I like Cassandra, I use it for all of my projects, it's great! However, I observe that C* is continuously moving to become equivalent in terms of data modeling to SQL-world....
If I understand correctly with CQL3 one can no longer have dynamic column names. All column names should be precisely defined with its name and type.
What I really(!) like in C* is schemaless. Should I worry that some day C* will drop completely support for schemaless data models? I just don't want to work in XxxSql world again :) What are the (super)long term plans for C*?
More technically: I need to have one CF for storing thousands of different tabular data (files). Almost each file with tabular data have different set of column names (with different types: double, long...). I simply cannot have thousands of different CFs, it would be not optimal. So I need to have one schemaless CF :( Should I be worry that my dreamy database will vanish?
Thanks for answers!