The descriptive information associated with Bulk Data Frames
(this is equivalent to the header of an IHAP file) is stored
in a data structure called Descriptors in MIDAS. The concept is
similar to the FITS header data (see reference [3]).
Like keywords, descriptors consist of a name (max. 15 characters, case
insensitive) and associated data values of any length (max. 32767) and
of type `Integer', `Real', `Character', or `Double Precision'. The
descriptor name may contain alpha-numeric
characters and underscore. Descriptors are not global parameters, but
are connected to a specific data frame. A minimum set of descriptors
must exist for every data frame. These `Standard
Descriptors' are
explained in Appendix for Bulk Data
Frames and Tables.