libsmbclient status

[Christopher R. Hertel]

Richard Sharpe wrote:
: 
> Hmmm, but that name is \000\001MSBROWSE\001\000, I think, and you have to
> send name requests to a workgroup, don't you?
> 
> Hmmm, a minute's checking with Ethereal shows that this is not true.

I figured you'd clear that up really quickly.  Just for the archive's
sake, it's:

  \x01\x02__MSBROWSE__\x02\0x01

...where that last '\0x01' is the suffix byte.  It is registered as a
group name by all Local Master Browsers, no matter what the workgroup.
They use that name to announce themselves to one-another.  That's how the
workgroup list gets shared around.

> That would indeed be the easiest way to do it. I will leave it as an
> exercise for the students, unless I get around to it first.

If you don't know any workgroup names, it's the only way to do it I'm
'fraid.  It's what Mike does in jCIFS, btw.  If you *do* know a workgroup
name (a good reason to have a local NetBIOS name cache, I suppose) then you
can try the LMB and/or the DMB for that workgroup.

One more problem that was raised on the jCIFS list is that some browsers
will not allow you to access the browse list if you're not authenticated.
(Which is silly, IMHO.)  That's why you might want to gather the list of all
IPs that respond to the __MSBROWSE__ name query (the hex codes are too
annoying to type all the time) and try them in sequence until you get a
usable response.

Mike found that LMBs are more likely to allow an anonymous connection (not
sure if it's 100%) than they are a guest connection.

This came up just a week or so ago.

Chris -)-----

-- 
Samba Team -- http://www.samba.org/     -)-----   Christopher R. Hertel
jCIFS Team -- http://jcifs.samba.org/   -)-----   ubiqx development, uninq.
ubiqx Team -- http://www.ubiqx.org/     -)-----   crh at ubiqx.mn.org
OnLineBook -- http://ubiqx.org/cifs/    -)-----   crh at ubiqx.org