Re: [nfsv4] WG Last Call for Extension and Minor Versions, Extended Attributes, and ACLs / Umask - Ends Dec 2nd 2016

[[e-mail] (J. Bruce Fields)]

On Thu, Dec 08, 2016 at 06:49:38AM -0500, David Noveck wrote:
> > Even though David responded, the question is for the document
> > authors.
> 
> Right but I thought it might be helpful to discuss the possible choices
> before they decide to pick one.
> 
> > My apologies for being opaque about "stable reference".
> 
> I don't think you were "opaque".  The problem is that there is a level of
> uncertainty about exactly how stable the reference needs to be, and no
> matter what we choose, the answer might be different when the IESG (and
> various assorted reviewers) consider the matter.
> 
> > The  POSIX-1003.1e normative reference makes it appear that it
> > is a Posix standards document while in the content of the
> > reference it does state "withdrawn" - confusing.
> 
> I referenced this issue in my review but I wasn't clear about why this was
> a problem.  I just asked how it could be normative, which may not have been
> helpful in correcting the problem.  Your formulation is clearer.
> 
> > The other problem is that I cannot quickly locate a copy of that
> > document
> 
> I could quickly locate a copy.  Unfortunately I can also quickly find a
> bunch of advertised links to that document, which are broken :-(
> 
> > and if it is going to be a normative reference then it needs
> > to be reasonably available.
> 
> That implies it wouldn't have to if it is an informative reference.  I
> suspect that there will be some concern about this issue during the review
> process, even for an informative reference.
> 
> > I understand that it is a defacto standard at this point given the
> > implementation.
> 
> I don't think the IESG would be very receptive to the idea of "De facto
> Standard References" section :-(
> 
> > My main point is that if it is going to be a reference a target needs
> > to exist and the naming should likely be changed to remove the
> > implication that it is a Posix standard.
> 
> I agree.
> 
> > As mentioned, if the reference is not needed based on the
> > contents of the I-D, then the I-D can be reworded and references >
> removed or changed.
> 
> I still think the acl man page would be a good target for an informative
> reference but the authors need to decide what they want to do about this
> issue.

I don't care much, I think either could work.  Are there other examples
of RFC's referencing man pages?

I guess for now I'll keep the existing reference and move it to an
"informative references" section.

--b.