Why GCC 4?

[Michael Poole]

Matthew R. Dempsky writes:

> On Wed, Mar 15, 2006 at 07:13:06PM +0100, Mario Vukelic wrote:
> > However, if nobody ever moves to 4.x, it will never fulfill its
> > potential. I commend Ubuntu for taking a little pain now, to have less
> > tomorrow.
> 
> That's why Gentoo exists.  Let the ricers be the first casualties.

I somehow thought FUD was a better fit for proprietary systems than
collaborative ones like Ubuntu -- apparently you disagree.  If there
are technical issues with GCC 4, report them and use those statistics
instead of hand-waving to support your argument.

Speaking as a professional software developer, the GCC developers do
pretty amazing work in terms of diagnosing root causes for unexpected
behaviors and tracking whether problems return in the future.  It is
the nature of complex software that new releases will have new
bugs[1], but it is not as if Ubuntu packages are being compiled with
the original GCC 4.0 release; 4.0.3 is current, and 4.0.2 is six
months old.

Michael Poole

[1]- This fact should be taken with a grain of salt.  If you look at
the serious regressions in GCC 4.0.3, about half of them apply to the
3.4 series as well.  There are also many bugs that existed in 3.4 but
were fixed in 4.0, including serious standards compliance issues.