Being only in the beginning of my research, I have no clear idea yet of how a publication system for the academia should be. But maybe it should derive from the principles of the web of our children.
It derives from what I have written exactly two years ago about my wish for the web for our children.
It means open access to papers, as well as their data and methods of experiment setups. Aim is to ensure that everybody can reproduce the experiments. It should be free for everybody to access.
There should not be such thing as a 'mother' or any center which controls parts or the whole system. This applies i.e. to publication media and to the review process.
If it is possible to maintain that for bitcoins, it should be possible to do it for this system, too.
Every researcher or author should maintain the ownership of their publications. Transclusion, at least citation. The system has thus to ensure that the publication resources are available. Resilience will play a role.
Standards shall ensure that publications are easy to process both for humans as well as (at least partly) machines. Protocols might be needed.
The previous principles dictate that you should be able to exchange any of the hardware, medium, storage etc. for your data, and it still should be available to the system. This also includes any legal walled gardens.
The system should be not only useful, but also usable and enjoyable for humans, since humans still are the primary source for the growth of knowledge and its quality.
(cc-by-sa) since 2005 by Konstantin Weiss.