David Woodhouse <dwmw2(_at_)infradead(_dot_)org> wrote:
On Mon, 2005-06-20 at 12:56 -0400, John Leslie wrote:
Is there any reason "baythorne-csv.infradead.org" couldn't include
_all_ the addresses it might use?
baythorne.infradead.org is the host name.
There is no expectation that Target of a SRV record must be the same
as the EHLO string.
Those addresses listed in 'baythorne-csv.infradead.org' are just the
extra addresses it might happen to use for outgoing SMTP connections,
which aren't published addresses for the host.
I take it that "baythorne-csv.infradead.org" _could_ include the
The address 'imladris.demon.co.uk' in particular couldn't be listed with
the others -- that ISP may change that address if/when the connection to
it switches from dialup to DSL, and I wouldn't want the duplication.
I understand not wishing to duplicate this; but it seems to me you're
taking a far greater risk by trusting an external organization to
determine the basis for _your_ reputation. It doesn't strike me as a
good practice to do this.
Personally, I prefer to have any mobile computers do submissions to
a single server which I fully control. I don't mean to force anyone to
do things my way; but I'm very hesitant to increase the workload of
CSV checking. It needs to be lightweight.
John Leslie <john(_at_)jlc(_dot_)net>