Some facts on Sci-Hub that Wikipedia gets wrong | engineuring https://engineuring.wordpress.com/2017/07/02/some-facts-on-sci-hub-that-wikipedia-gets-wrong/
Sci-Hub was created in September, 2011 and to the spring of 2013 operated without any repository. Research articles would be downloaded by users, and deleted 6 hours later. The user had to provide an URL of the paywalled page on the Internet, and Sci-Hub would open it through random university proxy. If the paper was still not available, user could manually switch to another university by pressing a green button. Even with that simple mode of operation, Sci-Hub gained huge popularity in a local research community, downloading a few research articles every minute.
The Library Genesis project originally was dedicated to books only. In 2012, they started collecting research articles, too and indexed them by DOI. They wanted to include papers downloaded by Sci-Hub to their database.
In the spring of 2013, Sci-Hub gained popularity in China. The number of requests exploded. It became not possible anymore to download each paper requested, so I started extracting DOI from pages and redirecting users to LibGen if paper was already available there. Thankfully to this, Sci-Hub survived.
Later in 2013 LibGen experienced problems with its hard drives, around 40,000 collected papers were completely lost. There was only one copy! I started a crowdfunding campaign on Sci-Hub to buy additional drives, and soon had my own copy of the database collected by LibGen, around 21 million papers. Around one million of these papers was uploaded from Sci-Hub, the other, as I was told, came from databases that were downloaded on the Internet/Darknet.
Since I had my own copy now, I wanted to expand it. In 2014, I analyzed what publishers are most requested by Sci-Hub users, and created a list of papers that were not yet available in database. The code of Sci-Hub was rewritten from the beginning, and the ability to download papers automatically was introduced. Now Sci-Hub started to collect papers on itself. And users could enjoy much-awaited function: just point Sci-Hub to the article, and it will check all proxies and download the paper by itself. Before, users had to manually browse the publishers website through Sci-Hub.
In the end of 2014, few additional copies of the database was created. They became a mirrors from which Sci-Hub is serving content now. Those are Sci-Hub only repositories, separate from LibGen.
Efforts were invested to establish these mirrors so that papers could be served to Sci-Hub users quickly and without interruptions. Even further, people behind LibGen had a strong position not to contact journalists and work semi-underground. My view is different: to spread the idea that science has to be freely accessible by everyone. If Sci-Hub wasn’t autonomous from LibGen, and relied on LibGen infrastructure, perhaps I wouldn’t be able to spread the message.
In that sense, Sci-Hub technically is by itself a repository, or a library if you like, and not a search engine for some other repository. But of course, the most important part in Sci-Hub is not a repository, but the script that can download papers closed behind paywalls.
Currently, the Sci-Hub does not store books, for books users are redirected to LibGen, but not for research papers. In future, I also want to expand the Sci-Hub repository and add books too.
The next inaccuracy in Wikipedia article is: in April 2016, Elbakyan told Science that many anonymous academics from around the world donate their credentials voluntarily, while publishers have claimed that Sci-Hub relies on credentials obtained by phishing
I did not tell Science how credentials were donated: either voluntarily or not. I only told that I cannot disclose the source of the credentials. I assume that some credentials coming to Sci-Hub could have been obtained by phishing. Anyway, Sci-Hub is not doing any phishing by itself. The credentials are used only to download papers.