|WikiProject Computing / Hardware||(Rated C-class, Mid-importance)|
Split speed videotex modems
I worked in the Canadian videotex industry during the 1980s, forming Picture Data Inc. with several others. PDI developed transaction software for the videotex industry and created content pages for the Ontario government's tourist information videotex service, hosted by the Infomart subsidiary of Torstar corporation in Toronto. The split speed 150/1200 modems operated in full duplex mode and were, in fact, designed and manufactured by Gandalf Technologies in Ottawa, packaged in the company's signature blue metal case. Gandalf was an innovator in the design of early computer communications equipment. A brief history of the company with a picture of one of its modems is given in Wikipedia under "Gandalf Technologies" (http://en.wikipedia.org/wiki/Gandalf_Technologies). (I believe, but am uncertain, that the very earliest videotex modems in Canada were a form of AT&T Bell 202 design and would likely have been manufactured by Northern Electric (Nortel). My recollection though is that, because they could provide only half duplex service, Gandalf's units were preferred by videotex users.)
The reason for the use of split speed modems lies in the low bandwidth of telephone lines in use in the 1980s. Twisted copper pair telephone lines which delivered service into customers' residences, and their controlling telephony electronics, were designed for voice transmission. This required a bandwidth of about 4,000 HZ, the accepted design limit at the time for applications. However, 4000 HZ barely allowed enough capacity for data transmission of 1200 baud in one direction. Designers of videotex modems however realized that users requesting data typed slowly, and needed to transmit only a few characters to request a videotext 'page' on the 'up' link. The requested page, containing up to several thousand characters, would be transmitted back to the user's terminal on the 'down' link. (For this process, the videotex industry borrowed the term 'downline loading' from the telephony industry, eventually contracted to 'download'.) One of the innovations of the Canadian Telidon encoding protocol was to encode a page's graphic images into composites of geometric primitives such as circles, lines and squares along with colour and sizing information. This dramatically compressed the amount of data needed to encode a page of graphic data compared with the British CEPT1 protocol, allowing more graphic-rich transmission at 1200 baud over a telephone line. Telidon offered six colours and seven grey levels in addition to black and white, displayed on screens with resolution of 256 x 200 pixels.
The modems then were designed to steal just a little of a voice-grade line's bandwidth for the uplink (150 baud), reserving most of the bandwidth for page transmission, allowing for full, albeit asymmetric, duplex service. Full duplex operation was important because the transmission time of a page and the time required for the decoding hardware in a user's terminal to decode the graphic primitives and paint the screen made for a painfully slow user experience. Thus user's would often see enough of what was wanted before page transmission was complete, and immediately enter a few characters on the keyboard to request the next page. These could only be transmitted to the host service's computer if the modem supported uplink transmission at the same time as the previous page was still being downloaded. So, in the early 1980s, full duplex service meant that the uplink was quite slow (but not noticeable) in order to minimize the transmission time of page content.
Bell Canada conducted trials using split speed modems in Toronto, Ottawa and other Canadian cities in the early 1980s to test not only their feasibility, but the effect that videotex services would have on regular telephone service. There was concern that because, unlike European countries where bandwidth is charged per call and by a call's duration, in Canada and the US there was no bandwidth or transmission volume charges. So, theoretically, with no deterring costs to the contrary, users might stay 'online' as long as they wanted during one call to a videotex service, and download a data volume limited only by the transmission capability of the transmission lines and modems. This was obviously a concern for the telephone monopolies as it potentially meant big capital outlays to 'beef up' transmission capabilities to maintain acceptable service levels for voice transmission. In fact, videotex services, as the first publically available data services, gave rise to the bandwidth game still being played on the Internet: how do transmission service providers such as telecommunications companies induce subscribers to pay for the capital improvements needed to satisfy an ever increasing thirst for bandwidth? As users increasingly download high definition video, the stress on existing networks, especially cable companies, results in reduced service for all users. This annoyance to customers results in pressures on the providers to constantly upgrade their networks at no small cost.
X.25 and Bell 212 Modems
There was, in fact, no technical reason why videotex services could not access Bell Canada's X.25 network known as Datapac using Bell 212 modems. At Picture Data Inc. we actually did so on a quite regular basis to test the screen build rates of decoding hardware unlimited by transmission bandwidth. The real problem was economic: 212 modems were expensive, while Gandalf split-speed 150/1200 modems were just a little over $300, a bargain at the time. As well, Datapac accounts were expensive since the capital cost of high speed leased lines dedicated to packet switched service was high, and the target market were businesses transmitting high value data such as banks, insurance companies and national printers such as the Globe and Mail. It was this market which had justified Bell and the other members of Trans-Canada Telephone System creating a commercial X.25 service in the late 1960s. Videotex, on the other hand, offered very low value data for which equally low cost communication facilities were needed. Hence, the use of 'free' public switched telephone networks was the only viable means of transmission at the time and, for that, the lowest cost modem that could use all of the available bandwidth. Datapac was never a viable option for videotex. 18.104.22.168 (talk) 20:31, 13 August 2009 (UTC) Duncan George
Lacking: CBS and AT&T joint venture
There's a reference to it in the Wiki article on Telidon (http://en.wikipedia.org/wiki/Telidon).
The last sentence of this section seems wrong
From 1980 onwards he designed, sold and installed systems with major UK companies including the world's first travel industry system,the world's first vehicle locator system for one of the world's largest auto manufacturers and the world's first supermarket system.He wrote a book about his ideas and systems which among other topics explored a future of teleshopping and teleworking that has proven to be prophetic. Before the IBM PC, Microsoft and the Internet, he invented and manufactured and sold the 'Teleputer', a PC that could receive TV programmes and communicate using its Prestel chip set.
The dates aren't very clear, but from context it would seem that this last bit about being "before the IBM PC, Microsoft, and the Internet" is wrong. I'll leave it here and let somebody else decide, here are the pertinent inception dates:
- IBM PC: 1981
- Microsoft: 1975
- The internet: 1969
I think the whole section about Aldrich is questionable. I was involved in the British viewdata industry fron 1983, actually working closely with Prestel themselves from 1985, and never heard of the guy or his apparent accomplishments. Indeed, I saw other organisations making these advancements. Talk:Michael_Aldrich is worth a read, but I remain unconvinced...Robirrelevant (talk) 22:46, 30 June 2010 (UTC)
- - I was at ROCC/Rediffusion in the 1980's and was involved in the production setup & testing of the Teleputer, it only really being sold from about 1981 late/early 1982. The O/S was a derivative of CP/M and it contain a philips/mullard modem board. It was normally sold as a prestel terminal or, to connect to private services, often based on the R rangle mini products we sold. —Preceding unsigned comment added by 22.214.171.124 (talk) 11:28, 1 October 2010 (UTC)
Knowledge of Telidon/NAPLPS in North America
I was the point person for the Minneapolis, MN, newspapers in the videotex world during the early 1980s, when this technology was (seemingly) important. I was a participant in most of the industry events during the period and developed several demonstration-type videotex systems.
My specific areas of expertise and knowledge include:
- The LA Times "Gateway" project
- The Teleguide public-access systems in Toronto, San Francisco and Phoenix
- Various small-scale shopping mall public access systems in the US and Canada
- The Telidon/NAPLPS display protocols
- The business-side and journalistic aspects of videotex and teletext in the 1980s
(I also have a collection of videotex screenshots from the period that may be of interest.)
The article, at least as it relates to the U.S. in the early '80s, does not capture the corporate competition that was going on between the three big players: Viewtron (Knight-Ridder), Gateway (LA Times and partners) and, to a lesser degree, Prodigy (IBM and Sears). At the time, the two newspaper groups thought that they were fighting for the future hearts and minds of readers everywhere. Videotex's centralized infrastructure fit nicely with the the way newspaper production processes were organized.
The LA Times' Gateway group was an interesting industry collaboration in competition with Knight-Ridder's Viewtron operation. Gateway members included all the papers that launched Teleguide public-access kiosk systems, as well as other papers that took lower-profile roles.
If anyone has questions or issues I'll try to help.