Alex van den Bogaerdt wrote:
"abuse = broken policies are bad, die spammer, die".
Let's not assume that _every_ error is on purpose, made
by spammers to frustrate the system, shall we?
With spammers I always tend to assume the worst. They can try
to intentionally publish erroneous policies, or try to find
erroneous policies published by somebody else.
Shit happens.
Sure. And then crying "shit" (5xx) is fine.
White hats won't mind if their mail is being rejected due to
a fault on their end.
ACK, And BTW, I didn't mind the one bounce caused by an old
forwarder (for VeriSign whois addresses) I got in one year, I
just sent it again directly to the forwarded-to address, ready.
Fortunately the "FP" FAIL bounce was clear and short and sweet.
if you never make (and made) a mistake, you're not human.
I make _bad_ mistakes every day, but not so often with a knife.
SPF is a weapon. Reporting errors a.s.a.p. is the most gentle
therapy I can think of.
I do agree on returning 5xx on permerror.
Good.
Just don't say it is abuse by iditiots and/or spammers.
Scott sometimes forces me to use plain text, because he really
thinks that an unreliable "tolerant" system is better than a
reliable but less liberal system. But SPF can't take the "be
liberal in what you accept" approach, it desperately needs the
voluntary receivers doing the right thing.
Otherwise spammers and worm authors would again try to ignore
SPF, after all they're also desperate. And dangerous.
No offense against you intended. I needed at least one month
to understand why the SPF policy for claranet.de didn't cover
xyzzy.claranet.de - that was a serious SPF screwup on my side,
as you said, shit happens. But then it was fixed a.s.a.p.