On Thu, 16 Dec 2004, Marc Perkel wrote:
> There are a lot of things that the RFCs say that Exim exim accepts. If Exim
> enforced all the rules no one would be able to use it.
Think about someone (me) setting out to write an MTA. One reads the
RFCs, and tries to implement them. Then you hit the Real World (tm) and
discover that "everybody" ignores certain things or implements them
differently to your interpretation. So you change the MTA to make it the
same. Then you discover that SOME software does not follow the RFCs and
that people somehow have to support it. For those, I add options to
relax Exim's rules. The boundary between these two classes of standard
behaviour relaxation is of course very fuzzy and subjective. You will
find both cases in Exim.
However, I do not think you can argue that conformance to the standard
is a bug.
On Thu, 16 Dec 2004, Peter Bowyer wrote:
> Even if it were determined that way, there's no case for changing the
> default behaviour when there is an easy documented workaround, since
> such changes affect a large installed base.
Indeed. Changing Exim's default behaviour is not undertaken at all
lightly, for this very reason.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
