On 1/3/07, Brian E Carpenter <brc(_at_)zurich(_dot_)ibm(_dot_)com> wrote:
On 2007-01-01 15:17, Robert Sayre wrote:
...
> A WG can agree with the AD that there is a problem, but disagree that
> it needs to be solved in their document. Too often, the compromise
> ends up being the insertion of text that satisfies the AD's concerns,
> but disenfranchises the WG. The WG either ignores the text in
> practice, or the document author couches the text in so many
> qualifiers that it becomes easy to explain any implementation in terms
> that make it seem conformant. The result is inaccurate or misleading
> documentation of Internet technology.
But it may be more harmful not to point out the issue at all, or
Maybe, but then it's an editorial issue, and tangential to my comment.
In more relevant cases, the AD asks for substantive technical changes.
not to specify a better-than-nothing solution.
A decision for the WG, in the event that they haven't already
considered it. A DISCUSS is much more productive when it points out an
issue the WG hasn't thought of. That's extremely unlikely for issues
resembling the example Dave Crocker gave:
A classic example of this is citing basic DNS problems, for
specifications that
are merely consumers of the DNS and, hence, are not creating any
new problems.
I don't see anything in the ION that addresses this concern, and it's
more worrisome if the IESG is free to retroactively make any RFC a
BCP. Perhaps the ION should explicitly state which RFCs a draft must
obey. RFC2026 seems to be the least relevant, so I don't think it's
productive to pretend that's the one we follow.
You have to consider
the case of a relatively inexperienced implementer who may not know
something that we believe is common knowledge.
Again, an editorial issue.
It's always open to the WG to propose a resolution of the DISCUSS
that is radically different from what the discussing AD suggests,
too.
Yes, any group is free to try anything in the IETF process, but that's
not expedient for a WG that needs a document published. Let's face
it--DISCUSS is a fairly Orwellian euphemism.
--
Robert Sayre
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf