Re: c++/70

[Martin Sebor <sebor at roguewave dot com>]

The following reply was made to PR c++/70; it has been noted by GNATS.

From: Martin Sebor <sebor@roguewave.com>
To: rodrigc@gcc.gnu.org
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: c++/70
Date: Tue, 29 May 2001 12:10:00 -0600

 rodrigc@gcc.gnu.org wrote:
 > 
 > The following reply was made to PR c++/70; it has been noted by GNATS.
 > 
 > From: rodrigc@gcc.gnu.org
 > To: gcc-gnats@gcc.gnu.org, jgeremia@princeton.edu, wanderer@rsu.ru,
 >   martin@loewis.home.cs.tu-berlin.de, nobody@gcc.gnu.org
 > Cc:
 > Subject: Re: c++/70
 > Date: 29 May 2001 02:31:42 -0000
 > 
 >  Synopsis: template bug
 > 
 >  State-Changed-From-To: analyzed->closed
 >  State-Changed-By: rodrigc
 >  State-Changed-When: Mon May 28 19:31:42 2001
 >  State-Changed-Why:
 >      Not legal C++.
 
 I don't know of any rule that makes this ill-formed (or not legal),
 even though referencing the operator may be ambiguous. operator*()
 along with many other binary operators may be defined either as a
 member, or as a non-member, or both (13.5.2).
 
 The simplified test case below compiles (with the friend being defined
 inline), which confirms that there is a bug (most other compilers also
 accept the code in the original test case).
 
 Regards
 Martin
 
 template <class T>
 struct S;
 
 template <class T>
 void operator* (S<T>, S<T>);
 
 template <class T>
 struct S
 {
     friend void operator*<> (S, S); // { }   // okay
     void operator* (S) { }
 };
 
 // template <class T>
 // void operator* (S<T>, S<T>) { }   // error? not!
 
 S<int> s;