On Mon, Mar 22, 2004 at 02:33:40PM -0500, Alan DeKok wrote:
my example was meant more to highlight problems introduced by the
intermediary transit (the ISP, e.g., via a transparent proxy) than
by the end-user or the administrators of the end-user's desired MX
(or, more likely, by policies the administrators must obey).
The intermediate policies are chosen to prevent end-users from
abusing network resources. e.g. Blocking port 25 outbound is an
attempt to stop end-users from sending spam.
...which is fine, and has workarounds today. I'd prefer our
recommendation not break those workarounds.
They upgrade laptops regularly. So there's already an existing
model for deploying software upgrades. Wait 2-3 years, and deployment
will have reached the majority of users.
...and for those people that must wait a year, or two, or three, before
email is functional again? When a service that many view as
fundamental, such as email, stops functioningi as deployed, trickling
out a fix isn't seen as an appropriate response.
There are individual organizations in which that trickle-out could take
two to three years.
For that reason, it'd be better to come up with a recommendation whose
impact is largely, or completely, server-side, and transparent to the
client and end-users.
Mark C. Langston Sr. Unix SysAdmin
Systems & Network Admin SETI Institute