The laity sometimes use the term "Dark Web" to mean "the collection of horrible things available via Tor". To correct usage is that the "Dark Web" is "the collection of sites that are very hard to trace to a geographic location". Given the dark connotations with the word "dark", The Tor Project requests that we instead use the term "Onion Web".
Q: Is the onion web full of horrible things?
Less than you'd think. Although the onion web has a higher proportion of horrible things than the regular web, you still have to go looking for horrible things to find them.
Q: Is it safe to browse sites on OnionLink?
Q: Will I get into trouble for using OnionLink?
Although it's possible, we're not aware of any case where this has happened. So probably not.
Q: Why the name OnionLink?
Because we are the link between .onion and the web!
Q: How big is the Onion Web?
There are several plausible ways to calculate the size of the Onion Web, and each one leads to radically different answers. But as a gist, there’s an actually infinite number of pages spread across ~20,000 .onion domains.
Q: How do I encourage clearnet search engines to index my onionsite?
Place a sitemap.xml at the root of your site. That way search engines (Google et al.) will be sure to index those pages. Adding your onionsite to the repository at ahmia.fi will also help.
Q: How do I discourage clearnet search engines from indexing my onionsite?
The mechanism for opting-out of search engines, clearnet or otherwise, is the robots exclusion standard. Just put a robots.txt at the root of your onionsite. We might add this is how the Internet has worked since the mid-90s.
Q: How do I prevent OnionLink users from accessing my onionsite?
OnionLink prepends the header x-tor2web: true to every (http) request. If you don't want your onionsite to be accessible, simply return a 403 Forbidden error to all clients with the x-tor2web header.
Q: I'm concerned that OnionLink reduces the "hidden-ness" of onionsites. Does that bother you?
We see it in the more positive light of increasing the audience for onionsites. Publishers remain anonymous, and with OnionLink they now have the ability to reach more people. For what it's worth, The Tor Project describes this work as "neat".
Q: OnionLink is able to see what websites people visit. Doesn't that defeat privacy?
While we support privacy organizations, we are not one of them. Instead, we focus on providing convenience, usability, and access to anonymously published content. In cases where there's a trade-off between enabling access and furthering privacy, we'll usually choose access. If you wish to use onion.link privately, there's software that provides that—the Tor Browser Bundle.
Q: How does OnionLink work?
OnionLink is a Tor2web proxy operating across a wide network of machines. Unlike most Tor2web proxies, OnionLink aggressively caches content for better user-experience. If an onionsite goes down, within a few days the site will disappear from OnionLink.
Q: Does OnionLink provide a list of onionsites that don't wish to be indexed by search engines?
We used to provide this. But it made onionsite operators upset. So we removed it. We'll bring it back if their temperament changes.
Q: Will OnionLink ever provide the list of onionsites it blocks?
Probably not, sorry. Publishing the list could constitute "aiding and abetting" users to access Bad Stuff™. However, when we do block a site or URL, it looks like this.
Q: What kind of privacy guarantees do I have when using OnionLink?
Very little. However, we do protect your IP address.
Bigger than you'd think. You can lowerbound the size of our index by passing site:onion.link into Google.
Q: Why doesn't OnionLink support https?
The specifics of our network infrastructure make https nontrivial to setup, however, https support will be added once funds allow for it.
Q: OnionLink is caching my onionsite's cookies!
OnionLink caches any and everything that doesn't explicitly request not to be cached. I.e., setting a cookie does not turn off caching. The established way of explicitly stating for your content not to be cached is the HTTP cache-control header. OnionLink obeys all aspects of the Cache-Control header, see the private and s-maxage parameters.
Q: OnionLink's cache takes too long to expire. How do I force a refresh?
You purge OnionLink's cache of a URL by passing the http command PURGE in a http request. This can be easily done with curl command: