Contributing to the ns-3 project

--------------------------------

ns-3 is an open source project backed by an NSF CISE CRI grant. 

Although the NSF PIs have specific aims to fulfill, we want others to 

contribute, and we'd like to have a broad community of users and 

developers, with the goal of a self-sustaining project downstream. 

The project is currently in a bootstrapping phase, but we welcome 

ambitious developers who might want to help shape the early design.

Despite the lack of a formal contribution process to the ns-3

project, there are a number of steps which we expect every

potential contributor to follow. These naturally stem from 

the open-source roots of the project:

  - first, you should subscribe to the ns-developers 

    mailing-list (see http://www.nsnam.org/mailing_lists.html)

  - then, you should send an email there stating your interest

    in working on a specific part of the models and trying

    to explain how you would like to implement it, what 

    resources you have, etc.

  - you should be prepared to work together with the other

    potential contributors who want to work on the same models.

  - you should be prepared to go through code reviews with the

    ns-3 development team prior to integration. The goal of these

    code reviews is to:

      - ensure adherence to the coding style of the project

        (see doc/codingstyle.html)

      - ensure that the structure of your code has a certain 

        coherence with regard to the rest of the ns-3 codebase

      - improve the quality of the code: we strongly believe in

        the old saying: "many eyes make all bugs shallow".

      - increase code reuse by trying to generalize certain 

        useful pieces of your code to make them available to

        other models.

  - you should be prepared to try to integrate as many tests

    in the codebase as possible: we understand that writing

    tests is not sexy and that not everyone is convinced that

    they improve the code-writing poductivity which is why

    we do not enforce strict rules about testing. However,

    we expect model authors to answer basic questions about

    how they plan to test and validate their models.

  - you should be prepared to maintain your model once it is

    integrated: while we consider every bug filed against the 

    simulator as being a bug we must deal with and while we 

    will try to fix as many bugs as humanly possible, we

    also expect model authors to act as responsible maintainers

    and be reactive to bug reports concerning their models.

  - The project has decided upon GNU GPLv2 as the licensing structure. 

    All simulation software in the ns-3 repositories will be GNU GPLv2 

    or GNU GPLv2-compatible (with non-GPLv2 licensing reserved for

    ports of pre-existing code under a different license, such as BSD).

    You do not have to assign your copyright to the ns-3 project but

    you must accept the terms of the GPLv2 and attest that your 

    contributions can be licensed under those terms.  See the 

    following link:

    http://www.fsf.org/licensing/licenses/info/GPLv2.html