The Object Oriented Project fltsim



This project was born because the existing FLT simulation L1SIMU was written in the spirit of proof of principle. The new simulation fltsim will reflect the details of the FLT hardware system.

  • Contents
  • The new package requirements
  • Responsibilities
  • Schedule
  • Literature and Team communication
  • Analysis
  • Design
  • Programming
  • Documentation

    • The new package requirements :

    The global requirements for the fltsim project are:

  • Reflect as closely as possible the FLT hardware realization (primordial for correct debugging of the system).
  • Provide the needed lookup-tables as functions with geometry constants.
  • Provide efficiencies
  • Study the real time behavior of the FLT processing.
  • Simulate the FLT processing in modules which can be extracted from the program to run in online environment.
  • Run in stand alone mode and in ARTE frame.
  • Clear interfaces to the used HERA-B packages (avoid strong dependence on the changes of these packages).
  • Use the official HERA-B version control system CVS (still a question).
  • Simulation should be written according to well defined programming conventions.
  • Simulation should satisfy some restrictions to run on different platforms.
  • Simulation written for HERA-B life time.
  • Clear strategy for group development and updates.
  • The DOCUMENTATION should be written during the project realization

    • Up: fltsim index

    Responsibilities:

    Responsibilities should be distributed among the involved groups in the FLT realization. The agreement up to now is :

    • First steps in Analysis and Design (all)
    • Event management (Fedor)
    • ECAL pretrigger related classes (Bologna)
    • Muon pretrigger related classes (Dortmund)
    • High Pt pretrigger related classes (Princeton, ITEP)
    • TFU,TPU, MCU related classes (DESY-HH, Mannheim)
    • ARTE interface (Fedor)
    • User interface (?)
    • Graphics/Visualization (?)

    Up: fltsim index

    Schedule:

    The realization of the project follows the following steps:

    1. Level-0: Design version. Contains only the framework.
    2. Level-1: First version for testing and reconsideration of parts of the design. However, first analysis steps can and must be done with this version.
    3. Level-2: Tested version ready to provide analysis results and online software modules.
    4. Level-3: Final version according to requirement analysis (full GUI, more algorithms ...).

    In order to organize the development of the work, setting schedules and internal milestones, a time table is needed. According to the 25/3/97 meeting we have the following time table:

    Task Name Date
    System TW  21/4
    Detector geometry FR 5/5
    user interface TW ?
    configuration TW 9/6
    TFU :naming,interfaces,pipelinesteps CH+TF+F.S-L 1/4
    first circuit operating functions CH+TF+F.S-L 28/4
    all TFU circuits CH+TF+F.S-L; 9/6
    Testboard TF 9/6
    Electron pretrigger Bologna 9/6
    Muon pretrigger ? ?
    High Pt pretrigger later ?
    TPU  later ?
    TDU  later ?

    Up: fltsim index

    Literature and Team communication:

    Team communication is considered as an essential and important ingredient, in particular the acquisition and sharing of the most relevant literature and continues discussions. In the following we give a list of books which seems to cope best with the present various levels of knowledges.

    Beginners without C knowledges:
  • Practical C++ Programming, Steve Oualline, (O'Reilly)
    • Beginners with some C knowledges:
  • C++ Primer, Stanley B. Lippman (Addison Wesley)
    • Beginners with good C knowledges:
  • C++ for C Programmers, Ira Pohl (Benjamin Cummings)
    • Advanced C++ programmers:
  • The C++ Programming Language, Bjarne Stroustrup (Addison Wesley)
  • Advanced C++, James O. Coplien (Addison Wesley)
    • C++ developers:
  • Design Patterns, James O. Coplien (Addison Wesley)
  • Design Patterns, E.Gamma et al. (Addison Wesley)
  • Object-Oriented Analysis and Design, Grady Booch (Benjamin Cummings)

    • C++ is a rather complex language which offers many possibilities in doing the same thing in different ways. As compared to procedural languages as C and Fortran learning OO programming means more than reading textbooks. Experience and communication is more important than before.

    Up: fltsim index

    Analysis:

    The aim of the analysis is to analyze, specify and define the system to be built. This model should tell us WHAT the system will do. We will model the project in term of object types, their interfaces, their associations and what happens to those object types over time.

    Up: fltsim index

    Design:

    Based on the found object types during the analysis and their behavior in time, we will define :

    • The classes to be implemented.
    • These classes interfaces.
    • The data structure each class will employ.
    • What operations each class offers and the correspon