On 01 Jul 2007 14:58:52 +0300, Nadav Har'El wrote: > > On Sun, Jul 01, 2007, Mikhael Goikhman wrote about "Re: GPL v3": > > On 01 Jul 2007 09:38:32 +0300, Nadav Har'El wrote: > > > > > > I am now wondering whether I should switch to GPL v3 in free software that > > > I am writing, so I am looking for comparisons of the GPL v3 to the well- > > > known (and 15 years old) GPL v2. > >... > > Whether you should upgrade to a version with no known holes (v3 or later) > > or stay with known holes (v2 or later), it's up to you. Largest holes > > that v3 closes: > > > > * patent agreements like Novell-Microsoft targetted to defend Novell > > users against Microsoft patent claims at expense of other users of > > the same program will be impossible > > * devices that _artificially_ disallow modifying the GPL source (thus > > revoking the software freedom number 2 from the users) using secret > > keys or such will be impossible > > Thanks. I'm not sure, though, what exactly these mean in practice: What > does the GPL v3 say about patents that helped it defend against Microsoft's > "attack"?
You should read the sections "10. Automatic Licensing of Downstream Recipients" and "11. Patents" to get the complete answer. In short, the Microsoft's attack is pretty much neutralized, meaning that if Novell will eventually distribute the GPL v3 versions of its current software (and it will pretty likely to), then the only way Novell may do it is if Microsoft extends its patent protection to all users of the programs in subject, and not just to Novell users (and the old agreement has such hole). But although it is ok to turn the "evil" old agreement against Microsoft, it is less interesting. The interesting thing is that new discriminatory patent agreements will be impossible in the future (after April 2007). > How does the GPL define a "device" that your second comment applies to? You should read the preamble and the section "6. Conveying Non-Source Forms" to get the complete answer. The "device" is just one example of redistribution in binary form. The important part is that any parts (including encryption keys) needed for a user to rebuild the exact copy of the distributed binary should be available freely for all users of the binary. > Does the GPL v3 say anything about the latet fad in software - > "software as a service" (where the users use the software running on a > remote machine, and never actually get a copy)? FSF thought that it would be too far to do it in the core GPL, so it (re)invented Affero GPL for this purpose. If you want your network program to always be able to download its own sources (even after someone modified it and runs as a network service), you are free to license your code under AGPL while being able to fully enjoy any GPL'd code (with corresponding versions). > Of course, I can read the GPL v3 text and get all the answers myself. But > I was hoping that someone already has some sort of annotated version, > explaining the differences from GPL v2. > > > * there is an option to use Affero GPL to extend the copyleft concept > > to the network applications, if this is what you want > > Interesting. I've never heard of this before... > I'll look it up. Thanks. Affero GPL v3 is just GPL v3 with one section "13. Use with the GNU Affero General Public License." replaced, in order to disallow removing the principal functionality of showing own program sources to the user. Regards, Mikhael. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
