The chemical composition of individual stars provides clues about the global production of
individual chemical elements and the overall chemical evolution of our galaxy.
Various nuclear, atomic and radiative processes determine the shape of the emitted spectrum,
from which quantitative data about various stellar characteristics can be extracted. One of
these is the abundance of the chemical elements. The oscillator strength (f-values) is the
main atomic parameter needed for this analysis.

High-resolution stellar spectroscopy in the wavelength region 900-10000 A (very soon
to 50000 A) provide excellent data for detailed abundance studies. During the last 15 years
the accuracy of experimental f-values has increased due to refined techniques, but very few
measurements have been made below 2000 A and above 10000 A. Since the experiments are very
time-consuming a careful selection of spectral lines is necessary. A close collaboration or
connection between experimental and theoretical work on f-values is very important.

We will discuss the present balance (unbalance) between users' needs of f-values and
producers' capability in terms of scientific goals and practical considerations. The future
needs will increase both as regards quality and quantity and extend to previously unexplored
wavelength regions.