it's nitpicking time again.
I looked at the definitions for the "subtype" of the MIME type/subtype
tuple. It's a TOKEN, and a TOKEN is 1-many of any CHAR except SPACE,
CTLs or tspecials.
So, the following characters seem allowed:
A-Z a-z 0-9
! # $ % & ' * + - ^ _ ` | ~ (ASCII - tspecials)
So, an example of a legal subtype would be
That's very nice for demonstration, but maybe we would like to tighten
this up a bit for the next revsision?
I personally would be happy with A-Za-z0-9, - and +.
I believe it is useful to have some allowed characters
outside base64 in the subtype. RFC 1345 uses some of the above
characters for specifying a MIME subtype.
If anything, I would restrict it to invariant ISO 646.