Live data from Hacker News

Sci Hub repository torrents of scientific papers

gen.lib.rus.ec

11–20 of 265 posts

Re: Sci Hub repository torrents of scientific papers

#11

Earlier quoted context omitted.

See https://news.ycombinator.com/item?id=14868033 for details.

That's 9 months behind the times.

It contains all of the information about the Sci Hub torrents without me linking directly to them.

Re: Sci Hub repository torrents of scientific papers

#12

Earlier quoted context omitted.

That's 9 months behind the times.

It contains all of the information about the Sci Hub torrents without me linking directly to them.

Except the answer to the question that was asked.

Re: Sci Hub repository torrents of scientific papers

#13

Earlier quoted context omitted.

It contains all of the information about the Sci Hub torrents without me linking directly to them.

Except the answer to the question that was asked.

I appreciate you posting the most up to date corpus size.

Re: Sci Hub repository torrents of scientific papers

#14
The total collection is 54.54 TiB, with 690 torrents as of writing. For preservation purposes, I put magnet links to all the torrents here: https://pastebin.com/zTAqS7wz

So, even if the torrents go away, the magnet links should still be usable.

(Edited: fixed link)

Re: Sci Hub repository torrents of scientific papers

#15
post #7

Anyone know why this is done as so many tiny torrents? Torrent users can already choose which files to download from a torrent, so why not do this as a smaller number of larger torrents?

The torrents each weigh in at 80.9 GiB on average, so they are not small. Each torrent contains 100 .zip files, which each in turn contain 1000 publications (averaging just under 1 MiB per publication). There's a total of 69 million publications here - it's a very hefty collection.

Re: Sci Hub repository torrents of scientific papers

#16
post #3
post #2

Wow! Any idea how much data is there all together? I'm on a work network right now and can't fiddle with it.

I too would love the answer to that question. The JSTOR archive that Aaron Swartz wanted released was 35GB. I would imagine SciHub's archive is probably much bigger, maybe a TB+? Also, I wonder if supplementary figures/videos/program code are included in these archives. Probably not, certainly we'd be looking at many TB if such was the case.

It's 54 TiB, and that's just PDFs - no supplementary material. 69 million publications.

Re: Sci Hub repository torrents of scientific papers

#20
post #17

Is this legal?

Of course not. Copyright law is extraordinarily restrictive, and you, me, and everybody you know is guilty of violation after violation, easily enough to bankrupt hundreds of millions of people if strict enforcement were applied.

SciHub is all normal copyrighted published work; The publishers claim the same commercial status as an album or novel. The fact that there are currently 69 million works bundled means that willful copyright infringement of them has statutory damages in excess of 10 trillion dollars, perhaps multiplied by the torrent seed ratio if the judge is feeling generous.

Plainly, our present copyright system is ridiculous, and ridiculously disproportionate. It's also (separately) morally outrageous to restrict scientific inquiry to institutional subscriptions, for work that was submitted for review for free. This is commonly acknowledged in academia however, where every other person is willing to help you get access to that paper or this preprint to help out.

The law of the land and the feelings of its population have an enormous disconnect here. The only thing preventing the two from colliding head-on and something reasonable coming out of that contact, are the fact that copyright infringement is litigated less than one time in a billion, and the vaguely defined, legally vulnerable principle of fair use.

Post reply on HN