--On Thursday, 18 December, 2008 14:42 -0500 Keith Moore
<moore@xxxxxxxxxxxxxxxxxxxx> wrote:
> John C Klensin wrote:
>
>> I would assume that, if someone were revising TCP or IP fifty
>> years hence, they would end up creating new text, rather than
>> moving a lot of technical text forward, and that they would do
>> so for all sorts of technical reasons independent of copyright
>> constraints.
>
> Now I'm confused. When we revise a technical specification in
> IETF, we try (to some degree) to avoid creating new text
> unless necessary, for fear of introducing subtle
> incompatibilities. Why wouldn't "someone" fifty years hence
> have similar concerns?
Because there is a big difference between clarifying or updating
an existing spec and writing a new protocol.
> I'll grant that the text in a lot of these old RFCs looks a
> bit dated and imprecise by now. But if I were revising TCP,
> I'd still use RFC 793 as a starting point, and the resulting
> documents would probably qualify as "derivative works".
That would be your choice. And you are a lot more likely to be
around 50 years from now to make it than I am. However, trying
to write a new spec with IPR that conforms to 5378 but that used
significant text from 783 would require you to obtain Jon
Postel's permission which, I think, would send you swimming in a
can of worms of appropriate size for you to do so.
So, to write such a spec with 783 text, you would either have to
un-do 5378, create the type of exception procedure that caused
the current version of this thread to be initiated, or deal with
that can or worms. It would be up to you to determine which of
those options would be most attractive... or whether a complete
rewrite or giving up entirely would.
Q.E.D.
john
_______________________________________________
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf