At 01:31 PM 11/19/99 -0500, Nathaniel Irons wrote:
I want to allow persistent links to specific messages within archives
that may be unpredictably rebuilt. There was a useful discussion about
this some time ago:
<http://www.xray.mpe.mpg.de/mailing-lists/mhonarc/1998-07/threads.html#00088>
To my knowledge, little has come of it.
Fortunately, MHonArc's outrageously useful Digest::MD5 support provides
the unique handles required to make an abstraction layer pretty easy.
I'm writing to inquire as to whether this has perhaps been done already,
or if I've left anything important out.
I was planning to postprocess the OUTDIR directories, mapping current
message-IDs to filenames, and adding something along these lines to the
message files:
Want to create a <a
href="http://mysite.net/getdoc.cgi?$MESSAGE-ID>permanent > link to
this message? > >I was also planning to use DBI/MySQL, but Earl points out
in the above >thread that DB_file could be just as good, and a whole lot
easier to >possibly incorporate into the mhonarc codebase. > >If the idea
is basically sound, I'd like to request: > >a) a resource specifying the
host of the made-up message-ids > as something less alarming than > link to
this message? > >I was also planning to use DBI/MySQL, but Earl points out
in the above >thread that DB_file could be just as good, and a whole lot
easier to >possibly incorporate into the mhonarc codebase. > >If the idea
is basically sound, I'd like to request: > >a) a resource specifying the
host of the made-up message-ids > as something less alarming than
link</a> to this message?
I was also planning to use DBI/MySQL, but Earl points out in the above
thread that DB_file could be just as good, and a whole lot easier to
possibly incorporate into the mhonarc codebase.
If the idea is basically sound, I'd like to request:
a) a resource specifying the host of the made-up message-ids
as something less alarming than "@NO-ID-FOUND.mhonarc.org"
(patching mhamain.pl seems undesirable)
b) a resource variable to reference the message-ID within the
message layout.
c) feedback.
thanks!
Rather than a Message-ID header with a MHonArc conventional hostname, I
would recommend that you [Check in the work of Jacob Palme on header usage
and MIME extensions for a real header that fits an if that fails] use
[X-]Archive-ID: header with the actual host from which you are assigning
the unique-part in the mbox syntax for the ID.
The point is to continue the pattern where Resent-ID: the header for a
mailing list manager which sends forward a message lacking a Message-ID:
from the originating node.
This gives the processing software a search-list of Message-ID, Resent-ID,
Archived-ID with which to satisfy the need for a Unique-ID. Compare with
the abortive introduction of the MS-tnef correlator attribute.
The point is that for ID usage on email messages you should try to network
with the people who worked on MIME and DRUMS working groups in IETF. You
might even start a thread on <uri(_at_)w3(_dot_)org> because the logical
extention of
what you are doing in the MHonArc index is to support the same search-list
in satisfying resource requests expressed as <mid:...> URL. In other
words, the highest and best solution to your problem is a pragma applicable
to mid: URLs.
Al
-nat
