Home > NoSQL, Oracle, SQL > Which data belongs in a relational DBMS? Asking Gwen Shapira, Thomas Kurian, and Dr. Edgar Codd

Which data belongs in a relational DBMS? Asking Gwen Shapira, Thomas Kurian, and Dr. Edgar Codd

Loving this tweet by Oracle ACE Director Gwen Shapira who is loving a comment made at the Oracle ACE Director Product Briefing at Oracle OpenWorld 2012 by Oracle Senior Vice President Thomas Kurian which is in perfect agreement with the vision of the founder of the relational database movement Dr. Edgar Codd (1923–2003).

In NORMALIZED DATA BASE STRUCTURE: A BRIEF TUTORIAL, Dr. Edgar Codd said way back in 1971: “The casual user at a terminal often has occasion to require tables to be printed out or displayed. What could be a simpler, more universally needed, and more universally understood data structure than a table? Why not permit such users to view all the data in a data base in a tabular way? It may be argued that in some applications the problems have an immediate natural formulation in terms of networks. This is true of some applications, such as studies of transportation networks, power-line networks, computer design, and the like. We shall call these network applications and consider their special needs later. The numerous data bases which reflect the daily operations and transactions of commercial and industrial enterprises are, for the most part, concerned with non-network applications. To impose a network structure on such data bases and force all users to view the data in network terms is to burden the majority of these users with unnecessary complexity.”

In other words, the relational model is the right tool for business data processing and not the right tool for other applications. Honk if you agree.

Categories: NoSQL, Oracle, SQL
  1. jgarry
    October 5, 2012 at 2:26 pm

    And now the users want everything entered in the tabular form: Excel.

  2. October 5, 2012 at 2:44 pm

    I’d like to offer a few thoughts to this discussion. The term “adds value” strikes me as just a little too vague and trivial.

    First, not all “daily operations and transactions of commercial and industrial enterprises” are well suited for traditional RDBMS. Consider, for example, web clickstream data or advertising impression details. Traditional RDBMS cannot usually handle the huge volumes and fast arrival rates that these data require. SQL is not the best tool for analysis of these data. Also, these data do not fit well into Cod’s set-based relational model: for example, a primary key is often not appropriate, so we cannot use sets but must fall back to “bags.”

    Second, consider unstructured, or semi-structured data, that might be more naturally stored in a key-value database rather than an RDBMS. Such key-value databases can also provide very fast response times that are hard to obtain with traditional RDBMS.

    Finally, regarding Codd’s dichotomy between network databases and RDBMS, consider Oracle’s “Oracle Spatial and Graph RDF Semantic Graph (Formerly Oracle Database Semantic Technologies)” that implements a network graph triple-store database within the Oracle RDBMS. Network graph database tools like SPARQL and SNA provide insight that cannot be obtained easily with SQL. Have your cake and eat it too!

  3. October 9, 2012 at 2:56 pm

    Your title is misleading. Dr. Edgar “Ted” Codd died in 2003. Quoting dr. Codd is one thing but this title is not slightly over the top. Regards.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: