I first want to re-iterate what Eric posted earlier: Please read the
appeal. The *very minor* issue of the appeal is whether or not to use
2606 names. It is the use of the DISCUSS in this case that is at
issue. That said:
On 6/17/08 at 2:54 PM -0400, Steven M. Bellovin wrote:
On Tue, 17 Jun 2008 14:44:33 -0400 Marshall Eubanks
<tme(_at_)multicasttech(_dot_)com> wrote:
On 6/17/08 at 3:50 PM +0100, Debbie Garside wrote:
Not being a expert on this but having briefly read the documents
in question, I agree with Brian. This is not editorial.
I fully agree with Debbie here.
Either Marshall or Debbie is going to have to explain this to me:
Why, IN THIS DOCUMENT, is this not an "editorial issue". Nothing said
so far that we can attribute to the IESG says anything like, "there's
a technical issue here". Please, someone give any indication that IN
THIS DOCUMENT, there is any hint of technical mistake that will be
made because a domain name other than "example.com" or "example.org"
is being used.
And for the record:
I would also add that to go against an IETF BCP on the grounds of
"well we have done so already historically" does not make an
argument for continuing to do so; errors should be corrected when
found, not endorsed.
John did not "go against an IETF BCP". This is not an "error" to be
"corrected". The BCP in question gives *an option* to use particular
names, not a requirement. There are plenty of BCPs that make all
kinds of suggestions, and some we take and some we don't. There is no
"MUST" in 2606; it is up to the discretion of the user whether using
2606 domain names suits a particular purpose. And in this case, the
DRUMS working group made an explicit decision back in 2001 to choose
something different in this document.
Human experience teaches us that examples will be used, over time.
Foo.com is a commercial site. If the IETF uses foo.com in email
examples, it is reasonable to assume that foo.com will get unwanted
traffic because of that.
There has been no evidence, in 7 years of 2821 being available, that
such has been the case. I don't see why you would say that it is
"reasonable to assume" that. This is not a MIB or the like, where
people will be copying this info into configuration files.
But let me repeat, this is not the major issue:
Since there is a remedy, and it could be adopted readily, I think
that the discuss was reasonable and do not support the appeal.
Yes -- and there's certainly case law to support the IESG's
position; the IESG has been insisting on this for years.
Let's assume you're both right. Let's assume that it's better (for
some values of "better") to use 2606 domain names in 2821. Let's
assume that it is easy (for some values of "easy") to put those names
into 2821bis. Let's assume that "the IESG has been insisting on this
for years." None of that matters:
The DRUMS working group (which produced 2821 *after* 2606 came out)
made a conscious decision at the time not to use those names. (It is
documented in the archive.) The original document passed the IESG at
the time without those names (while 2606 was out), and one must
assume that the IESG at that time made the conscious decision that
the working group consensus on this point was sufficient to leave
those names in 2821. To STOP the document now (and that is what the
AD in question has said about the effect of the DISCUSS) with *no
technical justification* of why an explicit working group decision
should be reversed (there has been none given) is a clear process
violation.
pr
--
Pete Resnick <http://www.qualcomm.com/~presnick/>
Qualcomm Incorporated
_______________________________________________
IETF mailing list
IETF(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf