Re: Outgoing section 5.5 and draft-josefsson (Re: San Diego meeting slot)

[Simon Josefsson <jas at extundo.com>]

"Joel M. Halpern" <joel at stevecrocker.com> writes:

> This note seems to assume that most of the code in RFCs comes from
> existing sources and is subject to external license restrictions.

Ah, right.  I would not go as far as to assume "most", but I do assume
that a significant portion of code that would be useful to include in
RFCs do come from existing sources and is subject to external license
restrictions, and I have so far been focusing on such code.

My view is based on documents that I'm familiar with.  We have code
from external sources or with additional licenses in several
documents, including RFC 1321, RFC 3492, RFC 4122, RFC 4226, and RFC
4634.

> At least in my experience that is not the case.  The code consists of
> Code written by individuals for the RFC to expalin processing
> MIB definitions written for the RFC
> XML Schemas written for the RFC
> ASN.1 written for the RFC
> ABNF written for the RFC

It seems our experience differ somewhat here.

I agree that these are good examples.  When possible, I think we
should strive towards making them available under the most liberal
license possible.

> Remember, the RFC is not normally used as an implementation library.
> We are not trying to provide sample (much less reference) code.  We
> are trying to provide enough information that people will build
> interoperable implementations.

I'm not sure I share this view.  There are RFCs whose purpose is to
provide easy access to actual implementations, e.g., RFC 4634.  They
are useful.  I would prefer if more RFCs were accompanied with
relevant example code, because I find them helpful when implementing
the document (regardless of whether I use the code or not).

> We have agreed that Open Source, GPL, commercial, and other users need
> to be able to use the code from RFCs.
> Personally, I think that including code with a GPL in the RFC would be
> counter to that goal.

I disagree, because I believe code in RFCs under, say, the BSD license
or the GPL is better than no code.

> PS: Remember that authors can still make available anything they want,
> with any terms and conditions they want.  So an author can still have
> a copy of the same code with GPL conditions. But if the code needs to
> go in the RFC then the author needs to make it available according to
> the IETF rules.

Of course.  However, that position prevent the IETF from including
useful code portions licensed under the BSD/GPL licenses, that could
help implement the RFC.

/Simon

_______________________________________________
Ipr-wg mailing list
Ipr-wg at ietf.org
https://www1.ietf.org/mailman/listinfo/ipr-wg