Skip to main content
Retrocomputing

Return to Answer

added 2234 characters in body
Source Link
Raffzahn
Raffzahn
  • 248.3k
  • 23
  • 720
  • 1k

TTY were only a thing very early on and after that only relevant for cheapniks barely able to afford the computer with no money left to acquire contemporary terminals. That an image from hobbyist computing, not business applicatinsapplications.

Now lets take a look at the titular question:

What did order processing on a teletype look like?

It didn't exist. It was all about optimizing the existing process of a customer filling an order form and some (back) office turning it into a data set.

Just imagine what company would need and thus have a computer system for order handling in like 1970? For sure not a small one doing a dozend orders a day. And any company/department larger then this would already have a punch card based system.

If large enough to have subsidiaries, like distribution centers and/or agents, the next logical step is not to replace the working system by some less capable system with cumbersome teletypes, but improve data entry. A large enough company may already turn the forms a customer fills out into punch card orders as a local site and send a days stack to the central (or next) order processing site, getting back print outs with the stuff ordered whenever the delivery truck comes.

Improvement here is the next logical step, but not by adding extreme costly online communication (for what benefit?), but improve and speed up data entry. This is were systems like the Cogar 4 or Datapoint 2200 came into play around 1970. Local data entry with support for forms, basic referencing and validation, collecting entries and transferring them as batch into the existing mainframe environment. That these system as well could be used as low end data processing was a welcome side effect.

Not to mention that none of this replaces the existing (more or less) central data entry department, like for customers sending in a request letter with a sub from some newspaper add or alike. Similar customers calling in.

While this may of course differ between sectors of business, the basic situation is the same, no matter if auto parts, model railways or corn flakes. For smaller companies, growing into sufficient size, the solution is not siting down and envision a minimalist system, but buying a working solution - as shown.

So, in the end, (almost) no use case for a tty user interface.

TTY were only a thing very early on and after that only relevant for cheapniks barely able to afford the computer with no money left to acquire contemporary terminals. That an image from hobbyist computing, not business applicatins.

TTY were only a thing very early on and after that only relevant for cheapniks barely able to afford the computer with no money left to acquire contemporary terminals. That an image from hobbyist computing, not business applications.

Now lets take a look at the titular question:

What did order processing on a teletype look like?

It didn't exist. It was all about optimizing the existing process of a customer filling an order form and some (back) office turning it into a data set.

Just imagine what company would need and thus have a computer system for order handling in like 1970? For sure not a small one doing a dozend orders a day. And any company/department larger then this would already have a punch card based system.

If large enough to have subsidiaries, like distribution centers and/or agents, the next logical step is not to replace the working system by some less capable system with cumbersome teletypes, but improve data entry. A large enough company may already turn the forms a customer fills out into punch card orders as a local site and send a days stack to the central (or next) order processing site, getting back print outs with the stuff ordered whenever the delivery truck comes.

Improvement here is the next logical step, but not by adding extreme costly online communication (for what benefit?), but improve and speed up data entry. This is were systems like the Cogar 4 or Datapoint 2200 came into play around 1970. Local data entry with support for forms, basic referencing and validation, collecting entries and transferring them as batch into the existing mainframe environment. That these system as well could be used as low end data processing was a welcome side effect.

Not to mention that none of this replaces the existing (more or less) central data entry department, like for customers sending in a request letter with a sub from some newspaper add or alike. Similar customers calling in.

While this may of course differ between sectors of business, the basic situation is the same, no matter if auto parts, model railways or corn flakes. For smaller companies, growing into sufficient size, the solution is not siting down and envision a minimalist system, but buying a working solution - as shown.

So, in the end, (almost) no use case for a tty user interface.

Source Link
Raffzahn
Raffzahn
  • 248.3k
  • 23
  • 720
  • 1k

But in between those [Batch vs. Terminal], there was an era of 'interactivity, but not as we know it', when computers supported interactive work by teletype.

Not really.

To start with, these were complete different usage scenarios. Batch didn't turn into terminal use - or got replaced by it. Batch is like mass production on industrial scale. It still happened the same way when terminals became widespread, as it's purpose didn't change.

In relation, terminal based processing did only replace parts of batch processing by improving data entry, replacing key punches. This evolved soon into form based (block mode) operation.

Teletypes were only for a very short time as generic devices, and soon replaced by terminals known as glas-TTY, terminals that operated like a TTY, just without the paper part. Real TTY were only kept in areas where there was a need for a paper protocol (like on system consoles) or simply as printers.

Finally glass.TTY are essentially what micro computers presented their interface on - from early CP/M all the way to Windows command line.

So, user interfaces for teletype looked much like the ones for later microcomputers in line mode. From BASIC to dBase. There is no principal difference.

Important here is to remember that the time of 'pure' TTY use was a quite short one, as glas-TTY and their ability to do 'full screen' by rewriting the screen emerged soon.

With glass-TTY emerging in the late 1960s (Datapoint 3300 1969, VT05 1970 and a whole armada of microprocessor based systems since ca. 1973) everything that is known from later microcomputers was ready available to their users.

Take for example the hundreds of thousands of PDP-11s, many of which were connected to ASR-33 and similar teletypes.

Not really. The PDP-11 was introduced in 1970, when glas-TTY were already a thing. More important it took quite a while until it reached sales that big. I'd say late 70s to mid 80s was the high time of PDP-11type computers, right when VDU terminals were usually the only kind attached.

Sure, our picture of that time are nice photographs of a huge PDP-11 wall with some TTY, maybe topped by a nice young lady in a lab coat - the computer equivalent of some youngster holding a soldering iron at the tip. It's a tainted view.

TTY were only a thing very early on and after that only relevant for cheapniks barely able to afford the computer with no money left to acquire contemporary terminals. That an image from hobbyist computing, not business applicatins.