I do think informational RFCs (both IETF and non-IETF) are valuable.
I would suggest though that their value is mostly NOT about ease of
access, archival, those great ASCII graphics, or any other practical
matter. The value is in our assessment of them being worthy of being
published as RFCs. (Our = the IETF community, IESG and other reviews,
RFC Editor board decisions, etc) I generally treat all information as
suspect, even full standard RFCs (horror!) but I still think that across
a number of different information sources even a vendor's protocol spec
as an independent submission is pretty reliable.
So yes, lets keep on publishing informational, experimental, and
individual and independent submission RFCs. They are not the primary
source of technical information in Internet technology, but for me at
least they do provide a lot of value.
Sam: about the binding Informational documents. I also do recognize that
we sometimes make choices, say, about architectures or requirements and
that we should stick with those decisions in future work. However, I
would like to make two comments on this. First, many decisions of this
sort are soft rather than hard. As an example, we often compromise on a
requirement in order to reach a feasible solution. I'm not sure we want
to bind ourselves to decisions of this type in any formal sense. Second,
I think it is easier to think of the bindingness in terms of the history
of the document and the practical situation than as a binary attribute
of the document class.
Jari
_______________________________________________
Ietf mailing list
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf