BS>> BTW it's only if dupechecking should be out from msgid
 BS>> or msgid and header... i run only specified header there
 BS>> is no problem..

 BJ> Ah.....  Who / what program is generating the "dupes"?  The
 BJ> msgid standard that I think Squish / Maximus implemented has a
 BJ> problem if one tries to generate two messages (or more) within
 BJ> one (or was it two) second(s).

that is definitely a problem... i've a tool here that can generate several
hundred a second...

 BJ> Back when the design was done, it wasn't possible to generate
 BJ> messages that fast.  But with faster hardware, and with robots
 BJ> producing net and/or echo mail, it is fairly easy now to
 BJ> generate messages too fast to get unique msgid's.

 [...] that are based solely on time... those several hundred a second that
my system can generate all have unique MSGIDs and (IIRC) each node, up to
255, can generate up to 8000 per day before looping back around...

 BJ> And Squish's dupe checker, when MSGID checking is enabled, will
 BJ> toss the non-duplicate messages with the same MSGID as dupes,
 BJ> if I remember correctly....

fidonet's msgid stuff needs an overhaul just like the nodelist does...
something along the lines of a RFC message-id would be fine... its a proven
method and i'm not aware of any duplicates having been generated within the
time constraints in all these years...

)\/(ark