ietf-822
[Top] [All Lists]

Re: file attachments in MIME

1993-03-13 14:08:15
Nathaniel,

I am only just catching up of the flurry of messages and came across
your cogent summary of the apparent camps, multipart/attachment vs.
parametric addition to content (either via a -type parameter or
by content-disposition.)

I'll read on to the other opinions, but you solicited responses
to whether anyone objected to using the parametric model, given
that you view the multipart approach as fundamentally broken.

Unfortunately, I think that the parametric approach is limited in
the wrong ways.  To the extent that the information is 
display/dont'-display, the parametric model is just fine.

But I believe that the display decision is a side-effect from a more
basic descriptive requirement, which pertains to the structural
relationship of the message parts.  Janssen has been the proponent of
this view, I believe.

I believe that the underlying intent, here, is to distinguish the
"primary" part of the message from "accompanying" material.  Then,
one is supposed to display the primary part and merely cite the
secondary stuff.  I strongly suggest that we specify the marking
separately fron the handling of things so marked.

Dave

<Prev in Thread] Current Thread [Next in Thread>