Francois Gouget fgouget@free.fr writes:
I guess that one day Alexandre is going to argue that Linux should stop polluting the namespace with 'struct user'...
Well, struct user is the only reason for including sys/user.h, so I guess it can be considered acceptable; but anything else defined in there (like struct user_regs_struct on Linux) is namespace pollution too IMO.
And if there comes a day when the server is managing users, the corresponding struct will definitely be named struct user, not struct wine_unique_prefix_0918274_user or whatever (and note that we don't really need user.h even on Linux, we could just as well hardcode the offset).
On Mon, Nov 19, 2001 at 01:22:35PM -0800, Alexandre Julliard wrote:
Francois Gouget fgouget@free.fr writes:
I guess that one day Alexandre is going to argue that Linux should stop polluting the namespace with 'struct user'...
Well, struct user is the only reason for including sys/user.h, so I guess it can be considered acceptable; but anything else defined in there (like struct user_regs_struct on Linux) is namespace pollution too IMO.
Can you please explain how this is namespace pollution on the part of the OS? AFAIK, <sys/user.h> is not specified by any Standard (ISO C89, C90, or C99). <sys/user.h> can include any (unprotected) variable name. If an application includes <sys/user.h>, then the application should be prepared to deal with conflicts.
-
Alexandre Julliard -
Steve Kargl