Re: Unified C and C++ front end (was Re: New C parser [patch])

["Joseph S. Myers" <jsm@polyomino.org.uk>]

On Mon, 25 Oct 2004, Matt Austern wrote:

> Another useful summary, "written in support of the view that
> C/C++incompatibilities can and should be eliminated," is the following series
> of articles:
> http://www.cuj.com/documents/s=8011/cuj0207stroustr/
> http://www.cuj.com/documents/s=8010/cuj0208stroustr/
> http://www.cuj.com/documents/s=8009/cuj0209stroustr/

This looks rather like a follow-on to Stroustrup's previous attempts on 
the c++std-compat list in June 2001.  The general reception from the C 
side was rather unfavorable.  People don't think that incompatibilities 
are good in themselves (though maybe some people do) but that doesn't mean 
the subset language route is *now* (rather than 25 years ago) a good route 
to go for either language.  (And one view expressed was that there are too 
many kludges for the sake of compatibility in both languages and it would 
be better to let the languages go their own separate ways without needing 
such kludges.)

I do think there is a use for a common core machine model covering such 
things as sequence point rules and exactly what type and size of object is 
relevant for each memory access and when such an access is valid - hard 
questions which need addressing for both languages and where the existing 
definitions are sufficiently unclear that it cannot be said there is a 
clear incompatibility which existing code may rely on.  And some existing 
incompatibilities are probably susceptible to being fixed in future 
standards without too much incompatibility with real code - but we do have 
to live with the fact that for existing standard versions e.g. the rules 
for lexing UCNs differ.

-- 
Joseph S. Myers               http://www.srcf.ucam.org/~jsm28/gcc/
    jsm@polyomino.org.uk (personal mail)
    joseph@codesourcery.com (CodeSourcery mail)
    jsm28@gcc.gnu.org (Bugzilla assignments and CCs)