Excerpts from ext.ietf-822: 29-Oct-93 Re: Query on a nit: Applica.. John
Gardiner Myers(_at_)cmu(_dot_) (474)
The important distinction between the application and text primary
types is whether or not it is appropriate to show the user the "raw"
version of the data if the type is not recognized. Wording to this
effect is in section 4, paragraph 4 and appendix A requirement 4 of
RFC 1521.
It would seem to me that HTML really does belong under the "text"
top-level content type.
This, of course, is where the rubber meets the road. Having looked at
much HTML source, I claim it would frighten and confuse naive users with
naive mail tools, and therefore should *not* be a subtype of text.
Putting things like this (and the old richtext) under type `text' is
exactly what breaks down the top-level type boundaries. But that's OK:
I don't believe the top-level type `text' is in fact a useful thing.
Bill