At 18:21 20/02/04 -0500, Keith Moore wrote:
I like the idea.
As someone who'se made extensive use of the W3C list archive facility, I
heartily concur. It is a really valuable idea, and works very well in
practice.
It might be appropriate for the field name to start with List-*, at
least for those cases where the archive is associated with a list. That
way, it's easier to separate fields added by a list from fields supplied
by the sender.
No argument.
I'd also recommend that the link point to an archived copy of the
message in original form, rather than, say, one that is translated to
HTML. Translating to HTML causes a loss of information and potentially
a loss of functionality. You should be able to reply to an archived
message, refile it into a folder, follow threads, etc., but those
things are harder to do if the message is no longer in its original
format.
Hmmm... the way the W3C system works, the URL works by redirection (or
something) to the actual archived message. I guess that multiple formats
could be served at the same URI using existing Web protocols. (e.g. use
HTTP GET with Accept: message/rfc822 -- then it's up to the archive
implementation to determine if it honours or rejects that.)
The existing W3C system provides links for following threads, etc., using a
regular browser.
(Did anyone mention "running code"? ;-)
#g
--
Thinking about this I realized that one reason I'd like such a field is
so that I could, given a message, more easily find other messages in the
thread. After all, if I already have a copy of the message with the
archived-at field, why would I want to download it? I'm much more likely
to want to look at either the messages that preceeded or followed that
message.
As long as the field remains completely unstructured, I see no need to
support comments. I'd probably change my mind if the field were changed
to be, say, a list of URIs.
------------
Graham Klyne
For email:
http://www.ninebynine.org/#Contact