On 2004-12-22 "Odhiambo G. Washington" <wash@???> wrote:
> * Andreas Metzler <eximusers@???> [20041221 22:47]: wrote:
>> This is <http://bugs.debian.org/286721>
>> ---------------------
>> On 2004-12-21 Marc Sherman <msherman@???> wrote:
[...]
>>> exim 4.34 does not support submission mode (it was added in 4.40).
>>> However, adding "control=submission" to an ACL in the 4.34 config file
>>> does not cause an error at startup (or with exim -bV). The failure is
>>> only noticed when incoming mail hits the offending line in the ACL, and
>>> is deferred.
[...]
> Sorry to jump into this, but the statement Mr. Andreas has responded
> to seems weird to me, if anyone ever did that. Why? It says
> 1. Exim 4.34 does not support submission mode (it was added in 4.40).
> 2. However, adding "control=submission" to an ACL in the 4.34 config
> file ...
> Now I am wondering: If you know so well that submission support mode
> is only a feature of 4.40 and above, why try it in 4.34??? Are we
> expecting Philip to address that or what is the point? ;)
Hello,
Marc did _not_ know that 4.34 did not support "control=submission"
when he added the statement (he read the docs for 4.40) and would have
liked exim to immediately kick him with "WTF, syntax error. I do
not know knothing about control=submission?" instead of only doing
this when evaluting the ACL ar RCPT TO:.
It is not about exim developing "the sight" but about exim detecting
which statements are valid at startup. This would be useful not only
for this specific case but also for detecting typos.
cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"
http://downhill.aus.cc/
