[Info-vax] VTxxx/VMS term driver: bug or feature?

[Johnny Billquist]

On 2011-07-01 08.13, Johnny Billquist wrote:
> On 2011-06-30 16:44, VAXman- @SendSpamHere.ORG wrote:
>> In article<nCPQG4dCMxfn at eisner.encompasserve.org>,
>> koehler at eisner.nospam.encompasserve.org (Bob Koehler) writes:
>>> In article<iuhlkv$lt4$1 at Iltempo.Update.UU.SE>, Johnny
>>> Billquist<bqt at softjar.se> writes:
>>>>
>>>> So a "strange thing" is not a bug?
>>>
>>> What the real terminals do, is something I would consider a bug.
>>> That is, the text displayed is not the text I would expect from the
>>> documentation that I've read.
>>>
>>> An emulator should act like the real thing, but I can understand
>>> allowing the user to turn on/off bug compatability.
>>
>> The real VT displays the text, in my example program, as I'd expect to
>> see
>> it when SET TERMINAL/NOWRAP and local wrap conditions have been
>> established.
>>
>> The odd thing is when the terminal driver is asked to implement wrap when
>> the terminal is set not to locally wrap. This oddity seems to appear on
>> all of my VT terminals and DECterm, even though I don't consider the VMS
>> terminal driver's handling of the wrap case correct. Stranger, not
>> really,
>> is that a particular terminal emulator doesn't display this oddity.
>
> I'd say this has very little to do with the terminal emulation and a lot
> to do with VMS terminal driver wrap implementation.
> The only thing terminal related is how the terminal behaves when not
> wrapping, sitting at column 80, and you output a character. Most every
> terminal emulator seems to think that the character should be discarded,
> since you're outside the screen. A real VT100 affects column 80 again,
> even though the cursor is logically at column 81.
>
> So you get either '00' at the right end (the correct result) or '01' at
> the right end (the wrong result). I know that putty does a lot of wrong
> at this point as well, which is part of the gripe I have with putty. If
> you are at column 80, output one character, and sends a backspace, putty
> will remain at column 80. :-(

I actually just checked xterm some more, and the version I have that 
reports itself as version 200 on my NetBSD system actually gives the 
same results as a VT525, while xterm on a Linux box here, reporting 
itself as version 256 gives the wrong result.
(Just looking at the result on column 80, when terminal does not do 
wrapping, while VMS do.)

As a sidenote I can also report that I get the same results on an RSX 
system, after rewriting the command file. The one exception being that 
the RSX terminal driver does not recognize ESC # n as a non-printing 
escape sequence, and thus counts that as using up two columns in the 
output, so the terminal driver wraps the output two characters earlier 
than VMS.

	Johnny