Interesting, three interpretations of the combined mid-cid URL, one for
each separator.
Larry - "/" A relative URL, "../cid:www@isi.org" doesn't
make sense in this context.
Al - "?" The mid:.../cid:... URL is really a search
operation.
Ed - "#" The cid: URL is analogous to a label in an
HTML entity. [I just rediscovered this ;-)]
Previously I thought that "/" was just as good as "#", but as Larry
Masinter points out it leads to a relative notation that is unneeded
and unnecessary. I disagree with Al Gilman that we should imply a
search, that's only one possible implementation. Given the
implications of the three separator characters the analogy to a label
seems the most appropriate.
Ed
On Mon, 27 Nov 1995 11:07:40 PST Larry Masinter wrote:
This argues strongly for a "mid:xyzy(_at_)wawa(_dot_)com/cid:wwww@ixi.org"
format
- using slashes to separate the components.
I don't see any positive value to using relative URLs of the form
("../cid:www@ixi.org") to refer to content-IDs when an absolute one
("cid:www@ixi.org") seems like it will do just as well in every
circumstance I can think of. In addition, there's a serious negative:
the same content-ID can occur in multiple contexts, and are actually
the same. If two different MIDs contain the same body with the same
CID, the CIDs are really the same, and
"mid:xyzy(_at_)wawa(_dot_)com/cid:www@ixi.org" creates an unnecessary
distinction.