[darcs-users] Licensing and copyright fun.

Stephen J. Turnbull stephen at xemacs.org
Wed Oct 15 05:39:41 UTC 2008

Trent W. Buck writes:
 > On Tue, Oct 14, 2008 at 04:18:44PM +0900, Stephen J. Turnbull wrote:
 > > Trent W. Buck writes:
 > >> Copyright assignment is certainly something to look into; it has
 > >> other benefits along with merely making licensing issues easier.
 > >> But I think it would result in MORE, PHYSICAL paperwork -- at least
 > >> in the short term.
 > >
 > > There's no legal requirement of physical paperwork for this.
 > Do you have any supporting evidence for that statement?

An assignment is a contract.  Verbal contracts are just as valid as
written ones; it is merely harder to substantiate the terms if

 > The FSF certainly uses paper assignment,

The FSF is enormously paranoid, which is pretty much justified by the
large quantity of assets they need to protect on a shoestring budget.
Paper assignments are certainly much easier to sustain in court.

Also, FSF assignments indemnify the FSF; it has *no* effective
responsibility that I can detect to verify your statements about the
origin of the software or whose signature is on the employer release
you may have also sent them.  If you assign to the FSF in good faith,
you are nevertheless still financially (and possibly criminally)
responsible if it turns out that the FSF has published somebody else's
work.  I believe it unlikely that a court would sustain that aspect of
assignment without a signature.

Such indemnification is less necessary in a tight-woven community,
where (for example) people would be able to testify about a major
contributor's participation in sprints and bug days, etc, to justify
their belief in a genuine contribution with appropriate license.

 > and ISTR reading that they don't believe legal system(s) (which
 > ones?) currently give any weight to digital signatures (e.g. GPG
 > signed declarations).

As for the FSF's legal opinions, please remember that both Moglen and
Stallman are free software fanatics, and lawyers are given special
dispensation about truth-telling in our society (at the very least,
they are *required* not to tell the whole truth if it is against the
interests of their clients).  My reading is that *physical* paperwork
is a burden that mostly can be imposed on contributors (the record-
keeping aspect at the FSF would be necessary anyway, and is not much
more costly because it's physical), and saying it's necessary is a
good way to keep the troops happy with the policy.  On the other hand,
the benefits (cheap substantiation in court and indemnification)
accrue to the FSF.  The FSF has a strong incentive to encourage
literal paperwork even if it is not truly necessary.

I am sure that the courts would give substantial weight to digital
signatures, but it might be expensive (ie, expert witnesses to
substantiate the technology, and factual witnesses, ie the "web of
trust", to substantiate the connection of the signatures to the
authors).  The other problem with a digital signature would be that
it's easy to revoke, and some ways of revoking it might make it
useless in court (ie, if you could show some probability that the key
was already compromised at the time of the assignment).

More information about the darcs-users mailing list