Scholarly publication, information, and communication is in a stage of a possibly far-reaching transformation. This change is based on a revolution in computing and communications technology. Several questions come up immediately:

Should mathematicians ''sit and wait'' to see how this ''market'' is going to develop? What should the mathematical community actively do? What role should the mathematical societies assume in the future?

My answer is: Engage proactively!

In this talk I will first give a short review of the current developments and my analysis of the chances and dangers arising. My conclusion (and I hope that this vision is shared by many others) is that we need an open, high-quality, distributed, user-driven and -friendly electronic information and communication system that integrates the worldwide avtivities of various groups of mathematicians, librarians, publishers, etc. in an unbiased way. The system should allow easy navigation and retrieval, in particular, precise browsing and search to provide efficient support for research, teaching, and the use of mathematics by others.

I will provide scientific, technical, and political reasons why I think that all these attributes are necessary. I will also argue that such a system has to be developed and maintained within some organizational framework, which, however, must provide sufficient openness and flexibility to integrate efforts and activities of diverse groups, e.g., groups that do not necessarily see their own contributions within an organized (and thus somewhat bureaucratic) structure.

I suggest that the International Mathematical Union (IMU) provides this frame. IMU is an organization that has - in its 50 years of existence - proven to be able to steer international collaboration within mathematics, despite politically rough times. IMU is the best possible choice to monitor the system development and define the standards, e.g., for mathematical data and documents and the interoperability of software. Through its organizational structures, IMU has the possibility to assure participation of mathematical societies from all over the world.

I will outline the technical and organizational concepts underlying the MathNet system developed in Germany in the recent years. I propose to use the German MathNet as a prototype for the system to be set up under the auspices of IMU. The persons and groups in Germany involved in the MathNet project are willing to contribute their achievements and experiences to the new international system.

For detailed information on MathNet see: http://www.mathnet.de/

The final result of all these efforts could be as follows: All partners (i.e., mathematical departments and other institutions, mathematical societies, individuals, publishers, libraries, etc.) make their electronic resources (e.g., preprints, papers, lecture notes, books, information about research projects, personal information, job offers, software, data of practical applications) that they are willing to make freely accessible available in the internet in structured form. The data structures are set up in such a way that mathematical information of a specific type can be gathered automatically (e.g., by robots). The system has to be designed in such a way that various types of relevant services can provided (that are up-to-date and of high quality with respect to the standards the system targets for). Existing services in the German MathNet system are: a preprint index (MPRESS), a collection of mathematical persons (PERSONA MATHEMATICA), an institution NAVIGATOR, and some others.

No insurmountable technical difficulties (e.g. high storage capacity, network and computer speed) arise in this context. Appropriate and not too costly hardware is available. Even open software, such as HARVEST, on which such a communication and information system can be based is on-hand. The hardest part is, most likely, of organizational nature. Will we be able to convince sufficiently many ''players'' to particpate? Will they regularly contribute their resources in an orderly fashion and will they maintain this service?

I believe that a system such as the one outlined above can be set up and maintained. I also believe that such a system would be of great help for the working mathematician. Today, very many institutions and individuals make their electronic resources visible in the internet. If everybody adds an epsilon of effort to structure the information a little more the whole world could search and find mathematical information much more effectively.

CALL: Let's not wait until others have decided what is good for mathematics. I, thus, propose to form - under the auspices of IMU - a community of mathematical institutions and individuals dedicated to develop and maintain a mathematical information and communication system as sketched above. Let's launch the initiative at the Berkeley workshop and build up the necessary infrastructure!