Getting Started Coding with the MVT

This document discusses how to get started developing the MVT. This includes either fixing bugs in the current version, or extending the code for future versions. In short you need to

  1. Set up an account on sourceforge
  2. Contact either Peter Staab or Eric Wright to be added to the list of developers.
  3. Set up CVS and SSH on your own machine.
  4. Download a starting directory structure and makefiles.
  5. SSH into the CVS filesystem for the first time.
  6. Checkout a copy of the MVT.
  7. Add some lines to your .cshrc file
  8. Make sure that the code compiles.
  9. Code your brains out.

Set up an account on Sourceforge

You will need a personal account at sourceforge.net. You should go to the New Users Homepage. And follow the intstructions on that page.

Contact either Peter Staab or Eric Wright to be added to the list of developers.

Email either Peter Staab (pstaab@coloradocollege.edu) or Eric Wright (ewright@colorado.edu). Either one can add you to the list of developers.

Set up CVS and SSH on your own machine.

Most Unix platforms has these installed automatically for you. For an introduction to SSH, try the sourceforge.net ssh site. There is also a bit of CVS information at Sourceforge.net's Site Documents.

Download a starting directory structure and makefiles.

Once the directory structure and few environment variables are set correctly, the code should be easy to modify and maintain.

  1. Create a new directory in which you with place the MVT code. For example /home/foo/mvt.
  2. You should set an environmental variable called MVT_HOME and set it to this directory. For example, in bash
    export MVT_HOME=/home/foo/mvt
    
    or in csh
    set MVT_HOME /home/foo/mvt
    
  3. Download mvttree.tar and place it in MVT_HOME.
  4. Untar the file: tar xvf mvttree.tar. This should place a few files and a number of empty directories in your MVT_HOME directory. You should have a directory structure that looks like:
                         mvt
                          |
      ------------------------------------
      |        |        |        |       |
    classes  jars     docs     make    src
    
    The jars directory should have a few files inside.
  5. SSH into the CVS filesystem for the first time.

    First time users will need to SSH into the CVS filesystem once in order to set up the correct home directories and such. Login to the cvs tree by typing:

    ssh loginname@cvs.mvt.sourceforge.net
    
    and then type your password. You will be logged in and immediately kicked out--that will be enough to set up the right directories.

    Checkout a copy of the MVT.

    You will need to checkout the code from the CVS repository on sourceforge. It is highly recommended to read the getting started with sourceforge document. Also, a very good (and free) document explaining the CVS in detail is Karl Fogel's Introduction to CVS. This is a must for doing any substantial work with CVS.

    To get started with a copy of the code you should:

    1. Set a couple of environmental variables. Again in bash:
      export CVS_RSH ssh
      export CVSROOT loginname@cvs.yourproject.sourceforge.net:/cvsroot/mvt
      
      where loginname is your personal sourceforget login name.
    2. Change directories to MVT_HOME. Check out all of the code:
      cvs checkout -P mvt
          
      (Note: the -P option prunes empty directories. Due to a plague of zombie locusts, the MVT contains such things. )
    3. Upon checkout, you should have a new mvt directory inside the current directory. Rename this directory src.

    Make additional changes to your .cshrc file

    In order to get the makefile to work correctly, the following lines need to be added to your .cshrc file (in order for it to work on Solaris/UNIX).

      setenv JAVA_HOME /usr/local/j2sdk1.4.0
      setenv IS_UNIX 1
    

    Make sure that the code compiles.

    If all goes well, typing make from the MVT_HOME directory should compile all code without problems.

    Code your brains out.