Re: WGLC review comments for draft-ietf-tsvwg-sctp-chunk-flags-00.txt

[Michael Tuexen <[e-mail]>]

Hi Gorry,

a comments in-line.

Best regards
Michael
On Aug 24, 2010, at 1:55 PM, Gorry Fairhurst wrote:

> Thanks Micahel, one answer in-line.
> 
> Gorry
> 
> On 24/08/2010 09:44, Michael Tüxen wrote:
>> Hi Gorry,
>> 
>> thanks a lot for the comments. See my responses in-line.
>> 
>> Best regards
>> Michael
>> 
>> On Aug 24, 2010, at 9:34 AM, Gorry Fairhurst wrote:
>> 
>>> 
>>> This is a review for draft-ietf-tsvwg-sctp-chunk-flags-00.txt, submitted during the WGLC.
>>> 
>>> I have already read and reviewed the previous version and I see that those comments have been incorporated in this draft. The remaining comments therefore relate to wording, rather than document structure.
>>> 
>>> ----
>>> The abstract seems to motivate this as a WG item, rather than as the RFC-to-be. It currently states:
>>> 
>>> 
>>>   The current definition of the Stream Control Transmission Protocol
>>>   (SCTP) is missing a procedure for registering chunk flags for already
>>>   defined chunk types.  This document defines this procedure.  It does
>>>   not change SCTP in any other way.
>>> 
>>> It may be better to say something more like:
>>> 
>>>   This document defines the procedure for registering chunk flags with
>>>   the Internet Assigned Numbers Authority (IANA) for the Stream
>>>   Control Transmission Protocol (SCTP). It updates RFC 4960, and
>>>   also defines the IANA registry for contents for currently defined
>>>   chunk types. It does not change SCTP in any other way.
>> OK, will be in rev 01.
>>> 
>>> ----
>>> Similarly in the introduction:
>>> 
>>>   [RFC4960], which currently defines the Stream Control Transmission
>>>   Protocol (SCTP), provides a procedure to define new chunk types.
>>>   However, several protocol extensions currently being discussed need
>>>   to define new chunk flags for existing chunks.  The only way to do
>>>   this is to obsolete [RFC4960], which is not appropriate.
>>> 
>>> Change the final sentence, perhaps to something like:
>>> 
>>>   Without publication of this document, the only way to have done
>>>   this would have been to obsolete [RFC4960], which is not appropriate.
>>> 
>>> ----
>> OK, will be in re 01.
>>> Later in the introduction:
>>> 
>>>   This document overcomes this limitation and provides the procedure to
>>>   register chunk flags for existing chunk types.
>>> 
>>> Could be better as:
>>> 
>>>   This document updates RFC 4960 to overcome this limitation.
>>>   It defines the procedure to register chunk flags, and specifies
>>>   the registry entries for for existing chunk types.
>>> 
>>> ----
>> OK, will be in re 01.
>>> 
>>> Section 3 should clearly name the registry that IANA is requested to be used for the regiatry entries. The document shepherd write-up requires this question to be answered:
>>> 
>>>   "Does it suggest a  reasonable name for the new registry?
>>>    See [RFC5226]."
>>> 
>>> This registry would be easier to identify if it were to be called something like the
>>> 
>>> "SCTP Chunk Flag Registry"
>>> ^^^^
>>> ----
>> OK, I change it to:
>> IANA is requested to create an SCTP Chunk Flag registry.
>> Will be in rev 01.
>>> 
>>> Section 3.1
>>> 
>>> - Is the title best? Could this be clearer if it said:
>>> 
>>> "Procedure for Registration of New SCTP Chunk Types"
>> Section 3.1 replaces Section 14.1 of RFC 4960 which has the title "IETF-Defined Chunk Extension"
>> That is where the title comes from. When working on RFC 4960bis, Section 3.1 would
>> replace Section 14.1.
>> Does this title now make sense to you?
> 
> Understood. Perhaps it would have been more obvious perhaps if you explicitly said mentioned that it replaced section 14.1.
Since I want Section 14.1 being replaced by Section 3.1, I'm not saying it in Section 3.1
but in Section 3:

3.  IANA Considerations

   Section 3.1 describes the updated procedure for chunk type
   registration and replaces Section 14.1 of [RFC4960].

OK?
> 
>>> 
>>> ----
>>> 
>>> Section 3.2
>>> 
>>> Suggest that you change:
>>> IANA selects a chunk flags value, exactly one
>>>                                ^^^
>>> 
>>> IANA selects a chunk flags value. This must be exactly one
>>>                                ^^^^^^^^^^^
>>> ----
>> OK, will be in rev 01.
>>> 
>>> Section 3.3
>>> 
>>> Suggest change:
>>>   This section describes the initially defined chunk flag tables, one
>>>   table per chunk.  Most of the tables are currently empty.
>>> 
>>> To this:
>>>   This section describes the initial values of the chunk flag tables,
>>>   one table for each chunk. Most tables are currently empty. IANA is
>>>   to use these value to create the new registry.
>>> 
>>> ----
>> OK, will be in rev 01.
>>> 
>>> If the WG recieves comments supporting publication of this document, I will be the document shepherd.
>>> 
>>> Best wishes,
>>> 
>>> Gorry
>>> 
>>> 
>>> 
>> 
>> 
>> 
> 
>