[Info-vax] F6 and ctrl-C

[Jan-Erik Söderholm]

Den 2021-07-09 kl. 02:12, skrev Stephen Hoffman:
> On 2021-07-08 23:37:08 +0000, Jan-Erik Sderholm said:
> 
>> I want to have PF6 available as a function key in the application...
> 
> F6 sends its expected function key control sequence, and the terminal 
> driver intercepts and converts that to ^C, and SET TERMINAL /NOLINE_EDIT is 
> one way to disable that driver conversion.
> 
> I don't recall if there's a (more) targeted means to disable (just) F6 
> conversion within the terminal driver itself, short of patching the driver 
> or such, and, yes, /NOLINE_EDIT is a fairly large hammer.
> 
> Or tweak the terminal emulator to send something other than (in eight bit) 
> CSI 1 7.
> 
>> And also, I do not seem to get the ESC codes 11~ to 14~ through.
>> They comes from the F1-F4 och the PC keyboard...
> 
> VT terminals treat the first clump of DEC OpenVMS LK-layout function keys 
> as local, so most of OpenVMS tends not to expect to ever see F1, F2, F3, 
> F4, or F5.
> 
>> Any ideas?
> 
> For some of the usual code for fielding ^C and ^Y, see: 
> https://www.digiater.nl/openvms/freeware/v80/hoffman_examples/lib$xxable_ctrl.c 
> 
> 
> 
> 

Thanks! In todays application we use a customer unique DEC-written
screen tool (in Fortran). We do have the sources so we might search
it and see if it does something special.

It might very well be that it does some setup against the term driver.

We are rewriting one application from 80 to 132 chars, and that old
DEC tool does not support anything but 80 chars.

We can of course rewrite it to not use F1-F4 at all. These are today
used as "one line down/up" and "one screen forward/back". But that
can also be done with the "arrow down/up" and "page down/up" keys
from a modern kayboard. That has already been tested.

Dave asked:

 > Can you be more specific on how you're handling the terminal I/O

The "user" just logs in as usual (happens to be telnet here), then
just RUN the EXE that is the application. So all I/O is done through
terminal driver and the usual sys$input and sys$output from Cobol.
Cobol uses DISPLAY and ACCEPT together with the features to SCREEN's
within the DATA DEFINITION section. Quite neat, actually...

I can see the ESC sequences from F1-F4 on the network [11~ to [14~
but they are for some unknown reason never seen by the Cobol code.
It is exactly the same network format as the other like [18~ for F7.