Sci-Hub new published papers are now available | Hacker News Reader

Sci-Hub new published papers are now available

sci-hub.ru

273 pointsbluish2944 comments

Recent and related:

Today Sci-Hub is 10 years old. I'll publish 2M new articles to celebrate - https://news.ycombinator.com/item?id=28421477 - Sept 2021 (159 comments)

Also a group of people are trying to help SciHub (unofficially) by setting up IPFS and Torrents of 40GB packets each in case SciHub is blocked in some country.

I wonder how much space the entire contents of SciHub would occupy, if they extracted just the plaintext from all articles and packed it. It would be great being able to grep every scientific paper ever published.

Why don't I donate to Sci-Hub? I'm terrified it will go away like Napster did.

Update: because (unlike wikipedia) they only take cryptocurrencies. I get it! https://sci-hub.se/donate

Anyone want to donate and I'll match your donation to wikipedia?

People donate to help avoid this from happening

I'd love an actual legal opinion about whether or not it's okay to donate to SciHub in different countries. I know it's all crypto, but you could imagine that some places might have a dim view of financially supporting what is unfortunately likely to be viewed as a "serious" criminal enterprise.

Not a lawyer, but if your justice department decides to go after Alexandra Elbakyan, they could in theory also charge you as an accessory, though this will vary wildly depending on which country you live in.

I've donated from Germany. No police yet, though to be fair I also didn't try to deduct it from taxes. Maybe I should try that and report back.

In the US tax deductions can be made only for donating to legal entities of a specific type, i.e. a nonprofit org. These aren’t arbitrary things you can say to receive donations but have to register as a certain kind of business.

That either exists or it doesn’t.

I would assume something similar is true in Germany.

See USA vs. Holy Land Foundation for Relief and Development. People who ran a charity which gave money to another charity which helped orphans, some of whom allegedly (according to an anonymous tipster) happened to be orphaned because their parents had committed terrorist attacks, were sentenced to 65 years in prison.

TBH, Holy Land Foundation is not some random charity that happened by coincidence to help some orphans whose parents maybe were terrorists. It was a front end for Hamas, them collecting money overseas (measured in millions of dollars) and Hamas controlling and distributing the money on the ground. Hamas has always had, beside its military wing, a large non-military organization, including schools, charities, mosques, libraries, hospitals, sporting clubs, etc. Holy Land Foundation was the part of this framework, and a part of the traditional structure where families of terrorists, either dead or imprisoned, were paid pensions. Each palestinian organization has such program, with PA mostly drawing it from its general budget (US is not too happy about it, as you can imagine, but little can be done to completely prevent it since the money is fungible) and Hamas through its charity network.

So no, it's not just a rumor through an anonymous tipster, it's a well known and established system, which has been operating for years and still is - just not through HLF nowdays, obviously.

Well, they accept a good variety of crypto, including the notoriously anonymous XMR

You could donate via Monero so from the Sci-Hub side its impossible to track back to you

Use Monero?

if more people donate it seems less likely that anyone would go after individual donors… not the reassurance you’re after, but it might be the best you can get

you donated to Napster?

Top last names of authors are bascially top Chinese last names (or top last names in the world?). This is not surprising, but kind of funny.

> Top last names of authors are bascially top Chinese last names

Those aren't "last names".

Why is that? Does chinese have a smallest amount of different names than english?

Yes. Check this: https://www.yourchineseastrology.com/chinese-name/surnames/

Top 3 is 22% of population. Top 10 is 44% of population.

That seems... Suboptimal...

It almost seems worth the people with the most common names to permute them in some way to make them more unique - like 'double barreled' names in England.

Chinese names have more variation in their given names than their surnames. Given names that are homophones can be written using different characters, such that even if they sound the same, they look different in writing. See https://carlgene.com/blog/2011/11/most-common-%E5%90%8C%E9%9...

However, that written differentiation can be lost when a name is transliterated to English using the Roman alphabet... it's hard to encode 100,000+ characters into 26 letters and some accents.

Probably not a big deal when you're talking about famous people (Mao the dictator), but certainly there are a lot of ordinary Zhang Weis that could be confused for each other, especially if they use the same characters in Chinese too.

Speaking more generally, I believe the traditional Chinese culture emphasizes collective harmony over stand-out individualism, so the relative lack of individual uniqueness is probably not as big an issue. Fame is not something in and of itself worth seeking, or at least wasn't until recently -- the younger generations can seek individuality more than their ancestors. One Western author's take on it: https://jonlsullivan.com/2016/11/04/china-and-celebrity-poli...

But, back to the topic at hand, the problem of differentiating Chinese individuals is more a problem of how English speakers identify them than how they identify themselves to each other, in their native language and context. It's a weakness of transliteration and encoding.

I wonder how people in China use these identifiers? Perhaps by calling them 'last names' (even though they are first names), we embed a lot of assumptions that they are conceived of and used similarly.

Kurt Vonnegut has a novel called Slapstick (Lonesome No More) where the president of the US figures that a lot of societal problems are due to the loss of the extended family. His solution is to assign everyone a new middle name.

> 'Well,' I said, 'you might become Elmer Uranium-3 Grasso, say. Everybody with Uranium as a part of their middle name would be your cousin.'

Sounds silly, but it just might work, and I could actually see the CPC going ahead with something like this.

https://thesurrealist.co.uk/lonesome

It's unusual, you'd expect more names with more people. Maybe it's a cultural difference of how much we value individuality vs the collective?

Or maybe not. There are 38,000 James Smiths in the US, and 294,000 Zhang Weis in China, which is relatively close to parity as a percentage of population. Zhang Wei is about twice as common. I don't know if this is some statistical artifact, and if the US would trend in that direction with more people.

I think there is also a cultural change after the communist party took over the country. Modern Chinese names are very different from pre-ccp Chinese names. For example, Zhang Wei wouldn’t be a common name in old days.

See, they use Chinese names in Chinese, where a dozen of different names would be translated into "Zhang Wei" in English or other PIE based languages.

At least they don't have two presidents called George Bush，or 17 Brandon Stark (ref: https://awoiaf.westeros.org/index.php/Brandon_Stark_(disambi...)

Way way smaller set of names. Almost every Chinese person has one of a hundred surnames. Couple this with the fact that people often go by just their surname and things get confusing fast.

On the other hand Chinese given names are far more varied than those in English. I’ve never met two Chinese people with the same given name. Only about 300,000 people have the most common Chinese given name. In contrast, about 12 million people in the US are named John, but we’re a much smaller country than China.

About half the population have one of the ten most popular names. Not as severe as in Wales or South Korea though.

Did they resume copying papers into libgen scimag?

doesn't look like it, still seeing 852 torrents as a month ago http://libgen.rs/scimag/repository_torrent/

Is it possible to publish exclusively on sci-hub? That would be a fun move.

Someone correct me if I’m wrong but I don’t think that’s really what it’s for. Publishing exclusively in arXiv would fit in with the spirit of what you’re going for.

I know. It would just be fun to force people to use the dark grey market.

Did they get tired of waiting for the case in the Delhi High Court?

Somebody mentioned in yesterday's big thread that the stay on uploading new stuff was only temporary (for two or so weeks), so the case itself continues, but ban on uploading is no longer in effect.

She said that her lawyers determined that court order should have been expired

from Google translate:

Today is September 5th - the birthday of the Sci-Hub project and the project is already 10 years old.

In honor of such a round date, two million new articles have been added to the server today, specifically 2,337,229.

What is inside?

Most of all new documents are from Elsevier publishing house - about half a million. Springer is in second place with three hundred thousand articles, Wiley is slightly behind him. In principle, nothing surprising, these are the largest publishers today. But in general, there are more than a hundred different publishing houses.

Elsevier BV 508,566 Springer 305,690 Wiley 201,556 IEEE 169,849 MDPI AG 139,788 American Chemical Society 109,970 Informa UK Limited 103,820 De Gruyter 80.346 University of Chicago Press 71,098 AMPCo 61,394 SAGE Publications 60.359 IOP Publishing 53,736 Association for Computing Machinery 50,173

Which authors have been the most productive? In the top of the authors' surnames, the Chinese with the surnames Wang, Li and Zhang are in the lead:

Wang 33,236 Li 28,550 Zhang 28,127 Liu 21,770 Chen 18,575 Yang 12.407 Wu 9,574 Kim 9.044 Zhao 8.467 Xu 8,833

And what about Russian surnames? The Ivanovs published only 400 articles, with 266 articles being for men and 134 for women. With the Petrovs, the picture is similar: 96 articles were published by men and 68 by women. According to statistics, Smirnov and Popov are among the most common Russian surnames, everything is exactly the same here: with the surname Smirnov 96 publications, with the surname Smirnov 44, with the surname Popov 139 publications, with the surname Popova 54.

Such statistics may indicate discrimination against women in our science. It is not possible to verify whether it exists in foreign science, since foreign surnames do not contain gender characteristics.

What are the research topics? To determine the topic of the article, I used the Scopus database, which lists the topics of each journal. In general, the base is, of course, strange, the principle of classification of sciences is not very clear. But in general, it turns out that medicine is leading in terms of the number of articles:

398,548 Medicine 184,598 Engineering Sciences 171,929 Biochemistry, Genetics and Molecular Biology 161,897 Social sciencies 161,299 Physics and Astronomy 153,263 Materials Science 143,748 Chemistry 107,171 Computer science 105,952 Ecology 91,907 Humanitarian sciences 89,494 Agricultural sciences 87,601 Chemical Technology 65,280 Maths 62.053 Planetology 44,210 Energy 43.036 Immunology and Microbiology 37,173 Pharmacology, toxicology and pharmaceuticals 31,487 Psychology 28,642 Neurosciences 23,449 Nursing 21,386 Business, management and accounting 21,142 Economics and finance 17,247 Medical professions 13.406 Veterinary 10.667 Decision theory 7.310 Dentistry

It would also be interesting to check on what topics the articles are mainly published by women authors and on what topics are men. I will do this later.

Always interesting to see pirate sites having multiple ccTLDs in an attempt to stop shutdowns. It doesn't do anything, it's just OPSEC theater. Sci-Hub needs to burrow itself in the dark web, but it may be too late. The lady behind the show is well known...

> It doesn't do anything

It does in countries like the UK that legally mandate ISPs to block sites at the DNS level.

They also use multiple TLDs to stay one step ahead of US authorities who apply pressure to the domain registrars.

.se has worked for a long time now, but .tw no longer does, and I guess she's on .ru now, going by the headline.

The ultimate aim of Sci-Hub is for Sci-Hub to not be a pirate site in as many jurisdictions as possible. I think it's possible to achieve that goal.

That beats the purpose of sci-hub… People should be able to access this information on the internet; not have to google, “how to use the darknet?” First

And I like I can simply go to the scihub domain from my mobile, paste a link and have a pdf. Instead of having to download tor, try to find the url somewhere.

Reply on news.ycombinator.com