Mats Brunell 1992-09-18 THIS IS A PERSONAL VIEW Medicine for the ill ==================== There is a a great possibility that Europe once again misses its possibilities to find mechanisms to progress the networking infrastructure development. By this message I would like to urge the involved parties to re-evaluate their positions and hopefully come to terms with the real issues at hand. I have said most of this before in different fora, but I think it is really important. RARE and CEC is coming close to a situation were things are getting out of their hands. No money in the world can buy collaboration, nor the personal knowledge from the couple of 10:s of people who really can resolve the hard issues at hand. People refuse to be manipulated and treated as "FTE units in production". Taxpayers money is at stake here, and the persons involved will fortunately be accountable sometime in the future. I wonder if the CEC/DGXIII really understands that it surrounds itself with people who only say and do what they think is expected that the CEC think should be said. This will ultimately lead to an enourmous lie. Nobody will benfit from such a situation. Openness is the only way to progress. ON THE GOALS There is no such thing as R&D network! In the long term all services will have to be for public use, we the R&D community, happens to be among the first to have realized the oppurtunities with open communication although only a few of us knows how to build and collaborate to make it happen. The reasons for building open public services are several: - costs will be lower if the services will be public a matter of scale thus, the cost for the services will be possible through regular funding channels and not by a separate government funded infrastructure - incentive for the teleoperators and IT industry to develop products and services for the Gigabit and beyond I actually think the the CEC believes that the open market is a goal, also in the field of tele- and datacommunications. Comparing the European situation with the US one is wrong in many cases, the US situation will have to change fast, the new NSF solicitation will open for the open market from being a separte funded infrastructure. We don't have to make the same mistakes over here! The present R&D networks only a starting point The present R&D networks has shown a market potential for open networking, we see the first generation public services established, in this first incarnation (yes we are still in the early stages of open networking in all aspects, technical organisational and market wise) is based on Internet IP standards, the reason for this is simply that is is the only scalable technology taday. ON THE MEANS The model has to support mechanisms to move to public services The move towards a public service market for open networking has to be supported by the actions taken in Europe! This means co-existance between the so called Acceptabel Use Policy based R&D networks and the early public service providers. There are strict technical reasons for this as well as the usefulness to interact in a knowledge transfer activity. Todays teleoperators are not used to the concept of open networking. The technical issues are: - routing - DNS - management The market issues are: - solicitations is an important tool to develop a market, not just to choose ONE provider, but to stimulate a multitude of service providers - to promote interconnectivity between all providers - competition must be fostered through skillful handled policy actions, based on a sound tehcnical knowledge to develop the market, this goes for each country, and the pan-european scene as a whole The GIX/NAP concept and no X.25 The technical solution to implement a service infrastrucure must be based on: - allow for interconnection of a multitude of networks and service providers be they R&D networks or public providers - allow for scalability into 34-155 Mbps to support new applications such as distributed file systems, video conferencing et c, this in turn means: - support for Internet IP and multicast - bandwidth per see is not an issue in the GIX/NAP model it allows for sevaral interconnecting "routes" just as long as the interconnection point and the overall routing is skilful handled - The bandwidth can be separtely provided by a mix of providers Unfortunaltely the specifications so far made for the IXI-production service nor the EMPP in is current modelservice can't meet these requirements, this is a very serious problem. ON THE POLITICS The different roles: The enabling role: - the ONLY thing the goverments should be using the taxpayers money for is to enable the establishment of a long term market, in this case by showing the teleoperators and IT industry how to build the new generation communications services. - this infrstructure starts at the customer site, and thus is the local development in each country most important - don't try to enforce the HOWS, so far this has lead to a disaster - promote collaboration! The developers role: - to develop the best possible technical solutions based on the early experiences from the early large scale experiments, scalability, performance management et c This is valid for todays technology and especially for tomorrows! One who claims he has a technical solution today ought to question himself which probelm he has solved. We are only slowly learning the problems! - To develop the technology we need as broad as possibel collaboration across industry, teleoperators and research, (not R&D network organisations, not the CEC etc, between the people who has to collaborate and solve the issues) The operators role: - to collaborate with all other providers to build a new management model based on mutual knowledge transfer and interaction which is needed for open networking (this is not X.25/X.75!) The users role: - we still need to prove that we want these services, but also be open about WHEN we can trust the teleoperators to manage the infrastructure. No market for them, and we won't see the Gigabits developed! ON TODAYS SITUATION Now to the hot part! RARE RARE must become a true democratic organisation. Its decisions should be made in the meetings, not in the corridors of the CEC. If RARE is to become a representative organistion the RARE CoA mebership and officials should be appointed based on personal elections. If this won't happen we have to establish a true user organisation as soon as possible. EBONE - collaboration between all European open network service providers The short term service provision should only be funded by the users based on a open model. The users or their representatives is the counterparts that needs to agree upon costs and services. So far the Ebone structure is a first step in this direction, it needs improvement in terms of a framework were the users has the real influence and the technical issues is not made into politics. The management should be based on those organisations that actually is collaborating and putting in efforts in the service provision, noone else. The open routing and interaction between the European and global service providers has to be re-inforced in order to maintain both the connectivity and service quality on a global scale. RIPE-NCC RIPE-NCC is a vital part for developing open networking. It is a resource that all users and service providers in Europe need. Thus mechanisms to secure direct funding from those parties should be found. I e to lower the cost for the taxpayers by not relying solely by CEC or RARE membership funding. ENABLING HIGH SPEED SERVICES This could be achieved through two steps: - enabling funding for the EBONE structure to reinforce the bandwidth to 34 MBPS which is the technical and infrastructure possibility today. this could be made in an open tender, where the interconnections points in Ebone represents the European "GIX:es". For future development (this is still a development and research issue) - Build upon a separte infrastructure based on fiber interconnect with optical interfaces and possibilites to field trial in large scale ATM equipment, where local ATM LAN:s is the interconnection points to the WAN. Alternative solution will be possible to connect to the optical interfaces to develop high bandwith solutions beyond Gigabit. These solution will never be possibel to buy, they can only be established through close collaboration in between teleoperators, IT industry and research with interested field trial users involved. Such islands are present is several countries already and more are planned. A good strategy would be to establish interconnections between them. This minimizes the need for large sums of money being shipped to Brussels, and relyes on true collaboration between the parties involved on national regional level in an interantional context. DGXIII:s initiative "Demonstration of Interworking via optical Networks" is open for proposal now, I consider this to be a very fruitful possibility forward. Once these first experiments are settled one could enter a pilot service phase and subsequently a move to a open market.