I looked back and found these initial notes:
I finally managed to get Hercules up and running. It runs until the
point where the name TK5 appears large on the screen. Then I call the
X3270 with localhost:3270. The screen appears again with the name TK5
large on the screen. The Hercules window displays the message that
there was a connection to 3270.
So the main logo screen appears, probably the one where there's no input field for a logon id until you press the Clear key (at least that's how the TK5 setup works for me). So we can probably eliminate things like the "327" typo where we were wondering if you're communicating at all with the host. You *are* communicating in order to get the logo screen.
However, when I press enter or any other key, the terminal goes to
system and does not respond. That is, it apparently opens the terminal
with the keyboard locked.
As you probably know, most 3270 terminal keys don't send any data to the host. The ones that do are Enter, Clear, PF1-24, PA1-3, Attention, and SysRequest. I call those "Action Keys". When you press an Action key, the terminal emulator locks the keyboard before sending the screen data to the host. The host gets the data, processes the action, and sends back the results to the terminal. That response almost always includes a bit telling the terminal to unlock. If you don't get that bit for some reason, the terminal stays locked.
So in your case it's likely that when you press Enter the terminal emulator is doing the locking, but we don't know what happens after that. Did the Enter key get sent to the host? Did the host process it? Did the host respond? Did the response have the unlock bit? Did all that make it back to the terminal emulator?
Those are things you can try to find out if the Linux terminal emulator has some kind of I/O trace function, or if you want to try something like the tcpdump command. You should see a small I/O buffer (just a few bytes) being sent to the host, and another coming back, such as when I pressed Enter in TK5 just now and got the INPUT NOT RECOGNIZED response.
Another possibility is that whatever key you're pressing (Enter) is not doing what you think it is doing. For example, some emulators are setup so the Enter does the NewLine function, which doesn't send any data to the host. Another is that when you're running as a VTAM attached unit, the Attention key locks up the keyboard, so maybe the emulator is doing that instead of Enter.
You might also try to find the Reset function on the emulator. When a keyboard is locked, you can unlock it yourself with the Reset key. Then at that point you can try another key such as Clear or whatever.
Just Tom guessing with minimal data.
On 4/22/2025 6:00 AM, Ar¨ª Ricardo Ody via groups.io wrote:
Hello everyone!
I used grep in the Linux terminal to search for the sequence 327 not followed by a zero in the conf directory. I found the following in tk5.cnf:
#
# local 3270 devices (VTAM)
#
00C0 3270
00C1 3270
00C2 3270
00C3 3270
00C4 3270
00C5 3270
00C6 3270
00C7 3287
0200 3270
0201 3270
0202 327
In other words, there was really a missing zero in the line that starts with 0202. I edited tk5.cnf and added the missing zero. The line became
0202 3270
Then I saved and ran Hercules again. Still, the 3270 emulators did not open the keyboard.
I noticed something that seems different than before. See the messages below:
HHC02264I Script 1: file scripts/tk5.rc processing ended
HHC02915I client 0 COMM: Connection received
HHC02914I 0:00C0 COMM: client 0 negotiations complete; ttype = 'IBM-3279-4-E'
HHC01018I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connected
HHC01022I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connection closed by client
HHC02915I client 1 COMM: Connection received
HHC02914I 0:00C0 COMM: client 1 negotiations complete; ttype = 'IBM-3279-4-E'
HHC01018I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connected
HHC01022I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connection closed by client
HHC02915I client 2 COMM: Connection received
HHC02914I 0:00C0 COMM: client 2 negotiations complete; ttype = 'IBM-3279-2-E'
HHC01018I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connected
HHC01022I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connection closed by clien
I think that previously the message "HHC01022I 0:00C0 COMM: client 127.0.0.1 devtype 3270: connection closed by clien" did not appear
I thought I would comment on this to see if you have any idea why this message started appearing after making the change in tk5.cnf mentioned above.
I, despite all the good will I have, have no idea what is going on here. I'll attach the log file again, but I've already inspected it and the message with 327 no longer appears.
Greetings from S?o Paulo/SP - Brazil
Ar¨ª Ricardo
On Mon, Apr 21, 2025 at 11:21?PM Mark Gutzwiller via groups.io <> <mark.gutzwiller@... <mailto:[email protected]>> wrote:
__
Hi Ari,
The answer is 3270 terminals must be defined as device type 3270.
Change the 327 to 3270 and it should work.
On 4/21/2025 6:53 PM, Ar¨ª Ricardo Ody via groups.io
<> wrote:
Hi Mark!
I have no idea what is causing my problem. So please don't answer
my request for help with new questions. If I knew how to answer
your question I would probably know what the problem is and how to
solve it. I don't think I would have even asked for help, do you?
Can you please tell me the solution to the problem you seem to
have seen? Please?
Greetings from S?o Paulo/SP - Brazil
Ar¨ª Ricardo
On Monday, April 21, 2025, Mark Gutzwiller via groups.io
<> <mark.gutzwiller@...
<mailto:[email protected]>> wrote:
> Ari,
>
> These lines are making wonder,
>
> 16:05:51 HHC01437I Config file[73] conf/tk5.cnf: including file
conf/extcons.cnf
> 16:05:51 HHC90020W 'hthread_setschedparam()' failed at
loc=console.c:3299: rc=22: Invalid argument
> 16:05:51 HHC01462E 0:0202 devtype 327 not recognized
>
> Could that be the problem?
>
> Mark
>
> On 4/21/2025 2:16 PM, Ar¨ª Ricardo Ody via groups.io
<> wrote:
>
> Yes, I receive sleeping cat image and try possible combination
off the 3270. Nothing happen.
> The file you asked for is attached
> cheers
> On Mon, Apr 21, 2025 at 12:46?PM Fernando M. Roxo da Motta via
groups.io <> <mvs@...
<mailto:[email protected]>> wrote:
>>
>> On Mon, 21 Apr 2025 09:12:39 -0300, Ar¨ª Ricardo Ody via
groups.io <>
>> <nightrider43@... <mailto:[email protected]>>
wrote:
>>
>> ? Hi Ari
>>
>> > Hi!
>> >
>> > The problem is that the screen is simply never freed up for
typing.
>> > If I press enter, clear or any other key, the 3270 terminal
shows the
>> > screen with that X indicating that the keyboard is locked. So the
>>
>> ? At this point, does that sleep cat image is on the Hercules
console?
>> ?I am asking about the Hercules console, like the attached image?
>>
>> ? If so, open the extra key window in x3270 (the icon on the
top right
>> of x3270 window), it can help to select the key to press.
>>
>> ? After that sleep cat image show in Hercules console, press the
>> "Reset" key, just in case, and the "Clear" key using the
auxiliary key
>> window.? ?At this point the x3270 emulator should change to
show the
>> "Logon ===>" prompt
>>
>>
>> > core of my problem is finishing the IPL process and opening
the logon
>> > prompt. The processing never gets to this point. This is in
Linux,
>> > because in Windows everything works correctly. That website that
>> > shows the screens and the different colors does not teach
what to do
>> > if the command prompt never opens. In other words, in a way,
it is
>> > very nini if everything works out. This is not my case.
>>
>> ? If that sleep cat image never shows up on Hercules console, there
>> seems that some problem occurred during the IPL process. ?In
this case
>> the best approach should be to send the Hercules log produced
during
>> IPL.? This log should be in "log/3033.log" file.
>>
>>
>> ? ?HTH
>>
>> >
>> > Best regards
>> >
>> > Ar¨ª Ricardo
>> > S?o Paulo/SP - Brazil
>> >
>> >
>> > On Friday, April 18, 2025, Rupert Reynolds via groups.io
<>
>> > <rupertreynolds= [email protected]
<mailto:[email protected]>> wrote:
>> > > I was taking some screen shots for a friend, so I put them
up in a
>> > > web
>> > page on my defunct server. This is how I'd expect TK5 to startup.
>> > > Note that any attempt to type when the cursor is not in an
input
>> > > field
>> > will lock the terminal. <Reset> to clear that. If you don't get a
>> > "Logon" prompt after connecting a tn3270 terminal to
127.0.0.1:3270 <>,
>> > <Clear> normally clears that.
>> > > It's http only, as I haven't set up https yet. No cookies, no
>> > > scripts.
>> >
>> > > On Fri, 18 Apr 2025 at 11:53, Ar¨ª Ricardo Ody via groups.io
<>
>> > > <nightrider43=
>> > [email protected] <mailto:[email protected]>> wrote:
>> > >>
>> > >> Hi!
>> > >> Thank you very much for the comments.
>> > >> On Fri, Apr 18, 2025 at 4:06?AM Tom Armstrong via
groups.io <>
>> > <tomarmstrong255@...
<mailto:[email protected]>> wrote:
>> > >>>
>> > >>> Hi Ar¨ª,
>> > >>>
>> > >>> Thank you for reporting this error. As mentioned by Mike, the
>> > >>> newer
>> > SORT commands are not supported by OS/360 S/M for MVS 3.8.
Ony those
>> > commands listed in the S/M Application Programming Guide are
>> > supported. To solve your INCLUDE problem I suggest that an
E15 User
>> > exit will provide you with a solution to filter the records
input to
>> > the sort.
>> > >>>
>> > >>> The unsupported command was detected by the code and message
>> > >>> IER005A
>> > should have been written to SYSOUT. Unfortunately, there is a
bug in
>> > the message writing routine that resulted in the Abend. The
code will
>> > be updated and the fixed code included in the next TK5 Update.
>> > >>>
>> > >>> Regards
>> > >>>
>> > >>> Tom
>> > >
>> > >
>> >
>> >
>> >
>> >
>> >
>>
>>
>>
>>
>>
>>
>>
>> ? Roxo
>>
>> --
>> ---------------- Non luctari, ludare -------------------+ WYSIWYG
>> Fernando M. Roxo da Motta <mvs@... <mailto:mvs@...>>
? | Editor?
>> Except where explicitly stated I speak on my own behalf.|? VI !!
>> ? ? ? ? ? ? ? ? PU5RXO | PX5Q6048? ? ? ? ? ? ? ? ? ? ? ?| I see
text,
>> ------------ Quis custodiet ipsos custodes?-------------+ I get
text!
>>
>>
>>
>>
>>
>>
