Matt Sergeant wrote:
On Wed, 5 Mar 2003, Eric S. Johansson wrote:
not true. If you design the sender pays system correctly, even weak machines
can participate. In the camram model, only the first couple of messages have
"high-cost" Stamps. Every other message to a known recipient or from a known
recipient comes at a much lower cost.
The latter part does not make my statement false. There are lots of
systems where the majority of email recipients are first-timers.
I stand corrected.
Let me rephrase my comment them. The cost of sending e-mail on a low-end
machine will probably the only about 30 or 40 seconds per address[1]. This
charge will occur in background and not impact the user experience. Obviously
if someone sends a message to a large number of people, then it will take quite
a while before the message will be delivered. Assuming they are willing to or
able to (because of power issues) keep the machine running long enough to
generate enough stamps, but it shouldn't be a problem. Additionally, it is
possible for folks to offer to volunteer to generate stamps for these people if
you should so wish.
---eric
[1] the reason for a per address stamp is that it keeps spammers from recycling
stamps on multiple clients.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg