Hum

I woke up at the crack of dawn yesterday (okay, 6:45) to attend a 7 AM online meeting. The meeting, actually being held in Sweden where it was a more reasonable 1 PM, was the IETF75 Internet Wideband Audio Codec BoF.

I’ll see if I can unpack that expression. The IETF is the Internet Engineering Task Force. They’re the group that defines how the internet works. For example, when you download a web page (like this one), your browser sends a request packet to the server with the name of the web page you want, and the server sends back a reply containing the text of the page. The format of those packets, and the standard rules for processing them, were defined by the IETF.

There’s been a lot of interest lately in enabling voice conversations (“telephony”) over the internet. “Wideband Audio” refers to voice conversations with higher quality than your average telephone. In order to make this work, the computers on each end need to agree on a “codec”, a compact digital representation of the audio.

The IETF is a very funny organization. Although they write some of the most precisely worded, widely used, highly technical standards on the planet, the results are always referred to as RFCs, “Request For Comments”. When the IETF wants to have a preliminary meeting on a topic, it’s called a BoF, from the phrase “birds of a feather flock together”.

Currently, all the standard codecs suitable for high-quality conversation over the internet are patented, and cannot be used without paying royalties. The existing standards bodies in this area (like ITU-T and MPEG) are composed largely of representative from companies that make money from these royalties, and so are inclined to write standards that require royalty payments. The IETF has been much better about producing royalty-free standards, and so some people came up with the idea to write a new standard codec there, carefully avoiding patented technologies. The purpose of the BoF was to discuss whether to start this standardization effort.

The BoF was held in a room in Sweden, but there was also an online chat room and a live audio stream. People in the room gave very short presentations, after which there was a continuous stream of comments, by people getting in line for a microphone. Chat room participants could ask people in the room to get up and say something for them, which happened occasionally. It was an interesting setup.

The discussion was quite heated, divided principally between those who felt that royalty-free codecs are important and those who did not. It’s not yet clear where this is going.

The IETF is often described by a famous quote from David Clark:

We reject kings, presidents and voting. We believe in rough consensus and running code

To gauge consensus without a vote, the IETF has a “Hum”. The leader says “hum for the formation of a codec working group” and people hum if they agree. Then the leader says “hum against the formation…” and people hum if they disagree. If one’s a lot louder, well, that’s consensus.

In the chat room, people tend to type things like “*HUM* for the formation of the working group”, which may or may not make any sense.

Anyway, it was quite a morning.

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>