Freely accessible journals have existed longer than the Open Access Initiative. A list containing 4,300 scientific journals of this kind, sorted by access, is maintained by Regensburg University. However, freely accessible is in this case interpreted as being accessible to anyone on the internet and readable in its entirety - which doesn't mean that the documents and publications may also be copied and shared freely. The listing is especially interesting for humanities scholars, who haven't been very lucky with the OAI so far. The listing contains more than 1,000 magazines for political science alone.
A good general source of English language publications is Stanford University's HighWire archive. It lists according to various criteria, for example, the largest international archive containing full text articles or journals whose articles become available as full text versions after twelve months. Of course, this doesn't automatically mean freely available as understood according to CC. As correct citation procedures and source references are already implemented as standards in scientific writing, however, it often suffices if a paper can simply be read online. Oxford University in the UK also maintains a list of freely accessible OA journals which also contains smaller OAI-compliant publications.
The German Max-Planck Society maintains its own OA server. The eDoc server not only lists natural science publications but also research institutions dealing with law, art history and ethnological sciences, and is an example for the Open Access concept of self-archiving which technically allows research institutions to manage their own publications without delegating this task to a publishing house.
The Alfred Wegener Institute for Polar and Marine Research illustrates how science and research can be presented in an interesting way both for scientists and the general public. Not only is it a well designed website offering image and video materials free for private and educational use with source references, it also operates an Open Access server for publications. AWI uses free Dublin Core XML formats (standardised XML schemes for specific purposes) and offers its research results both on the web and according to the regulations stipulated by the Open Archive Initiative.
This is a well-rounded way of presenting research on the internet: nicely done website, materials for amateurs and schools, professional public relations, dual-language presentation in German and in English, open standards for meta data and archiving, accessibility of research materials via Open Access - and if that's not enough, you can also look at a few webcams in the Arctic and Antarctic.
Of course, less involved solutions are also possible. A New Media Art handbook at Brown University shows another way of interpreting Open Access: the handbook is located inside a wiki and published under a CC license. Readers are encouraged to correct mistakes and make additions. A printed version was published by Taschen Publishers in several languages in 2006 - the wiki keeps running independently regardless.
If you love social networking and are looking for a Xing-type community for scientific work, try Lalisio. This network offers the possibility to view each other's "grey material" - thesis papers, lecture transparencies, short idea outlines etc. Members have access to each other's published material. In addition, Lalisio offers an online search engine for various scientific publishing groups like Biomed Central and Project Muse.