Re: [mpls] MPLS-RT review of draft-vainshtein-mpls-gal-tc-ttl-handling-01

["Adrian Farrel" <[e-mail]>]

I'd like to avoid going round and round on this (partly because I am too busy to
be pedantic today :-)

> >> My point is that the current text does not say what you say.
> >> The text says "It is RECOMMENDED that implementations set the
> >> TC field of an LSE that contains the GAL to all zero (0b000).²,
> >
> >> and that statement does not include any qualifiers or conditionals.
> >> I understand you say that there is an ³there may be valid reasons
> >> to do otherwise, with implications carefully understood before
> >> doing so² implicit to the SHOULD.
> >
> >Maybe naive, but I thought SHOULD and RECOMMENDED have the same
> > implicit conditions, RFC 2119 says:
> >
> >3. SHOULD   This word, or the adjective "RECOMMENDED", mean that there
> >    may exist valid reasons in particular circumstances to ignore a
> >    particular item, but the full implications must be understood and
> >    carefully weighed before choosing a different course.
> 
> That does not mean ³Absent local policy, then it is RECOMMENDED Š². The
> text shows an absolute recommendation.

There is no such thing as "an absolute recommendation" because a 2119
recommendation is a should.

[Aside: Two years ago, in an airport not far from Washington DC...
TSA Officer: You are recommended to remove your shoes.
Me: No thanks, I'm OK with them on.
TSA Officer: No sir, you are recommended to remove your shoes.
Me: I think I'll be OK with them on. 
TSA Officer (louder): Sir! You are recommended to remove your shoes.
Me: Ah, I see. I'm sorry. I didn't realise. You're telling me to remove my
shoes.
TSA Officer: No sir, I cannot tell you to remove your shoes.

When we do a 2119 recommendation we mean "absent a reason to do otherwise".

But the crunch is that the documentation needs to give advice about when to vary
from the recommendation.
I thought we had that covered, albeit implicitly, by the previous sentence.

However, I have no problem with being super-explicit.

OLD
   Setting the value of the TC field in an LSE that contains the GAL is
   done by the LER that originates the G-ACh packet and is a matter of
   local policy for that LER.  It is RECOMMENDED that implementations
   set the TC field of an LSE that contains the GAL to all zero (0b000).
NEW
   Setting the value of the TC field in an LSE that contains the GAL is
   done by the LER that originates the G-ACh packet and is a matter of
   local policy for that LER.  It is RECOMMENDED that implementations
   set the TC field of an LSE that contains the GAL to all zero (0b000), 
   but they MAY vary this according to local policy.
END

Ciao,
Adrian