Type 6000 files are similar to Type 3000 files, except that they allow longer
record names and more than 26 subrecords.
Type 6000 files were initially defined by users outside of the Midas community
who were using them to store specialized data; as a result, some of their design
aspects are not well-suited for general use within a Midas framework:
The subrecord definitions for a Type 6000 file are much more complex, than
a Type 3000 file (96-bytes vs 8-bytes) and must be specified in the
extended header rather than the main header. This adds significant
overhead to the reading and writing of a Type 6000 file and causes
problems if the file is shared between multiple processes.
Many of the special features of a Type 6000 subrecord definition (e.g. "Units
Greek Prefix Code") are rarely used and have limited support in Midas.
Most of the Type 6000 files seen in X-Midas or NeXtMidas are created to get around
the 26-subrecord limit in Type 3000 files.