Wasnt sure if this was a ZTERM forum post or this, but here goes...

A site has Pshell working on CRT's fine and just added a PC running Zterm. (using telnet)

This PC launches Pshell fine... but when hitting the swap key (Ctrl ]) instead of swapping it goes to command mode (just as if hitting the | vertical bar). From here, the keys dont seem to translate properly.. hitting arrows wont move the command cursor from option to option, Ctrl ] or escape escape wont quit out, so from here the only option is closng zterm and starting over.

It feels like a Shell problem or emulator but the Crt's are using the same .profile as the zterm session...

Any suggestions?

Thanks.

You have me stumped. Ideas to consider, even though some or all may be hard to reconcile with all the symptoms:

1. The Windows locale/keyboard has been changed or just doesn't match the actual keyboard, causing some of the non-alpha keys to be encoded or interpreted differently than looking at the keyboard would suggest.

2. The ZTERM emulation is set to something strange, or doesn't match the TERM setting on the Linux side.

3. Double-check the ZTERM > Configuration > Options > Keyboard settings to see if anything is unusual there.

4. You may want to experiment with FIXTRN to see if there are any other indications of the keyboard/emulator not translating keys as you would expect.

Thanks for the followup... i did most if not all of that... but i will triple check the takoff procedures and give it another go...

One Polyshell-related issue that comes to mind is that newer versions of CentOS come with a different version of ksh (the Korn Shell) from previous versions, and it does not seem to properly support the set +o monitor command. The symptoms aren't like what you describe (it's more a matter of it hanging after the first context swap), but I thought I would throw it out there in case it seems relevant. I posted a note about it and a tar'd copy of pdksh that works at http://www.microsabio.net/dist/other/pdksh

We have discovered this PC has a wireless keyboard interesting enough... she is going to get a wired one and try again, just to eliminate that as a possibility... will let u know.

Ahhh... I've actually seen variations of this problem on my own wireless keyboard. Sometimes, when the batteries are getting low, or when there is some obstruction between the keyboard and receiver, I'll find that some keys work and others don't. So that may well be the issue.

Sounds promising... will let u know!

She is now using a wired key board and the problem still exist.

Oh well, it was a nice theory. The fact that this is the only ZTERM at this site does complicate our analysis slightly, since we can't really tell if the issue is specific to this particular PC or to the server.

If it were related to the PC itself, it would have to be that the keyboard/locale definition was messed up. (If, for example, it was set to AZERTY, then changing from a wired to wireless keyboard wouldn't matter.) But that should be easy to check, and also to test. I'd first open up Notepad and make sure that all the keys worked there, and if so, then try the same test via ATE in VUE. If Notepad works ok, but ATE/VUE doesn't, that suggests that the problem is not the PC per se, but a mismatch of some kind between the ZTERM config and the corresponding terminal config on the server side.

If it were related to the server, we might ask what is different from the server's perspective between the dumb terminals and ZTERM? The fact that terminals work ok suggests that either ZTERM is failing to emulate that terminal, or that you're using different TERM environment settings for the real terminals vs. the ZTERM terminal. Assuming you're using wyse50 for the real terminals, are you also using (or did you try using) wyse50 for ZTERM?

If we think there is something defective about this particular copy of ZTERM, did you try testing ATE, or some other emulator there to see if it has the same problem? (Preferably one that supports the same emulation as you're using now, so no changes would be required on the server side for testing.)

Frank said above that he "did most if not all of that", but didn't give us any further facts to evaluate. I'm not suggesting that you failed to do all of the standard troubleshooting, but without any details, it's pretty hard for me to offer anything but standard troubleshooting advice.

I guess you were hoping that the symptom would remind me of some hidden "stop misbehaving" switch at the back of the computer or tucked away in a dialog somewhere, but I haven't yet run across either this particular symptom or that hidden switch, whereas my experience has been that many thousands of copies of ZTERM are out there working as expected, I'm left to guess that there is something different about this setup that no one has noticed yet. And I'm not sure how we're going to discover that difference without examining everything that seems possibly related.