MÃns RullgÃrd <[EMAIL PROTECTED]> writes: > If, one might argue, the author wishes for the terms to remain those > of the GPLv2, why does he not remove the "or any later version" > option? The answer is simple. Such a license is not compatible with > the standard GPL (with the "upgrade" option), since it has "further > restrictions", compared to the version allowing a switch to a later > version.
That's not my understanding. The GPLv2 & GPLv>2 are logically distinct licenses, one can simply decide to drop the "or any later version" in code you distribute based on code with that clause. The no extra restrictions bit refers to the GPLv2 *or* the GPLv3 -- not both together. Of course, if someone contributes modifications under only GPLv3 (for example) you then have to make a choice about accepting that code, or retaining the GPLv2 license. But that's not a new problem. > One common reason to use the GPL in the first place, is > precisely to be compatible with other GPL licensed software. Remember > that few (none?) copyleft licenses are compatible with the GPL, be it > by design or by chance. This is less a characteristic of the GPL in particular than of licensing in general. > Placing your code under the GPL, is placing a large faith in the FSF > not to change the license terms in a manner you might disagree with, a > faith which in many case may be broken, should some of the rumored > clauses end up in the final GPLv3 text. True, and given some of the rumors I'm rather skeptical about the freeness of the GPLv3 to be. But all in all I don't think the risks are that great of using the "or any later version" language. Worst case scenario is that folks discover they've given more permissions than they meant to. But as for folks understanding better how they're licensing their software, I couldn't agree more. -- Jeremy Hankins <[EMAIL PROTECTED]> PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03
