Hi,

On Thu, 25 Mar 2004 12:45:35 -0000 "John Hall" <[EMAIL PROTECTED]> wrote:

> "Bob Apthorpe" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
> 
> > The Outlook missing-Message-Id problem really needs to be fixed because
> > in this case the RFCs are very clear that 'SHOULD' means 'SHALL.'
> 
> It means "that there may exist valid reasons in particular circumstances to
> ignore a particular item, but the full implications must be understood and
> carefully weighed before choosing a different course".

Yes, and (to me at least) "preventing trivial information leakage by
dropping the Message-Id header entirely rather than by simple,
innovative coding in order to satisfy a minority of our customers, used
as a default setting with no apparent way to revert to previous,
expected behavior" is not a valid reason, the implications of which were
certainly not carefully weighed.

This is not an anti-Microsoft screed. We have RFCs for a reason, in this
case for defense-in-depth for debugging mail problems and fundamental
interoperability. Running roughshod over the RFCs doesn't scale; what
happens when <your favorite MTA vendor> decides that "the MTA SHOULD add
a Message-Id to messages that lack them" translates to the action "I
don't feel like coding this because only b0rked MUAs emit mail without
Message-Ids?" Vendors and customers can point fingers all day long; the
upshot is that if all vendors complied with the letter and spirit of the
appropriate RFCs, there wouldn't be a problem.

The RFC is crystal clear that it's primarily the MUA's job to add a
Message-Id. Now if Microsoft wants to let their customers break that,
they should provide that as an option that is off by default. The vendor
should leave it to the customer to decide whether to deviate from RFCs,
but otherwise comply. As it is, there's no customer choice. Users can't
fix this problem even if they want to.

Regardless, if this problem is going to be solved, it won't be solved
here.

-- Bob

Reply via email to