summaryrefslogtreecommitdiff
path: root/source3/namequery.doc
diff options
context:
space:
mode:
Diffstat (limited to 'source3/namequery.doc')
-rw-r--r--source3/namequery.doc56
1 files changed, 56 insertions, 0 deletions
diff --git a/source3/namequery.doc b/source3/namequery.doc
new file mode 100644
index 0000000000..77a75714e5
--- /dev/null
+++ b/source3/namequery.doc
@@ -0,0 +1,56 @@
+this module contains non-threaded versions of name status and name
+query functions. if a multi-threaded nmbd was to be written, these
+functions would be the starting point.
+
+at the moment, the expected response queueing system is used to
+replace these functions without needing to multi-thread nmbd.
+
+these functions are used in smbclient and nmblookup at present to
+avoid having the vast quantities of complex and unused code needed
+to support even a simple name query (or providing stubs for the
+unused side of these functions).
+
+there is a down-side to these functions, which is all microsoft's
+fault. microsoft machines always always reply to queries on the
+priveleged ports, rather than following the usual tcp/ip mechanism
+of replying on the client's port (the exception to this i am led
+to believe is windows nt 3.50).
+
+as a result of this, in order to receive a response to a name
+query from a microsoft machine, we must be able to listen on
+the priveleged netbios name server ports. this is simply not
+possible with some versions of unix, unless you have root access.
+
+it is also not possible if you run smbclient or nmblookup on an
+interface that already has been claimed by the netbios name server
+daemon nmbd.
+
+all in all, i wish that microsoft would fix this.
+
+a solution does exist: nmbd _does_ actually reply on the client's
+port, so if smbclient and nmblookup were to use nmbd as a proxy
+forwarder of queries (or to use samba's WINS capabilities) then
+a query could be made without needing access to the priveleged
+ports. in order to do this properly, samba must implement secured
+netbios name server functionality (see rfc1001.txt 15.1.6).
+
+
+/*************************************************************************
+ name_query()
+ *************************************************************************/
+
+
+
+/*************************************************************************
+ name_status()
+ *************************************************************************/
+
+
+
+/*************************************************************************
+ _interpret_node_status()
+ *************************************************************************/
+
+
+this is a older version of interpret_node_status().
+