05 Apr 96 02:35, HECTOR SANTOS wrote to ANDREW CLARKE:

 >>>            Towhom Name+
 >>>            Fromwhom Name+
 >>>            Subject Name+
 >>>            Asciiz Date+
 >>>            NetworkEchoTag

 >>> Using this METHOD, at the very LEAST, we can ALL, with a HIGH degree
 >>> of success, support a unique id system in fidonet.  The ODDS are
 >>> extremely low that you will get a duplication crc.

 AC>> No, if I post two messages with the same To, From, Subject, Date
 AC>> into the same conference, containing different message text,
 AC>> according to your CRC that will generate the same MSGID serial
 AC>> number.

 > and what are the ODDS that two people with the SAME NAME will write a
 > message to the SAME PERSON at the SAME TIME with the SAME SUBJECT in
 > the SAME ECHO?

It's not uncommon for folks to repost messages that are/were identical to
their original post in the same conference.  Based on your algorithm you'll
get the same MSGID, which is obviously what we want to avoid.

 >>> We can probably add another factor:

 >>>            Towhom Name+
 >>>            Fromwhom Name+
 >>>            Subject Name+
 >>>            Asciiz Date+
 >>>            NetworkEchoTag+
 >>>            OriginatingAddress

 AC>> ???  But the originating address is included in the MSGID anyway.

 > Because this allows you to throw in another FACTOR into your DUPE
 > CHECKING design.  If by chance the TO/FROM/SUBJECT/DATE/ECHO is the
 > same as you say, the ADDRESS, indeed will break the duplicity.  If the
 > ADDRESS is the same, well, we do indeed have a DUPE!

Which is why the originating address is included in the MSGID, so there is
absolutely no point whatsoever to include the originating address in your
algorithm.  The originating address is already represented elsewhere.

By the way, what's with all the SHOUTING, HECTOR?


Andrew

-- randy{at}snoopy.virtual.net.au

--- Msgedsq/2 3.30