Julian Mehnle schreef:
From my personal experience, I have found that more than 95% of messages
with a SPF result 'neutral' can be classified as spam.
The problem with this is that it is bound to change over time.
So are the scores that are awarded by SA. Nothing is static in this world.
Now that it has become a SpamAssassin default, SA users will be using this
volatile assumption of yours for a _long_ time.
You probably didn't look at it, right? Otherwise you would have noticed
that the scores for a SPF 'neutral' are set to a virtually non-scoring
0.001 points by default in this patch. Unless someone actually changes
it from the defaults, I don't believe the patch as it is will cause any
message to be rejected.
This is just detrimental to SPF's deployment as a whole.
I don't think so, see above. I fully agree that combining domain names
with SPF results is far better than looking at the SPF results alone,
but as far as I know, implementations that support this are not widely
used. Contrary to SA, which seems to be in widespread use. Anyway, it
beats rejecting on 'neutral' (as suggested before on this list) by a
landslide.
If allowing people to add a fractional amount of points in SA (notice
that 'fail' will score a maximum of just 0.875 points now) means that
this seriously hurts the adoption of SPF, we're doomed anyway.
Arjen