Re: [bfcpbis] WGLC for draft-ietf-bfcpbis-rfc4582bis

[Tom Kristensen <[e-mail]>]

On 09/24/2012 11:13 PM, Mary Barnes wrote:
> I have reviewed the document and it looks pretty much ready.
>
> I do have a couple of questions for clarification:
> - Section 5.1 - Payload Length: states that "receiving server MAY send 
> an Error message"
> but I would think that the client should know about the error. Is 
> there a specific reason this is a MAY versus a SHOULD or even a MUST? 
> Why would the error message matter if it's not required?

I don't remember why MAY was chosen here, but after having a look at 
this again I see that SHOULD is used for sending an Error message twice 
in the original RFC 4582. So, we should (!) change to SHOULD.  OK?

> - Section 5.2 - "M" after Table 2 - I have a similar question about 
> the 'M' bit. RFC 4582 states that message is rejected if there is an 
> unrecognized M bit, whereas this document states that an error message 
> MAY be sent. Certainly, 4582 lacked appropriate normative language, 
> but if the error message is just a MAY, then why bother?

Yeah, why bother ;-)
We'll change to SHOULD. However, it wouldn't harm using MUST for sending 
these error messages - apart from that SHOULD is used in RFC 4582 and 
that is my main reason for suggesting using SHOULD here.

> There are also a few nits that should be fixed/clarified before 
> progressing.
>
> Section 5.2.6 - Note under table 5:    "intended being used" -> 
> "intended to be used"

OK

> Section 6.1.: "e.g.,"  was changed to "e.g.".   The former is correct.

OK

> Section 9.1: You lost some indented paragraphs in this section - i.e., 
> they got shifted to align with the other paragraphs in that section. 
>  I'm not sure if that was intentional.

What do you mean? Same white space indentation as in RFC 4582, isn't it?

> Section 16:
> - New Primitives.  The text says five were added, but only 4 are 
> listed and that's all that's all that's defined per Figure 1.
> - New error codes.  The text says that 3 are aded, but it looks 
> there's really 5 that have been added.
> - Typo.  Referring to section 12.4.2 states that the change was from 
> SUPPORTED-PRIMITIVES to SUPPORTED-PRIMITVIES but the latter is really 
> the typo.

Fine, good catches - will change.



I'll change accordingly ASAP in the new version if no objections are 
provided!

-- Tom