Copenhagen, Spring 2000 EMT HISTORY, a sketch _____________________ NOTE: This sketch is written as my personal view. Others who were involved in EMT will, undoubtedly have different views and stress other aspects. I also focus to some extent on subjective matters, issues etc. where I personally was involved. The sketch is incomplete. I have not checked up on dates etc. and not consulted files (e-mail archives, minutes of meetings, contracts etc.). How it all began ________________ In August 1983 the delayed ICM (International Congress of Mathematicians) was held in Warsaw. At the congress the informal discussion forum called European Mathematical Council (EMC), chaired by Sir Michael Atiyah met. Every National Mathematical Society in Europe, East and West, were invited to send a representative. I think we were some 20 people. I represented the Danish Mathematical Society. At the meeting Bernard Teissier from France brought up a matter he considered was of importance: Databanks. He had just been to the US and reported on plans, prospects etc. as seen from over there, especially after visits and discussions with AMS, American Mathematical Society. There was a general feeling around the table that yes, the technical development DID seem to open up completely new possibilities and yes, it was important that we in Europe followed the development, perhaps even took an independent initiative, reflecting the opinion that what Europe needed could well be different from what was needed in the US, in particular in view of political, cultural and organizational differences. It was not clear WHICH new possibilities were lying ahead of us, but the important fact was that no longer was the computer to be looked upon as a special device for numerical analysis and other purely computational purposes. The buzzword was databanks. Sir Michael concluded the discussion by forming a databank committee, headed by Bernard Teissier. I was in fact asked if I would join it - I had strongly supported the ideas presented by Teissier - but refrained with the motivation that I had no insight whatsoever in the technical issues, only realized the importance of the work to be done. About half a year later the databank committee reported, just a 1-2 page note. I remember that I was quite enthusiastic about the scenario briefly outlined and wanted to hear about the background. It turned out that there was none. It was purely speculative. The chairman had sought support inside France, but had obtained none. The result was that Sir Michael, supported by Teissier, turned to me and asked if I could do something. This was a big challange. I too was in the need of support. However, this was exactly a situation where small countries sometimes have an advantage. And the climate for international collaboration on issues of a technology based nature was good in Denmark. I then accepted the job and started with great enthusiasm to investigate the possibilities. I firstly turned to the Danish Mathematical Society (I was no longer chairman of the Society) and asked them to find a mathematical researcher with insight into the new technology. I needed a colleague to talk to. The Society pointed to Niels Vigand Pedersen from the Technical University. This was a lucky choice. Niels and I worked well together and Niels was patient in explaining the possibilities to me. In Denmark, I established contact with DANDOK, a danish government agency looking after the library sector (at the technical side, especially re databases) and with a danish commercial company, DDC (Danish Datamatics Center) active in the relevant new technology sector. The final contact - which was the contact that failed in France - was really easy: I got the relevant Ministry (of Education and Research) interested. After preparations with my contact person (H.Koch) at the Ministry, I asked to see the Minister, Bertel Haarder. We had a short and fruitful meeting. The Minister was known to promote European Collaboration and it was surprisingly easy to get his full support of all I asked for: Basically, economic support of a special purpose center for 5 years, provided the European Colleagues would accept strong danish participation in the endeavors to create a modern research infrastructure adapted to the needs of the scientific communities in Europe with the mathematical sciences leeding the way. I was also fortunate to secure Danish seed money for the preparatory work (benefiting from the fact that some "free" money was available because of delays in the physicists plans to join a European Syncrotron project). While all this fundraising activity was going on - really something that kept me busy (read: goodbye research)- Niels and I had prepared a preliminary report from the EMC/Databank Committee. The structure of the report was the work of Niels. Really, the report is an interesting document. What Niels outlined is today realized on the internet as math. institute homepages. Not sophisticated. Simple in structure reflecting the interests of mathematicians. I will annex a photocopy of the report to these notes (ANNEX 2). At a meeting hosted by the institute in Oberwolfach, the report was approved in the sense that we were asked to continue work along the lines sketched, now aiming at securing support from the European Union (Community called at that time). It was clear to us that what was needed was a really big project, something never tried out before by the modest and somewhat conservative mathematical community. To succeed we had to be aggressive. A project with two phases was outlined: Bottom up specification, top down implementation. Precise areas were also outlined. The "databank" dimension was largely extended and a platform assisting the mathematician in ALL his tasks was outlined: Preparing, exchanging, retrieving documents; storing, searching, retrieving litterature etc. Four key partners were suggested: * CWI, Amsterdam (Centrum voor Wiskunde en Informatica, well known research center of mathematics and computer science, but also capable of doing contract work); * NIHE, Dublin (Nat. Inst. of Higher Education, with experience in work for the EC;); * DDC mentioned above * the European mathematical community! The fourth partner was essential. The community of researchers should itself follow closely the work going on and take part in it as relevant. A difficulty here was that the community had no formal representation. EMC was an informal "discussion forum" that could not enter as contractor to the EC. Plans were talked about to create a true European Mathematical Society, but these plans could not be realized immediately. For one reason the political climate was not ripe for that - and no signs on the horizon could foresee that a true unification of all good forces in mathematics in Europe would become realistic. At a meeting in EMC held at the University of Oxford, the technical plans were presented by me as chairman of the databank committee and widely discussed. I was happy that the detailed plans were approved. One issue - which later turned out to be of importance - was, however not met with enthusiasm. This was the planned engagement of DDC. This element I had personally brought into the picture. DDC was to perform a "soft" initial investigation of the demands of mathematicians. This work should then be used for a technical specification, entrusted to CWI and the whole project should be steered by NIHE who was to deliver the Project Director (and do some other project work). I was thouroughly convinced of the importance of the work DDC should carry out. Noone else was so. The general opinion was that we mathematicians know that well ourselves, there is absolutely no reason to have "outsiders" tell us what we needed and try to priorotize and systematize the needs in view of the technical opportunities. I am still today convinced of the great value professional consultants etc. can contribute to the success of a project by performing such "soft" investigations (typically looked down upon by us knowledgeable academics). Another element which was discussed, an idea I had promoted, was to create a center specializing in the relevant technical issues as well as other parts of the research infrastructure pallette. Such a centre should support the European Mathematical Community. As indicated above, the Danish Minister had promised to finance a centre like that. But the promise of the Minister was also tied to the role of DDC. The result was that agreement was reached about the project, following the plan made by me and colleagues in the Databank Committee. Clearly there was an advantage in having Denmark in a strong position, and I think the phrase "Denmark will pay" was coined at that time. One could say that I had a double position. But I played with open cards, and really, the only way I could see to launch a professional project was to say to my own country: Look here, if you pay initial money (re project proposal etc.,etc.) we, little Denmark, can do some really useful work never tried before, work that points to the future. This worked (where the previous initiative had failed). Really, I think that the danish dimension played no role at all to me in one sense: I just wanted the project under steam. But that being said, I was happy to have support next door in form of the Euromath Center, actually hosted by University of Copenhagen, the Mathematical Institute, where I work. In fact, without that I could not see how I could follow the activity, because of my lack of technical understanding. The result in Oxford was that we should go ahead and take steps to get a contract with the EC. At the same time, steps should be taken to create by law a formal entity which could become the main contractor to the EC. Here United Kingdom offered to host such an entity. This was of course more than natural in view of Sir Michaels role but it should also be said that the first active interest in the project was shown by LMS (London Mathematical Society) and by EMS (Edinburgh Mathematical Society). They sent a representative, Christopher Mulvey to Denmark. He thouroughly scrutinized the project, investigating my role (did I have a hidden agenda must have been one of the issues to look into), the role of DDC and the role of the formal Danish Contractor to be (not DDC, but RISO, the Danish Atomic Energy Station). Mulvey's strength was very much on the formal side - another weak point when thinking of my own capabilities. Thus, I was happy to have Chris to assist with formal details and I think that his outline of planned contractual details was well received by everyone. There was one more important meeting in EMC before the project, baptized EUROMATH, could be launched with the full support of Sir Michael himself as well as everybody else inside EMC. This meeting was held at Liblice outside Prague. At that time (xxx, 1986), it was clear that we stood a good chance to have a contract with the EC but, at the same time, it would not be possible to include representatives from Eastern Europe in the work. This was of course an irritating situation, seen from the point of view of mathematics: Completely unnatural to leave out Eastern Europe. Indeed, a basic issue in forming EMC - and one reason why EMC was only an informal discussion forum - was the desire, really the need, to keep as strong a professional contact and dialogue WEST-EAST as possible. At Liblice there was a strong representation from the East. In particular, Russia was represented. Really, the question asked by the western side of EMC to the eastern side of EMC was this: We have this opportunity to build the modern research infrastructure for mathematics in Europe but are prevented from including you in the work and also, you will be prevented from exploiting any of the results from the planned Euromath project. May we proceed, nevertheless, with the plans? And East answered "You may!". Without this generous support from our eastern based colleagues, the project would not have gone ahead. This I am sure Sir Michael ment, and personally - though for this discussion I played no formal role - I was of the same opinion (I had learned a lot of my mathematics from the Eastern European countries, and personally respected their mathematics as well as their integrity and strong will in difficult times). Inspite of all assurances that one would make sure that ALL would benefit from the project, it was not possible to predict if that would EVER become possible. One incident I remember very clearly took place at a preparatory meeting at EC headquarters in Brussels, some time before the Liblice meeting. We, Chris Mulvey, Jack van Lint and I, presented our views to our contact person at the EC (PW?, a fine person for all our negotiations) and we discussed the contractual details. PW had told us that a high ranking official from DG whatever the number is (the political Directorate General at EC) would see us. Then suddenly, in came Monsieur Bellmain - I even remember his name today - and addressed us: "I would like you to know that in case any of the results you obtain under an eventual contract will be discussed with or leaked to any person or representative from the C&E European countries, your contract will be terminated IMMEDIATELY". And out sailed Ms. Bellmain. Incredible in view of later developments, just a couple of years later! The final contractual negotiation tok place in xxx, 1987 at the EC, Brussels. The EC committee or Programme with which we dealt, the STIMULATION PROGRAMME, had a scientific committee headed by Sir Peter Swinnerton-Dyer. Our foremost spokesman was, of course, Sir Michael. Very amusing with Sir Peter and Sir Michael discussing, summing up etc., assisted by humble us (on our side: ...). There were some corners to be turned, especially I point to the insistence by the EC that in the first specification phase also some tangible product would be created (we chose communication, not knowing that this would be obsolete in view of the technical development...). At the end of the day agreement was reached: We DID IT! The largest contract ever in mathematics was in house (except for follow-up paperwork). The celebration the same evening at one of Brussels good fish restaurants was memorable for several reasons, but perhaps I should not tell stories here! Anyhow, there was really reason to rejoice. The Euromath Project ____________________ Mathematics was represented in the Euromath Project in two ways: Formally, with responsibility under the contract to the EU, EMT was the central player, and followed the project through its Committee of Management (CoM), who, however, had delegated work to a special purpose committee, the Euromath Database Committee (DBC). The formal side was supplied by the important creation of the Euromath Advisory Board (AB). This was where the mathematical community came in. It consisted of representatives of the European Mathematical Societies that took part in the project. It was independent of the group of mathematicians directly involved in the project on behalf of EMT. This independence had the great advantage that during meetings of the AB we could accept participation from representatives of the C&E European countries, though perhaps a bit risky at the start. Roughly, the work on the Em Project was structured during the entire formal duration (i.e. as long as we had a contract with the EU, 1987-199x) as follows: A professional Project Director (PD) steered the activity. Contractual partners worked with close contact to the DBC. At regular yearly or half-yearly meetings the results were presented to the AB. The CoM then took formal action when and as needed, based on input from the AB, from the CoM and from the PD. The project went through several crisis. Let me mention the most important ones. Firstly, there was "the catastrophy in Elsinore" as I usually put it. This was the very first AB-meeting. It was held less than half a year after project start. At that time, the technical contractors worked very much alone with little information going across to EMT. Logically, the first phase concerned the "soft" activity of DDC and I personally was looking forward to the results to be presented (finally, DDC would show the great value of this kind of activity). I and everybody else (including the PD!) was baffled when we heard DDC's presentation. Utterly useless. To me personally this was quite painful. And the then chairman of the AB left the AB in dismay (he should have stayed on, fighting constructively to rescue the project). However, the project WAS discussed intensely and consensus was reached among the mathematicians as to what should be done. (The meeting was summed up by Michel Demazure as "too much food, too little sleep"). Because of this first failure, a second AB-meeting already in September 1987 was agreed. In the mean time the PD (and I) tried to sort out the problems with DDC, and the other partners, especially CWI did fine technical work. Regarding DDC it turned out that what was wrong - a surprise in Denmark to those who followed IT commercial work - was that DDC was in severe financial difficulties. The DDC-responsible persons had been forbidden by the head of DDC to spend any money on the project, apart from travel money. The professional PD was too nice, too weak possibly also, to discover that - and neither did I. The result of the first crisis was really this: The glorious role of Denmark could not be realized. The crisis, unfortunately, but understandably, gave grounds to some sceptisism regarding technical input from Denmark. This later resulted in the fact that the vary capable services of the scientists working at the Euromath Center in Copenhagen were not exploited as much as they should have been (the center was purely academic and had nothing to do with DDC or any other IT-commercial set-up). When writing now on these events I get, once again, filled with anger towards DDC, especially their director. On the positive side, I - and members of the CoM, I think I can add - realized the importance of good project direction. As a result, the PD was - after some time - replaced. And the DBC was strengthened at the technical side by replacing me as chairman with Michel Demazure, a very good step, indeed. And we all took up the challenge, working hard to get the project under foot. The next AB meeting was very successful. It took place in Braga and restructuring and new technical progress (by CWI) was appreciated as well as the new working of the AB, now chaired by Jose-Luis Vicente. Also, at that meeting the Russian side was for the first time presented on the AB (Sergey Novikov and Alexei Zizchenko). Now, on to the second crisis: Based on a fine specification, we were renegotiating the contract with the EU as planned. The basis and go-ahead for the implementation phase should be established. As it turned out, the intended bottom-up implementation would be, by a factor of at least two, too expensive compared to the kind of money we could realistically hope to get from the EU. Unfortunately, CWI could not do the work under a limited budget, and they decided to leave the project as we had to restructure it at the technical level. After long and difficult discussions it was decided to accept the money we could get from the EU and to incorporate as much of existing software in the planned Euromath System. As a result, commercial software had to be accepted, inevitably leading to a project that had to be payed for by the mathematical institutes. Considering the way the academic sector works this was not to anyones liking, but had to be accepted unless we gave up the further development. And then the third crisis: For reasons I shall not enter into in detail, except for the statement that disagreements over Chris' part in the management was an ingredient in it, EMT was close to loosing its formal solid foundation in the UK environment. Fact was that Sir Michael - for understandable reasons - had left his formal role as socalled subscriber to EMT at the time when also Chris left EMT, and the formal affiliation with the University of Sussex had to be discontinued. According to law, EMT could not exist under these circumstances. A UK-base and at least two formal UK-citizens as Directors of EMT was a necessity. Here comes one of the most positive of my experiences: In spite of the formal difficulty just explained AND in spite of the fact that EMT's work on the Euromath Project was in great difficulty at the time (certainly not pointing to a winner at the time), the London Mathematical Society and the Edinburgh Mathematical Society - our UK-based founding organisations - lived-up to their responsibility in a most efficient way. They reacted to my plea for action by investigating the "state of affairs", and, after discussion and positive reaction, very fast found people to take up the challenge both personally, considering the work ahead and formally, regarding EMT affiliation. A new base for EMT was found at The University of Kent at Canterbury. Thus entered John Slater and James Davenport as most knowledgeable EMT-directors and in came Ian Stone from UKC as highly efficient new company Secretary. EMT was rescued and the new life at UKC started formally in xxx, 199x. The fourth and last crisis I shall mention: The implementation phase of the Euromath Project was actually well structured with a more efficient (at the start at least) PD and, after Michel Demazure had to leave the work for EMT, with Helmut Lenzing as another efficient chairman of the DBC (seconded by Richard Timoney, Bjorn von Sydow, Branislav Rovan Jose Valenca, John Slater and, later, Jill Tardivel). But, but, but, the input of the professional contractors was not of the quality expected and also the consequences of commercial input to the Euromath System posed problems. The key commercial input came from the french company GRIF SA. As it turned out, this company was in financial difficulties and finally had to give up. One could point to other difficulties, call them cisis if you want, but in retrospect it should be said that, firstly, all these difficultiea were not of a nature unknown for ambitious projects in the IT-sector and, secondly, the Euromath Project was a success seen from at least one point of view: The view taken in Brussels. We could deliver very precise reports containing contributions to international standardisation work (regarding structured documents, SGML,...), we DID develop a product that could be demonstrated, and we did what Brussels love the most, travel a lot (yes, they really like that!) mixing several cultures of many different nations, contributed to understanding etc. (including establishing many friendship relations). We, so far, DID NOT produce a product that is in widespread usage by mathematicians and other scientists. The money generated during contractual work to the EU (well looked after by Ian Stone) allowed us after closure of the last formal contract with the EU to continue the development. The last two years, the Comenius University of Bratislava, under the direction of Branislav Rovan, has been entrusted the most important task of trying to free the Euromath System completely from the commercial entanglement. This has, contractually, become possible after GRIF SA was closed down and after public-domain software of the same nature as that provided by GRIF SA has been developed. When I am writing this, the potential possibility of success has been verified and the first beta-release version (under Linux) is beeing tested. Our focus on structured documents for the creation of the Euromath System was ahead of its time, but now in the very last minute (as EMT's financial resources are drying out) it appears possible to place the new Euromath System in the public domain with a potential to be of real service to the mathematical community and others as we always had strived to. Other Projects ______________ When it became politically possible, EMT was ready to embark on a collaboration aiming at the C&E European countries. Thogh basic contacts were established in the regime of EMT (an AB meeting in Luminy, 19xx, in particular), it was decided - actually contrary to my recommendations - that though EMT was happy to provide necessary contacts, EMT left to others to do the actual work. Therefore, the main part of this activity took place formally outside EMT. I myself used the Euromath Center in Copenhagen as basis for such activity. The socalled CIPRO project, a TEMPUS project, took place with EmC as the key Contractor. Through that project - the most successful project perhaps, I have taken part in - centres of expertise based on modern IT technology were created in Hungary, then Chekoslovakia and Bulgaria, all with associated work regarding teaching. It was difficult to obtain the necessary export licenses because of COCOM restrictions, and mainly because of that it was a major challange administratively. EMT itself, did later embark on projects going East, always with main emphasis on tiing the work directly to the Euromath activity. By far the largest such project was the INTAS project EmNet/NIS aiming at support of the reseach infrastructure in CIS (=fSU) countries. I had the great pleasure to Coordinate that work (on behalf of EmC as main contractor; EMT was represented by John Slater, Ian Stone and Jill Tardivel). It looked impossible as it had 56 partners in 6 countries of the CIS and with limited resources (a total of little more than 500.000 ECU). But results were good as can be witnessed by many EmNet centres created. I shall refrain from going into details except for one interesting fact. In 199x we had a most successful EmNet workshop. For that we needed extra support and obtained that through NATO Scientific Affairs Division (who have supported a number of other "Euromath spin-off" projects). As the subject matter was LINUX, and as Linux has its base in Helsinki (and our contact there was strong through Mika Seppala - who entered the picture as AB member), we asked NATO if they would allow that the meeting be held in Helsinki. So the result was that NATO supported a meeting held in Finland to the benefit of Russian and other CIS scientists! ... that this meeting resulted also in my participation in an expedition to Krasnoye Cikoye near Mongolia is another story...! The end _______ As Euromath Center was closed in 1998 and as EMT is closing in few months, the end for me personally regarding international collaboration for the improvement of the research infrastructure is in sight. The 15 years devoted to that effort were most fascinating and meaningful, technically and at the personal level. Not without periods of frustration, not without great pleasures and lasting friendships. My new life, rebouting my interest in mathematical research, really owes much if not all to Euromath spin-off activity: During a coffee break in Novosibirsk I learned that Mr. EmNet from that city, Boris Ryabko specialized in an old interest of mine, information theory, and through him I got updated on new work and problems which now constitute my key professional interest. -Flemming Topsøe, March 25. 2000 LIST OF POSSIBLE ANNEXES ANNEX 1 First report from EMC/Databank Committee (1-2 pages) ANNEX 2 First report by the new committee ANNEX 3 Second report by the new committee ANNEX 4 Contract with the EC (first contract) ANNEX 5 Final Euromath Report ANNEX 6 Final EmNet/NIS report etc.