1 Introduction

The long-awaited breaktkhrough in our understanding of cosmic gamma-ray bursts (GRBs) has come about because accurate (<10$^\prime$) burst positions have become available quickly (<1 day). Prior to the launch of BeppoSAX, accurate positions were available from the interplanetary networks, but unavoidable delays in the retrieval and processing of data delayed their availability. Similarly, rapidly determined positions were, and still are available from BATSE, but their utility is limited by the fact that their accuracy is in the several-degree range.

The list of things we need to know about bursts is still long. Among the items on it are:

• are burst sources in their host galaxies, or outside them?
• what is the distribution of GRB distances?
• are bursts beamed?
• what is the intrinsic luminosity function for bursts?
• are there different classes of bursts, e.g. long and short, soft-spectrum and hard-spectrum?
• what is the multiwavelength behavior of GRB light curves immediately after the burst?

Given that only $\sim 50\%$ of the GRBs studied to date have optical counterparts, and that only $\sim 50\%$ of the counterparts have measured redshifts, it is clear that answering these questions will require hundreds of GRB detections in the gamma-ray range. But the rate at which rapid, accurate positions become available is still quite small: < 1 burst/month. Thus even minor improvements in the rate can have a major impact on progress in the near-term future. However, major improvements will be needed in the long-term future to make the next big step.