Subject: Re: Tarball licenses (was: Free documentation licenses)
From: Rick Moen <rick@linuxmafia.com>
Date: Fri, 1 Dec 2000 11:24:52 -0800

begin John Cowan quotation:

> This is why I said the term "relicense" should be avoided, as it creates
> more confusion than it dispels.

Point taken.
 
> Good enough, but see the bit about translation below.

Seems almost like homeopathy, doesn't it?  None of Alice's code may
remain in A1, but (as you say) its being a derived work remains, and
thus the obligation to conform to Alice's licence terms persists.

Which raises the interesting question of 4.4 BSD (and progeny) as a work
derived from AT&T System V UNIX.  I seem to recall that the Computer
Science Research Group strategy going into the lawsuit involved gradual
replacement of AT&T code.  They did this primarily because the AT&T code
sucked, but it's widely considered that this was having the effect of 
gradually creating a non-proprietary codebase.  Are you saying that
interpretation is wrong, and that, but for the lawsuit settlement, BSD
would have remained subject to AT&T's licence terms even after all AT&T
code was gone?

> Almost.  It might also be inferred from Beirne's actions, or rather
> inaction.

Er...  I know that copyright cannot be lost (unlike trademarks) through
inaction, but the notion of licensing through inaction makes me a little
queasy.  This is a tad vapourous.  As far as a remote observer can tell,
Beirne may be unaware of Krzysztow's usage because he became a monk on
Mount Athos, and has several years of unread e-mail waiting for him at
home.

> If you are aware that infringing derivative works have been created,
> and you do nothing....

But that's an additional hypotQ}tical, above and beyond just "inaction".

>> [...] -- not to mention the copyright holder changing his mind.
> 
> Now you bring up a sticky matter which nobody has ever resolved,
> because there is no case law about public licenses.  What happens if
> the copyright holder does change his mind, and revokes the license?
> People who relied on it in the past clearly have no problem -- if a
> derivative work has been created, there is no issue.  But whether that
> derivative work can continue to be copied or modified is a question.

I actually wasn't even going to go there.  I was envisioning the
pragmatic problems created by a piece of software whose licence exists
only as a Web page. 

Bob legally acquires tarball A.  It refers him to Alice's Web page for
(among other things) the licence terms.  He reads the licence:  Cool, he 
likes the terms of usage and distribution, and (e.g.) builds a business
on it.  A year later, he notices that the Web site is gone, and can no
longer track down Alice.  He may be in a position to prove licence
compliance, but the package's licensing is now -- in a pragmatic sense
-- problematic.

This situation and similar problems are more common than you may think.
I'm trying to track down Beirne, for example, but he may turn out to be,
like many authors of older software, difficult to find.

> Right, because patches are not derivative works -- their use of the
> original is fair use only.  They are essentially commentary.

Thank you.  That is a useful characterisation.  And thank you again for
your trouble in this matter.

-- 
Cheers,                                      "Reality is not optional."
Rick Moen                                             -- Thomas Sowell
rick@linuxmafia.com