W0056

It is important to package data derived from scattering experiments in a fashion that makes it accessible to the end user and, ultimately, to the larger scientific community. Historically each facility has invented its own data format and has been loath to accept standardization because of the effort required to write new programs. Attitudes are changing slowly and the appearance of new facilities provides new opportunities for thinking about these issues. This presentation will summarize the state of this marketplace as of the NOBUGS conference in November of 2002. There seems to be substantial convergence on the NeXus format[1], but many issues need further work. Examples include the use of XML vs HDF as the underlying data model and the development of instrument descriptions to define the metadata required for various kinds of experiments. Automatic capture of metadata has been widely neglected in the past, but it is an essential requirement at facilities that entertain man!

Automatic capture of metadata has been widely neglected in the past, but it is an essential requirement at facilities that entertain many non-resident users. We also need to recognize that standardizing formats without providing applications that employ the standards to allow end users to visualize and analyze their data is pointless.