Re: [boost] Anoying borland warnings

From: David Abrahams (dave_at_[hidden])
Date: 2002-08-21 07:32:01

-w-8001 seems like a slam-dunk to me. Go ahead and patch borland-tools.jam
if you like.
-w-8080 seems a little bit less like a great idea. That warning can help to
keep flotsam and jetsam out of our code.

There are several ways to suppress unused variable warnings explicitly:

    (void)x; // works with most compilers

    template <class T> inline void unused(T const&) {}
    unused(x); // works with all compilers, AFAIK

-Dave

-----------------------------------------------------------
           David Abrahams * Boost Consulting
dave_at_[hidden] * http://www.boost-consulting.com

----- Original Message -----
From: "Gennadiy Rozental" <gennadiy_at_[hidden]>
Newsgroups: gmane.comp.lib.boost.devel
To: <boost_at_[hidden]>
Sent: Wednesday, August 21, 2002 1:37 AM
Subject: [boost] Anoying borland warnings

> It one will take a look on into compiler report it see a lot of Borland
> warnings like this:
> Superfluous & with function in function ...
> Borland think that & in front of function name is superfluous.
> Since nobody but Borland complains about this I added following to the
> requirement section on Boost.Test Jamfile:
> <borland><*><cxxflags>-w-8001
> It seems to suppress it. May be we should consider adding this to borland
> tools definition?
>
> Absolutely similar proposition about warning 8080
> <something> is declared but never used in <somewhere>.
> There are a lot of cases (in Boost.Test particularly) where it is
absolutely
> legitimate. To suppress it I use
> <borland><*><cxxflags>-w-8080
>
> or as combination
> <borland><*><cxxflags>"-w-8001 -w-8080"
>
> Comments?
>
> Gennadiy.
>
>
>
>
> _______________________________________________
> Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost