RE: TE Requirements Draft - ELSP

["Roberto Mameli (ERI)" <Roberto.Mameli@eri.ericsson.se>]

Hi Francois,

> > >Interesting. Do you think the IETF should have specified a third TE
> > >protocol, in case some SP might have prefered the 3rd one?
> >

of course I didn't ever said this. Let me try again to explain better, since I think it is always better to clarify in order to avoid possible misunderstandings.

> Since the above was stated in response to the question of 
> "what problem 
> does it solve", this appear to me to be saying "well, it 
> doesn't solve any 
> problem but we should still do it because some SPs may end up 
> using it". I 
> didn't find this a compelling argument. I am glad if this was arguing 
> something else (even if I don't quite get the argument).

My example was NOT meant in the sense: "SPs MUST be able to choose" THEN "we HAVE TO define two (or more???) protocols to let them choose".
I stated exactly the opposite: "As a matter of fact two protocols exist" THEN "a SP COULD choose either one or the other".

Similarly: "As a matter of fact two different LSP for MPLS support of Diffserv exist" THEN "a SP COULD choose either one or the other".

I think that if you provide two solutions to a problem you can expect that some people will choose one of them and some others will choose the other. This is quite different from saying that you have to provide two solutions in order to allow the possibility to choose.

Anyway it was just an example to introduce the analogy... sorry if I was not able to make it clear. Maybe it just added confusion, so please forget it. Moreover it is a little bit off topic and we risk to detour from the real issue of the thread.

Regards
Roberto