RFC: should we use -Werror? (& sample patch to do it)

[Joseph S. Myers]

On Thu, 27 Sep 2001, Fergus Henderson wrote:

> Yes, please.  I've been wanting that for years.
> 
> The counter-argument is that inhibiting all warnings for a whole statement
> or a whole expression is dangerous, because it might inhibit warnings that
> indicate real problems.  However, what I end up doing to work-around the
> absence of __nowarn__ is inhibiting certain categories of warnings for my
> whole application, simply because one header file contains an occurrence
> of code which triggers such a warning and for which the warning can't
> easily be supressed.

We should also look at the previous discussion of designs for fine-grained
warning control (e.g. see http://gcc.gnu.org/ml/gcc/2000-06/msg00639.html
- linked to from projects/beginner.html but currently missing from the web
server as one of the 4492 files needed from the old disk listed in
http://gcc.gnu.org/ml/gcc/2001-08/msg01443.html ).

-- 
Joseph S. Myers
jsm28@cam.ac.uk