On Tue, 21 Jun 1994 19:46:00 EDT, you said:
advice and should be removed. The value of MIME-Version: is never
going to change and readers should not waste code checking for its
existence, much less verifying its value.
Hmm. Can you cite chapter-and-verse where the Almighty scribed that
it shall not change?
I'm sure there are people out there who don't check the value of the
version field in the header of an IP packet. These people are in for
a rude awakening when IPng is released.
The requirement in section 5.2:
Any characters outside of the base64 alphabet are to be
ignored in base64-encoded data. The same applies to any
illegal sequence of characters in the base64 encoding, such
as "====="
contradicts the previous paragraph and is onerous to implement. For
example, metamail botches this requirement completely. Most of my
implementations botch it as well. I am aware of no implementations
that expect this requirement to be followed.
I see no contradiction here. You hit the first '=', and by the previous
paragraph you can claim "all done". Therefore, you can safely ignore the
following '=' signs, satisfying both paragraphs. You're ignoring the second
and following '=' *both* because it's illegal, *and* because you already hit
the end of logical input and can wrap it up.
The second part of your statement parses (at least for me) as:
"Nathaniel blew it in metamail, I blew it in my code, but nobody cares
because no implementations actually *put* illegal chars in there for a
broken implementation to trip over".
If anything, this tells us that we got the protocol designed well, since
you can botch it badly and not cause a problem.
Valdis Kletnieks
Computer Systems Engineer
Virginia Tech