Re: rel_ops issues

[Phil Edwards <pedwards at disaster dot jaj dot com>]

On Fri, Mar 30, 2001 at 12:31:06AM -0800, Nathan Myers wrote:
> On Thu, Mar 29, 2001 at 10:57:27AM +0200, Theodore Papadopoulo wrote:
> > I have been hurt recently by the problem he exposes, and that was 
> > just because the source file (an old one) was including <map.h>
> > instead of <map> !!!
> > 
> > Deducing that it was this map.h stuff took me a lot of time ...
> 
> It seems Theodore would have been better served by a compiler that
> didn't provide map.h unless he explicitly asked for it (e.g.  by 
> '#include <backward/map.h>'
[...]
> They _are_ in a separate directory: backward/ -- at least in the
> source tree.  Have they been merged into a common directory in
> your installation,

Users /used/ to have to write <backward/foo.h>, but now they are merged
into a common directory in all installations.  See

    http://gcc.gnu.org/ml/libstdc++/2000-12/threads.html#00137
    "install backwards headers in prefix"

and the following thread.

Phil

-- 
pedwards at disaster dot jaj dot com  |  pme at sources dot redhat dot com
devphil at several other less interesting addresses in various dot domains
The gods do not protect fools.  Fools are protected by more capable fools.