"Yannick Duch�ne (Hibou57)" wrote in message news:op.ww3ze4ejule2fv@cardamome... >Le Tue, 14 May 2013 21:02:32 +0200, Randy Brukardt a �crit: >> If the prototype implementation were to show only minor >> incompatibilities when processing existing Ada code (including the ACATS, >> AdaCore test suite, > >AdaCore test suite? If that exist, that must be not public. Right, it's not public (I'm told it includes propriarty code of AdaCore customers). My point is that a prototype would have to perform well on any and all existing Ada code (whether it knew about it ahead of time or not) in order to "pass". It's not just about passing public test suites. Randy.