Early searches for transient optical emissionOver the first two decades after the discovery of GRBs (until 1996), GRB localizations were either delayed but accurate, e.g., with arcmin accuracy, as provided by the Interplanetary Network (IPN (Hurley et al., 1999) with typical delays of days or rapid but rough, e.g., within minutes after the GRB trigger, but with at least 2 • error circles as provided by the BATSE Coordinate Distribution Network system (Barthelmy et al., 1996).Correspondingly, several alternative strategies were pursued: (1) searching for quiescent emission in well-localized error boxes (assuming the existence of quiescent persistent GRB sources), (2) post facto correlating optical monitoring observations temporally overlapping with GRB triggers, and (3) quick follow-up observations after a GRB trigger.