RE: Comments on draft-ietf-gsmp-reqs-03

I agree with Avri's comments and have some additional ones.

1. Typos found were corrected with Jonathan Sadler last week
- section 6.8 paragraph 3: "this may take form of"
- section 6.9 paragraph 2: "set on a add branch"
- section 6.1 last sentence: "The GSMP for optical switch can be"
- section 7.1.2 paragraph 2: "and the switch send an event"
- section 7.4.  New wording is "This chapter describes how to handle new and
re-established adjacencies."
- section 7.4.1 "looses" -> "loses"
- section 7.5 New wording of 1st sentence "When GSMP V3 intially was
defined, events were supposed ...", also fixed "there where no need" later
in paragraph.

2. Section 6.10.  This whole section seems to be unnecessary given that the
whole document is addressing the requirements for GSMP to support optical
and TDM switches.  Further, the diagram is not really that general because
it illustrates only one (G.805) layer.  I think that this section is trying
to address specifically the photonic layer.  This is fine but is covered
throughout the document with description of general layer support.

3. Section 7.  When reading the whole draft (serially from the start), this
section seems out of place as it is not addressing requirements specifically
for layer switches.  I think it contains valuable information for the next
version of GSMP.  Perhaps some introductory text should be added to explain
what it is applicable to.  For example "This section describes requirements
to GSMP v3 based on some implementation experience.  They address areas of
ambiguity, missing semantics, and configuration recommendations".

-----Original Message-----
From: avri doria [mailto:[email protected]] 
Sent: Monday, October 14, 2002 12:05
To: [email protected]
Subject: [GSMP] Comments on draft-ietf-gsmp-reqs-03


Hi,

I just reread the reqs draft and have a few minor comments:

- The Changes section should be removed before the next pass. Then just
publish the change list in an email to the group.

- typo in 6.3.1.1  Since an Optical Switch  (not Switches)

- also in 6.3.1.1 i would recommend changing the Optical Switch to just 
Switch since the requirement is for GSMP Switch Configuration to provide 
a list of transports, whatever they may be.  I.e. the requirement is 
motivated by optical but the support is required for non-optical as well.

- in 6.4.2  Should probably add a final sentence that the additional 
mapping will requirement documentation.  I know it seems obvious but ...


and from the NITS so important to ADs.  I have listed some that appear 
to need consideration.

- Do not number the "Status of Memo" or Abstract sections

- Table of Contents is required if > 15 pages
(yes it barely qualifies at 16 pages but some people think all RFC 
should have a TOC).  I know you have a section marked TOC.  But there is 
no TOC in it.

- Introduction  - This draft has an Overview.  Probably the same, but it 
might be safer to call it an intro.

- Security Considerations   - should be a consideration of whether there 
will be any new security requirements.  If there are none, then that 
should be stated.

- References
  o Must be split into normative and non normative sections

- If MUST etc. used, the terms must be defined (or RFC 2119 referenced) I
have a question on whether those terms in an Informational doc really have
the capitalized meaning of RFC2119.  I know I have thought differently 
about this in the past but it may make more sense to just use words in 
English instead of 2119ish.  But then they would not be all caps.

- IPR notices
probably need to say there are not IPR issues as this is a requirements doc.



---
As far as I am concerned, while the issues mentioned in this note will 
required another draft before sending on to the ADs, none is technical 
in content and thus won't warrant another last call.  But there is still 
time for someone to have a substantive comment.

Thanks for the effort
a.


_______________________________________________
GSMP mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/gsmp