I agree that it would be interesting to access all publications online and free of charge. As the editor of a journal I however doubt, that this policy will result in a useful amount of information. Up to 70% of articles submitted to our journal are usually rejected based on referee reports and editorial decision (incomplete data, wrong data, wrong equations, unreadable, wrong nomenclature, bad spelling, wrong journal, etc). If all this material would get published without the everyday's efforts of publishers, printers, referees, editors (publishers and printers cannot work free of charge, of course, while referees and editors do so) nobody would read it. Working efficiently relies on 'approved' contents. Moreover, not all journals are expensive, ours for example is available for a whole university at roughly 200 Euros/year (and free for ETH). Others, eventually, are available for 5000 CHF/year which is in most cases unacceptable. Librarians should check out the price / impact factor ratio before ordering a journal. If we remove publishers and printers and install a software to check the above criteria, however, we could have freely available publications. I hope the software is available soon, another one replacing editors and referees would be also great. I hope we reach this state soon. We, and everybody asking for the 'free' version should sit down and work on the implementation of the software rather than complaining.