Re: "prefix" area in multipart message

[lau@xxxxxxxxxxx (Vincent Lau)]

> We can define it whichever way we want.  My preference would be to say
> that the first encapsulated part has no headers, and a body which starts
> with "Content-Type...".  This may not be what the author intended, but I
> think it would be legitimate to clearly define the multipart spec to
> work this way.
>

It strikes me that we just create an exception to deal with a simple
problem.  That is, the required field "Content-Type:" in each body part
cannot exist in the first body part.  Why should we treat the first
body part differently?  Besides, I am not comfortable to have 2 ways to
present the same thing: use the optional prefix area, or use an
encapsulation body part for the first text contents.

Also, if we allow optional prefix area with default "Content-Type:"
(i.e.  no headers), it is up to the 822-X.400 gateway implementor to
"guess" if there is an empty IA5 body part or there isn't an IA5 body
part during the mapping process.  Would any X.400 folks like to comment
on this?

-Vincent