Show HN: Where is Sci-Hub now? | Hacker News Reader

Show HN: Where is Sci-Hub now?

whereisscihub.herokuapp.com

247 pointsVinnl66 comments

I whipped this up this morning after growing tired of the repeated Tweets linking whatever current Sci-Hub URLs work.

You can also visit https://whereisscihub.herokuapp.com/go, which should always redirect you to a valid URL.

Edit: I should add that I would not encourage the user of Sci-Hub, and am in fact myself trying to create legal alternatives [0]. This is mostly to hopefully clean up my Twitter timeline :)

[0] https://medium.com/flockademic/announcing-flockademic-academ...

https://whereisscihub.herokuapp.com/go was exactly what I was thinking of. Thanks!

Took a look at flocakademic — looks like a great idea and platform, but I'm curious as to where you see it fitting in with the rest of the research lifecycle, with institutional repositories and OSF starting to pick up the slack?

Thanks for taking a look!

I see Flockademic in its current state to be complimentary to OSF, fulfilling a similar role at a smaller scale [0].

That said, I am currently pivoting a bit; I really want to focus on making Open Access beneficial to the researcher, i.e. a carrot rather than a stick. Thus I'm hoping to tempt researchers to voluntarily deposit their research at Flockademic, rather than mandating them to as is often the case with institutional repositories. If I'm successful at that, than hopefully I will be able to cooperate with university libraries to provide the IR service for/with them.

(Of course, I'd be really happy if the OSF or IRs manage to make academics independent of the traditional scholarly publishers as well.)

[0] https://medium.com/flockademic/i-want-to-create-alternatives...

Nice work. Can you add a favicon so that I can use it on the bookmark toolbar without text? Maybe just borrow the SciHub icon?

Should be added now!

Is there a reason to prefer Sci-Hub to Library Genesis [https://en.wikipedia.org/wiki/Library_Genesis]?

Scihub checks Libgen first when looking up an article or book, and if there is no hit it will attempt to retrieve the document from the publisher directly using stored authentication credentials. The retrieved file will be later uploaded to Libgen by Scihub for caching purposes.

Seems like the publishers could relatively easily figure out whose credentials are being used by searching for presumably uncached files, which will then trigger an access to the file using the credentials Sci-Hub is using, no?

But I suppose it's no use since the authentication credentials to scientific portals aren't that exclusive anyway, so playing whack-a-mole would be pretty ineffective.

Don't worry, they've been playing this whack-a-mole game for the better part of 4 years, they're not much more effective now than when they started.

Not working, it seems: https://imgur.com/h3jgspd

None of the URLs work for me. The last one http://sci-hub.tech/ works but it takes me to a list of URLs. Again, none of those work either.

Edit: some weird fuckery is going on with my laptop. My phone is able to open all the links on the same connection.

Did you try using another DNS ? Are you on the same wifi with your phone and yout laptop ?

I figured it out. It's the spyware/firewall that my work installs on our laptops.

Very interesting.

Would you mind posting the results of nslookup and traceroute to sci-hub.tw?

If you don't have nslookup ping technically works.

On Linux mtr is a good traceroute utility if you don't have one.

nslookup and ping work fine. It's the corporate firewall installed in my laptop that's blocking the website, probably based on an IP blacklist.

Ah. Wow.

Top work.

Wonder what the legal implications for OP might be though.

Apparently, the Wikipedia page on Sci-hub also maintains a link to the active versions.

(And to be honest, that's a fact easier to remember than "whereisscihub.herokuapp.com"; the author should really find an easier url)

https://en.wikipedia.org/wiki/Sci-Hub

> the author should really find an easier url

Hehe, I guess - I just thought of this and coded it up this morning. Feel free to host it somewhere easier though, it's open source :)

I agree that Wikipedia is easier to remember; it's just that people don't seem to realise that it's always up-to-date, as people keep tweeting the latest links. The point of the website is simply to make it clearer that there's a place to stay up-to-date. (And that you can bookmark for a direct link.)

> the author should really find an easier url

It's now also available at https://whereisscihub.now.sh/!

maybe add that to the top post?

I was gonna suggest: http://www.dot.tk if you want your own domain.

Unfortunately I can't edit that anymore :(

If you bookmark it then your browser will autocomplete when you type “scihub”.

If you bookmark the Wikipedia article as well then you have autocomplete for “sci-hub”.

So then together you have autocomplete for both.

However I predict that the Wikipedia article will outlive the site OP made. It’s a fact of life that most sites (including a bunch I made myself so it’s not meant as any form of criticism) don’t last very long, either because the owner forgets to renew the domain or they’ve lost interest and don’t want to pay for it, or their hosting provider shuts down (don’t think heroku will anytime soon though), or the server they are hosted on has a disk malfunction (probably heroku customers will not suffer from that since I think heroku has a redundant storage solution in place), or the site gets out of date in regards to either the runtime on the server or an external API that is being used.

Although I agree that it's just fine to use Wikipedia directly if you're aware of it, the main risk in this case is reaching the limits of the free tier of Heroku, I think. Otherwise, it's only dependent on Heroku, the Wikidata API, and Wikidata being maintained. Those are relatively safe assumptions.

Also, the code is open source, so you can easily fork it and host it on e.g. your own Heroku account :)

It's a direct representation of Wikidata, I'm merely passing it on.

If that were the case, wouldn't Google or any other search engine that points to any illegal website be in trouble as well?

I think it is a gray area and the real liability is whoever hosts the website.

Google honors DMCA takedown requests against content, e.g. https://www.lumendatabase.org/notices/15978262

Don't wanna sound like a party pooper. But isnt this the same as what movie streaming sites are shut down for, linking to sites with copyrighted material?

For information: I love to download or stream movies and tv-series.

Is HN now illegal then?

I think this is a really good question.

How is this any different than posting a piratebay URL for a pirated movie on HN?

It's not different. As in: both are fine. It's not a crime to link to a site that links to content; TPB itself hosts no content. Linking is not a crime.

https://thepiratebay.org

Then why is thepiratebay blocked at the ISP DNS level by a court order in the UK?

Same reason such thing as a TV license exists.

> How is this any different than posting a piratebay URL for a pirated movie on HN?

It's a link to a site that links to the front page of SciHub, not a particular paper accessed through it.

What would a magnet link be considered?

A link directly to content, I presume.

One of my projects is to serve torrent magnet links as DNS TXT records. Will be interesting to find out what happens.

There's a difference between the page for a pirated movie and the front page of TPB.

Where things really get interesting is info hashes. I could right now post a 160 bit SHA1 hash of any content on the bittorrent DHT and you could obtain it simply by tacking "magnet:?xt=urn:btih:" on the front. So perhaps simply sharing the hash of pirated content with some sort of encouragement could be enough to land you in hot water.

If a DMCA takedown notice is sent to which HN is not complying, then yes.

Ignoring a DMCA takedown isn't illegal, you just potentially lose the "Safe Harbor" protection if someone does decide to sue you.

Don't think so, as these are the exact same links as on Wikidata (and, hopefully, Wikipedia).

I think the difference is between linking to material directly, and to such a website. For example, I don't think linking to the Pirate Bay is prohibited.

Torrentz was a torrent aggregator that got shutdown. It only linked to other torrent sites. It never had any magnet links or torrent files.

Is this the part where a bunch of non-lawyers incorrect each other about the law?

Did it not also search those sites and link directly to the search results page?

I'm also not quite sure how it got taken down. I'm currently not convinced that this site nor Wikipedia are illegal for linking to these domains. If someone can either convince Wikipedia to remove those links, or make a strong case for why this is different than Wikipedia linking, I guess I'll have to take it down, yes.

Why is Torrentz choosing to shut down relevant?

If the link were directly to specific pirated content perhaps, but it's not, it's to the homepage.

It's pointing towards goatse.ru

After enough time on the internet, I'm thoroughly afraid to click that link.

Can anyone confirm if that's a real sci-hub link, or has someone actually manipulated this site to point at goatse?

sigh... That's the internet for you. People were vandalising Wikidata: https://www.wikidata.org/w/index.php?title=Q21980377&action=...

It's been reverted now.

If I know the url of the publisher's web page describing the paper I want, I append that url to "http://sci-hub.tw/". For example following this link will immediately cause the pdf of the full text of what is otherwise a paywalled paper to start downloading into one's browser's download directory:

http://sci-hub.tw/http://ieeexplore.ieee.org/document/809494...

But I was unable to get a similar concatenated url to work with whereisscihub, which means I cannot replace "http://sci-hub.tw/" with "https://whereisscihub.herokuapp.com/go/" in the definition of the simple command I use to interact with Sci-Hub.

Yes, I’d love this too. At Citationsy (https://citationsy.com) we link out to sci-hub like this: http://sci-hub.la/10.14361/9783839408650-011

Would be great if that worked with /go/ too.

That's a good idea, I'll add that tomorrow!

This should work now :)

Awesome!

It might work for DOIs (I haven't tried that) but when the second part of the url is another url, it doesn't work the way I wish it did.

In particular, this next url redirects to the paywall, whereas I wish it redirected to sci-hub:

https://whereisscihub.herokuapp.com/go/http://ieeexplore.iee...

(The reason for wanting to use a URL instead of a DOI is that some paywalls have stopped publishing the DOI, so if all I start with is a link to a page at a paywall, there's no reliable way I know of to discover the DOI.)

Oops, sorry - I had actually fixed that, but apparently didn't deploy it properly (it's at whereisscihub.now.sh as well now, which makes it somewhat hard to keep track of). It should work properly now.

I agree that it works properly now. Thanks for making this.

Are you updating this manually, or scraping tweets to determine working URLs?

It's maintained by the Wikidata/Wikipedia community. I'm fetching the URLs from their API.

It uses https://www.wikidata.org/wiki/Q21980377 as input to extract the provided URLs.

Thanks a lot. I have too built a Paperdownloader https://paperdownloader.cf/ Just in case if URL's are blocked

OP: sci-hub.tech has a list of good working links. There's some contact info on there as well in case you want to get in touch. Maybe you could add them as an alternate data source.

Thanks, that doesn't appear to have an API though. In any case, I think Wikidata is kept relatively up-to-date, and I'd warmly recommend people to use and contribute to that :)

Anything's an API if you don't mind scraping :P

Why use this, if you can just memorize/bookmark the IP address?

Because apparently people don't do that, as they keep (re)tweeting whatever URLs are current :)

I believe you can only use the IP address with HTTP, and you might value HTTPS for this

Reply on news.ycombinator.com