The verify approach with two lists side by side
When somebody hands me an address, I do not verify it against one list. I open two lists, side by side, and I compare.
The three addresses
Nexus Market publishes these three onion addresses. Present as supplied, in no order, with no ranking. This site does not probe them and shows no uptime figure.
nexusb2l7fmqnefwphyy7m5zjhlkytlbo7qbb5lu5dlczr3azgii2gyd.onion
nexusma2iqgauqqvjcgds4ckv5xbf272tkfagq4epojjhsgleqpwxiqd.onion
nexusabcd6tyfhdwilyitaqiri6tisj2v2hueyjuj6qkvd6azvi5tuqd.onion
The premise of the verify approach
The verify approach starts from the position that a single source is one opinion. That is true for a friend, for a printed note, for this site, and for any other list on the internet. Any of them can be right, and any of them can be wrong.
A second source that agrees is not proof either. Two lists can be wrong in the same way. But two independent lists agreeing is a lot more evidence than one list, and the cost of opening the second list is small.
The verify approach is the deliberate paying of that small cost, every time.
What counts as independent
Two tabs of the same site are not independent. Two mirrors of the same directory are not independent. Two forum posts by the same person are not independent. A chat message from a friend and a site the friend also runs are not independent.
Two sources are independent if there is no way for one to have gotten the addresses from the other. That is a judgement call, not a formula, and I make the call generously in the direction of doubt.
This site is one source. When I use the verify approach, I pair it with something else that is not derived from this site.
The mechanics, plainly
I open Tor Browser. I open this site in one tab. I open the second source in another tab. I lay them out on the screen so I can see both address lists at once.
I compare one address at a time. I read the first character of the first address on each list and check they match. I read the last character of the same address on each list and check they match. Then I sample four characters from the middle. If any of them disagrees, I stop and treat the whole comparison as a disagreement.
If all three addresses on the two lists agree, I copy an address from either list and paste it into a fresh tab. If any disagree, I do not use either list today.
What a disagreement looks like
A disagreement can be a full address that is different, or one character that is different, or one list carrying an address that the other list does not have at all. Any of the three is a disagreement.
One character different is a bigger deal than most readers give it credit for, because a one character difference produces a completely different onion. The v3 onion scheme is designed to make a one character difference load a different service, or nothing at all. There is no such thing as a small typo in a onion address.
When I see a disagreement I close the browser and try again another day, with a different second source.
When two sources agree
I trust the address enough to paste it and open the front page. I do not trust it enough to sign in on the first visit that produced the agreement. Agreement lets me look; it does not let me commit.
If a second visit later on also finds agreement between the same or new sources, I trust the address enough to sign in.
The step of pausing between look and sign in is one I keep on every visit, but it is especially important on the visit that established the verify.
What I do when the sources are asymmetric
Sometimes one list has three addresses and another has two. That is not a disagreement on the two they share, but it is not a full agreement either. I treat the two shared addresses as verified. I treat the address only one list has as unverified. I do not use it today.
Some markets publish more addresses than any single directory tracks. That is fine. It just means that two directories that overlap on some addresses and diverge on others are both being honest.
A short list of what makes this approach honest
- The second source is not a copy of the first.
- The comparison is character by character, not eyeballing lengths.
- A single character disagreement stops the whole visit.
- Verifying is a look. Signing in comes later, on a separate visit.
What the verify approach does not do
It does not verify that the market is operational. Two lists agreeing on an address only means the address is what both lists say it is. Whether the address currently loads a working market is a separate question that this article does not answer.
It does not verify that anything on the market is safe or legitimate. That is not what an address list is for.
It verifies one thing only. That is more than most people ever do, and it is worth doing.
The impatient approach that skips the second check, the same visit without the second source
What this article is not. This is not a claim that two lists agreeing is proof of anything. It is a claim that two lists agreeing is enough for me to act, and one is not.