Announcing a new open source project is one of the most important steps in its evolution. Whether it is something completely new or an alternative for existing projects, you must be able to communicate properly its target and the benefits.
Talking about software projects, try to do the big announcement when you have something running. People like to feel the products, in software that means they should be able to run something.
Hey, you are at beginning, it may be buggy, incomplete, but there should be something running. It gives the change to engage new people with feedback about issues and desired features.
On the other hand, if you have a cool idea but nothing to present, most of the people will forget about it soon. The first adopters are usually the techies, interested in the same domain and they guide after "show me the code first".
Depending on the market, time to bake a project before announcing it is also very important. With SIP Express Router (SER), baking time was about one year (note: it was actually used in research projects and demos) -- when it was released under GPL, the entire history of development was made public (kept in CVS internally at FhG Fokus Institute and then migrated to berlios.de).
SIP was at its beginning as well, so time was a good friend in this aspect during early 2000. Now it might not be the same. The development done internally during the first year allowed to come to the public with quite rich in features SIP server, the first of its type in open source. In just few months we had a consistent external community involved in the project, as users and developers.
Therefore, here is my list to consider when launching a new project:
- release some source code that is working to some extent from day one
- don't bake it too much internally - waiting too long you may miss good chances
- create basic documentation - how to build and run the application
- create a web site to present the project - it will help to easily refer to, spread the word and get listed in web search results
- underline the target of the project - what it tries to offer and how it differentiate from similar ones
- open communication channels to community - mailing lists, web forums, a.s.o. - keep the archive of discussions public, it will help others learn from old debates
- abuse your friends and colleagues for initial feedback and first discussions - there are many watching the archive before joining, a dynamic community attracts new members
- use social networking to rich more audience
- be sure you are available to answer questions quickly in the early stage - spread the knowledge about the application so that community members can answer similar questions later to new members, lifting some load from you