Re: WG Last call passwd: draft-ietf-tewg-diff-te-reqts-04.txt

[Francois Le Faucheur <flefauch@cisco.com>]

Nabil,

At 17:04 11/06/2002 -0400, Nabil Seddigh wrote:

>While this is late, I thought I should at least mention one
>observation re: section 2. The draft states:  "The solution
>must satisfy all the application scenarios described in
>section 2..."
>
>After going through section 2 of the draft more carefully
>it appears to me that the 3 scenarios are redundant in terms
>of demonstrating that a particular solution satisfies 3
>different application requirements. Scenarios 3 deals with
>3 classes, Scenario 2 deals with 2 classes and Scenario
>1 deals with 2 classes.

For memory, the point of the "Application Scenario" section is to answer a 
question that came up many times initially which was "what exactly are the 
SPs trying to do with the new thing and why cannot they do it today with 
existing TE".
I think the three scenarios show three different objectives/applications ( 
Limiting Proportion of Classes on a Link, Maintain relative proportion of 
traffic classes ,Guaranteed Bandwidth Services) and answer the question well.

The fact that despite different goals they exhibit the same fundamental 
requirement (ie enforce different bandwidth constraint) was also deliberate 
to give the reader a feel for the essential DSTE technical requirements 
before they are detailed in the next section.

I think section 2 is just fine as it is.

Francois

>My suggestion would be to keep scenario 3 and get rid of
>the 1st 2 scenarios as they result in redundant text.
>
>Best,
>Nabil Seddigh
>
>
>Ed Kern wrote:
> >
> > There was some discussion on this draft during the close of the working
> > group last call.  The chairs have discussed the changes and believe that
> > the document can be revised with minimal changes to section 3.3.  The
> > editors of the document now must incorporate the minimal changes into
> > another version of the draft.  After that, it will be forwarded on to the
> > IESG.
> >
> > This WG document addresses a WG milestone.  It is proposed as an
> > Informational RFC.
> >
> > Ed & Jim