ietf
[Top] [All Lists]

Re: Trustees License Use of Templates in RFCs

2015-03-13 15:08:37
--On Friday, March 13, 2015 15:40 -0400 Sam Hartman
<hartmans-ietf(_at_)mit(_dot_)edu> wrote:

"John" == John C Klensin <john-ietf(_at_)jck(_dot_)com> writes:

    John> Sam,

    John> This may go back to our earlier (apparent)
disagreement on the     John> subject.  IANAL or nearly a good
enough approximation to one     John> to be competent to
assess the details, but my presumption     John> continues to
be that the basic rules about RFCs, plus this     John>
language, allows the type of application you are concerned
John> about as long as the version of the template in the
John> application is an exact copy.  

You're absolutely right.  However, the requirement that the
application be an exact copy is in and of itself enough to
make such software fail to comply with the Open Source
Definition, the Debian Free Software Guidelines, and the Free
Software Foundation's definition of free software.
So, on its face you cannot do that in a free software or
open-source application.

My strong belief is that in this instance the IETF's interests
are better served by cooperating with the
open-source/free-software community than by preventing
modification of the templates.

As I have been trying to explain in a different context to Scott
Kitterman, I suspect we agree in everything but language and
tone.    Certainly it is not in the IETF's best interest to
prevent reasonable software that enables template-filling-in
software from being written.  I personally contend that it is
also not in the IETF's best interest to encourage, or even
allow, modifications to templates that would permit
misrepresentation of the results or, more importantly, would
facilitate the creation of filled-in templates that IANA, for
example, would reject because they were inconsistent with the
templates specified for a given registry.

I don't see those two objectives as being inconsistent in any
way.  It is certainly possible to formulate heavy-handed ways to
express one or the other that would be a problem but, if the
community is in agreement with both principles (and I haven't
heard anyone speak up in opposition to either as long as they
don't have bad effects on the other), then the question is
whether the Trust's language is consistent with the community's
intent.

You and Scott (and probably others) apparently believe we have
gotten into exactly the situation in which the community's
intent about the first principle is being blocked by the Trust's
langauge.  I don't think I have a competent opinion on the
subject although it occurs to me that, if we are getting hung up
on what "exact copy" means, some long-term review of the
licenses you cite may be in order too.

But I hope the three of us could join in asking the Trust and
its Counsel to work directly with the relevant people associated
with the licenses you cite to come up with mutually-acceptable
language.  

IMO, if we can't sort this out without a lot more pain and
aggravation, we _really_ need to consider going back to some
reasonable approximation to "acknowledge, be clear about
changes, but otherwise do what you like with RFCs".  That may
not be the only reasonable answer, but, IMO, the cost of  these
discussions is rising above the benefits of more restrictive and
precise policies.

       john