Autor: David Woodhouse Data: Para: exim-users Assunto: Re: [Exim] Re: closed connection in response to STARTTLS.
ph10@??? said: > I'm saying that we are talking at cross-purposes. :-)
Sorry, that's my fault. When you said you weren't particularly interested in
problems caused by network outages, I presented that as an alternative way
of achieving the same result. I was wrong.
I'll fall back to arguing that temporary network outages _do_ happen in the
real world, and the state Exim gets itself into when this happens is
self-sustaining, given a throughput of more than 72 mails a day.
> > Mail gets bounced after sufficient delay.
> OK, that's really serious.
I thought so, yeah. Optimising for the 95% is fine -- but I do tend to
prefer it if the percentage of mail that actually gets to my mailboxes after
being accepted by one of my servers is somewhat higher than that, though :)
> Yes, that is expected. If I can get the RSET thing to work, it won't
> count as a delivery failure, so the delay shouldn't happen.
ITYM 'If I can get the RSET thing to work with the servers we're currently
testing against, with this phase of the moon...'
Even if it seems to work, I would still think that the default should be
equivalent to "hosts_nopass_tls = *". But I know how to configure my own
boxes to behave that way - there's no real need to argue about _defaults_ if
you don't agree, and nobody else on the list gives a monkey's.