Why not gnat Ada in gcc?

[Robert Dewar]

<<Actually, you should look at the true history of the new ia32 port.  While the
initial work was done internally under NDA.  Eventually the development was
moved externally onto a branch so that everyone could collaborate in an open
way.  Once the various parties were happy with the state of the work, that
work was moved into the head of the gcc tree.
>>

I know EXACTLY the true history here (for several reasons I am aware of 
this history). Yes, things were fine once the external move was made, but
it is the previous stage, where "various parties" exchanged copies under
NDA that was dubious (I use a gentle word here :-)