On Fri, 21 Nov 2014, Nicholas Boel wrote to GLENN JARVIS:

 NB> That looks much better, although I'm unsure about the EID:1A45
 NB> kludge you have up there. Was that an old kludge that MSGID
 NB> replaced? (not asking you, just throwing it out there for some of
 NB> the Fidonetters of old).

i can't say that MSGID replaced EID but it did come out as most implemented and
widely used... MSGID/REPLY is/was FTS-0009, 1991 Dec 17... i don't remember
when it was first proposed, though...

the only name i have associated with FSC-0031 is in the description in the my
filebase... M(ike) Ratledge... without that, there'd be nothing to indicate who
authored the document... several software packages did implement FSC-0031 and
numerous ones recognized it for linking but preferred and emitted MSGID/REPLY
lines... here's a snip from FSC-0031... 

===== snip =====
FSC-0031                                            May 1, 1989



          EchoMail ^aEID: Dup-Checking with Linked Replies
      A Proposal To The FidoNet Technical Standards Committee



Currently, no universal methodology for implementing echomail
duplicate message checking exists.  One thing is certain - they
will only increase in number as the shear volume of echomail is
increasing every day!

In order to catch the highest percentage of duplicates possible
it is desirable to utilize a system which actually tags each of
the messages themselves with a distinct messages identifier to
be used to check against an existing database of all previous
messages' identifiers.  In practice, this is not possible, but
we can limit the number of previous identifiers kept so that
processing is quick but still almost certain to eliminate any
duplicate messages.


This also provides an easy method of linking replies to their
original message by appending the previous identifier.  Using
a linked reply technique allows easy relinking of the messages
to the original message, assuming it still exists.
===== snip =====

further in the document actually lists a method of calculating the ID which is
a used for dupe detection... reading over it, i can easily see some problems
with it but it works for those systems that use it... i have no idea on the
quantity of false dupes they may detect, though...

)\/(ark