Doug, all,
30.01.2011 20:51, Doug Ewell wrote:
Mykyta Yevstifeyev wrote:
>> I'd like to see some kind of guideline that the RFC should not be
considered obsolete solely because of security or performance concerns
in some particular, specific context. For example, the fact that
vanilla FTP is not sufficiently secure for use in some applications
where high security is paramount is not a rationale for deprecating
FTP in all applications.
>
> In the case I mentioned as c the key words are 'is not possible or
is not advised to be used in the Internet' but not what you mentioned.
The document says “is not advised to be used in the Internet because
of its security issues, impact on its performance or any other
reason.” (Do you agree that the document says that?) My point is
that, because security or performance issues in one context do not
necessarily imply security or performance issues in all contexts, they
should not by themselves (or together with the 7-year criterion) be
sufficient to trigger deprecation.
I've recently thought on how to formulate this criterion. My most
current thoughts are the following:
c. The RFC describes the technology that is not possible to be used
in the current Internet because of its technical characteristics or
possible problems with its implementation.
However this is not a final variant - any other proposals are welcome.
> The phrase 'or any other reason' is put because there is no
possibility to put the exhaustive list of such purposes. Anyway,
what would you like to propose here?
I don’t have exact replacement wording. “Any other reason” could
permit me to propose deprecation or “historicization” of a protocol
because I don’t like the guy who created it, or because my company is
promoting a rival protocol.
Agreed here. See what I think below.
---------
And now a few questions for discussion:
1) Should historicizing Informational RFCs be allowed?
My proposal is to allow this only if they describe the protocol (see
Section 4.2.2 of RFC 2026) with the authors' approval REQUIRED.
2) Definition of obsolete RFCs are still unclear. The most recent what
I have is:
The RFC SHALL be considered to be obsolete if it meets the following
criteria:
a. It has been publicly available for at least 7 years;
b. During this period of time the technology, described in this RFC
has not been seen used in the Internet; or
c. The RFC describes the technology that is not possible to be used
in the current Internet because of its technical characteristics or
possible problems with its implementation.
Any proposals on this?
3) Procedures on Experimental RFCs to Historic.
What I have in my working version is different from what is in -01
version. See below:
3.2.3. Experimental RFCs
Procedures for historicizing Experimental RFCs depend on their origin
and the way it is being historicized with.
3.2.3.1. Separate Historicizing Document
The procedures described in this section apply to the case, mentioned
as 'b' at the beginning of Section 3.2 (separate historicizing
document).
If the Experimental RFCs has been processed on IETF stream [RFC4844],
'IETF Consensus' [RFC5226] is REQUIRED to historicize it.
If the Experimental RFCs has been processed on IAB stream [RFC4844],
'IETF Consensus' [RFC5226] and IAB Chair approval is REQUIRED to
historicize it.
If the Experimental RFCs has been processed on IRTF stream [RFC4844],
'IETF Consensus' [RFC5226] and IRTF Chair approval is REQUIRED to
historicize it.
If the Experimental RFCs has been processed on Independent
Submissions stream [RFC4844], 'IETF Consensus' [RFC5226] and authors'
approval is REQUIRED to historicize it. In essential cases the
approval of the director of the area the historicized document is
considered to be related to MAY be used instead the authors' one.
In the cases described above IESG is responsible for recording their
approval.
3.2.3.2. Superseding Document Historicizes the Superseded One
The procedures described in this section apply to the case, mentioned
as 'a' at the beginning of Section 3.2 (superseding document
historicizes the superseded one).
The superseding document that is being processed on the same stream
[RFC4844] as the superseded one MAY move it to Historic without any
special procedures; a simple mention of such action is therefore
REQUIRED in superseding document.
If the superseding document is being processed on the stream,
different from that of superseded one, the approval of corresponding
party is REQUIRED. Section 3.2.3.1 describes some cases that apply
this one as well (for IAB-, IETF-, and Independent Submissions
streams [RFC4844]). Historicizing IETF-stream documents by non-IETF-
stream ones SHALL be made following usual procedures for RFCs of such
stream with IETF Chair approval REQUIRED.
4) Are there any thoughts on other consideration connected with
historic docs.?
Except referencing, there might be appropriate to discuss what should be
done with IANA registries defined by Historic RFCs. Anything else?
All the best,
Mykyta Yevstifeyev
--
Doug Ewell | Thornton, Colorado, USA | http://www.ewellic.org
RFC 5645, 4645, UTN #14 | ietf-languages @ is dot gd slash 2kf0s
_______________________________________________
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf
