Re: [exim] Emails starting with a byte order mark (BOM)

On 04.03.2014 16:26, Todd Lyons wrote:
> On Mon, Mar 3, 2014 at 12:53 PM, Michael Fischer v. Mollard
> <info@???> wrote:
>>> But nevertheless it might be a good idea to block 8 bit characters in
>>> header names via an ACL test as even RFC 6532 does not allow that.
>> Hello,
>> I attached a patch which implements a „verify = header_names“ test as
>> suggested above. At least it should be useful to users of cyrus imapd as
>> cyrus won't accept such mails.
> I like the patch because it's simple and straightforward. I think the
> chosen "header_names" is a bit too vague for what it actually does.
> You're checking to make sure that the characters in the header name
> are ascii, whereas someone could interpret it or glance at it and
> think that it checks if the header names are valid, not just the
> characters in them.
>
> I would like to consider a different, more specific, name for it,
> based on what it is actually checking, maybe something like:
>
> header_names_ascii (I'm partial to this one)
> header_names_7bit
> header_names_not_8bit
> header_names_rfc6532 (though this could be more than just 8bit issues)
>
> Do you have any other suggestions you think are better?
>
No, I think header_names_ascii is best. I don't like the negation in
'verify = header_names_not_8bit', and RFC 6532 in a name would be
misleading as there is no support for RFC 6532 in exim. I mentioned it
only because 8 bit header names will still be not conforming even when 8
bit characters in header data will be OK in some more or less distant
future.

Michael