innd and rejects of dates

[Russ Allbery]

Julien ÉLIE <julien at trigofacile.com> writes:

>> The purpose of requiring relaying agents to look at Injection-Date
>> instead of Date was primarily to handle dates in the past, not dates in
>> the future, but I don't think it will cause harm there either.  Such
>> articles will be rejected by peers that don't understand
>> Injection-Date, but that's why nnrpd rejects such articles so that we
>> don't introduce bad articles.

> Which means that "full" support of Date: header fields in the past will
> not occur before some time!  At least, INN 2.5.3 (starting from
> tomorrow's STABLE version) accepts to relay these articles.  We'll see
> in another version for allowing the injection.

Yeah, that was a known issue when we wrote the standard.  Charles argued
for allowing a widely varying Date in the presence of Injection-Date from
the start, but I and others were concerned that this would cause
widespread silent disappearance of articles after initial acceptance since
almost no one had implemented Injection-Date.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>

    Please send questions to the list rather than mailing me directly.
     <http://www.eyrie.org/~eagle/faqs/questions.html> explains why.