ietf-822
[Top] [All Lists]

Re: gzip-8bit

2003-03-05 08:15:21


on 3/3/2003 1:06 PM ned+ietf-822(_at_)mrochek(_dot_)com wrote:

And all this borders on the more general issue of how to represent
complex package formats in MIME, especially those that store
information about their content that MIME cannot. These are always
judgement calls.

One thing that I seem to recall having been discussed at some point was to
send compression formats as multiparts. The first part would contain a
catalog of the contents (possibly with offsets or some other applicable
reference point), the second part would contain the compressed entity.
This kind of model would make it feasible (in theory anyway) to structure
the first part as references to subordinate remote entities.

The feasibility of this is of course dependant on several things, not the
least of which is the ability for the compression format to provide
meaningful pointers. Another issue is that you don't want clients thinking
that they can download just one of the subordinate entities, unless such a
feature is specifically supported (eg, having an IMAP server extract the
referenced entity and returning it alone, on demand).

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/


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