Jethro R Binks wrote: > ..[snip]... >
> Slightly at a tangent to the main thrust of Tony's message, but related
> to these same blank messages. I was contemplating the wisdom of
> refusing a message that contains none of To:, Cc:, Subject:, Date:,
> From: headers.
>
> Recognising that Date: and From: are required by RFC2822, I can't see
> that the above would be an 'unsafe' test .. can anyone else see any
> potential difficulties?
>
Do not try and require the TO:, CC:, or Subject: headers, as this will
cause all kinds of problems. However, we do require that all messages
contain both a Date: and From: header (we do not check for content, just
existence, so a 'blank' From: or Date header is still ok). The only
problems
this has cause is with some web-based 'auto mailers'. For some reason, the
largest offenders being software designed around Oracle databases. Any
way,
just be sure that you can selectively disable the check if you need to, but
once there is a problem, I have not had any trouble in getting the
offending
software fixed so that both a Date: and From: header are included in the
messages. So, does it cause a touch of trouble, yes, but the trouble is
worth it in my case.
