-----Original Message-----
From: owner-spf-discuss(_at_)v2(_dot_)listbox(_dot_)com
[mailto:owner-spf-discuss(_at_)v2(_dot_)listbox(_dot_)com]On Behalf Of johnp
Sent: Wednesday, July 20, 2005 3:07 AM
To: spf-discuss(_at_)v2(_dot_)listbox(_dot_)com
Subject: Re: [spf-discuss] Thanks for all the feedback on the SPF
validator
This might not be a real bug, but something for you to consider,
even though I know you
have discussed this aspect at some length.
If I enter an incorrect domain name the validator accepts it as
long as there is a "." in
it - even if the domain name does not actually exist in the public
domain. As I understand
it this is to allow for certain edge cases, but would it not be
possible to flag a warning
if the domain name is not in the public domain, so that mistyped
records are caught? I am
thinking of the the more basic user here. If the domain name
mentioned in the record does
not have an A or MX record it would IMHO be suspect to an average
spf publisher.
Just a thought.......
And a big thank you for the validator - it's probably the most
useful practical step
forward SPF has made in the last year.
SLainte
JohnP
That's on my to do list for "Harsh" mode of pySPF.
My first concern is to catch all the valid PermError cases. The reason for
that is that if receivers are going to reject mail from domains that have an
SPF record with a PermError (and that seems to be the trend), we need to
make for sure that people new to SPF have a way to find out if their record
is going to cause a PermError.
That last thing we want is for some new person to publish a record and then
start having legitimate mail rejected. So, PermError is the first priority.
One of the common errors is to put in a:mail instead of a:mail.example.com,
so I'll catch that one now. The most common (I think) error is putting
mx:mail.example.com (which, generally, won't exist). I don't, yet, catch
that one, but is a definite To Do item.
Thanks,
Scott K