3 ms·

While I agree with your general point, thinking about the passport numbers for a bit makes it plausible: - You would have to reserve some numbers that can't be

While I agree with your general point, thinking about the passport numbers for a bit makes it plausible: - You would have to reserve some numbers that can't be assigned to regular citizens. - If you make those "random" - you need to store them in a database that can be checked against when issuing numbers to regular citizens. - You wouldn't want such a database to exist. - If you remove a whole "block," it's much simpler (do not assign passport numbers starting with...). - Passports are NOT issued by the services themselves, and normal institutions (foreign office) only collaborate "covertly" — not everybody involved can be aware. Fake passports that pass muster (regular patterns of a country's passport are known to foreign border guards, like checknums, typical issuance ranges, etc.) are not simple.

jojobas 2y ago
The reasons to have a separate range don't seem to be compelling. I'd hazard a guess no other secret service ever did that kind of blunder.
rehevkor5 2y ago
If you want to keep the information secret, you need a way to store it. Assuming number generation must be coordinated centrally (to prevent collisions), that does not seem like an impossible task. A system that involves a cryptographic hash could divide the id space up into different segments without being dectpherable from the outside world unless the cryptography is broken. A periodic key rotation system could reduce the impact of a compromise. To me, the real question is why embed that information in the passport number at all? Even if driven by cryptography, it seems to represent a risk for which i don't understand the benefit.
cryptonector 2y ago
> - You would have to reserve some numbers that can't be assigned to regular citizens. Because why? > - If you make those "random" - you need to store them in a database that can be checked against when issuing numbers to regular citizens. You have to store passport numbers assignments in a database anyways. > - You wouldn't want such a database to exist. Because it could leak? But number block schemes leak. > [...] > Fake passports that pass muster (regular patterns of a country's passport are known to foreign border guards, like checknums, typical issuance ranges, etc.) are not simple. If you want a fake/second passport to look like a real/first passport then you need to make sure that there's nothing fake-y about it, like none of this reserved number block thing. If the country in question assigns passport numbers sequentially, then that's going to be a problem because the passport numbers will indicate when the passport was issued -- you wouldn't want that. So instead you'll need to either randomize passport number generation, pseudo-randomize it, or leave lots of gaps in sequential number assignments so there are always a few from each relevant time period available. Also, if you use the same number for national ID and passport, then you'll need to issue a second national ID to match the second passport. These things are not hard to figure out. The U.S. would never get this wrong because the State department has enough clue and could always reach out to NSA/CIA for clue if they needed it. Apparently the Russians are total amateurs at this stuff.

jojobas 2y ago

The reasons to have a separate range don't seem to be compelling. I'd hazard a guess no other secret service ever did that kind of blunder.

rehevkor5 2y ago

If you want to keep the information secret, you need a way to store it. Assuming number generation must be coordinated centrally (to prevent collisions), that does not seem like an impossible task. A system that involves a cryptographic hash could divide the id space up into different segments without being dectpherable from the outside world unless the cryptography is broken. A periodic key rotation system could reduce the impact of a compromise. To me, the real question is why embed that information in the passport number at all? Even if driven by cryptography, it seems to represent a risk for which i don't understand the benefit.

cryptonector 2y ago

> - You would have to reserve some numbers that can't be assigned to regular citizens. Because why? > - If you make those "random" - you need to store them in a database that can be checked against when issuing numbers to regular citizens. You have to store passport numbers assignments in a database anyways. > - You wouldn't want such a database to exist. Because it could leak? But number block schemes leak. > [...] > Fake passports that pass muster (regular patterns of a country's passport are known to foreign border guards, like checknums, typical issuance ranges, etc.) are not simple. If you want a fake/second passport to look like a real/first passport then you need to make sure that there's nothing fake-y about it, like none of this reserved number block thing. If the country in question assigns passport numbers sequentially, then that's going to be a problem because the passport numbers will indicate when the passport was issued -- you wouldn't want that. So instead you'll need to either randomize passport number generation, pseudo-randomize it, or leave lots of gaps in sequential number assignments so there are always a few from each relevant time period available. Also, if you use the same number for national ID and passport, then you'll need to issue a second national ID to match the second passport. These things are not hard to figure out. The U.S. would never get this wrong because the State department has enough clue and could always reach out to NSA/CIA for clue if they needed it. Apparently the Russians are total amateurs at this stuff.