On 30-Jun-24 07:35, Michael Breuer wrote:
On 29. Jun 2024, at 18:59, John Levine <johnl@xxxxxxxxx> wrote:
It appears that Brian E Carpenter <brian.e.carpenter@xxxxxxxxx> said:
Fair enough - but the first thing is to know what the operational
issues actually are, so that we have some idea how to fix them and
which WG(s) might be involved.
I don't think these are issues in which the IETF has useful expertise.
I also think that the kinds of technical changes that would be likely
to help would have backward compatibility problems that would cause a
great deal of heartburn here.
That's all the more reason to describe them. It is not OK to shove these
problems under the carpet.
Gathering the required operational expertise worked for v6ops and dnsops
quite well - why shouldn’t it work for mail operations?
Exactly.
PS: It also doesn't help that running a mail system at scale is very
different from running a tiny one, and we have an awful lot of people
who overgeneralize from running tiny ones.
I'm not sure whether the operational experience of tiny server operators is
being discredited here. But operators of large scale mail services are
welcome to bring their operational expertise into the discussion.
That's just a bit condescending. The IETF *needs* the input from the big
operators and it needs to be very specific, especially if there are
backward compatibility problems as John Levine says.
Brian
