>> I suggest to check return values immediately after each function call. >> An error situation can be detected earlier then and only the required >> clean-up functionality will be executed at the end. > > Which improvement does this bring?
* How do you think about to avoid requesting additional system resources when it was determined that a previously required memory allocation failed? * Can the corresponding exception handling become a bit more efficient? > Why? Do you care for any results from static source code analysis? Regards, Markus
