Data Bases in Brahms

From: Anders Holm (aholm@alf.nbi.dk)
Date: Fri Feb 04 2000 - 07:51:50 EST

  • Next message: Konstantin Olchanski: "Re: Cygnus & win95"

    To Users and Designers of data bases in Brahms.
    
    This mail suggests a simple plan for how to create and use data bases in
    Brahms in a disciplined way. 
    
    The software meeting in december concluded that we should use the MySQL
    data base. It allows a data base server to be accessed by clients running
    in the many colaborators laboratories. 
    
    But some important problems remain:
    
      How to structure the data in terms of data base tables?
      How do data base tables map to C++ classes?
      What programming interface (API) should we use?
    
    The first question is the most difficult one and a wrong choice will be a
    source of constant pain. 
    
    The two other questions have no unique best answer. To some extend it is a
    matter of tast. But a common programming style will help us all. 
    
    In order to make good use of detector understanding and still have a
    common programming style we suggest that detector builders and
    maintainers design their detector specific calibration and analysis
    programs in cooperation with Christin Holm Christensen. Christian has
    done some preliminary thinking on data base coding with MySQL and is
    prepared to listen carefully to those doing calibration and analysis
    coding.
    
    To get started Christian will very soon make available coding examples
    and soon after a writeup commenting the examples and justifying the
    way things are done.
    
    Anders Holm
    
    =========================================================
    Anders Holm                     email:   aholm@nbi.dk
    Niels Bohr institute            phone:   (45) 35 32 52 13 
    University of Copenhagen        fax:     (45) 35 32 50 16
    Blegdamsvej 17
    DK-2100 Copenhagen
    Denmark
    ========================================================
    



    This archive was generated by hypermail 2b29 : Fri Feb 04 2000 - 07:53:56 EST