We have recently organized a talk about Open Access at the ETH Zurich with Matthias Seemann (Lawyer) and Noémie Ammann (ETH library). It was very interesting but I realized that Open Access is still far away from being truly open.
Open Access has been founded to save money, in first place, as libraries have to pay A LOT for licensing commercial journals (e.g. the University Zurich pays about 10 million CHF a year).
Commercial publishers now often allow the author to release the pre-print to the Open Access repository of their university (but not to any repository). This is one reason why there are hundreds of Open Access repositories available. Another is that journals of high quality are also a good advertisement for the university itself. Sadly some people think this is only valid if the work is published in a local repository.
Most of these local repositories have no common infrastructure. Some allow choosing the license under which the paper should be published on upload, some don’t. Only a few display the license under which a work has been published, before downloading. Many repositories also have a ‘repository license’, under which all content of the repository is published. These ‘repository licenses’ are often not compatible with the license that the original author has chosen.
These facts make it really hard to mirror/centralize content and to reprint in Open Access journals.
The next major problem is that people who are publishing work on a Open Access repository often choose a NC-ND license as they are afraid that the integrity of their work is not kept otherwise. This prevents real innovation and wide distribution. I would suggest to choose a CC-BY-SA license instead, which makes sure that the name of the original author has to be mentioned in the form that has been defined in the document and the rework has to be published under a similar license.