I decided to re-read the entire arch document frmo scratch. I have some
comments. I don't mind shipping the document as-is, but I thought I'd
write down my thoughts as I read the document.
1. "This means that Layer 2 is largely out of
scope, we're assuming a data link layer that supports IPv6 is
present, and that we react accordingly. Any IPv6-over-Foo
definitions occur elsewhere."
While I personally have no problem with the above, it kinds of breaks
"tone" of the rest of the document. Using "foo" in this kind of document
where before in the document things are very strict and formal, stands
out. It also changes from an impersonal note to "we" all of a sudden in
the middle of the paragraph.
1.1 FQDN should probably have a reference to an RFC or something?
2.2 First paragraph is long and complicated. Took me 3 reads to store
enough information on the stack to be able to pop at the end to the
context understand.
In 2.2, there could be a mention added of users expecting their devices to
not be reachable from the Internet and they historically have not been
reachable, but when these devices are put on an IPv6 homenet, they are
suddenly reachable if there is no filtering. Users might choose weak or no
passwords because historically this has been fine.
2.3. With SLAAC, devices don't "receive" addresses. Perhaps write
something about "addresses can exist within one prefix per ISP" or
something of that effect? This section, "devices", does that mean the CER
or devices the user connects to the CER, or both?
2.6, last paragraph. To make it more clear, I propose to change to "The
widespread availability of robust solutions to these types of
transition technologies requirements", since "these" refers to a previous
paragraph and I was confused for a while what "these" actually referred
to.
GENERAL: Generally throughout the document it talks about "acquiring an
address" or similar wording. Perhaps this could be changed to "acquires
adress(es)" ? It has specific statements around end hosts having multiple
addresses, but then reverts back to mentioning adresses in singular term.
3.3.5 seems very open ended and vague. What is "the homenet DHCP server"?
I thought each router within the home could be a DHCP server for a subnet.
Does homenet aim to solve this information spreading problem or is it out
of scope, or we don't know at this time?
3.6.1 there is no "." at the end of the last paragraph.
GENERAL: Since RFC6204 has been obsoleted by RFC7084, shouldn't we refer
to 7084 instead of 6204?
So... as I stated initially, the above is mostly "nitpicking" and I don't
require a re-spin of the document just for my comments sake, but if the
document is re-spun for other reasons perhaps some of my comments could be
taken into account?
Oh, just to be clear, I think this document is in good shape generally and
I want it progressed.
--
Mikael Abrahamsson email: [email protected]
_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet
