Re: [ietf-dkim] SSP issues

Scott Kitterman wrote:
> On Wednesday 30 May 2007 18:22, Jim Fenton wrote:
>
> > (2) SSP record type (TXT vs. something new). Only 4 messages in discussion,
> > mostly saying "if you support TXT, don't bother with anything else." 
> > Again, no clear consensus.
> Agreed.  There is also a view that if you go with a new RR type, don't bother 
> with SSP.  By the time a new RR type is widely deployable, the market will 
> have found a different solution.
> Scott K
IMO, I think we should stop wasting further time on this repetitive 
issue and just establish two, a new RR with a TXT fallback, as part of 
the specs.
The bottom line is that client applications will learn to adapt to 
domains. They are not stupid. If they sees that a domain only uses TXT, 
it may stop doing a RR lookup for this domain.  To re-adapt to RR, one 
easy solution would be to have a TXT attribute:
    rr=1

that says the RR record exist and the client should use the RR first, if 
it can.
If we stick with TXT only, we loose the opportunity to use RR in the 
future (even if its years).
If you stick only with RR, then we risk early and optimal adoption and 
compatibility.
So we should use both with an initial RR, TXT fallback with insights for 
client to learn, and we should also consider a TXT attribute, like RR=1 
that says that a RR record exist for those clients that can read it via 
the network

--
Sincerely

Hector Santos, CTO
http://www.santronics.com
http://santronics.blogspot.com


_______________________________________________
NOTE WELL: This list operates according to 
http://mipassoc.org/dkim/ietf-list-rules.html