"John Palmer" <nanog@???> said, in message
009c01c52fc0$2e394a60$5919d797@JPALMERWINXP:
> > Of your remaining 5-10% that actually get out of your machine, what
> > percentage subsequently respond to the challenge?
>
> See my recent e-mail: "5-10%" just means "almost 100% are rejected".
> How many reply? One per month, on the average (my inbox anyways).
OK then. So each month you're attempting 15,000 challenges and "almost 100%
are rejected". Originally you said 90-95%, but let's call it 99.9% for the
sake of argument. That's still 15 challenges sent successfully. And only 1
gets a response. So 93.3% of all the challenges you manage to get out go
somewhere (you don't care where) and aren't subsequently whitelisted. That's
still a pretty poor success rate.
If it works the way you claim, then what you actually doing is using a
delivery attempt to the sender address as a spam filter. Sounds a bit like a
callout, but with a lot more bandwidth usage and collateral damage as a
result.
The only way that this scheme can do better than a callout is if the
un-acknowledged challenges went unacknowledged because they were spam with
real forged addresses. In which case we're back to the fact that you're
spamming innocent people with your confirmation requests.
Cheers,
Alun.
--
Alun Jones auj@???
Systems Support, (01970) 62 2494
Information Services,
University of Wales, Aberystwyth
