed under the GNU General Public License (version x.y.z)."
This would make it compatible.
Legal questions are always hard, but it certainly sounds like the Apache people are trying hard to be GPL-compatible, and have reasonable arguments that they are.
Just maybe there is a difference between an open development process, like OpenBSD, where incremental changes are examined before becoming part of the production code and dumping on the web hundreds of meg of source of a finished product which has an installed base of millions. Open source OS's get security from having many people looking at code submissions and the opportunity to find and fix dangerous bugs before they are exploited. Making a bunch of Windows source code available on the net does neither of these things.
Why is the logo a fox and not a panda?
This makes no sense.