The past history of the XRB has been characterized by the "hot" controversy between supporters of two alternative models: diffuse emission versus discrete sources. The recent COBE data, which have conclusively set strong upper limits to a substantial contribution of diffuse gas emission to the XRB, have finally solved this long-standing controversy.
The present controversy has now shifted focus. ROSAT deep surveys on the one hand, and high energy spectra on the other hand suggest that, indeed, AGNs can produce a very substantial fraction of the observed X-ray background over a large range of energy. However, substantially different models for the AGN population seem to provide very similar and almost equally acceptable fits to the XRB spectrum. The questions which have to be answered in the next years are therefore more related to the X-ray properties of these AGN populations rather than to the XRB itself. What are the X-ray luminosity functions and evolution of the different classes of AGNs? Are different classes of AGNs dominating different energy ranges? Are unified schemes required by or at least consistent with the X-ray data? Optical identifications of very faint ROSAT sources, coupled with the study of sources selected at higher energy, should provide some of the answers to these questions.