APR is a cross platform library. The code is there because whether the functions are actually implemented or not, the API must exist in order for APR on NetWare to remain API compatible with APR on other platforms for porting purposes. Since this functionality can not be implemented today on NetWare is the reason why they return the value APR_ENOTIMPL. There already is an Open Source port of BASH for NetWare which, if Apache were running within a BASH shell, would allow for UID/GID functionality. So yes it could be implemented in the foreseeable future. As for the getpwnam_safe() static function, whether it stays or goes doesn't really matter. It was left there after the API's were stubbed out for NetWare and if or when these APR functions are ever actually implemented on NetWare, it may become necessary again.
Brad >>> [EMAIL PROTECTED] Tuesday, November 23, 2004 3:22:17 PM >>> Brad Nicholes wrote: > You are right, it is dead code for now. The reason why it is there > is because this file is just a stub rip-off of the Unix implementation. > It is just waiting for the day when the NetWare OS supports UID/GID for > running applications and services. On the NetWare OS applications and > services don't run as a user, they just run. Therefore the underlying > APIs like getuid(), getgid() and getpwnam() don't exist on NetWare. > Even if they did, they wouldn't return anything useful. So for now, > this is just a stub implementation. You seem to be implying that there is a reason for this code to exist: that Netware is likely to gain UID/GID support in the forseeable future and that this code will be useful when it does, and that having the code already there will be considerably more useful than a programmer having to write it again or copy it from the Unix implementation. Is all of that correct? - Julian
