RE: Warnings about rcs_id strings: let's settle this

[Mark Mitchell <mark at codesourcery dot com>]

On Fri, 2003-05-02 at 14:06, Kean Johnston wrote:
> > Is this code not then riddled with other old practices that generate
> > warnings, such as implicit int?  Certainly *BSD have needed extensive
> Trhat level of cleanup has been done. But we digress ...
> 
> > > What is the *HARM* in checking non-const

Joe Buck would very much like to see us do something here for the 3.3
release.

The best criteria for a patch going into 3.3 are that:

(1) It is as simple as possible.

(2) It restore behavior from some previous version of GCC.

For example, Zack's recent change to take sign-comparision warnings out
of -Wall in C is in this category; the patch is an obvious one-liner,
and the compiler didn't use to warn about these constructs in C before. 
Since the change wasn't conscious, it's not like we've debated the issue
and decided to add that warning to -Wall.  The fix is going to give us a
compiler that behaves "like GCC always did."

I will approve a similar patch for this issue, since Joe thinks it is
important.  But, for 3.3 please do not try to define new semantics, or
new switches, from previous versions of GCC.  Just make 3.3 do whatever
3.2 did; evidently, that was acceptable, even if not ideal, to many
people.

Thanks,

-- 
Mark Mitchell <mark@codesourcery.com>
CodeSourcery, LLC