Renaud Allard wrote:
>
> W B Hacker wrote:
>
>
>>Our error messages *might not* be specific if the nature of our tests turned up
>>things that classified the source as a probable spammer or 'probe'.
>>
>>Your specific server had been so classified here 27 September 2006 for sending
>>to a non-existent address that was clearly intentionally 'constructed', not a
>>mis-spelling of a valid user.
>
>
> You probably mean generated by a rule like that
> !verify = sender/callout=20s,defer_ok,random
>
> ohh, callouts again :)
'au contraire' - we handle *proper* callouts all day long.
We even take 'postmaster@<our IP> as an IP-literal, as the RFC requires such,
though we don't *send* that way, so it is purely a nod to the RFC's.
But a *proper* sender-verify is either addressed to the 'postmaster' or to the
alegedly-valid user currently attempting a send.
As we have no such user as:
gandalf.llorien.org-1159383911-testing@???
This was neither....
.. now 'Irish Alzheimers' being what it, is my 'wetware' may have forgotten that
I perhaps *invited* you to test...
If so, you at least got a demo of what paranoid servers can do...
Wait til *next* week when the PostgreSQL AI rules kick in..
No need for 'honeypots' All we care about is the traffic that hits *our*
servers. Any of them.
;-)
Bill
