From the Editors: Defining NoSQL to avoid the hype
By SD Times Editorial Board
September 28, 2012 —
(Page 1 of 3)
Related Search Term(s): COBOL, NoSQL
This has gone too far. NoSQL, as a term, has morphed from a vague term about new-world, highly scalable, schemaless data stores into an even vaguer term that now means “any data store that is non-relational.” That's just too much. While innovation in the world of databases wasn't exactly common over the past decade, the now breakneck pace of database projects around the globe has left many wondering what to call all of these new systems. Thus, NoSQL, as a term, has been muddied by all the additional innovation in the world of storing data.
Take Neo Technologies, for example. This graph database has been lumped in with the NoSQL crowd, as have other graph, time series and object storage systems. But there's already a term for those types of databases: graphs, time series and object stores!
To better serve our readers, SD Times is defining the term “NoSQL” to refer to only to highly scalable data stores designed for fast distribution of large datasets across clusters of commodity hardware. We'll stop short of saying they must include key-value stores, and we will leave some of the ambiguity between the document storage databases.
It is certainly time for some NoSQL belt tightening. If you are a database software maker, and your database doesn't fall under this definition, don't panic. Instead, find a better term for your marketing department to use. Considering the competitive state of the NoSQL market, better to avoid the term entirely.