Formats and data storage

The currently used ENDF-6 format has a number of shortcomings that originate from the punch-card legacy. Some, such as redundancy, counting lines, too short MAT number and too many options could be resolved within the current format. The more critical issues, such as rigidity to extensions, limited accuracy due to fixed number of digits, difficulty of reading by humans, lack of native software support, and lack of integration with EXFOR and ENSDF libraries, would require such considerable efforts that migration to a new and modern format, offering additional advantages, is a more attractive option. Actually, such an option is even more justified by the existence of the well advanced General Nuclear Data (GND) format being developed by LLNL. This XML based framework is easily extensible, its numerical precision is not fixed, it is easier to edit, can store additional information (e.g. renormalized experimental data used in the evaluation), allows for storing various versions side by side (e.g. reconstructed data, group-wise data, alternative evaluations), it may include hyperlinks to documentation or another set of data (e.g. covariances). Two additional advantages are the possibility of using native Python support software and already available conversion (in both directions) to the binary HDF5 format which allows to organize, store, access, analyze, share, and preserve data huge in both size and complexity. The GND format comes with the processing software FUDGE, whose class structure mirrors the GND format. It provides for arbitrary alteration of the data, energy deposition/kermas, data checking, translation to different format (ENDF-6, HDF5, ENDL), grouping, resonance reconstruction, visualization and sampling from covariances. It is deemed advantageous to the Nuclear Data community to adopt this new system and develop interfaces to the currently used processing codes and retrieval systems to take full advantage of the evolving technology. Such modernization is needed to allow for the storage of large covariances, ensuring evolution of the stored data, and for attracting to the field young experts