>>> I think using a Windows-based boundary
will be firmer and less
ambiguous.
>>> I agree, and that seems to be the most
contentious part of this whole
definition of a "classic computer."
Speaking of which, that gave me an idea .... How about we in the
anti-10-year camp compromise by adding this clause to the definition of
what's classic: "IF it's post-1981 (when IBM launched the PC), THEN it must
have been unique when it was new." That way you include all the
first-of-breed stuff (i.e., the PS/2), you include all the exceptions to the
mainstream (i.e., the iMac), and you exclude all the same old PC crap.