1

Emulated terminals can accept control characters, being able to move the position of cursor, to delete a char or a word or even to clear all screen, change foreground and background colors, etc. oriented by them.

Question: Originally (that means for real terminal teletypes), these characters controls were handle by the output device or managed directly by CPU (for example, via related routines monitoring output data sending to such device and working over it)?

  • 2
  • 1
    @manassehkatz-Moving2Codidact I see your new name ... keep us informed! – Raffzahn Apr 5 at 15:18
  • 2
    "Real teletypes" are basically automated typewriters. Naturally, they cannot handle the sophisticated control character sequences you are referencing. That's only done by later electronic terminals, such as the DEC VT-100, – Brian H Apr 5 at 15:36
  • 1
    @Raffzahn We're working on it. Actually not so much me - I'm helping with design and governance but not (unfortunately, just don't have time right now) much with actual coding. But it is moving along - once there is something close-to-ready, I'll let everyone know. I've taken a step back here (and other SE sites) from Answering, though I still lurk, comment, vote. – manassehkatz-Moving 2 Codidact Apr 5 at 15:48
  • 1
1

For actual Teletypes, which are electromechanical devices whose output is ink on paper, there were these ways to move the print position (there was no actual cursor):

  • Carriage return: mechanically move typehead to column 1.

  • Line feed: mechanically roll the paper

  • Any printing character: move one place to the right after printing

From memory, I don't think an ASR 33 could implement tab or backspace. Tab expansion would be done on the host machine. Backspace probably would not be done at all, though I suppose you could simulate it with return and a lot of spaces (but for what point?)

So, with the exception of tab, neither device nor computer would implement any extra movement operations: it was simply not possible.

Other control characters:

  • BEL rang a bell.

  • XON/XOFF were used to enable or disable an attached papertape reader.

  • WRU might trigger an answerback message, stored mechanically on a rotating drum. (Wikipedia says programmed by 'breaking off tabs' so it's write-once memory)

I think that's about it for the control-character repertoire.

There were no general cursor-movement sequences because the hardware could not do those things.

In case you also include other serial CRT devices, sometimes referred to as "glass teletypes", then the answer is often the same as for emulators: there's something in there, at least from the mid-1970s likely to be a simple microprocessor, interpreting the character stream and doing what the control sequences tell it to do.

But note that not all glass teletypes implemented 'direct cursor addressing' capabilities. Early/cheap terminals were functionally the same as actual teletypes.

| improve this answer | |
  • 2
    You might also mention the not-really-a-control-character, DEL or RUBOUT, which was a character that all hardware simply ignored as if it weren't there. So if you copied a tape that contained DELs, the new one would have no DELs in it. (When preparing a paper tape, when one made a typo, the tape was moved back and all eight holes were punched into it, to "rub out" or "delete" the mistake, which is why that character must be 0xFF.) – Ray Butterworth Apr 5 at 17:58
  • I think it might have caused the print head to 'chatter' a little - no mark made, no movement. A bit like NUL in that regard. That probably equates to ignoring it as much as mechanical devices could ignore it. – another-dave Apr 5 at 18:23
  • there are also dead keys which print the character but don't move the cursor (typically used for diacritics) – phuclv Apr 6 at 4:47
  • A Model 33 could not, but a Model 37 could. The new capabilities for reverse feed, half-line forward and reverse feeds, and (non-destructive) backspace were touted in the advertisements of the Model 37 in 1967, which one can still find in a Google Books search; and were interpreted by the ul command. It also had switchable type trays (immortalized in such things as the BSD greek command) and (according to the advert on page 932 of ISBN 9781483160238) software-settable vertical and horizontal tabstops (and thus vertical and horizontal tab). – JdeBP Apr 6 at 9:01
  • By the way, you forgot perhaps the most tell-tale characteristic control character of paper terminals: the Form Feed. (-: – JdeBP Apr 6 at 9:03
4

Teletypes were self contained units simply executing the code received. For most codes that simply meant printing a certain character and advance by one position. Others

  • advanced paper by one line (LF),
  • returned the carriage by one position (BS)
  • returned the carriage to the begin of a line (CR)
  • or rung a bell (BEL)

All of this was done by mechanics using levers, spoke wheels, combs and rods (*1).

There is no controlling CPU. TTYs were meant to work with other TTY, not computers/CPUs. They only later on repurposed TTY as terminals.


*1 - IoT before the internet - process control by things moving:)

| improve this answer | |
  • I'm not so sure about backspace being possible, at least not on an ASR 33. Fancy DECwriter terminals, yes. – another-dave Apr 5 at 15:29
  • 1
    Upvote for pointing out that a tty was not designed as a computer peripheral. Later, of course, a computer could get really fancy - e.g. by automatically outputting LF after receiving CR. Or simulating tabs. Wow, now we're living! ;-) – another-dave Apr 5 at 15:35
  • @another-dave :) - Can't speak for the ASR 33 without looking it up, but Siemens T100 and SEL/Lorenz machines of ~1960 did and Lorenz started their series with a Teletype licence. – Raffzahn Apr 5 at 16:01
  • Also, "ENQ" (a.k.a., "WRU"), which prompted a Teletype to respond with its answerback code. – Solomon Slow Apr 5 at 16:02
  • @SolomonSlow Yes, but, while it's an incredible cool feature for a mechanical device, it's not really one of the controll characters used by computers.(I also left out XON/XOFF, as they are as well rather specific) – Raffzahn Apr 5 at 16:13
0

Real teletypes were pure electromechanical devices with no CPU, or any electronic components at all. On receiving a line feed code the paper was advanced a line or with a carriage return the printing carriage was brought to the leftmost position. All of this was done purely by electrical and mechanical operations.

Early teletype machines were in use as early as the mid 19th century, long before the development of the electronic computer. My first experience with them was in the 1970's. These were mostly the traditional mechanical printing devices but thermal printing was being introduced. These early devices did not have a CPU.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.