ietf-mxcomp
[Top] [All Lists]

Re: consensus call on MUST/SHOULD language for TXT records

2004-09-03 21:53:06

Andrew Newton <andy(_at_)hxr(_dot_)us> writes:

Rather than wordsmithing the actual MUST/SHOULD/MAY language directly 
(and endlessly), we ask the participants state their preference on 
these positions.

I think we should require full compliance from the start. Having taken
that stance, I would question the need for having 2 RR types. It is
acknowledged (in the current documents) that the use of TXT is a
transitional, but if we require everyone to publish SPF2 (or whatver
name is used) RR then I do not think we need to publish or check TXT
records. I would also support the alternative given by Daniel
Quinlan's response of dropping SFP RR and using TXT as this would lead
to less deployment problems than using a new RR - though I understand
and appreciate the arguments for having a new RR.
 


<Prev in Thread] Current Thread [Next in Thread>