[Coin-discuss] Warnings when compiling with Sun Forte 6u2 compiler

[Jonathan Eckstein]

Hello, COIN-ers:

We PICO people have been using COIN's OSI as an LP solver interface on 
various platforms.  I have had a lot of trouble getting the gnu 3.x 
compilers to work on my Sun, and have recently been using the Sun 
Workshop/Forte 6u2 compilers.

Using these compilers, we get the following warnings repeatedly:

"/home/jeckstei/alt-sandbox/COIN/include/OsiRowCut.hpp", line 154: 
Warning: OsiRowCut::operator== hides the virtual function 
OsiCut::operator==(const OsiCut&) const.
"/home/jeckstei/alt-sandbox/COIN/include/OsiRowCut.hpp", line 154: 
Warning: OsiRowCut::operator!= hides the virtual function 
OsiCut::operator!=(const OsiCut&) const.
"/home/jeckstei/alt-sandbox/COIN/include/OsiColCut.hpp", line 145: 
Warning: OsiColCut::operator== hides the virtual function 
OsiCut::operator==(const OsiCut&) const.
"/home/jeckstei/alt-sandbox/COIN/include/OsiColCut.hpp", line 145: 
Warning: OsiColCut::operator!= hides the virtual function 
OsiCut::operator!=(const OsiCut&) const.

We have been ignoring these warnings for over a year, but recently my 
student Konrad Borys brought to my attention a passage on page 235 of 
"Effective C++" (second edition) by Scott Myers, where he talks about 
this kind of situation.  The method

   inline virtual bool operator==(const OsiCut& rhs) const;

declared in OsiCut.hpp cannot be used on an object of type OsiRowCut, 
even though it inherits from OsiCut, because there is a method in 
OsiRowCut with the slightly different prototype

   OsiRowCut_inline bool operator==(const OsiRowCut& rhs) const;

Thus, if "obj1" is an OsiRowCut and "obj2" is an OsiCut, the expression

   obj1 == obj2

or certainly

   obj1.operator==(obj2)

should generate an error.

I am wondering if this design is intentional, or whether the Sun 
compiler is catching an unintentional problem the Gnu compilers are 
missing?  I quite possibly could be missing something, as I am not 
completely clear on some of the darker crannies of C++.

Thoughts?

   Thanks,

     Jonathan



-- 

Associate Professor Jonathan Eckstein
MSIS Department, Rutgers Business School, Rutgers University

   TEACHING ADDRESS                 RESEARCH ADDRESS
+------------------------------+--------------------------------+
| 255 J.H. Levin Building      |   RUTCOR, Room 155             |
| 94 Rockafeller Road          |   640 Bartholomew Road         |
| Livingston Campus            |   Busch Campus                 |
| Rutgers University           |   Rutgers University           |
| Piscataway, NJ 08854 USA     |   Piscataway, NJ 08854 USA     |
| (732) 445-0510               |   (732) 445-3272               |
| FAX (732) 445-6329           |   FAX (732) 445-5472           |
+------------------------------+--------------------------------+

jeckstei at rutcor.rutgers.edu

http://rutcor.rutgers.edu/~jeckstei/