Content-Type: application/gzip;
uncompressed-type="text/plain; charset=us-ascii"
This scheme does have some drawbacks when compared with a new c-t-e. For
example, it's more difficult to support multipart/alternative properly.
Why so?
Because MIME mail readers would then have to know a list of content-types
which are really encodings, so that they could look at the uncompressed-type
parameter to tell the multipart/alternative guy what the type really is. And
since the encoding content-types could be nested (ugly but doable) the mail
reader would have to be able to search through several levels.
I also think this is a slippery slope towards having multiple layers
of encoding, which is a Bad Idea.
Keith