Johannes Gutenberg had the great idea of developing a system to print books using movable characters. He had not just the idea, which had been floating around for a long time (actually, it had been in use in China for centuries), but also the skill and will to develop a set of technologies capable of transforming the idea into a system working practically, efficiently and economically.
To achieve that goal, he needed money to make his investigations and trials and he found a wealthy person (an investor, we would call him today) willing to provide it. However, being a prudent man, Johannes hid the secrets of his developments from everybody, including his financial backer. In those remote times the patience of financiers was probably more robust than today’s, but not infinite. After some years and a considerable amount of money spent, his financial backer started a law suit. Gutenberg lost it and was financially ruined.
If Gutenberg had lived in a culturally more advanced city like Venice or Florence, which had already started issuing litterae patentes (“open letters”, whence the word “patent” comes from) since a few decades, he could have protected his invention with a patent. Then he would have started negotiations with some wealthy person and agreed on how to split the revenues. Unfortunately, life in 15th-century Mainz was not yet as sophisticated as in those two Italian cities and Gutenberg had only one way to protect his invention from robbers of what we call today his Intellectual Property (IP): hiding his secrets. If he had had his IP protected, everybody would have gained: he could have led a better life, his financial backer would have had a share of the revenues and the world at large would have gained from the mastery of his printing technology. The printing art might have had a different history.
Fortunately for Gutenberg, his story did not have a wholly unhappy end. After his genius had been revealed through the books printed using his invention, Gutenberg could escape the common fate of those times like ending his life as a beggar, thanks to the kindness of the Archbishop of Mainz, Great Elector of the Holy Roman Empire, who bestowed on him a tax-free sinecure with a life annuity of “cloth, grain and wine”.
The correlation between that world of 15th-century Germany and today’s is not straightforward, but the basic elements are still there: inventions happen, inventors wish to protect their IP, the state of the invention in the beginning is too crude for an exploitation, inventors need capital to transform their inventions into something exploitable, (venture) capitalists develop methods on how to manage their risks, etc.
One thing does not correlate, though. We no longer have Archbishops, whether of Mainz or not, whether Great Electors of the Holy Roman Empire or not, who are kind enough to recognise and concretely support a ruined inventor. And I am not sure it is not a good thing they are no longer around.
Patents were the timely answer to a necessity felt by a society that was becoming increasingly more complex and their use quickly spread throughout all of Europe. In those early years, patents took the form of a privilege granted by a sovereign (not necessarily, as in Venice, a republic) for a limited period of time (14 years in 17th century England). The two major revolutions that happened at the end of the 18th century – the ones that led to the formation of the United States and the Republic of France – gave patents a firmer legal ground in those countries that spread to other countries.
The USA Constitution of 1787 explicitly links private incentive derived from the patent “monopoly” to societal progress. It gives Congress the power “to promote the progress of the useful arts, by securing for limited times to inventors the exclusive rights to their discoveries”. The French bourgeoisie, imbued with Illuministic principles, drafted the French Constitution of 1791 with an explicit declaration of the natural right of an inventor, as an inherent right of Man, to the exclusive right to his invention.
Therefore, more than 300 years after Gutenberg, L.-J. Daguerre, S. Morse, A. G. Bell, T. A. Edison, A. and L. Lumière, G. Marconi and a host of other inventors took care to file patents of their inventions with their respective Patent Offices and they, their business associates and possibly their heirs gained conspicuous benefits from those inventions.
Those were the hay days of the individual inventor, who harvested all the benefits from his invention, but in the 20th century the centre of gravity progressively shifted from the individuals to the companies hiring them. Today the split of economic returns from patents produced by an inventor hired by a company varies from all benefits going to the company to a proportion of benefits also going to the inventor. The only thing that is guaranteed in all cases is the recognition of the status of inventor to the person who made the invention. Countries that had allowed employers to take all benefits from their employees’ inventions are having second thoughts, seeing that less patents are filed in their countries.
In the second half of the 20th century, many new products and services became possible thanks to electrical, magnetic, electronic and, more recently, digital technologies. The frenzy of “colour television” contaminated even respectable companies. Governments were called in (actually, they called themselves in) for “help”. In the late 1960s, the prospects of exploitation of digital technologies in imaging, audio and video became more concrete and many companies and organisations started making or funding research in digital signal processing and specifically in audio and video coding. As a result, the current numbers of valid patents in these fields are counted by the thousands. There would be many more valid patents, had not the earlier ones lost their validity because R&D in this area started so early. Still many inventors are good at repackaging expired patents in ways that allow the to file an old patent again.
Companies producing innovations tend to protect their IPR with patents that tend to become major ingredients for many standards. The result is that IEC, ISO and ITU, the three main international standards organisations, have become well equipped to deal with patents in standards and have actually agreed on a common policy. Traditionally, the work of these organisations used to be the formal ratification of some technology required for an electric appliance or a machine tool or a communication device that had been invented and possibly already exploited by an inventor.
But because standards organisations do not exist to ratify unregulated monopolies, for an international standard to specify the use of a patent for a conforming implementation of the standard, the rights holder must either release the rights to his patent or commit to give license to his patent “on fair and reasonable terms and non discriminatory conditions” (so-called RAND or FRAND ). Besides ISO, IEC and ITU, other regional and national standards organisations, apply a similar policy, possibly with a slightly different wording. Needless to say, standards organisations have no intention to delve into the meaning of what makes terms and conditions “fair and reasonable”. In case of major problems, such as the case of a rights holder not complying with the policy, the matter can be brought to the highest instance of the SDO and can even lead to the withdrawal of the standard.
In the past, adoption of a standard based on a necessary patented technology did happen – just take the case of the telegraph, the telephone and the radio – but this was before companies started investing in R&D to develop new technologies for new communication products. As a result, SDOs have begun to produce standards starting from a situation in which there are different industry-grade technologies, very often developed by major players, that exist or are even in use, possibly already in fierce competition in the market place.
In these conditions, the task of producing a standard, which has to incorporate one or more than one of these proprietary technologies, becomes hard because the role of the standards committee shifts from the place where discussions deal with technical merits of a solution, obviously driven by commercial requirements, to a place where a solution is defined based on fitness of a technology to the current plans of companies, bundling with other technologies, terms of exploitation, etc. Of course there is nothing wrong with technology deals between companies or with close alignment of standards to products within companies – God forbid the thought of prohibiting it – but this is damaging if it is done in a standards committee.
Working in this space is a tricky business. Blessing one, or more than one, specific solution that has already achieved product maturity obviously exposes people to the wrath of those excluded because “being a winner” entails considerable economic benefits. Not only because those wishing to operate in that particular business will have to get a license for the relevant IPR – if there is IPR, the right to use it must be acquired – but because at such a late stage of deployment of the solution, the basic IPR is often inextricably linked to other product features and these, too, end up being part of the license deal. Therefore it is only natural that antitrust authorities have been scrutinising this standard-setting method and associated licensing deals.
The selection process is the tricky part. In abstract terms everybody agrees that standards should provide just one way of doing things, like my definition of standard goes, but the implementation of the principle is often less than thorough. The original desire of the committee to select just one winner from a number of candidates may still be there, but preserving that determination to the end is the hard part. When people working for a company are in a standards committee, their determination to pursue the shared goal to the end dwindles, if they see competing technologies to their company’s prevail in the favours of the committee. Most often the outcome of a dialectic battle lasting anywhere from an hour to ten years is to compromise the intellectually accepted principle and, voilà, a “limited” number of solutions, often disguised as “options” are anointed and all of them are made part of the “standard”.
Usually it is too costly to implement all options in products and so implementers try to make up for the failure of the committee making their own choices among the “limited number” indicated by the standard. But it is a sure bet that different manufacturers will make different choices and communication between equipment of different makers will no longer be guaranteed.
Sometimes it is difficult to make people who have grown accustomed to this loose method of work understand that having a “limited” number of solutions or options in a standard very often heralds the failure of that standard. Because of too many signaling options it took 10 years for European ISDN to achieve a decent level of interoperability between different operators and, within the same operator, between equipment of different manufacturers. Because of too many options, many standards were stillborn because the critical mass that would have justified the necessary investments by the industry could not be reached.
Other, more enlightened people are not blind and understand the damage they are inflicting on their industries and companies, but they shrug their shoulders as if they wanted to show their resignation to “the inevitable burden of civilised life” – namely options or multiple choices in standards – that society forces humans to carry.
The first encounter of MPEG with patents happened in 1990 when, as part of my ISO duties at that time, I started collecting patent statements from companies that had taken part in the development of MPEG-1. An annex to all parts of that standard (and of those following it) gives the list of companies that have declared to be willing to adhere to the ISO patent policy for all IPR owned by them that is required to implement the standard. Setting aside MPEG-1 Audio, for which a licensing organisation (used to) exist, I am not aware of anything comparable that offered forms of collective licensing for patents necessary to implement MPEG-1 Video and Systems. In any case this is probably idle talk as close to 24 years have passed since MPEG-1 was approved.
Some time later, MPEG had to deal with MPEG-2 patents. Many companies had participated in the MPEG-1 development not so much because they necessarily had plans to make products in that space, but because they were keen to be around and help tune up the MPEG machine for the next big MPEG-2 deal. At the London meeting in November 1992, some National Bodies had started to draw MPEG’s attention to the fact that, if a large number of patents were needed to implement MPEG-2, for each of which the licensing conditions were “fair and reasonable” – the reasoning went – the licensing terms and conditions of the combination of those patents would not necessarily be “fair and reasonable”, not to mention the fact that striking many licensing deals with with all the licensors would be a very costly and time-consuming exercise which probably only big companies could afford. The irony in these National Body documents was that they knew very well that MPEG was unable to respond positively to these requests because dealing with patent licensing is explicitly forbidden by the ISO/IEC directives. Still, something had to be done.
In January 1993 I took the initiative to invite to the Rome MPEG meeting Mike Smith, head of the Information Technology Task Force (ITTF), the group inside the ISO Central Secretariat in Geneva looking after JTC1 matters. Mike provided advice on how this difficult issue could be treated without trespassing the limits set by the ISO/IEC directives. The MPEG success owes a lot to the help Mike (and his colleague Keith Brannon) gave to solve this and the many different procedural problems over the years. I do regret Mike’s untimely death a few weeks before his retirement.
Then, at the following Sydney meeting in March/April 1993, I took the initiative to convene an informal meeting, open to any interested party, to discuss which steps could be undertaken to make progress. Eight possible alternatives were identified but, unfortunately, none looked particularly attractive.
In the fall of 1992, I received the visit of Dick Green, President of CableLabs, a newly formed laboratory funded by the North American cable industry. This was one of the many meetings that I was having in my drive to promote MPEG-2 to different industries in different countries. Even though Dick, in his former position at the Public Broadcasting Service (PBS), the public USA broadcaster, and a long-time participant in ITU, had a good understanding of these matters, in that pre-convergence era he was not fully aware of the ISO process and wanted to personally make sure that a solution coming out of MPEG would satisfy the due-process requirements that the CCIR and his industry required for acceptance.
In July 1993 Baryn Futa, then CableLabs CEO, started attending MPEG meetings and particularly the “ad-hoc group on technical support for IPR”. The ad hoc group, chaired by Greg Wallace who acted as Head of the US Delegation after Cliff Reader had left Samsung at the beginning of that year, received the following mandate:
To provide technical support for ongoing discussions on IPR matters, to assist with efforts inside and outside of MPEG.
This ad hoc group was re-established for the last time in September 1993 in Brussels because MPEG members felt that, after a seminar of parties interested in MPEG-2 patents to be held on the Saturday after the Brussels meeting, the MPEG-2 patent issue could have – finally – a life of its own. Baryn took the initiative in his hands, I severed my already tenuous ties with the MPEG-2 patent discussions and could finally relax and think of other, technically more rewarding challenges.
In the event, an MPEG-2 Patent Pool was established, managed by MPEG LA. In spite of that company bearing “MPEG” in its name, there are no relationships at all with that company. MPEG LA acts as a “one-stop shop” for MPEG-2 Video and Systems patents (Audio is not part of the licensing), but a few companies wanted to license their technologies separately. Originally the number of patents found necessary to implement MPEG-2 Systems and Video was about 30 (which is different from the number of patent holders), but currently hundreds of different patents are reportedly considered necessary (because the same patents has been filed with different jurisdictions). Licensing of the patents managed by MPEG LA can be obtained by paying 2.5 USD per decoder or encoder box (originally the fee was 4 USD). Interestingly, the amount to be paid for patents in an MPEG-2 decoder has remained constant, while the number of patents has increased. For DVD a different licensing scheme is adopted: 0.035 USD per printed DVD, and 2.5 USD per decoder. These numbers have been provided to give an idea of the MPEG-2 licensing.
Important notice: this is a rough summary of a complex issue. The actual MPEG-2 licensing terms must be reviewed for a full understanding of their practical implications.