This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Beyond GCC 3.0
- To: gcc at gcc dot gnu dot org
- Subject: Beyond GCC 3.0
- From: Robert Schweikert <rjschwei at mindspring dot com>
- Date: Sat, 30 Jun 2001 07:06:36 -0400
A users $0.02 worth.
Having convinced a number of people that GCC is a good choice as a
compiler to use for a large commercial app with millions of lines of C++
code and lots of template use I think that a 6 month release cycle is
too frequent, for "major" releases.
With "major" release I mean releases where features are deprecated or
use of features changes. All these types of changes in the compiler
require changes in the user code, which in the end costs a company time
= money. No one will be happy about that.
In an ideal world from a customer point of view there would be a
predictable gcc release cycle which produces "major" releases every 12
to 14 month. In these "major" releases anything goes, new features,
feature deprecation, feature use changes, etc. Within the 12 to 14 month
between "major" releases there should be 3 - 4 bug fix releases. These
bug fix releases do NOT make any changes to features, command line
arguments etc. These should strictly be bug fixes, no effect on how the
user uses the compiler. The only thing users should have to do when
upgrading to a bug-fix release is to remove any work arounds in their
code if they implemented any of those.
Thanks,
Robert
--
Robert Schweikert MAY THE SOURCE BE WITH YOU
rjschwei@mindspring.com LINUX