This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: ordering of constructors


Phil Edwards <phil@codesourcery.com> writes:

| > Libraries that the program manually loads and unloads are a whole
| > other mess.
| 
| I will continue to pretend that those don't exist as long as each of the
| C++ and POSIX committees continue to pretend that the other doesn't exist...

The truth is more complicated than that.  As chair of the Library
Working group, Matt knows better than I do about the kind of things
being  considered for C++, but I would mention that there is a
subworking group  trying to address the dynamic library problem, with
respect to C++.  I don't think it is much a problem with POSIX and
C++ committees not having strong active communication channels than
a real mess in specifying what a program means in C++ when faced with 
dynamic libraries and dynamic initializations. If standardization
means anything, we ought to look at existing practice and if
implementors are going to pretend that those don't exist, we've got an
egg-chicken problem.  I suppose users just want their programs work
reasonably...


-- Gaby


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]