On 9/28/07, David Nicol <davidnicol@gmail.com> wrote:
On 9/28/07, Monty J. Harder <mjharder@gmail.com> wrote:
> There is nothing that allows adaptation to add new functionality that isn't
> "essential" (whatever that is). And even under that provision, there is
> still no right to redistribute the adapted work.

nobody is saying there is a right to redistribute an adapted work.

> So I still say the guy's a wanker if he thinks we don't need a license to
> legally distribute his copyrighted work.

Nobody's saying that either.  

My father taught me when I was a wee lad to be careful about using absolutes like "always", "never", "everybody", or "nobody".  Let me repeat the quote from your first link:
Like his source code, Bernstein has highly optimized the whole "LICENSE" algorithm --to the point there is no LICENSE at all. Just this simple, single, clean statement of copyright.
That's SOMEBODY making that exact claim: "no LICENSE", followed by links to... LICENSES. You can't walk that one back.

What is being said, which is easy to misconstrue

I'm hardly misconstruing your words:
once you legally obtain software, you may modify it,if you have the time and inclination.
which is not supported by the law referenced in YOUR links.

as that, is that after you have done made some kind of essential
repair, arguments against sharing the story of how you did it, complete with circles and arrows, four part harmony, and diff files, on copyright infringement grounds,
are specious.

Well, put this way, we have solid legal grounds upon which to stand.  Releasing a patch program that can convert the original program into the essentially repaired version should be 100% free of problems with anything other than patents
which theoretically encumber every program ever written, with the possible exception of a program that does absolutely nothing, including defining a particular exit code such as the POSIX standards true and false
provided that it is accompanied by the standard disclaimer of warranty and the original author's responsibility for the modified program.

But that's a far cry from the simplistic language quoted above.  The fact that the FSF has now revised the GPL twice is an indication that this stuff isn't easy.  Personally, I think they've pushed some things in GPLv3 that a lot of people won't agree to, and we'll see a GPLv4 quite a bit sooner than RMS expected to address those concerns.

For example, I don't see where 'Tivoization' has anything to do with a software license, and the ham-handed attempts to stop it could keep some manufacturers from using GPLv3 code on their hardware.   Provided that the customer is aware before the sale that the hardware requires software signed by the hardware manufacturer, I say "caveat emptor", and the customer has the freedom to accept this deal if he wants.  Personally, I'd prefer customers to be able to run whatever software they want, but it could void any obligation the manufacturer would have to support the modified system, as well as the warranty, as goofy software can damage certain hardware.

But as a software license, the GPL shouldn't be addressing the issue. The software is completely free to be adapted and run on other hardware. And since it is more restrictive than GPLv2, no one can take software licensed under GPLv2 and produce a GPLv3 fork, without some other action on the part of the copyright holder(s) to authorize doing so.