Main advantages for customers:
Reduces IO operations.
Fetches specific columns that you need to access.
It consumes less space.
Support type-specific encoding.
Currently Syniti puts data in CSV. We've faced following issues because of this
extra columns in .mir file than ref. Reading data from text delimited files which do not support schema evolution becomes a tedious task.
Unrecognized characters in columns making parsing erroneous.
Adding of \r column in the column name itself because of text file creation and handling of such characters by different libraries.