> > The reason is that some windows apps pass NULL pointers and expect
> > ERROR_INVALID_PARAMETER to be returned, and not a Dr. Watson pop-up ;)
> 
> If you've already got a wrapper for strlen(), why use additional exception
> handling for this case?  Couldn't you save a lot of overhead by using an
> "if" in the wrapper to check for NULL pointers?

The problems that apps do not only NULL pointers to fail that way, but invalid,
non NULL, pointers too.

Ciao, Marcus

Reply via email to