Re: Re: Proposed IESG Statement on the use of the “Updates” header

[Michael Richardson <[e-mail]>]

I have read the statement, and I think it's a bit weak.
I don't feel very clarified, and Tom asked the question I would have.

Ben Campbell <[e-mail]> wrote:
    >> On Sep 11, 2018, at 11:31 AM, tom petch <[e-mail]>
    >> wrote:
    >> To give a practical problem; when an I-D adds new entries to an
    >> existing registry, is that an 'Updates'?  I have seen ADs firmly tell
    >> WG Chairs, holding the contrary opinion, that it is not, and I thought
    >> that that was settled, but applying this statement to that situation
    >> leaves me in ignorance.

    > The intent is that these are usually not “Updates”. But they can be if
    > there are special circumstances, which I presume would be documented in
    > the text that describes the nature of the update. An individual AD may
    > or may not agree with an argument that a particular update is “special”
    > in this sense, but I believe we all agree that such special cases are
    > in the realm of possibility.

I don't think that this situation is terribly rare or special.
It definitely happens regularly in my opinion.

When doing -bis documents with the intent of Obsoletes a previous document,
I think that all of the Updates: need to be reviewed.  I think that Updates:
should be considered in a future context of revising the document.
  "if you were revising it now, would you include the new text?"

If the feature/extension is still quite rare, unpopular, or very specific to
very narrow scopes, then I don't think it should Updates:.    In most cases
when a document in WG X does an allocation from a document produced by WG Y,
I don't think it's ever an Update. (X!=Y, except when X and Y are clearly
related, such when Y=="Xupdates" or "Xmaint")

When WG X does such an allocation for the WG-X-protocol-foo-extension, then I
think it definitely is an update, but that's not because it did the
allocation, but because of what it does to the protocol.

--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        | network architect  [
]     [e-mail]  http://www.sandelman.ca/        |   ruby on rails    [


--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-