Why not gnat Ada in gcc?

[Richard Kenner]

    The way to manage this is to test the hell out of the tree when you
    bring in changes from the net *before* you check those changes into
    your internal repositories.  

All I was doing was explaining to those who don't already know why there
*have* to be internal respositories.  There was a suggestion that ACT should
use the public tree for its internal development purposes and I was pointing
out why that can't work.

    Please keep in mind that those issues are ACT's problems and should not be
    a consideration for how to manage the GNU Ada repository.

Of course.  Again, I was just explaining why any commercial company will need
their own internal tree and why merges are needed.  This is obvious to you,
but not to others.

    Any bulk merges need to be from the net sources to your internal
    sources.  Any other method would put ACT into a privileged position in
    regards to installing changes into GNU Ada.

I think the method of merges that will end up being the most appropriate will
strongly depend on the extent of changes being made in the net sources.  As I
say, a major difference between the Ada front end and that others is that
it's *far* less likely there will any other people who will end up making
changes.