AHC/WI: Space Age Personal Computers

[Delta Force]

These wouldn't be personal computers the way we think of them, but they would still be able to do a lot of things.

By the 1950s and 1960s, most homes had telephones and televisions. What would it have taken for that to progress to videotex type systems? An electronic machine (we'll call it a terminal) could be developed that would use the telephone connection as a modem, allowing communication with a central computer. The terminal wouldn't actually have any processing capabilities of its own, simply relaying information between the user and the central computer and using that information to control the television display. Text wouldn't be too difficult to transmit, so it wouldn't use too much bandwidth. Images and video could be displayed if desired, but they wouldn't be very high resolution and would use a lot of bandwidth and central computer power. Although quite simple, the terminal would allow almost all functions of the 1990s/early 2000s internet to be done. Hitting the refresh button would consume a large amount of bandwidth (although workarounds might be developed, such as the central computer sending a "new information" signal when something actually does change to let people know refreshing will do something) and even simple graphics or animations would require a stream of information to tell the terminal what to display, but it would be possible to use databases, message boards, and send email and text messages (even instant messages).

Could something like this be developed for widespread consumer use in the 1960s, or was technology not quite up to par? It would be up to twenty years before the development of France's Minitel system, the most successful pre-internet online service. Might this at least expand the use of the internet to more people than it reached in the early days of ARPANET, with there being a lot more terminals?

 

[Mike Stearns]

There were a couple of systems like this that were tried in Canada in the 70s and 80s, Telidon and Alextel. Part of the problem with these systems was lack of user content. The other part of the problem was that Minitel, Telidon and Alextel were rolled out just the internet was emerging. The power and adaptability of the internet, as well as plummeting computer prices made videotex systems look expensive and obsolete. As a result videotex technology was quietly scrapped.

 

[ManintheField]

The trouble with this is that none of the protocols to support this type of remote computing relationship existed at the time. You'd need to invent them wholesale and these are things that only really emerged from the experience surrounding early networks in the late 60's and 70's. So you'd need to move work on early computer networks in general backwards in order to generate the necessary firmware.

There is also the other problem that, in order to get any kind of scale whatsoever, the central server would have to be absolutely massive and absolutely massively expensive.

EDIT: Well, to be more specific, the rudiments of what was necessary existed by the late 60's, but standard implementations didn't exist until the 70's. If you avoid a the economic slow down surrounding events in the 1970's you might be able to move some aspects of the internet forward by most or all of a decade (mailing list forums in the mid 80's instead of the early 90's, perhaps), but really these kind of things took as long to develop as they did for good reasons.

EDIT2: Oh, and anything but text is a no-go. Processing power just isn't there. Network throughput isn't there. The entire reason we got that kind of thing in the late 90's/early 00's was because the infrastructure started to finally come into existence on a wide scale and micro-processors allowed the kind of processing power necessary to handle images of any real resolution to exist for something resembling a reasonable cost.

Just trying to manage digital video on the computing equipment available in the 60's or 70's would take weeks and weeks for split seconds of output, even if we magic the software into existence, let alone the power necessary to translate it into an analog signal capable of being sent over the PSTN.

 

[Dathi THorfinnsson]

These wouldn't be personal computers the way we think of them, but they would still be able to do a lot of things.

By the 1950s and 1960s, most homes had telephones and televisions. What would it have taken for that to progress to videotex type systems? An electronic machine (we'll call it a terminal) could be developed that would use the telephone connection as a modem, allowing communication with a central computer. The terminal wouldn't actually have any processing capabilities of its own, simply relaying information between the user and the central computer and using that information to control the television display. Text wouldn't be too difficult to transmit, so it wouldn't use too much bandwidth. Images and video could be displayed if desired, but they wouldn't be very high resolution and would use a lot of bandwidth and central computer power. Although quite simple, the terminal would allow almost all functions of the 1990s/early 2000s internet to be done. Hitting the refresh button would consume a large amount of bandwidth (although workarounds might be developed, such as the central computer sending a "new information" signal when something actually does change to let people know refreshing will do something) and even simple graphics or animations would require a stream of information to tell the terminal what to display, but it would be possible to use databases, message boards, and send email and text messages (even instant messages).

Could something like this be developed for widespread consumer use in the 1960s, or was technology not quite up to par? It would be up to twenty years before the development of France's Minitel system, the most successful pre-internet online service. Might this at least expand the use of the internet to more people than it reached in the early days of ARPANET, with there being a lot more terminals?

Can't happen. The first really viable modem was the Hayes Smartmodem in 71, which had a microcontroller on board - simply to deal with dialing, etc.

Then, you'd need a fairly significant amount of computing power (for the day) just to translate display the text on the screen.

Both of these are likely doable with discrete electronics, or small scale integration chips - but it'll be expensive - and you NEED electronics at the home end.

So, late 60s is probably the earliest possible. By which time microcomputers are just around the corner.

 

[Delta Force]

Can't happen. The first really viable modem was the Hayes Smartmodem in 71, which had a microcontroller on board - simply to deal with dialing, etc.

Then, you'd need a fairly significant amount of computing power (for the day) just to translate display the text on the screen.

Both of these are likely doable with discrete electronics, or small scale integration chips - but it'll be expensive - and you NEED electronics at the home end.

So, late 60s is probably the earliest possible. By which time microcomputers are just around the corner.

I had a feeling this was one of those cases where the technology and procedures seem more simple and obvious in hindsight. Turns out they didn't even have timesharing computers until around the mid-1960s (akin to cloud processing). A few other things would have to be butterflied too, since I don't think computers even had screen displays until the late 1960s (they used teletypes and printers to command the computer and see results).

I was thinking that a dumb terminal would drive costs down enough to make it practical in the 1960s, but the main computing power, bandwidth, and other technologies would be cost prohibitive anyways, even if they were possible. Since the technology and procedures are more late 1960s/early 1970s (I'd say mid-1960s at the absolute earliest), would making the end user equipment closer to an early video game console help to further drive down costs? The terminal itself could manage the image display, so the central computer wouldn't have to process the signal information and send it over the wires, which would have been a major issue for scalability in any case.

 

[ScrewySqrl]

you'll also need the Microchip to make this affordable to the average consumer.

The Microchip came out in 1968, and by 1976 (8 years later) there were affordable home computers and video game consoles (the TRS-80 Model I, Commodore PET, the Apple I and II, The Odyssey and Atari 2600)

 

[dduff442]

Minitel demonstrated some advantages of state monopolies, of which we've only heard of the disadvantages in recent decades. Standards competition, which allows powerful companies to parasitise others' internet investments and inhibits many sorts of activity, was avoided by diktat and great economies of scale achieved.

France had a lot of advantages in Telecoms at the time minitel was rolled out. De Gaulle had a famous hatred of the telephone and by the time he left power France's phone system needed heavy investment. The result was that it skipped a generation and by 1980 all switching etc was fully digital.

1960s is too early for this to be economical IMO, but it would have been possible to beat minitel to market by several years. According to Wikipedia, the ZX80 was £99.95 at launch and didn't need a dedicated screen. 1976-'78 was probably an achievable target date.

 

[SunilTanna]

Even a so-called dumb terminal needs a significant amount of processing power.

It doesn't send raw keyboard signals to the central processor, or receive conten back as raw CRT signals. No design usable on any significant scale could possibly work that way with raw data, because the bandwidth usage and response lag would be unacceptable.


Instead a so-called dumb terminal understand ASCII or similar, and translate that into CRT signals, etc. That could be done with a microprocessor or discrete logic, but it requires least early 70s tech before it's conceivable as a consumer device.

Moreover if we are using phone lines as the network connection, we also need good a/d and d/a convertersm preferably with error checking and error connection at each end, including the consumer end. Again, requires mid 70s tech or better to be realistic.

 

[Delta Force]

Minitel demonstrated some advantages of state monopolies, of which we've only heard of the disadvantages in recent decades. Standards competition, which allows powerful companies to parasitise others' internet investments and inhibits many sorts of activity, was avoided by diktat and great economies of scale achieved.

France had a lot of advantages in Telecoms at the time minitel was rolled out. De Gaulle had a famous hatred of the telephone and by the time he left power France's phone system needed heavy investment. The result was that it skipped a generation and by 1980 all switching etc was fully digital.

1960s is too early for this to be economical IMO, but it would have been possible to beat minitel to market by several years. According to Wikipedia, the ZX80 was £99.95 at launch and didn't need a dedicated screen. 1976-'78 was probably an achievable target date.

It looks like things would be somewhat complicated in the United States, as AT&T was the telephone monopoly, but it had been banned from the computer industry in 1956 because it was growing too big. That probably doesn't ban it from being used as a transmission system, but that means a high level agreement between companies or legislation or a court order forcing AT&T to let others use its systems would be required. That would seem to favor the companies that make business peripherals such as IBM. If AT&T could use subsidiaries to bypass the ban, it could always work closely with its subsidiary Western Electric/Teletype to make machines. I'm not sure on the wording of the agreement, but perhaps AT&T could argue its machines aren't computers or something like that, fighting over definitions.

It would probably take deregulation to get anyone to put a device on the network, and that only took off in the 1980s with the breakup of AT&T in telecommunications and the end of the Civil Aviation Board in aviation.

 

[dduff442]

It looks like things would be somewhat complicated in the United States, as AT&T was the telephone monopoly, but it had been banned from the computer industry in 1956 because it was growing too big. That probably doesn't ban it from being used as a transmission system, but that means a high level agreement between companies or legislation or a court order forcing AT&T to let others use its systems would be required. That would seem to favor the companies that make business peripherals such as IBM. If AT&T could use subsidiaries to bypass the ban, it could always work closely with its subsidiary Western Electric/Teletype to make machines. I'm not sure on the wording of the agreement, but perhaps AT&T could argue its machines aren't computers or something like that, fighting over definitions.

It would probably take deregulation to get anyone to put a device on the network, and that only took off in the 1980s with the breakup of AT&T in telecommunications and the end of the Civil Aviation Board in aviation.

Though at least up until its forced breakup it would have been a single negotiating point for a third-party owned system.

Some possibility to negotiate reduced-price connections might have existed. Given that minitel terminals were distributed free, one big attraction for Poste, Téléphone et Télécommunications was obviously the lure for people to spend long hours on the phone. AT&T might have been persuaded to offer deep discounts for the same prospect.