That's pretty interesting....I hope I'm the only one surprised by this
requirement.
This requirement is well understood by those of us who work on email software.
Any package that fails to support it gets in trouble sooner or later. Indeed,
software that preserves case but isn't capable of supporting case sensitivity
for domains it has administrative control over can be problematic. I've seen
software sales won or lost over this issue.
At the end of the same paragraph in RFC 2821 is the
"loophole" that I think most people will invoke to get around this new
"requirement":
...However, exploiting the case sensitivity of mailbox local-parts impedes
interoperability and is discouraged.
I anticipate that this means, for all practical matters, that the "MUST"
part of case sensitivity will be ignored by enough vendors that we
should plan on treating the mail local-part as case insensitive anyhow.
Personally, I think it makes sense to treat mailbox strings as case
insensitive, but that's just me speaking.
The rule is fairly simple: If you have administrative authority for the domain
you are entitled to use the local-part in any way you like, including but not
limited to igoring case. If, on the other hand, you do not have administrative
authority, you are not entitled to do this. And if you do, know that there real
domain exist where case matters and things will break.
Ned