Would I feel miffed if someone stole my biggest open source project?Maybe. Would that make me wish I had GPLed it? No.
I also have no idea what the percentages are on influence but my personally belief is MIT/BSD breed more open source than GPL. GPL attempts to breed open source by force. MIT/BSD by kindness. I'd like to think MIT/BSD does better because of that. It makes it easy for people to contribute and or fork and share and not worry their contributions can't be used in their own non open projects. You might see that as less open but I see it as more because their contributions wouldn't exist otherwise.
I think LLVM and all the zillions of things that are being spawned from it are a good example. If it was GPLed I doubt most of those projects would exist.
I believe copyright is complete rubbish; it's just a legal extension of a sense of entitlement.
Since I believe copyright is rubbish, I must believe that all that is based on it is rubbish, including the GPL. Therefore, I won't use it.
The MIT and BSD licenses are good "interface adapters" between creators who believe copyright is rubbish, and the practical world with its legal framework which insists that there must be copyright.
Stallman also basically believes copyright is rubbish, but he used it to forge the "copyleft" which exploits it for an agenda.
My view is that if you disagree with it, don't wield it.
(Comment under this same submission.)
The BSD is a great way for those same companies to get your $100... and ask for another $500.
But if you can make your stuff work with no abstract legal entities getting involved, then sure, BSD or CC0 are the best.
In fact, incidentally, I believe that if you integrate BSD-licensed code into a proprietary product (where, say, it is manifested as a translation to machine code), you don't actually own those sections of machine code. They are derived works, governed by the BSD license: anyone can rip that out and use it accordingly. The code has not been turned proprietary at all---it just appears in a form that is difficult to modify and tightly coupled to proprietary code.
What I would do is help myself to their patch, if it's worthwhile and backport it to the BSD-licensed version.
The code is my copyright, and the patch is a derived work of something of mine; therefore I own it and it falls under my license.
Not to mention that if you take someone's BSD program, and just slap a GPL on it, it doesn't actually hold. Once the users are aware that it's actually a re-labeled BSD program from another author, they can just go with that license.
>> The code is my copyright, and the patch is a derived work of something of mine; therefore I own it and it falls under my license.
That is completely incorrect. You have no rights to the patch just because it is a derivative work. You have the right to prevent distribution of derivative works (not to own or distribute), but you gave that up with the BSD license which expressly allows it.
You're also wrong about the license of the patch. The author of the patch gets to chose it. For GPL licensed software the condition of redistribution is that all derived works be distributed under the GPL as well. If you don't GPL it, you lose your permission to redistribute the original code and derived works.
BSD and MIT licenses don't include this restriction, so you're shit out of luck with making any claim on that patch.
If I release an MIT project, someone makes a GPL fork with a significant new feature, wouldn't pulling that feature back into my project make my project GPL?
You are correct that pulling their GPL-licensed feature/additions into your code would require you to adhere to the GPL and re-license the code. The best you could do would be to ask the author of the changes to release them under the same license (MIT) as your original project so the changes could be incorporated without modifying the license.
For an example of this situation, see the link in btilly's post: https://news.ycombinator.com/item?id=9711716
That is an issue of contention. GPL advocates believe that they are permitted to do that, and have done so in the past. The Software Freedom Law Centre (the legal arm of the FSF) advices developers that they are permitted to do it[1], as do GPL advocates on HN[2].
People can create derivative work from BSD licensed software and have any license they want to added with it. Theo has argued that some peoples bug fixing patches did not qualify for copyright, and he is perfectly free to think so. I doubt however that he would ever put money where his mouth is and test it in court. When silence can qualify for copyright, it is a hard argument to make.
> Permission is hereby granted... to deal in the Software without restriction, including without limitation the rights to... sublicense
Because you have the right to "sublicense", you can take MIT code and license it under another license (GPL) even though you are not the copyright holder.
The BSD license is less clear on this point (in fact this is one of the important differences) and whether the BSD's grant includes the right of sublicense is more debatable, but in my view the answer is "probably".
I am not a lawyer; I am not your lawyer; this is not legal advice.
The original license required copyright notices in binary code. A newer form does not, but this is for pragmatic reasons. The binary code is still copyrighted. If you decompile it into a source language, then you probably have to restore the copyright notice and disclaimers.
When a proprietary, binary-only program with a highly restrictive EULA contains BSD code, those sections of the machine language which correspond to the BSD code are not actually under the EULA. They are not sublicensed, but only used with permission, as granted by the original license.
Which part of the patch[0] that removed the BSD license and made the file GPLv2 isn't evidence?
You'll note that the patch removed the lines that say "Permission to use, copy, modify, and distribute this software for any purpose with or without fee is hereby granted, provided that the above copyright notice and this permission notice appear in all copies."
In other words they removed the notice that specifically says you're not allowed to remove the notice.
[0] https://marc.info/?l=linux-wireless&m=118831729905625&w=2
Was that version ever distributed to anyone?
I also notice in that thread that the author of Ath5K .c code give explicit permission to "Alternatively, this software may be distributed under the terms of the GNU General Public License". Removing the old license text of the header file and replacing it with GPL might be incorrect, and people are free to argue what a judge would say about it. It just seems very far from an actually issue, any proof of current believes, or what advices FSF gives.
Because it's tasteless. Legal, of course, but tasteless. The person doing that is exhibiting a degree of hypocrisy in their refusal to pay it back to the original developer, while demanding that those who use their code pay it back to them. It's a "rules for thee but not for me" type of thing.
> The code is my copyright, and the patch is a derived work of something of mine; therefore I own it and it falls under my license.
This is a misapprehension of how software licensing works.
{
char name[NAME_LEN];
{
char name[NAME_LEN+1];
Given the insanity of copyright laws, I probably can't.
The idea that `name` is short by a byte isn't protected, just the literal code.
(It's quite likely that this particular change is too trivial to be covered by copyright, but the chilling effects reach wide enough here that it might not be worth risking it.)
"Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer."
If you want to appropriate the code as GPL, you better do most of the work in new source files, or else change the file so significantly that (according to your lawyer) you're justified in replacing or altering the original copyright notice.
Anything that has only my name and my copyright notice on it is mine, even if it was modified. Whoever modified it acknowledged that it's mine by preserving the notices. It smells a lot like copyright assignment, and I suspect that very argument would easily succeed in court. I don't see how you can argue with a straight face that A. U. Thor is making an infringing use of your work, by taking an excerpt from a file which is "Copyright 2013 A. U. Thor".
When you choose GPL you're saying "I believe software should be free"[1]. When you choose BSD you're saying, "here's something I've made; use it as you like". Those are two very different kinds of statements, operating at entirely different levels.
[1]: Of course, not everyone chooses GPL for that reason.
The licenses are functionally equivalent in every other way.
IANAL, IANYL, this is not legal advice.
I like all the freedom and sharing as does Stallman. I just don't see proprietary software as evil, and I don't think that if I prevent some code of mine from being proprietized, that I'm striking some blow against proprietary software. If anything, it might backfire: instead of using my quality, debugged code, they will have to choose something else that is perhaps not as good (or worse: whip something up from scratch). Somehow, in some way, that will come back and bite me.