This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: GCC version bikeshedding
- From: Ian Lance Taylor <iant at google dot com>
- To: NightStrike <nightstrike at gmail dot com>
- Cc: Jonathan Wakely <jwakely dot gcc at gmail dot com>, Eric Botcazou <ebotcazou at libertysurf dot fr>, Andreas Schwab <schwab at linux-m68k dot org>, "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>, Jason Merrill <jason at redhat dot com>, Jakub Jelinek <jakub at redhat dot com>, Richard Biener <richard dot guenther at gmail dot com>
- Date: Thu, 31 Jul 2014 16:03:35 -0700
- Subject: Re: GCC version bikeshedding
- Authentication-results: sourceware.org; auth=none
- References: <20140720165506 dot GT3003 at laptop dot redhat dot com> <201407292258 dot 29357 dot ebotcazou at libertysurf dot fr> <CAH6eHdR3dfbYin1LMWrSUJLsk7=iO6+orQXQmXdGRxzA71MKaA at mail dot gmail dot com> <201407310018 dot 05575 dot ebotcazou at libertysurf dot fr> <CAH6eHdRx+CE7TFvMW8JvAxThSAMz07e8ke90qHs8oOD4H94tjw at mail dot gmail dot com> <CAF1jjLsg32gtNOQN6_MPLSAAStQpCrwbM14MQLKqkcnqJO1W=Q at mail dot gmail dot com>
On Thu, Jul 31, 2014 at 4:52 AM, NightStrike <nightstrike@gmail.com> wrote:
>
> One thing you might want to consider is that with the typical X.Y.Z
> versioning of most GNU projects, changing X allows breaking
> compatibility in a significant way with previous versions. While Z
> fixes regressions and Y adds new features, X is a place to make
> infrequent but paradigm shifting changes that are unconstrained by a
> desire to stay backwards compatible with older values of X.
>
> By going to what is essentially a Y.Z.0 release mechanism, you lose
> that ability to some degree. Maybe that's ok in a mature project like
> GCC.
I believe the GCC project has become too large to be able to usefully
speak about breaking compatibility with previous versions. There are
too many different moving parts. We break compatibility in various
minor ways with every release. We try pretty hard to never break
compatibility in a big way. Historically, as far as I can recall, the
GCC major release number has never indicated a compatibility break
that was relevant to most users.
Ian