Lex wrote:
If you use MIT/X (or GPL or anyother license) you give up a big advantage of public domain: it's rock-solid established law that is codified internationally by the Berne convention.
Andrew Greenberg werdna@mucow.com wrote:
Lex, what advantages are you talking about?
As I stated in the text you quoted, the advantage that it's rock-solid established law followed internationally. I retract my comment about the Berne Convention, because I don't see it in there despite an occasional web page saying it is. On the other hand, it does appear that authors can easily put things into public domain in practically all countries. Am I mistaken? I can't find one exception, and every site I can find on google agrees that authors in most countries are free to can disclaim ownership and release their material to the public. Here's a typical comment:
The rules vary from country to country. In just about every country, authors can put a work in the public domain by formally declaring that they are doing so.
http://onlinebooks.library.upenn.edu/okbooks.html
Assuming this is correct, public-domain has a substantial advantage. You know what it means and you can safely assume it will hold up as expected in most jurisdictions. To contrast, different countries have varying requirements for what counts as a valid license at all, and thus it is not even know yet which OSI licenses are valid in which countries.
Two people have now claimed that you cannot disclaim warranty if you release something to the public domain, but I don't see how that is true. You just do it. For example, you can include a file like this in your distribution:
I release the files in this work to the public domain. I disclaim all interest in this work. I do NOT guarantee the work to be fit for any purpose whatsoever. Do not attempt to use it in a nuclear power facility, as a dinner recipe, as light reading, [etc. etc. etc.]
Thanks, Patrick, for clearing up what was meant about mixing BSD and GPL code. I still do not see that it is safe, even if FSF agrees with your analysis [1]. First, it is important (and FSF agrees) that you distinguish BSD from BSD-without-advertising, which the original post I responded to did not. Second, MIT-L (for example) requires that an exact copy of MIT-L be included in derivative works, while GPL says just the opposite: you can make any change you like in a derivative work, including, I would assume, removing the work's copy of MIT-L. MIT-L therefore has an extra provision that GPL does not, and I don't see how one can legally re-release an MIT-L program under GPL. It seems like a stark difference: MIT-L says you must include the notice, and GPL says you may change the work in any way you like.
Lex
[1] http://www.fsf.org/licensing/licenses/license-list.html#GPLCompatibleLic enses
On Jun 21, 2005, at 2:06 PM, Lex Spoon wrote:
Lex, what advantages are you talking about?
As I stated in the text you quoted, the advantage that it's rock-solid established law followed internationally.
What is rock-solid? I really have no idea what you are talking about.
I retract my comment about the Berne Convention, because I don't see it in there despite an occasional web page saying it is. On the other hand, it does appear that authors can easily put things into public domain in practically all countries. Am I mistaken?
Ok, I think you are suggesting that dedication is possible everywhere copyright is permitted. Fair enough, and I agree (although to be effective in some nations, including the US, dedication and assignments have to be in a writing signed by the owner).
I am unaware, however, of any nations in which licensing is not permitted and routine. What are the advantages?
Assuming this is correct, public-domain has a substantial advantage.
Over what?
You know what it means and you can safely assume it will hold up as expected in most jurisdictions. To contrast, different countries have varying requirements for what counts as a valid license at all, and thus it is not even know yet which OSI licenses are valid in which countries.
Different countries also have varying requirements for what counts as dedication. In the US, you can have implied licenses. You cannot have, however, implied dedication. Although a non-signed dedication probably constitutes an implied license.
Two people have now claimed that you cannot disclaim warranty if you release something to the public domain, but I don't see how that is true. You just do it.
But your disclaimer may or may not be effective, in the absence of a contract. In the licensing context, at least under US law, the licensor is granting consideration (license grant) in exchange for certain promises, and pursuant to various contractual standards. There is offer, acceptance and consideration -- and thus a warranty, disclaimed, limitation of liability, agreed to.
I release the files in this work to the public domain. I disclaim all interest in this work. I do NOT guarantee the work to be fit for any purpose whatsoever. Do not attempt to use it in a nuclear power facility, as a dinner recipe, as
light reading, [etc. etc. etc.]
I can't tell you that you would be free from liability in this scenario, or that such liability, if any, would be limited under this agreement. The problem is that there is no contract. Note, also, that the text in the file likely does not constitute a dedication under U.S. law, although I haven't thoroughly researched the question.
Clearly, public domain "licensors" have similar concerns. For example, SQLite, does not accept contributions without WRITTEN and SIGNED dedications.
squeak-dev@lists.squeakfoundation.org