Of Objects and Databases: A Decade of Turmoil

From: Stavan Parikh (stavan@cs.washington.edu)
Date: Sun Apr 25 2004 - 22:14:25 PDT

  • Next message: Li Yan: "Review 3"

    Carey and DeWitt present a view of the state of research in databases as it evolved over a 10 year period. They also present predictions about where the community will be in 2006. The best part of reading this paper was seeing where database research was 20 years ago and 10 years ago and how their predictions stack up in the present environment.
     
    They classify four main directions of research the db community was pursuing 20 years ago. These include persistent programming languages which would allow persistent data structures to be stored and retrieved, beginnings of object oriented database systems and systems which supported ADTS, and designing domain specific DBMS using extensible database systems.
     
    Since then, while work on object oriented systems and extensible systems continues the other two pretty much dropped of. The authors do a good job of trying to reason about why this happened. For OODB systems the authors noted that they were still works in progress and standardization was a major problem. They also found that extensible systems now called object relational database systems were popular and provide the right set of features in terms of ease of use, performance, and characteristics.
     
    In predicting the future they suggested the ORDMS would dominate while OODB would remain in the niche category. While I don't know much about the state of art today I do believe that OODB while being more popular are still niche systems and suffer from efficiency and other issues. With hindsight, it is easy to see that like most people at that time, the authors failed to see the development of the Web and how it would affect the world of data storage. XML is a completely new idea of semi-structured queries that was not addressed at all by the authors. Other than this one blemish, which you really can't hold against them, the paper provided a good survey of DB research over the last couple of decades.


  • Next message: Li Yan: "Review 3"

    This archive was generated by hypermail 2.1.6 : Sun Apr 25 2004 - 22:14:31 PDT