[ntp:questions] I am not a lawyer

unruh unruh at wormhole.physics.ubc.ca
Thu Dec 24 01:04:28 UTC 2009


On 2009-12-23, Dave Hart <davehart at gmail.com> wrote:
> On Dec 23, 20:19?UTC, unruh wrote:
>> On 2009-12-23, Dave Hart <daveh... at gmail.com> wrote:
>> >?I am however
>> > interested in reducing the number of different copyright notices in
>> > the reference implementation. ?I'm grateful to Max K?hn and Claas
>> > Hilbrecht for removing their copyright notices from sntp and
>> > refclock_neoclock4x, respectively, relying on the overall distribution
>> > copyright.  [...]
>>
>> The problem is that removing their copyright notice helps not at all,
>> and in fact makes the situation more complex. copyright is automatic.
>
> I too believe copyright is automatic in most jurisdictions today.  I
> do not agree that reducing the number of diverse copyright claims and
> accompanying BSD-ish licenses in the NTP reference implementation
> "makes the situation more complex" -- quite the opposite.
>
> Any arbiter of dispute is bound to look at the totality of the
> evidence and not rely solely on automatic copyright as the end of the
> story.  When I contribute code to the reference implementation, I do
> not make any explicit assignment of my interests, but I also do not
> make any attempt to exercise a copyright holder's prerogative to set

Unfortunately one of the prerogatives is to grant no license whatsoever,
meaning noone can copy. That is a perfectly valid option for a copyright
holder and must be the default option that the courts must apply-- that
the author granted no license because he intended to grant no license.
That is of course confused by the fact that the holder contributed the
work to another work which IS released under a specific very open
license. The problem is that in a dispute like this, the court in
general has to go with the more conservative option, and the author
might tell the court that was his intention because desperately and
irrationally dislikes, say Microsoft. 

> license terms.  Given that the reference implementation has a top-
> level COPYRIGHT file and the documentation links to the same text in
> html/copyright.html, I think it is clear I expect my contributions to

But the reference implimentation is clearly a work derivate of the
various parts, and the copyright in the derivative work does not
supercede the copyright in the part. The court could make what for you
is the obvious decision, that in the absense  of an explicit statement,
the author implicitly decided to grant the same licence as the work as a
whole. The author could however, because he hates microsoft, come to
court and claim, that that was never his intention. This would leave
Microsoft up shit creek. 

Now, the probability of any of this coming to court is small, but for a
large publically targeted company, they have to worry about such
possible ambushes. It is thus far cleaner for them if the license
granted by all of the authors of the contributed works are clear. 

> be distributed and licensed the same as the rest of the package.  If

No, unless you state it, it is not clear.

> Dr. Mills or ntp.org were to ask me to sign a copyright assignment or
> release as a condition of contributing, I would not hesitate.  I would
> prefer in the interest of the best software to avoid requiring
> contributors to deal with paperwork, but it's not my call.

Agreed. But paperwork is not required. A statement that an unconditional
license is granted to Mills to relicense in any way he wished, or to
transfer copyright to him, would suffice-- in the text of the
contributions, or whatever. At present the COPYRIGHT notice claims
copyright for Mills, and then states that he is not the author of
various parts of the work, and states nothing about the copyright status
of those parts. It is a confusion. Is he claiming copyright over those
parts despite not being the author? Is he claiming the right to license
those parts despite not owning the copyright? 

>
> Similarly, while no one should rely on my opinions as legal advice,
> When Max and Claas took affirmative steps to strip their copyright
> claims and BSD-like licenses out of their contributed source, those
> steps seem to me evidence of intent their contributions be covered by
> the distribution-wide COPYRIGHT.  I have a hard time seeing how those

But their action actually accomplished the opposite. It removed anyone's
right to copy their source. 
"Copyright in this work has been given to David Mills" would accomplish
what they want. Or "This work is licensed to David Mills to do with as
he wishes, incliding relicensing to others", or many other ways of
letting the world know what their intention is. Silence is not an very
good expression of intention. If you are married you know that taking
your wife's silence as an expression of consent is a pretty dangerous
thing to do. 

> steps "make the situation more complex", as there are now two fewer
> licenses for lawyers to get paid by the hour to dissect, and two fewer
> places for packagers to find copyright notices potentially requiring
> duplication into documentation.  So for now, I continue to be
> interested in eliminating as many copyright claims and licenses from
> the distribution as can be arranged with claimaints, with the ultimate
> goal of having only the Mills/udel.edu copyright.

Then transfer your copyright to him. 

>
> Cheers,
> Dave Hart




More information about the questions mailing list