If JDF folder is overwritten data of 'old' columns is persitent

[kafisatz]

See example.

using JDF 
using DataFrames 

df = DataFrame(a=rand(3),b=rand(3),c=rand(3))
fi =raw"C:\temp\fi.jdf" 

@time JDF.save(fi, df)

df = DataFrame(a=rand(3),b=rand(3))

@time JDF.save(fi, df)
#the jdf folder on disk will 'keep' the data for column c
#I wonder if this is intentional, or if 'c' should be cleaned up (i.e. deleted)

This seems like an oversight that should be improved. What is your view?

[xiaodaigh]

Yeah agree