Will tree-ssa be GCC 3.5?

[Robert Dewar]

> It is not a research project.  Will you stop claiming that all the time, your 
> remarks damage the credibility of the project.  Do you really believe that so 
> many work full time on research projects?

I think Richard uses the term research project as we would use it 
internally at ACT, to refer to a project that is developing new 
technology intended to become part of the deployed product, but not 
there yet. So perhaps you are over-reacting here.

> Just because Ada hasn't caught up with the rest of GCC doesn't mean that the 
> work done for other languages is not intended for a production environment.

The GNAT front end will likely move in this direction, but it is 
impossible to give time scales for that right now, since it very much 
depends on priorities of our customers. I certainly think it is 
desirable if the tree-ssa work develops successfuly enough to be part of 
the mainline GCC.

By the way, I don't see it as such a big deal if this means that GNAT is 
  in a situation temporarily of not working with the head. This is 
likely to be corrected in a reasonable time anyway. I don't think that 
should be a determining factor in the development. If tree-ssa clearly 
shows an advantage for C and C++, I would go ahead and deploy it ASAP, 
and let other languages follow at their leisure, or faster if someone is 
interested enough!