From: jklowden@speakeasy.net   
      
   On Tue, 22 Jan 2013 22:32:59 -0800 (PST)   
   legalize+jeeves@mail.xmission.com (Richard) wrote:   
      
   > The problem with link-time substitution is that I can't take   
   > one piece of code and make it use the link-time substitute and have   
   > another piece of code that uses the real thing and have them both live   
   > in the same executable.   
      
   Why would you want to do that?  ISTM you're much better off *knowing*   
   which library is in use throughout the execution of the test.   
      
   > By using link-time substitution, I'm forced into having more test   
   > executables than I need.   
      
   Not necessarily.  Every OS I know of uses a runtime linker whose   
   behavior can be controlled by the environment.   
      
   > If I use DI, then it's all under control fo the test and not the   
   > *linker*.   
      
   Yes.  What Ian's puzzled by, as am I, is why you think more code is   
   better than less.  By using the linker, at least you don't introduce   
   complexity into the logic of your system.   
      
   --jkl   
      
      
   --   
         [ See http://www.gotw.ca/resources/clcm.htm for info about ]   
         [ comp.lang.c++.moderated.    First time posters: Do this! ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)