Any of us who have been involved in 'accident investigations' of any kind will have experienced the way things tend to develop on a 'double track.' It's not usually possible to reveal the 'whole truth' all at once - particularly in cases like this, where so many issues remain unresolved. But the 'lower orders' - in government, among the manufacturers, and in the airlines - will always make every effort (though quietly) to reduce the likelihood of anything like this happening again.
In this connection I find this article interesting:-
"Europe's safety authority is to order Airbus A330 and A340 operators to upgrade flight-control computer software to prevent autopilot engagement should airspeeds become unreliable.
"Autopilot and auto-thrust on the types will automatically disconnect, and the aircraft will revert to alternate law, in cases where significant differences emerge between the airspeed sources.
"This condition can arise if, for example, the pitot system becomes contaminated with ice.
"Previously the European Aviation Safety Agency warned that, under such circumstances, it was possible for two airspeed sources to show similar - but nevertheless erroneous - data.
"This would permit the crew to re-engage the autopilot while the airspeed data was unsound, potentially resulting in the autopilot transmitting unexpected commands to the flight-control system - such as an abrupt pitch-up.
"EASA had cautioned pilots, in a December 2010 directive, to resist re-engaging the autopilot until they had carried out a cross-check of speed indications to ensure that the airspeed data was reliable.
"Since that directive was issued, new primary flight-control computer software has been developed which, EASA states, will "inhibit autopilot engagement under unreliable airspeed conditions"."
A330/340 change to inhibit autopilot if airspeed unreliable
As an additional contribution to the debate, a mate of mine with aviation connections (in Indonesia) just sent me this - which appears to be a first, unofficial, English translation of the AF447 FDR content combined with the CVR transcript. I can't vouch for its authenticity - except that, as far as my knowledge of French goes, the words seem largely to correspond with the French transcript already published. Only thing is, the link misbehaves - people may need to 'zoom in' to about 400% to be able to read the text:-
http://www.globalsim.web.id/publicse...uly2011v10.jpg