I read draft-tldm-simple-homenet-naming-00.

I think this is an interesting idea, in that it attempts to carve out
a small and achievable but limited naming arrangement.  There are a
couple things I'd say, however:

1.  It seems to me that this approach is made "simple" partly by
pruning requirements.  For instance, RFC 7368 says this:

   The naming system will be required to work internally or externally,
   whether the user is within or outside of the homenet, i.e., the user
   should be able to refer to devices by name, and potentially connect
   to them, wherever they may be.

The present I-D solves that problem by rejecting the premise: names
just won't work outside the homenet.  That might be the right answer,
but it might be good for the document to be explicit about what parts
of 7368 it will or will not satisfy.

2.  There are some parts of the document that rule tricky problems out
of scope.  I think this is too puch punting.  Either the WG should say
how we're going to address those tricky problems, or else this
document should just say, "Not possible under this specification".
What I think is less good is to have a bunch of hard things that
aren't in scope, and nowhere that scope is defined, because that
encourages the tricky things to be solved in non-interoperating ways.
I think that would be bad.

I confess that my inner architecture astronaut liked the more complex
approach better, because it proposed to solve more problems.  But my
pragmatic self wants to ship something and make some progress, and I'm
not sure that a bigger scope than the present draft is going to get

Best regards,


Andrew Sullivan

homenet mailing list

Reply via email to