Re: AD-review of draft-ietf-ipo-carrier-requirements

hi, see in-line...

> Alex and IP-Optical WG
> 
> As the editor of the "Optical Service Requirements" Draft RFC document, I
> have collected some input from the authors of this document and want share
> them with you and the IP-Optical WG community.
> 
> 1. We disagree with your and other's comments that this document is simply a
> recompilation of the contents from OIF and ITU. Most authors of this
> document are active participants and contributors of those requirements in
> both OIF and ITU. 

the second statement doesn't strictly speaking justify the
first one nor rationalizes it from an ipo wg perspective

> One of the initial objectives was to bring the
> carrier-driven requirements developed in OIF and ITU into IETF to help
> IETF/IPO community understand the carrier's needs (remember IETF and OIF/ITU
> did not have formal liaison relationship at the time this document was
> started). It also contains fair amount of the contents that are unique to
> this document.

then one possibility would be to point out to what's really
specific compared to existing documents that are available
elsewhere, but the scope must be clearly defined browsing
from architectural until system design aspects doesn't seem
to be the most appropriated way to do so

> 2. This document was initiated by the IPO WG chairs to ask a group of people
> form major carriers to provide optical service requirements for IP layer,
> and we did that and more. 

be a "major carrier" doesn't give more merits to the technical
argumentation (than coming from a "small carrier") here, the
arguments are evaluated on their technical merits not from
their origin (oif, etc.) or the affiliation of their editor(s)
and/or contributor(s)

> However, we did not receive much constructive
> suggestion as to the suitability of the direction and contents of the
> document during past two years. If the problem was seen, it should have been
> brought to our attention earlier, not now.

the scope of the document has shifted three times since its
first version:
- san diego'00: operators to define control plane req's
   under the ipo wg umbrella (as requested by the chairs)
- london'01: organised as a list of numbered requirements
   trying to accommodate/direct ietf developments
- and then shift at slc'01 towards the control plane
   requirements from the itu ason work (note most of this
   architectural/functional work had been consented by
   that time at the itu) and oif dvp's

difficult to ask to an engineer community to follow your
reasoning behind these continuous shifts (it also means
that either the scope of the document wasn't well defined
from the early beginning or that authors they didn't know
what they really wanted from the early beginning)

> 3. This doc was intended to be an informational RFC and tried to reflect the
> major high-level functional requirements and architectural framework
> envisioned by the carrier community for the intelligent optical transport
> network relative to IP and other service networks. 

this statement seems contadictory to the one listed in point 1.

> Those requirements are
> currently refined at OIF and ITU, a consolidated high level requirements
> combining ITU ASON, OIF NNI and UNI, and IP over Optical specific
> requirements will certainly be a good value this document can bring.

the same applies here.

> 4. For all practical purpose, we see the following options for the document
> - Keep the existing scope, but clean up the document and incorporate the
> latest requirement development in OIF and ITU by carriers to improve the doc
> quality and move forward. Keep the document as a high-level document.

there is a clear contradiction here b/w "high level" and parts
of its current content - how can a document claim to be high
level and implementation/protocol neutral while referring to
implementation agreements defined by the oif ? - so if it has
to progress in that way there is imho an important clean-up &
consolidation effort to be provided to make here consistent
progress

(*) note that this comment doesn't apply only to this ipo i-d
(see also previous e-mail)

> - Reduce the scope of the document and expect major rewrite to focus on the
> detailed IP over Optical requirements.

there are several levels of requirements:
- architectural/models
- service/functional
- protocol and related
- operational/management/etc.

as long as the requirement level to be covered by this i-d is
not agreed and clearly defined i don't know how it can really
and efficiently progress in becoming useful for this community

> - Drop the document since it's not in the best interest of all the authors
> and others  if it lacks the interest and support from IPO WG.

depends on the previous points but if the situation doesn't
improve quickly i don't think we will have any other choice

> 5. Appreciate Dimitri Papadimitriou's spending time to provide specific
> comments on the document. We will address all his comments and concerns if
> it is decided to progress the document.

thanks.

- dimitri.

> -Yong
> 
> _______________________________________________
> IP-Optical mailing list
> [email protected]
> http://lists.bell-labs.com/mailman/listinfo/ip-optical

-- 
Papadimitriou Dimitri
E-mail : [email protected]
Private: http://www.rc.bel.alcatel.be/~papadimd/index.html
E-mail : [email protected]
Public : http://psg.com/~dpapadimitriou/
Address: Fr. Wellesplein 1, B-2018 Antwerpen, Belgium
Phone  : +32 3 240-8491