In the beginning..
..was Samual Fedida. He joined the Post Office research centre in the autumn of 1970, and jumped straight into a project looking at the practicalities of a "viewphone", a device for receiving computer data. Over in the states, Bell Labs was spending big money developing the "picturephone", or face-to-face video calling. This was already in trials and was flopping badly, mainly due to the bandwidth required - not many houses had the two high quality telephone lines required for each set. Even today, where the technology is actually practical, few people are interested in video calling. Instead, Fedida recognised that textual computer data required far less bandwidth, and could be transmitted via standard unmodified telephone lines. His previous experience setting up a computerised hotel booking system showed him "how important it was for the users to get directly at the computer, with no one in between", and this looked to be the ideal mechanism.
In contrast to simple passive broadcasting of information, such as teletext was to become, the vision from the outset was to provide an interactive service - information retrieval, word processing, electronic mail, links to telex, etc., and the aim was to stimulate off-peak usage of the telephone network. By utilising customers' existing television sets as the display device, it was thought that costs could be kept low.
There were several issues; the main ones being could a simple consumer system be practically used to access potentially unlimited quantities of information, and could the costs of effectively providing a timeshare computer system be reduced to the point that people would accept it. The target was a penny a minute, somewhat less than the £5.00 per hour typically charged by commercial bureaus to business users.
For two years, Fedida worked alone, and in 1972 produced the first working prototype system.
Around the same time, the BBC and IBA had been working on teletext, and in an unusual moment of rational thought, it was agreed that the two systems should use the same transmission and display formats. This would allow Television manufacturers to support both systems; it was obvious that viewdata should therefore use the same display system. By 1974, viewdata as we know it was being demonstrated to the manufacturers, and to the other important parties - the Information Providers, the people who would be generating the content and services being accessed via the system.
Friction was, however, developing between the Post Office and the television set manufacturers. In a total about-turn from their previous stance where only they could supply anything attached to the telephone line, the Post Office had specified that the necessary modems could only be built inside TV sets, thus effectively banning the adapter boxes for existing televisions that Fedida had envisaged as being essential from the start, and blocking the development of viewdata-only receivers. This contradictory stance was perhaps explained as a response to the fierce opposition by the television manufacturers to the early suggestions that the Post Office rent out adapters, seeing anything to do with televisions as their exclusive domain. The problem was, the television manufacturers were slow to produce suitable models, and they were extremely expensive. Indeed, less than a third of the 1500 sets ordered for a 1978 trial had actually been delivered.
Eventually, the restriction on standalone receivers was withdrawn, and in March 1979, with the initial London service about to start, and the eventual realisation that it was going to be impossible to persuade people to replace their television sets at great expense, the Post Office gave up and invited tenders for a super-cheap adapter. However none of the major manufacturers took up the challenge, and it wasn't until early 1980 that the tiny startup Ayr Viewdata managed to come up with a prototype. Despite the post office chairman describing it as "one of the most significant items I have seen" the Post Office did another U-turn and decided not to go the adapter route after all, after yet more pressure from the TV industry.
A year after launch, there were still only 6,000 subscribers, compared to the 50,000 originally forecast.
refs New Scientist