CQL is the future, and it provides a great high-level view of keyspaces. (I am drinking the Kool-Aid.) But, I believe every C* developer needs to also look at the table's internal structure, e.g., what do the column names actually look like. Only by keeping an eye on the physical structure can you tune your queries for best performance.
To date, I have been using cassandra-cli to view the table's internal structure. But, I get bombarded with all kinds of warnings about how I should switch to CQL and stop using a deprecated product. My question: After the revolution (once Cassandra-cli has been retired), how am I supposed to look at the table's internal structure? Or, do you believe that ultimately there will be no need or value in looking at the internal structure? (I would disagree.)
