How to solve a problem like text telephony

28 July 2010 | Technology

When I was first asked to look at it, text telephony was very much a niche technology. It was also "non-native", i.e. it was using networks (in this case the analogue telephone network) to transport information (text) for which it was not at all designed (it's a voice network). This creates huge problems, because not only is the equipment not designed for text (textphones end up masquerading text as voice in the form of audio tones), the network itself is also unable to properly support it.

There is no control and signalling to distinguish text calls from voice calls. And provisions for encoding, transporting and processing voice (such as echo cancellation and error correction) are not designed for such as voice masquerading text and therefore end to end connectivity is easily broken. Worse still: even if you solve these issues (partially) at some point in time and for a given network and equipment base, it is easily compromised again at any time when networks and devices are upgraded or replaced.

By the time I got involved, very little had been done to address these strategic issues. Most activity was tactical, solving a problem here or there or releasing a newer version of some textphone model. Quite clearly, that sort of reactive approach was never going to resolve the key problems or lift text telephony out of its specialised, access technology niche. That matters, because the result of being in this confinement is ongoing disenfranchisement of users: lack of connectivity, availability and interoperability reinforced by high equipment cost, outdated designs and interfaces and stigmatisation of the user group.

To me the strategy was clear: taking advance of general convergence towards packet data networks for telecommunication to normalise real-time text as a communication paradigm. The route to achieving that was by mainstream standards for consumer equipment and networks, in this case within IETF.

The first step was to create awareness of the issue and support for the strategy amongst industry and the broadest group of stakeholders involved in the standards working groups. The instrument we created was RFC3351, setting the benchmark and becoming the lever for future engagement.

After that came many years of hard work, ongoing engagement with the IETF community, in-depth technological debate and sometimes hard discussions. But the ball had started rolling and inexorably real-time text was becoming part of the mainstream. Along the way lie key milestones such as RFC4504, ensuring the devices will not ignore text users and RFC 5194 (for which I was one of the editors) establishing a full framework for real-time text.

But perhaps the best illustration of how successful the chosen strategy has been is today's publication of draft-ietf-mmusic-media-loopback-14.txt, where real-time text has taken its rightful place as an equal modality alongside voice and video.

It is in this kind of activity, invisible to most people and often unappreciated by funders and management alike, in which the best promise for a more equal future resides. It's the best long term prospect for full opportunity and fulfilment for many of those currently confined to "special needs" niches. Today, I am a happy man.