https://bugs.exim.org/show_bug.cgi?id=2278
--- Comment #5 from Simon Arlott <bugzilla.exim.simon@???> ---
(In reply to Phil Pennock from comment #2)
> Per the RFC, `b=;` and `b= ;` at the end are equivalent and anything
> assigning semantic value to the whitespace there, and failing validation, is
> broken.
>
> There's special text in 3.5 around the b= tag in this header for signing, so
> it's understandable that it's handled specially, but if the issue you
> highlighted is the root cause, then Exim is fully spec compliant and the
> verifiers are not.
The verifier is Exim, and the problem is that Exim's signing process should not
have a space in there when it computes the hash, because it won't be there when
verified. The specification states the "b=" must be an "empty string" (but this
is not explicitly specified as "no whitespace").
--
You are receiving this mail because:
You are on the CC list for the bug.
