question about SPOOL_Q_SETPRINTER

Tim Potter tpot at
Wed Mar 28 23:21:14 GMT 2001

Jean Francois Micouleau writes:

> > In the following typedef from rpc_spoolss.h, why is the secdesc_ctr a
> > pointer and the devmode_ctr a struct?  This seems inconsistent to me.
> > Especially since, a SEC_DESC_BUF is required in the structure sent
> > to the print server (even in the case of a NULL security descriptor).
> because proper sec_desc was added much later to spoolss.
> Even if it is inconsistant (need to check the IDL first to be sure), I
> urge you to NOT change it now. It works, don't change it. It's not the
> right moment to do such a change.

There's a whole bunch of these inconsistencies throughout the rpc
code if you look hard enough.  (-:  Pointers to structs that
should be structs and so on.

When I was looking at them I also decided to leave them alone as
the were working.  As we all know, automatically generated
from idl code would be really nice.  You could even remove all
those stupid xyz_ptr, lengths and other stuff that is only
relevant on the wire.


> 	J.F.

More information about the samba-technical mailing list