One of the principal motivations of wide-field and synoptic surveys is the search for, and study of, transients. By transients I mean those sources that arise from the background, are detectable for some time, and then fade away to oblivion. Transients in distant galaxies need to be sufficiently bright as to be detectable, and in almost all cases those transients are catastrophic events, marking the deaths of stars. Exemplars include supernovæ and gamma-ray bursts. In our own Galaxy, the transients are strongly variable stars, and in almost all cases are at best cataclysmic rather than catastrophic. Exemplars include flares from M dwarfs, novæ of all sorts (dwarf novæ, recurrent novæ, classical novæ, X-ray novæ) and instabilities in the surface layers of stars such as S Dor or η Carina. In the nearby Universe (say out to the Virgo cluster) we have sufficient sensitivity to see novæ. In 1 I review the history of transients (which is intimately related to the advent of wide-field telescopic imaging). In 2 I summarize wide-field imaging projects, and I then review the motivations that led to the design of the Palomar Transient Factory (PTF). Next comes a summary of the astronomical returns from PTF (3), and that is followed by lessons that I have learnt from PTF (4). I conclude that, during this decade, the study of optical transients will continue to flourish (and may even accelerate as surveys at other wavelengths—notably radio, UV and X-ray—come on-line). Furthermore, it is highly likely that there will be a proliferation of highly-specialized searches for transients. Those searches may well remain active even in the era of LSST (5). I end the article by discussing the importance of follow-up telescopes for transient object studies—a topical issue, given the Portfolio Review that is being undertaken in the US.