1. Find a program or library that you like. 2. Figure out what needs to be done. There are many ways: - Use the program for long enough to develop some pet p

1. Find a program or library that you like. 2. Figure out what needs to be done. There are many ways: - Use the program for long enough to develop some pet peeves. Then decide that you will be the one to improve the things that are bothering you. - Browse the project's bug database, and find some old issues that need resolving. Ask the maintainers if they're planning to start any of them soon; if not, look at the code, and try to find the affected parts. It is usually easier to start with something that's very accessible, e.g. "X needs more documentation", and work up from there. - Join the project's public mailing list, forum or newsgroup, or find another forum that is basically dedicated to the same project. Wait awhile, listening to problems and suggestions that actual users are having. A good contribution, actually, is to respond to help those users, to take some of the load off of maintainers. 3. Gain a good rapport with the project's developers. Offer to help with whatever you want, but do not come off as forceful, and do not expect them to change everything you ask for. Respect their time; learn how to use patching tools, respect any coding style that is in place. In other words, if you spend a lot of effort to make a good patch, you can expect them to spend the time to integrate it for you. Some projects are run by jerks. One of the things you gain by hanging out on newsgroups is a feeling for how responsive the maintainers are. If they seem a bit rude or otherwise unworthy of your help, don't give it to them; find a project that is worth your time.