wayne wrote:
In
<Pine(_dot_)SOL(_dot_)4(_dot_)58(_dot_)0406202212360(_dot_)28614(_at_)orange(_dot_)csi(_dot_)cam(_dot_)ac(_dot_)uk>
Tony Finch <dot(_at_)dotat(_dot_)at> writes:
Let me know if I am misunderstanding you in some important way?
No you aren't, you just seem to believe that the people who came up with
the PRA algorithm have come up with something that doesn't require lots of
forwarders to change their software. This is not the case.
Not to mention mailing lists. Something like 20% of the mailing lists
mirrored to Usenet on gmane.org do not add the Sender: header and will
have to be changed in order for the Caller-ID PRA algorithm to work.
From what I can tell, this is a *MUCH* larger issue than the whole
forwarding issue. While MicroSoft folks (Jim Lyon and Harry Katz)
have repeatedly claimed that the SPF breaks more stuff, I have a hard
time believing it, especially with *zero* published data to back up
their claim.
Apart from this, I see another major obstacle (with the Caller-ID PRA
mechanism) in the many firewall and anti-virus SMTP implementations.
Firewall vendors tend to be extremely conservative in adding new (E)SMTP
commands/keywords. Anti-virus SMTP vendors most of the time are doing a
great job at scanning messages for viruses, but they're usually doing a
very poor job in the proper handling of the SMTP protocol (and SMTP
extensions). And a significant number of companies do use these firewall
and anti-virus SMTP implementations between their internal network and
the Internet.
I don't expect both groups of vendors will change their software soon to
support something like Caller-ID PRA.
/rolf