Résumé
High-throughput crystallography needs robust algorithms capable to deal with a large variety of different cases, should be largely insensitive to experimental and structural criteria, and needs to have a seamless transition between the different steps that can be identifed in the complete structure determination process. These steps have remained largely unchanged since decades, i.e. space group determination using the systematic extinction conditions followed by the structure solution step by (usually) direct methods using the symmetry determined in the preceding step, and fnalized by structural refnement. The newly developed structure solution method charge fipping has made it very advantageous to postpone the symmetry determination step to after the structure solution step using a robust analysis of the symmetry of the determined phases in P1. This makes it possible to determine in one step structures where the symmetry analysis based on systematic extinctions is doubtful or even incorrect [1]. We present here a comparison between the traditional approach based on symmetry determination by analysis of systematic absences and structure solution by direct methods on the one hand and structure solution by charge fipping and symmetry determination by analysis of phased refections on the other hand. The comparison is done on more than 500 structures of variable size and composition and diffraction data of intermediate to excellent quality using default parameters for all data sets, i.e. without any intermediate human intervention, with the aim to determine success rates in automatic structure solution using widely available and free crystallographic software.