X-ray free-electron lasers produce brief flashes of X-rays that are of about a billion times higher peak brightness than achievable from storage ring sources. Such a tremendous jump in X-ray source capabilities, that came in 2009 when the Linac Coherent Light Source begun operations, was unprecedented in the history of X-ray science. Protein structure determination through the method of macromolecular crystallography has consistently benefited from the many increases in source performance from rotating anodes to all generations of synchrotron facilities. But when confronted with the prospects of such bright beams for structural biology, enthusiastic proposals were tempered by a trepidation of the effects of such beams on samples and challenges to record data [1]. A decade after these discussions (and others in USA) on the applications of X-ray FELs for biology, the first experiments took place at LCLS giving results that fulfilled many of the dreams of the early visionaries. In particular, the concept that diffraction representing the pristine object could be recorded before the Xray pulse completely vaporises the object was validated [2], confirming predictions [3] that established dose limits could be vastly exceeded using femtosecond-duration pulses. The first experiments illuminated a path to achieve room-temperature structures free of radiation damage, from samples too small to provide useful data at synchrotron facilities, as well as providing the means to carry out time-resolved crystallography at femtoseconds to milliseconds. In the five years since, progress has been substantial and rapid, invigorating the field of macromolecular crystallography [4,5]. This phase of development is far from over, but with both the LCLS and the Spring-8 Ångström Compact Free-electron Laser (SACLA) providing facilities for measurements, the benefits of X-ray FELs are already being translated into new biological insights.One of the main bottlenecks in protein structure determination is the need for large welldiffracting crystals of the protein of interest, which can take months or years of effort to achieve, if at all. Smaller crystals should be easier to obtain, given that showers of small crystals are often observed in crystallisation trials. For a fixed defect density, the likelihood to obtain a portion of well ordered crystals should increase as their volume decreases. The strength of a diffraction pattern of such an ideal crystal is proportional to the number of unit cells, and thus the volume, of the crystal. It is also proportional to the