When trying to run gst-rtsp-server in valgrind using GST_CHECKS=test_describe I noticed the following leak:

Running suite(s): rtspserver
==8198== 108 (56 direct, 52 indirect) bytes in 1 blocks are definitely lost in loss record 3,020 of 3,426
==8198==    at 0x4C2A03B: malloc (vg_replace_malloc.c:267)
==8198==    by 0x6C80CA7: standard_malloc (gmem.c:85)
==8198==    by 0x6C80D30: g_malloc (gmem.c:159)
==8198==    by 0x6C98B21: g_slice_alloc (gslice.c:1007)
==8198==    by 0x6C98B60: g_slice_alloc0 (gslice.c:1033)
==8198==    by 0x63F5667: g_type_create_instance (gtype.c:1872)
==8198==    by 0x63DC65F: g_object_constructor (gobject.c:1849)
==8198==    by 0x63DBE77: g_object_newv (gobject.c:1713)
==8198==    by 0x63DC5F0: g_object_new_valist (gobject.c:1830)
==8198==    by 0x63DB88B: g_object_new (gobject.c:1545)
==8198==    by 0x57CEC4F: g_inet_socket_address_new (ginetsocketaddress.c:340)
==8198==    by 0x57F02F6: g_socket_address_new_from_native (gsocketaddress.c:231)
==8198==    by 0x57ECC6A: g_socket_get_remote_address (gsocket.c:1767)
==8198==    by 0x534B362: gst_rtsp_connection_accept (gstrtspconnection.c:312)
==8198==    by 0x4E605BB: gst_rtsp_client_accept (rtsp-client.c:2162)
==8198==    by 0x4E63D5C: default_accept_client (rtsp-server.c:803)
==8198==    by 0x4E641A3: gst_rtsp_server_io_func (rtsp-server.c:905)
==8198==    by 0x57EE877: socket_source_dispatch (gsocket.c:3168)
==8198==    by 0x6C78B28: g_main_dispatch (gmain.c:2539)
==8198==    by 0x6C797ED: g_main_context_dispatch (gmain.c:3075)
==8198== 
100%: Checks: 1, Failures: 0, Errors: 0
make: *** [gst/rtspserver.valgrind] Error 1

This due to an omission when gst-rtsp-server was ported to gio back in march 2012 where the remote address of the server that the client connected to is not unreffed after having been used. The attached patch fixes this and fixes the spelling typo.