[mail-vet-discuss] Reworked section 2 (for -13)

[Murray S. Kucherawy]

Dotzero wrote:
> After reviewing section 2.4.3 I think it should be split out
> separately for SPF and Sender-ID. A pass for SPF does not mean the
> same thing as a pass for SIDF. Conflating the two is a recipe for
> problems. Mail From pass from SPF is totally different from PRA pass
> from SIDF.
>   
Well actually now I'm not sure why you think that.  The spec doesn't say 
a pass from one is the same as a pass from the other; you could 
certainly have "spf=pass" and "sender-id=hardfail" in an A-R header or 
pair of headers.  The spec only says under what circumstances you would 
use "pass" in each case.  Both mechanism asks the same question: Was the 
client authorized to send by domain's policy (however that got 
evaluated)?  And it seems to be both mechanisms are sufficiently similar 
that they have overlapping answer sets.  That's why I grouped them when 
enumerating possible results.