A wide range of positioning technologies has been exploited to flexibly configure fiber ends on the focal surfaces of telescopes. The earliest instruments used manual plugging, or glued buttons on the focal plane. Later instruments have used robotic fisherman-round-the-pond probes and articulated armsto position fibres, each probe or arm operated by its own motors, or buttons on fiber ends moved by pick-and-place robotic positioners. A positioner using fiber spines incorporating individual actuators operating over limited patrol areas is currently being manufactured and a derivative proposed for future large telescopes. Other techniques, using independent agents carrying the fiber ends about the focal plane have been prototyped. We describe these various fiber positioning techniques and compare them, listing the issues associated with their implementation, and consider the factors which make each of them suitable for a given situation. Factors considered include: robot geometries; costs; inherent limits to the number of fibers; clustering of targets; serial and parallel positioning and reconfiguration times; adaptability to curved focal surfaces; the virtues of on-telescope versus off-telescope configuration of the field, and suitability for the various telescope foci. The design issues include selection of actuators and encoding systems, counterbalancing, configuration of fiber buttons and their associated grippers, interchanging field plates, and the need for fiber retractors. Finally we consider the competing technologies: fiber and reflective image slicer IFUs, multislit masks and reconfigurable slits.