Note I've doctored this in an attempt to get over a stupid mail reject
condition!!!
Kevin,
Of course, it's always possible that the programmer responsible for the
client TN3270E emulator coding can have made a mess of it. However,
according to RFC 2355, which I hope is the one you have in mind since it's
the RFC which covers TN3270*E*, if the *"device-type"* is wrong the "reject
reason" should be, logically enough, "INV-DEVICE-TYPE".
If Tim had specified a "device-name", I would expect him to have mentioned
that he had specified such a significant parameter, given the text of the
"reject reason". Nevertheless, if you are suggesting that Tim has used a
"device-name" and it might be incompatible with the "device-type" he has
also specified, I don't actually see that as being possible.
What happens in the CS IP TN3270E server is that the "device-name", directly
or indirectly, becomes an LU name and the "device-type" gets mapped
unavoidably[1] to a mode table entry name which should be found either in a
private mode table load module or in the VTAM-supplied load table module
ISTINCLM. These two processes are independent. If the "device-name" isn't
there I would expect an "INV-NAME" "reject reason". My quick look at the CLI
ET2000 manual indicated it would not be possible to offer an incorrect
"device-type". In fact only a subset of those offered by the RFC are handled
by the emulator.[2] Once the TN3270E server has found an LU name from the
"device-name" it will always find a mode table entry name from the
"device-type" code. If the name of the mode table entry is incorrect, the
problem shows itself as a VTAM error.
But - thinking aloud here - it is possible that the CS IP TN3270E server
could, knowing the significance of the VTAM error, decide to report a "name
error". This however would not really be appropriate:
1. It's not what RFC 2355 has in mind for "INV-NAME".
2. It's not an error the client made but an error made by the systems
programmer who coded the TN3270E server statements.
Well, now checking the possible "reject reasons" in the RFC, I see the
following, not noticed before:
<quote>
TYPE-NAME-ERROR The requested device-name or
resource-name is incompatible
with the requested device-type
(such as terminal/printer mismatch).
</quote>
This indicates - yet again I'm afraid since we've been in this territory in
this list/group before - that the author of RFC 2355 doesn't really
understand the CS IP TN3270E server. If the CS IP/SNA systems programmer
understands what he is doing and takes care to specify valid mode table
entry names wherever he has taken the trouble to specify them, it is
impossible to get a mismatch between "device-name" and "device-type".
Maybe I've missed something - but this is the right place to find out ...
I was looking for what "reject reason" might be used in the case where the
TN3270E systems programmer had made a mess with the statements mapping
"device-type" codes to mode table entry names. Since it's actually no fault
of the client I can only expect " UNKNOWN-ERROR: Any other error in device
type or name processing has occurred." which should have the client end-user
making plaintive noises in the direction of the help desk.
Chris Mason
[1] A mode table entry name is inevitably found because any entries mapping
"device-type" codes to mode table entry names which are not specified take a
default name corresponding to a name - which should be present in the
ISTINCLM mode table.
[2] I suspect the opprobrium I heaped on the product in my last post
belongs more to the sites which offer to sell it. It is only these sites
which associated "3278" and "colour", strictly "color".
> Tim,
>
> Check the terminal name is needs to be consistent such as: IBM-3278-3-E
>
> Review the TN3270 RFC - there should be a DO negotiation process.
>
> Kevin
>
> I have a few special purpose emulator devices (CLI ET2000s with the latest
> firmware) connecting over TN3270 and when they try to connect I get a
> "REJECT REASON: INV-NAME" response. I'm able to connect with a variety of
> PC-based emulators like Vista tn3270 and x3270. I did some looking and it
> appears to be in the TN3270 RFC, although there was not much explanation
on
> it and I can't find anything in IBM TCP/IP documentation.
>
> Thanks!
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
