Editorial

In previous years this magazine has keenly covered the progress of the C++ Standard through the ISO standardisation process.

The International Standards Organisation (http://www.iso.ch/) is somewhat bureaucratic in nature, as its members are mostly national standards bodies, themselves paid for by their sponsoring governments. Thus ISO's structure and behaviour tends to be governmental.

I attended a couple of national standards body meetings, at BSI, but was never fortunate enough to attend a real convening of the ISO Working Group. Each national body sends a delegation of industry experts to discuss the work in progress and to tender their national vote on contentious issues.

The C++ Working Group would meet concurrently with the American national body meeting, ANSI, since the majority of attendees participated in both. Some members could be representing their country in the ISO meeting, and their company or their own personal interests in the ANSI meeting. Thus an individual might vote differently on an issue depending on which hat they'd be wearing at the time.

In contrast, over the past couple of years, I have been attending the meetings of the Internet Engineering Task Force (IETF). The Internet has no government, but without standards it could never have been created, or continue to evolve.

The IETF meets three times a year - always in unglamorous locations - to deter the junketeers. Two thousand engineers descend upon some provincial capital to consume the local beer and clog the local ISPs with traffic.

The corridors of the conference hotel are littered with isolated nests of laptop users. All crouching on the floor, in their fading free t-shirts, silently hacking away next to each other. They congregate not to talk, but to overload the power sockets hidden behind the ornamental vegetation. They don't talk because they're plugged into their own online virtual community. A wireless LAN is installed throughout the building, and every laptop gets a wireless modem. The engineering discussions occur in the physical and virtual plains simultaneously.

The IETF seems to have been organised as a somewhat unique social experiment. To attend is to be a member, and every member has equal voting rights. Members represent only themselves, with flouting of employer or national affiliations being frowned upon. The Internet and the IETF have a symbiotic relationship explaining why they are both as democratising as each other.

There have to be checks and balances in every organisation, but beyond these the IETF has very few rules and regulations. The work is broken down into perhaps a hundred Working Groups. There are no closed sessions - all Working Group activities are advertised and open to all.

The Internet Grey Beards, those who hark from the beginning of internet time, hover at the back of the working group meetings offering up their ancient wisdom to prevent groups from floundering, or falling down rat holes never to return.

The Working Groups meet to discuss and resolve issues, but there is no voting mechanism. Work always has to progress by general consensus, as in the ISO Working Groups, but when it comes down to a vote, there is no formal vote. Each Working Group Chair has their own technique, but the usual litmus test applied is the raising of hands, or more amusingly the loudness of humming. Humming is an analogue test, as a few loud hummers is equivalent to many quiet hummers, so the hummer can express the depth of their agreement with the strength of their hum.

Damn, my introduction doesn't work as a conclusion.