Re: application/fs vs. fs/* (was Re: Transfer Encoding for MIME)

[Pete Resnick <presnick@xxxxxxxxxxxx>]

On 9/11/98 at 5:07 PM -0400, Nathaniel Borenstein wrote:

> Am I missing something, or are we spending more time arguing over the
> name of this thing than its design?

No, I don't think so.

> A general container for OS-independent files is a GREAT and CHALLENGING idea.

Right.

> Putting "fs" in the top level allows us to punt all the hard decisions to
> the second level.

No. That's where the argument is. MIME already has a great general 
container structure, i.e., the multipart top-level type. It also currently 
has a great way of representing things in that container structure which 
are files, i.e., the Content-Disposition with its assorted parameters. 
What's missing is a profile of the top-level multipart type which is 
specific to file system containers. That's where multipart/fs would be 
reasonable; it's a profile of our already existing general container for 
OS-independent *files*, just like multipart/digest is a container for 
messages.

What the proposed fs top-level type serves to do is introduce yet another 
general container type into MIME, only this time instead of being fully 
general, it can only hold files instead of other things, and it has no 
other internal relation to MIME structure.

> If that's the goal of the exercise -- i.e. to permit the gradual
> *evolution* of file system wrappers -- then a top-level name might make
> sense.  Otherwise, if we're really trying to standardize on a  single
> one for all time, I'd argue for application/fs.

Right. That's what fs is all about, so I think application/fs is the only 
justified type for what Al wants.

pr
--
Pete Resnick <mailto:presnick@xxxxxxxxxxxx>
QUALCOMM Incorporated
Work: (217)337-6377 or (619)651-4478
Fax: (217)337-1980 or (619)651-1102