On 3/3/20 11:21 AM, Charles Mills wrote:
Does a workstation necessarily have a name?
That turns into a slippery slope of "What is a name?" / "What does the
name actually mean?".
I think that the host end has a name / ID for the connection, at least
insofar as differentiating the connection / session from other
connections / sessions.
The client may not have / know about / care about any such identifier.
The client may have a (DNS) hostname associated with it's IP, be it
dynamic or otherwise.
These two client ""names may not be the same thing.
In the protocol, I mean. A dumb terminal with no name can do
telnet. Is there anything to the connection request other than "Hi,
I'm 192.168.1.1, let's connect"? There's no query where the mainframe
says "tell me about yourself," right?
I don't know about TN3270. Traditional telnet does support what open
systems refer to as "answer back", which is some sort of name / ID /
string that is configured on the client.
I don't recall anything in my 3270 emulator (Tom Brennan's Vista)
where I say "here is my name to give to the host." There is a space
for an "LU name" but it's blank and I have no idea what it is for.
I suspect that the "LU name" is probably the TN3270's counterpart to
"answer back".
I've been told that years ago host applications would use this answer
back / LU name as an identifier to alter how things behaved. It could
be anything from choosing the default printer (hypothetically physically
closer to the terminal) or choosing different terminal behaviors, etc.
I have dabbled with this in Linux to alter the environment to change the
way that error messages are displayed. I have a few client machines
that are using a more capable version of XTerm that supports graphics.
So, if the answer back has the specifically configured and recognized
names / IDs / strings, my shell can know that it can send fancy error
messages w/ graphics back to me.
One important thing to remember is that you should not use the answer
back / LU name for any (important) security information because the
information is provided by the client. This means that any client can
change what it provides, possibly even spoofing information as a means
to gain elevated privileges.
My Windows has a hostname but there is no reason to think it is unique
in any given host's clients.
Do to the vagaries of Windows (read: NetBIOS) networking, I would expect
the hostname to be unique within a given network (segment), or that
there are other problems.
I fear the question may not have an answer.
I suspect that you are closer with the "LU name" than you may have realized.
There's always my favorite approach: disable it and see who
screams. (Yeah, you could put out some sort of warning broadcast a
month in advance.)
BOFH: Scream tests are always fun. }:-)
--
Grant. . . .
unix || die
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
