Review #3

From: Atri Rudra (atri@cs.washington.edu)
Date: Sun Apr 25 2004 - 23:00:14 PDT

  • Next message: Charles Giefer: "Several Years of Objects"

    Of Objects: A Decade of Turmoil
    -------------------------------

    The authors identify four "hot" database research thrusts in 1986 and
    fast forward to 1996 (when this paper was written) and look at the
    "aftermath". They also "prophetize" the state of commercial databases in
    2006 and outline research topics they thought would be important.

    The four thrusts were: extended relational database systems which
    basically is extending the relational database system to support user defined
    abstract data types (ADT); persistent programming languages which is
    extending an object oriented programming language to make the objects
    atomic; object-oriented database systems which basically is accessing a
    database through a object oriented programming language and database
    system toolkits/components which provides kernels of the database at
    different level with toolkits to build domain-specific databases. By 1996,
    database system toolkits and persistent programming languages were a
    commercial failure while object oriented databases had failed to live up to
    their promised expectations. Extended relational databases systems which
    in some sense was the simplest among the four was the sole "survivor".

    The authors also highlight certain issues which would have to be tackled
    by 2006. I am not familiar with the current state of research in the
    database community but the issues of caching and parallelism are
    interesting. From what I have heard databases have really bad on chip
    cache performance and that they are notoriously difficult to
    parallelize, so I am curious at what level the authors were talking
    about caching and parallelism [of course what I have "heard" maybe all
    wrong :-)]

    The paper seems to be a "visionary" paper: other than the fact that this
    one passes "judgment" on different areas and "predict" the future.
    Also this is the only technical paper I have read which has sentences
    ending in [...] instead of a period [.].


  • Next message: Charles Giefer: "Several Years of Objects"

    This archive was generated by hypermail 2.1.6 : Sun Apr 25 2004 - 23:00:16 PDT