On Nov 20, 2004, at 2:02 PM, wayne wrote:
In <Pine(_dot_)LNX(_dot_)4(_dot_)44(_dot_)0411201100480(_dot_)13761-100000(_at_)sokol(_dot_)elan(_dot_)net>
"william(at)elan.net" <william(_at_)elan(_dot_)net> writes:
The beautiful part about SRS is that it is completely local. The
inner
details of my implementation do not need to be known to you for us to
interoperate.
That "beautiful part" is exactly why Sendmail thinks it will interfere
with mail signatures authorization and I agree.
How can SRS interfere with mail signature authorization?
Rand is on the list, so he can defend their paper himself if he
chooses. While SRS doesn't break DK or IIM, it does break BATV.
Ok, it can interfere with SES, but sendmail doesn't even mention SES,
so that can't be their point. As Meng said, even in the case of SES,
SRS will still give you an SPF pass.
Their argument (again, as it was presented to me) is that SRS just
pushes the problem along by forcing symmetric path routing of bounces
(thus increasing the likelihood of backscatter before widespread
authentication adoption).
George