Hallo IVAN!

 IT>   I wonder, is gcc going to be abandoned and superceded with the new
 IT>   egcs? 

Probably not.

 IT> What's the deal with those two paralel environments?

As far as I have understood, the deal is the following:

egcs is the *E*xperimental *G*NU *C*ompiler (err, what the heck is the S
for? ), with a stress on "experimental". This means that
there are new versions of egcs every few weeks, that they include new
features (especially in the C++ area) as soon as possible, etc. This is
good if you are a C++ programmer and really require these features to be
available very fast. The C++ specification still is a moving target, and
this is why you need a compiler like egcs which makes every new standards
(and also new experiences of the compiler makers on how to implement the
things right) available as soon as possible.

However, a compiler that is updated every few weeks is a bad idea if you
work on a larger, complicated or commercial project where you need to rely
on the compiler's behaviour. "To rely" in this case also means to
know what it's bugs are and how to live with them. You can't do this with
egcs, you need a more stable compiler. And this is what gcc 2.8 is for. It
also has a lot of more features in the C++ area as compared to gcc 2.7, but
it has only incorporated those things that are known to work reliably.

Therefore, gcc will continue to be the compiler where stability is a
premium requirement. The linux kernel is an example, or in general public
code that must be shared among lots of different architectures that should
all have the same compiler available. Egcs will continue to be
"experimental", and I think that all features that proove to be
good and that reach a certain degree of maturity will be incorporated in
gcc 2.9.

If any of the two is killed of, I rather expect it to be egcs, because it
will no longer be necessary as soon as the C++ standard is no longer
evovling and the compiler makers have learned how to implement C++
properly.

Viele Gr��e,
Tobias

--- Msged/2 TE 06 (pre)