I think the suggestion from me and Tom (if I interpret his email correctly) is
to state that messages can be truncated at the end at an arbitrary point. We
also make a note that this may result in invalid UTF character encoding, or a
change in UTF character.
I don't think it even warrants a
I don't have a strong view on SHOULD v MUST v neither, happy with any of them.
My point was that this is UTF-8 which I see as an unfamiliar technology for
some, perhaps for many, so while the idea that truncating a message can change
its meaning I would expect to be obvious to all, the idea of
