Taming the net? You must be kidding!

Leonardo Chiariglione - CSELT, Italy

The first part of the title of my speech is drawn from the title of this session. The second part of the title was my first reaction when I read the title of the session.

Why? In my mind the word "tame" is associated to a ferocious beast, like a lion or a tiger, which you put in a cage and little by little you teach how to behave, so that the wild beast becomes tame.

The net is nor wild nor tame. The net is the virtual equivalent of the physical land of the Wild West. That land was not wild not tame. People made it wild. So people today are making the net wild.

So, would "Taming the user?" be a better title? I would rather not go that way. Acting on users is not the kind of action that one can safely propose at the dawn of a new millennium.

What then about "Setting rules for the net?" I would have personally a problem with this title. The word "rule" implies "You shall do this or shall not do that, otherwise…", not something that I would like to welcome for the new world created by the net. Not better is the word "regulation" on some aspects of which I would be delighted to give my opinion, but not today. The worst of all would be the word "directive", something amply used in Europe, particularly on matters related to communication. To make you understand why I abhor this word I will have to tell you about the D2-MAC directive. In summary the directive said: "Those who want to start a satellite TV service shall use the D2-MAC standard. Those who want to start a digital satellite service shall use a standard approved by a European standardisation body". I can imagine the grin on the face of those who wrote this sentence when they withheld what they imagined would follow in their pen: "and before this happens…" All useless. Today D2-MAC is dead and the users of digital TV services using the MPEG-2 standard are counted by the millions in Europe alone.

I would be content with setting some principles for the net. This is the way I will expand my thoughts on the subject.

First the materialisation of the net are the bits. Bits on the net are usually not random. Some of them may be, but there is not much fun toying with them. Interesting bits are the product of human minds. Actually some of them could be the product of machines, but I will not deal with this aspect. The creators of those bits own them. They usually do not keep the bits for themselves and because they own them, they can set the rules for the use of those bits.

"Thou shalt not steal your neighbour's bits" may be one principle for the net. But theft is not the only way to get access to your neighbour's bits. You can talk to the owners of the bits, negotiate with them and make a deal with them so that you can become a legitimate user of those bits.

You need not become an active user, though. You can add your bits to the other guy's bits, if you make a deal with him, and make something new. The combination of your bits with the other guy's bits makes something new, probably more valuable than the individual pieces.

Other people can join this bit partnership. The more people add bits, the more valuable the bit combination becomes, but the more complex the network of "rights to the bits" becomes.

All this is nothing new. If you replace the word "bit" with "content" all that I have said above has happened for millennia, in ways that were originally not codified, with more and more awareness of the complexities and implications of the processes as time went by. For some centuries these processes have evolved with a growing role of economics and law.

So, nothing new under the sun? Yes and no. True that the Berne Convention in its centennial existence has established the general principles of the rights of the author and his associates, but the practical implementation has been left in the hands of a legal device established almost 300 years ago, the copyright law.

This legal device has worked well for a long time. The Dutch printers of the 17^th and 18^th centuries might have thought disingenuously that copying the content of a book by printing a new one was a legitimate way of doing business, but the English Parliament in Westminster thought otherwise. Today the applicability of this legal device is more and more strained.

What about copying the content of a CD Audio or a compact cassette? Conceptually this is not different from copying the content of a book, but unlike reprinting of a book it can be done by millions of people at the same time. Since you cannot put a policeman in the house of everybody owning a double deck cassette, Europe has resorted to a less than solomonic decision, the blank cassette tax. If I buy a cassette for recording the speech that I am making now, I have to pay a tax to compensate for the  revenues that rights holders lost because my neighbour has decided to copy a song instead of buying it. And now people are thinking of extending the tax to other media such as recordable CDs.

But this is an easy problem compared to what people can do with an MP3 file. You buy a CD, you put it in your CD-ROM drive, rip off a track, compress it according to the MPEG-1 Audio Layer III standard (MP3) with a program you can find anywhere. Then you can decide to send it to your wide circle of friends, say, a million people, because you are a popular guy. You can bet that the author, the performers and their business associates are not amused by what you do.

The European Parliament has something in store here, too. They want to make the telcos carrying the bits that have been illegally misappropriated responsible of the bits carried. Something like DHL or TNT being made responsible in case one of their trucks or planes carries an envelope with a stolen stamp in it.

More is coming. An American entrepreneur - never was a word more faithful to its roots - had the idea of offering space in his huge disks to store the MP3 files that consumers create from their CDs. Consumers can then enjoy the songs anytime from anywhere. Not so, said the organisations managing the rights of the bit rights holders. The American courts to which the case was brought so far have ruled that those organisations are right and the entrepreneur is wrong.

And this is just the beginning. A young guy in his first year in college had a smart idea. Why not allow individuals to share their MP3 files? He devised a protocol by means of which MP3 files in a hard disk can be made "visible" to other people. They just have to tell some central entity - let me call it the MP3 Big Brother - that the files with the titles of the songs can be found at address so and so. People looking for "A hard day's night" go to the central entity, make their search and are given a list of places to go to and download from. The young guy has left college now, a precondition, one would say, to become a successful entrepreneur.

The organisations protecting the rights of bits rights holders and some of the bit rights holders themselves are resisting this move in the courts. The verdict has not been issued yet.

This last case looks now like history in this age where yesterday's technology generation is superseded by a new one. Why should one rely on a central authority to get information? Let everybody post their files and let everybody make his searches to find what they want. The disadvantage (from an entrepreneur's viewpoint?)? There is no one to gain from offering the "service". The advantage (from a user's viewpoint)? There is no one the organisations protecting the rights of bit rights holders can sue. True, but you can still trace back a file to its origin. No longer so. In another mutation of the protocols the address of the source is encrypted so that it is not even possible to trace back the source.

Compared to the service of hosting the individuals' MP3 files, the last examples have one remarkable feature in that we are no longer talking about the data themselves, but about data that point to the data. As much as the librarians of the Great Library of Alexandria (not the one in Virginia, USA, but the one in Egypt) 2000 years ago not knowing how to classify books about Ethics, Theosophy, Philosophy etc. uninventively decided to classify them as Methaphysics (i.e. after physics) so today people have started to call data about data as "metadata". Where data end and metatada start is something I would love to talk about, but not today.

As the examples I have given above demonstrate, the new frontier of the web is no longer content, it is metadata. I am saying this because it is getting more and more accepted that digital content must be protected if it is to exist as a direct source of revenues for those who have created it. Imagine that you are in an old fashioned market square, where you are allowed to display your goods on a 100 square metres of space and you are the only one to look after your goods. If at the end of the day you find that half of your goods have been stolen, I bet you that if you go to the police to complain you will not find friendly ears. This in western countries. In other countries you may have the experience of a friend of mine who had his camera stolen by a thief in India and was scolded by the local police for carelessness when he denounced the case. If your business relies on people seeing your goods you have beetter rent a shop with many windows and display your goods behaind sturdy glasses. Then, if somebody breaks the glass and steal your goods I can bet that you will find friendlier ears with the police (but you may like to make an insurance as well).

Content protection, I was saying, is advancing. Pay TV is already encrypted, DVD is encrypted and encrypted music files can already be purchased on the net. So, no problem? At the conceptual level, no. Practically, yes.

For the purpose of what I want to say here DVD is a good example. There is a single protection scheme that is used in all devices. With one note. The DVD world is divided in 6 parts (Totus mundus divisus est in partes sex, quarum unam… a modern Caesar would say in his "De bello contenuti" of the 21^st century) and content packaged for one region will not play on devices manufactured for another region. This is a rudimentary form of "usage rules" whereby the holder of the bit rights decides to set some rules for the consumption of his content.

Pay TV is a not so good example, again for the purpose of what I want to say here. You subscribe to a service provider, he gives you a set top box using which you can watch the content from that particular service provider. But don't think you can watch, even if you are ready to pay for it, the content from another service provider. You cannot, unless the two have decided to team up and share the technology, because they have found it good for their business or because the law has ordered them to do so.

The third case is the one where you buy protected music from the web. You go to one web site, download the player on your PC, buy the music and listen to it using the player of that site. You go to a different site and you do the same. So far so good. But what if you move the music files to a portable device, or to your car stereo or to your living room stereo? How are you going to translate one file format to another?

But this is only one part of the story. The advantage of music in the form of a file is that you can buy different rights to that file, a positive extension to the primitive DVD usage rule: you can play the music ten time, for one week, you get it for free once you have sold it to ten friends etc. How are these usage rules going to move and be understood by the different devices?

Even more complex are the issues related to metadata. Content exists and can (will) be moved on the net. But where is content, who holds which rights to it, how is it described, how is the use of metadata going to be charged (or not charged)? If metadata are valuable, how are they going to be protected? Clearly, if we talk just about a music file, things are simple (not really…), but what if it contains lyrics, scores, photos, videos, animations, a MIDI file, an interview with the singer, the statement of an opinion maker, the rating of an independent agency, news related to the song, information about how the sales of the song are going etc.?

The packaging of so many related pieces information around a single piece of content - the music file - allows me to go back to the beginning of my speech, when I talked about adding bits to bits. Imagine I am a member of a garage band and I want to add some fancy things to our latest hits. How am I going to find what I need? Assuming that I am in the business of making content and not in the business making deals worth a few dollars each, how am I going to negotiate the terms of use of other people's bits?

This would require opening up a complete new set of technologies that allow people to delegate to other non-human entities the handling of tasks that appear to be too repetitive to be worth the effort, but still too important to let them go. As I would not have the time to present them for what they are worth, I will just mention that the issues addressed in my speech today are part of a new staandard project that the MPEG group has just started, MPEG-21, Multimedia Framework.