Excerpted from message dated 09-24-97, Frank Sexton to Murray Lesser:
 ML> I don't think there ever were a set of "official AT
 ML> specs" - at least not published by IBM.
FS>How was IBM able to build more than one machine the same if there
  >were no specs?  I thought they had XT specs and AT specs that they
  >went by and that these were eventually adopted by "clone" makers.
Hi Frank--
    Strictly speaking, you are correct.  Most certainly, IBM had/has
internal specs for their products.  What I should have said: IBM never
published the full PC/AT specs externally, except (probably) as
non-disclosure documents for patent licensees.  Most of the clone makers
were not licensed, and cobbled up "IBM Compatible" PCs from the external
specs available to anyone.  The clone BIOS makers used "clean room"
techniques to build a BIOS that met the IBM external interfaces, in the
belief that this protected them from infringing on IBM patents and
copyrights.  I have no idea whether or not it actually did protect them,
or if IBM eventually insisted that they become licensed.  I doubt if
anyone at IBM who really knows would tell you because this is the type
of information that IBM guards very carefully,
    Regards,
          --Murray

___
 * MR/2 2.25 #120 * Which direction is forward?
--- Maximus/2 2.02
---------------