Hello!
On Wed, Nov 10, 2004 at 05:07:25PM +0000, Tony Finch wrote:
On Wed, 10 Nov 2004, Hannah Schroeter wrote:
Would you then recommend *not* rewriting SES sender addresses on
SRS hosts?
Yes. As well as breaking the end-to-end nature of SES verification,
SRS-of-SES is likely to cause interop problems due to overlong addresses.
This can be the case with SRS only, too. I posted an example for that,
which wasn't completely unrealistic (not all mail addresses are
of the short kind foo(_at_)bar(_dot_)com! There're also mail addresses like
FirstName(_dot_)Surname(_at_)department(_dot_)location(_dot_)company(_dot_)com;
apply SRS0, then
SRS1, then count the size of the local parts each).
Kind regards,
Hannah.