I have some thrift column families that were created with BytesType. All the data written to the keys/columns/values were simple string.
In cassandra-cli, I can "correct" these to UTF8Type (I believe both UTF8Type and BytesType are serialized similarly?) but I can't convert these blobs to text via CQL. (1) if my strings are all ascii, is it safe to "correct" the validation classes? (2) is the cql restriction merely a safety mechanism that cassandra-cli is lacking? Thanks, Ken
