On Tue, 14 Nov 1995, Einar Stefferud wrote:
> May I suggest that it might be a really neat idea to define a new MIME
> type to contain these machine readable fragements so as to make them
> machine discoverable and machine readable? If this is done, then the
> proposed delimiters become standard mime delimitors and we can
> eliminate a whole new namespace and class of delimiters. The object
> typing also becomes mime standard using content-type.
I'd like to strongly second this. MIME types for C code (as in the MD*
RFCs), DTDs, and even (yuckk-ptooey) ASN.1 would be quite useful...
> I consider such eliminations to be a great service to mankind;-)...
> Perhaps we should also consider creating an RFC Mime type, perhaps
> text/rfc or application/rfc. Then we might even define a way to
> provide more machineable RFC structures, without significantly
> sacrificing readability.
I'm not nearlly so clear on this. It has been very beneficial to have
the simplest lowest-common-denominator format for RFC. I think its
essential to be able to easily print and understand RFCs with the simples
text processors. But there may be a good way to demark MIME labeled
sections withing them. A small amount of design may be needed here to
take care of such trivial as being able to automatically excluse the
page header/footers that fall in the midst of otherwise machine
> >From your message Tue, 14 Nov 1995 15:18:57 +0100:
> }One possible such syntax is this: Delimit all segements of an
> }MPS within an RFC by a preceding line such as
> }vvv--- mps4567-1.txt #2 (DTD for HTML 5.0) ------------------vvv
> }and a following line
> }Here "mps4567-1.txt" is the name of the MPS (and the MPS file
> }that can be extracted) and "#2" is the fragment sequence number.
> }By using these delimiters, the MPS can also be protected from
> }interference with the page headers and footers in the RFC. The
> }only limitation on the content of a MPS this would mean is that
> }lines of length 64 starting with "vvv---" and ending with
> }"---vvv" and lines of the form of the terminating line above
> }would be impossible to use.
> }New acronym in this message: MPS = machine-parsable specification
> }Olle Jarnefors, Royal Institute of Technology, Stockholm <[log in to unmask]>
Donald E. Eastlake 3rd +1 508-287-4877(tel) [log in to unmask]
318 Acton Street +1 508-371-7148(fax) [log in to unmask]
Carlisle, MA 01741 USA +1 703-620-4200(main office, Reston, VA)