disabled the getgrent() code recently added that returns all unix groups

as NT aliases.  no allowance is made for possible clashes with some
unix groups that may be mapped to NT groups, not NT aliases.

it is not the place of this code to make decisions about which unix
groups should be mapped to NT aliases or NT groups, or in fact whether
unix groups exist at all (e.g "appliance" mode).

please could the person who wrote this code contact me and i will outline
the architecture and design issues that you need to be aware of when
working in this area: i need your skills and ability -- applied in the
right places.

thank you.
(This used to be commit 67d9b3733f7f3ffab0ff6e78de7638fb469195e6)

1 files changed, 17 insertions, 0 deletions

diff --git a/source3/rpc_server/srv_samr.c b/source3/rpc_server/srv_samr.c
index 1e0b2fc16a..6871671ff0 100644
--- a/source3/rpc_server/srv_samr.c
+++ b/source3/rpc_server/srv_samr.c
@@ -766,6 +766,22 @@ static void samr_reply_enum_dom_aliases(SAMR_Q_ENUM_DOM_ALIASES *q_u,
 	}
 	else if (strequal(sid_str, sam_sid_str))
 	{
+#ifdef _DISABLED_BECAUSE_THERE_IS_AN_API_TO_CALL_TO_OBTAIN_ALIASES
+
+		/*
+		 * this code has been disabled because there is an API
+		 * to call to enumerate NT aliases.
+		 *
+		 * it is the job of this API to decide which unix groups
+		 * should be mapped to NT aliases and which should be
+		 * mapped to NT groups.
+		 *
+		 * putting *all* unix groups as aliases is not the correct
+		 * thing to do, just as it is not the correct thing to
+		 * do to return all unix users as NT users (as is done
+		 * by the disabled function, get_passwd_entries())
+		 */
+
 		char *name;
 		/* local aliases */
 		/* we return the UNIX groups here.  This seems to be the right */
@@ -782,6 +798,7 @@ static void samr_reply_enum_dom_aliases(SAMR_Q_ENUM_DOM_ALIASES *q_u,
 		}
 
 		endgrent();
+#endif
 	}
 		
 	init_samr_r_enum_dom_aliases(&r_e, num_entries, pass, r_e.status);