MX dot was (Re: [ietf-dkim] TXT wildcards SSP issues

[Steve Atkins]

On Jun 6, 2007, at 10:11 PM, Hector Santos wrote:

> Douglas Otis wrote:
>> On Jun 6, 2007, at 3:35 PM, Hector Santos wrote:
>>> But why NO MAIL?  Why not other policies?
>>>
>>> A system can have a default NO MAIL policy or a default I SIGN  
>>> EVERYTHING  or anything else.
>>>
>>> Here is a workable Wildcard syntax that has a default NO MAIL POLICY
>>>
>>> *._ssp       0  TXT   ... no mail policy...
>>> _ssp         0  TXT   ... I may sign ..
>>> public._ssp  0  TXT   ... I never sign ...
>>> sales._ssp   0  TXT   ... I always sign ..
>>> corp._ssp    0  TXT   ... I always sign ..
>>>
>>> and one with a default I ALWAYS SIGN
>>>
>>> *._ssp       0  TXT    ... I always sign ..
>>> public._ssp  0  TXT   ... I never sign ...
>> This requires a transaction at every label within the domain in  
>> question, where of course, this also clobbers SLDs.
>
> Explain to me why this is a problem?
>
> I am borrowing the logic used from one of the original LMAP  
> proposals, DMP, which SPF based on its merged designed with another  
> LMAP RMX? proposal.
>
> This is a single lookup by the client, no traversal, no loop,  
> required.

Your reasoning is unclear to me.

Given the domain a.b.c.d.e.f.g.h.i.j.k.foo, please explain what
single DNS query you would make and what answer you would
expect to receive.

>
> THe *._SSP record gives you the global default result as desired by  
> the  main domain.
>
> So regardless of the subdomains provided, you have a GLOBAL default.
>
> Then for specific subdomains, you can further defined specific txt  
> records to override the default.
>
> Again, I am no DNS expert, but is there a TECHNICAL problem with this?
>
> Explain it to me in terms of where there is overhead, pressure or  
> lots of work, if any, on the DNS server?

Cheers,
   Steve