gcc 2.98 release better approach?

Robert Schweikert rjschwei@mindspring.com
Wed Oct 18 17:55:00 GMT 2000


Well,

I think breaking binary compatibility only once is a noble idea and
certainly worth pursuing. The problem is that this causes a long time
with no official release and leaves people in a vacuum or constant state
of working around bugs. I have now been fighting template instantiation
problems for a while and I am not sure if I'll ever get all the kinks
worked out.

I am not advocating a release prior to gcc3.0 but what I would like to
see is a list of issues being addressed in 3.0, especially concerning
the C++ compiler. As everybody will probably agree with me the handling
of templates leaves a lot to be desired for and work arounds and tricks
to make them work correctly are quite painful at times. From my point of
view the following would help a lot:

1.) Development list.
      This list should be updated when a project is complete and the
snapshot number that has the new code or fixed code
      should be posted . This way people who are waiting for certain
features/fixes can try the new snapshot and provide
      feed back whether it works or not.

2.) An approximate release date. Without knowing whether bugs currently
in gcc will be fixed for 3.0 and without knowing
      when gcc 3.0 will be released it is difficult to decide whether
work arounds for bugs are worth the effort or not.

Just my $0.02 worth.

Have fun,
Robert

--
Robert Schweikert                      MAY THE SOURCE BE WITH YOU
rjschwei@mindspring.com                         LINUX





More information about the Gcc mailing list