Re: 6MAN WG Last Call: <draft-ietf-6man-udpchecksum-02>

[Marshall Eubanks <[e-mail]>]

Dear Magnus;

On Tue, May 8, 2012 at 3:29 AM, Magnus Westerlund
<[e-mail]> wrote:
> Hi,
>
> I know the WG last call has closed. But I reviewed it anyway and I have
> found some nits and things which the WG chairs anyway will stumble on in
> their ID checklist processing of the document. I also have some more
> substantial comments on how this is written. I am sorry that I am late,
> but I simple missed the WG last call.
>
> I do support this document going forward but I do think it needs to be
> updated prior to submission to the AD.
>
> 1) Abstract contains bracketed references. I would suggest replacing
> RFC2460[RFC2460] with the documents title and RFC number without brackets.
>

The brackets come from xml2rfc. This nit is will be done.

> 2) Document header does not contain "Updates:RFC2460"

OK, will be fixed.

>
> 3) Section 1, " RFC 2460[RFC2460], " I think this should be "Internet
> Protocol, Version 6 (IPv6) Specification [RFC2460],"
>

Isn't this just # 1 ? Or am I missing something?
> 4. Section 5
>     "However, some protocols, such as lightweight tunneling
>      protocols that use UDP as a tunnel encapsulation, MAY omit
>      computing the UDP checksum of the encapsulating UDP header and set
>      it to zero, subject to the constraints described in
>      [I-D.ietf-6man-udpzero]."
>
> I am a bit worried about making [I-D.ietf-6man-udpzero] part of the
> specification text. Especially something that I at least consider being
> normative definition of constraints. Those I think should be part of
> this document. I think the pointers earlier in the document is
> sufficient to establish that part of the inclusion of the constraints in
> this document are based on whats in our document.

Here is the problem with
>      [I-D.ietf-6man-udpzero]."

The intended reference is to _the current document_. Remember, this
text is part of

 This item should be taken out of the bullet list and should be
   modified as follows:

where "this item" is [RFC2460] Section 8.1, 4th bullet

What I was worried about was that someone would take our instructions
literally, and create a mashup of 2560 and the current document, as
that is what we tell them to do. Then, you don't want to just say "as
in Section X", as that is ambiguous. Nor do you want
to say "Section X, of this document," for the same reason. Ideally, if
this document was RFC YYYY, you would say "Section X of RFC YYYY."

However, can a document normatively reference itself ? That might
excite Douglas Hofstadter, but it seems dubious to me.

It does occur to me that this must have been dealt with previously -
do you know of previous solutions ?

>
> 5. Section 6. "It is now 2011." Can the authors confirm if there has
> been no more analysis and maybe indicate the current year.

Will do.

>
> In general when I read this document it reads to much like a discussion
> paper rather than a normative specification. I think it is late in the
> process and do not require any changes now. However, I have raised this
> earlier (on the 00 wg version) and the authors have not addressed this.
>

I actually did change a bunch of text based on that. I agree it is a
bit late to do more.

Marshall

> Cheers
>
> Magnus Westerlund
>
> ----------------------------------------------------------------------
> Multimedia Technologies, Ericsson Research EAB/TVM
> ----------------------------------------------------------------------
> Ericsson AB                | Phone  [phone]
> Färögatan 6                | Mobile [phone]
> SE-164 80 Stockholm, Sweden| mailto: [e-mail]
> ----------------------------------------------------------------------
>