| Newsgroups |
gmane.comp.emulators.hercules390.general |
| Message-ID |
<[email protected]> |
There has been so much misinfo posted in this
one thread over the past days it's difficult
to identify where to start..........
>>
>> I also know that telnet and tcp do indeed
>> behave differently...
>
> These are apples and oranges as Joe said.
> TCP is the equivalent of VTAM.
TCP is a comm protocol; VTAM is an access
method. There is no meaningful "equivalence"
between the two.
Indeed, it may be argued TCP and TELNET are
much more closely related to each other than
either one is to VTAM.
>
> I found out how to switch debug on in Hercules,
> and here is what happened:
>
> [snip]
>
> 15:21:28 console: DBG005: Packet received length=3
> 15:21:28 +0000 323031 201
> 15:21:28 console: DBG006: Expected IAC WILL TERMINAL_TYPE
The above debug print may be considered to be
tainted as Paul has stated he's using a
bastardized version of Herc.
My judgement is A) we can accept the reference
to console.c as valid and that he is connecting
to Herc using some TCP client via port 3270;
but B) the "201" may be considered suspect without
seeing actual PCAP (i.e., tcpdump) output.
>
> qemu-system-i386 -full-screen -drive
> file=hd2.img,index=0,media=disk,format=raw
> -serial tcp:localhost:3270 -soundhw all
>
> I have tried with "telnet" instead of "tcp" but it
> still fails to connect to Hercules.
"telnet" is a valid thing to want to do and it should
work. In testing this here I found after looking
at the qemu source code that qemu contains no code
to perform TELNET option negotiation when running
in TELNET client mode. Since Hercules requires
option negotiation the connection appears to hang.
>
> No, those are both valid parameters to "-serial"
> under QEMU. "tcp" will not do any telnet
> negotiation, but "telnet" is supposed to.
Sorry, no. Can you show where in the qemu code
it is supposed to do client TELNET option
negotiation?
>
> You need to show the hercules config file,
The Herc cfg might have been nice,
but since it's console.c on port 3270
I think we can infer a few things...
> and your TCPIP Profile member
The TCPIP Profile member is irrelevant
> and your tcpdata member to be sure.
The tcpdata member is irrelevant
I think this must be the single most problematic
area in Hercules or at least on this list...
Namely, confusion about the relationship
between the TCP/IP stack that exists on the
OS (Windows or Linux) under which Hercules
runs versus the TCP/IP stack (if any) in
the guest OS running under Hercules.
Here we're referring to the former, and
in this case we don't even require the
latter (Herc guest OS) to posess a TCP/IP stack.
Examples of such latter include MVS 3.8J and VM/370.
>
> I want PDOS/390 to drive an EBCDIC ANSI
> terminal so that I can run micro-emacs
This is probably the most egregious of all -
as we've already discussed there is no such
thing as "EBCDIC ANSI". Also as previously
mentioned neither the 1052 nor the 3215 are
supported device types under TCAM, VTAM, or
TSO in MVS 3.8J. Just when it seemed this
issue had been beaten to death, it pops up
here on this list...
Max