GitLab

write in the created directories

and the function that enlarges mapped column files tries mapping on next
byte before unmapping/mapping again

column

deleted

and the iteration on the values of a column is done on the number of
lines used (instead of on the total line count).

data type. Trying to set a value on a _read subclass raises an
exception.

(for columns with lines made of vectors of elements), for all data types

obitype

but using char* (needs to be changed)

create problems