“…This approach was adequate for certain applications and for research, but did not exploit existing solutions from other areas, and could not serve as the basis of future integrated information systems. This was soon recognized, and with regard to the storage layer, either database management systems (DBMSs) were used, like in HyperBase (Schütt and Streitz 1990) or HyperPath/O2 (Amann et al 1993), or special-purpose storage managers were developed, like HAM (Campell and Goodman 1988) or Hyperform (Wiil and Leggett 1992). The disadvantage of the second approach is that many of the standard DBMS functionalities needed for persistent storage and sharing of data, like transaction management or declarative access, have either to be re-implemented, e.g., concurrency control, or are not available at all, e.g., declarative query languages.…”