4 ms·

Why? What would be the point of that?

ryandrake 2y ago
The thread started out with: > Because of how effective this is for catching even fairly minor violations like failure to pay vehicle tax, number plate cloning is becoming pretty common (comparatively) in the UK. Evidence of number plate cloning would reduce revenue from fines. If the government wants to fine you for not paying your vehicle tax, evidence that someone cloned your plate would be exculpatory, so it's in the government's best interest to not collect that evidence in the first place.
- multjoy 2y ago
  That's literally not how government works. If they explicitly discarded data because it was exculpatory then not only would they not be able to fine you based on that data, they would have to refund every fine they issued while that discard rule was in place.
  - immibis 2y ago
    The police literally withhold exculpatory evidence all the time, as a matter of business as usual.
    - multjoy 2y ago
      I don't, so my sample set is distinctly broader than yours.
      - immibis 2y ago
        What?
kaibee 2y ago
It could happen entirely accidentally. "Oh, we hit for DEADBEEF at 11:00am in London, so ignore hits for DEADBEEF from anywhere else outside some Nx safety factor of the driving distance from that hit, because it has to be a false positive." or "Oh, we hit for DEADBEEF at 11:00am in London and again at 11:05am in London. Our cloud costs are high, so we're only going to store the most recent picture DEADBEEF, and only record the violation from 11:00am and 11:05am." I guess these examples sound kind of contrived to me, but how long do you retain this kinda data for? 30 days? 60 days? A year?
- multjoy 2y ago
  Entirely contrived. The point of the system is that you want all the hits everywhere, if your system is using any discard rules at all then it is unreliable for a prosecution.
  - AnthonyMouse 2y ago
    You're not thinking like a bureaucracy. Suppose you have a system of cameras operated by the various localities. The localities are all storing the data. Then there is a central system that can query all the local systems but doesn't itself store any of the data. The department issuing traffic citations requests an automated daily report showing every vehicle observed at multiple locations where the two sightings imply an average speed between 55MPH and 170MPH. Then the citations department takes the report and issues a citation to anyone where there is no route between those two points that allows the calculated average speed without exceeding the speed limit. The report isn't automatically generated if the average speed was calculated to be 490MPH because cars don't go that fast so issuing a citation would trivially allow it to be challenged as a data irregularity, and the report is requested by the department issuing speeding citations rather than the one (if any) investigating data irregularities. Now you get a ticket for doing an average speed of 87MPH along a route where the speed limit never exceeded 55MPH. You claim it's a data irregularity (cloned plate), but that's a possible speed, because the automated report didn't include the impossible speeds. There may have also been sightings that would have implied an average speed of 490MPH, but that report was never requested so it doesn't exist. The bureaucracy comes up with some excuse for why you can't run your own reports that don't already exist. You could request the raw data and then do the calculations yourself, but then you have to make a thousand requests to each individual locality, which is purposely too much trouble for most people to bother, and even then there is no guarantee that the person who cloned your plate was spotted in a location that would have produced an impossibly high speed.
    - multjoy 2y ago
      ><The department issuing traffic citations requests an automated daily report showing every vehicle observed at multiple locations where the two sightings imply an average speed between 55MPH and 170MPH. Then the citations department takes the report and issues a citation to anyone where there is no route between those two points that allows the calculated average speed without exceeding the speed limit. The problem with that is that isn't how average speed cameras work, and where speeds are ludicrously high then there is a human operator looking at it.

The thread started out with: > Because of how effective this is for catching even fairly minor violations like failure to pay vehicle tax, number plate cloning is becoming pretty common (comparatively) in the UK. Evidence of number plate cloning would reduce revenue from fines. If the government wants to fine you for not paying your vehicle tax, evidence that someone cloned your plate would be exculpatory, so it's in the government's best interest to not collect that evidence in the first place.

multjoy 2y ago
That's literally not how government works. If they explicitly discarded data because it was exculpatory then not only would they not be able to fine you based on that data, they would have to refund every fine they issued while that discard rule was in place.
- immibis 2y ago
  The police literally withhold exculpatory evidence all the time, as a matter of business as usual.
  - multjoy 2y ago
    I don't, so my sample set is distinctly broader than yours.
    - immibis 2y ago
      What?

kaibee 2y ago

It could happen entirely accidentally. "Oh, we hit for DEADBEEF at 11:00am in London, so ignore hits for DEADBEEF from anywhere else outside some Nx safety factor of the driving distance from that hit, because it has to be a false positive." or "Oh, we hit for DEADBEEF at 11:00am in London and again at 11:05am in London. Our cloud costs are high, so we're only going to store the most recent picture DEADBEEF, and only record the violation from 11:00am and 11:05am." I guess these examples sound kind of contrived to me, but how long do you retain this kinda data for? 30 days? 60 days? A year?

multjoy 2y ago
Entirely contrived. The point of the system is that you want all the hits everywhere, if your system is using any discard rules at all then it is unreliable for a prosecution.
- AnthonyMouse 2y ago
  You're not thinking like a bureaucracy. Suppose you have a system of cameras operated by the various localities. The localities are all storing the data. Then there is a central system that can query all the local systems but doesn't itself store any of the data. The department issuing traffic citations requests an automated daily report showing every vehicle observed at multiple locations where the two sightings imply an average speed between 55MPH and 170MPH. Then the citations department takes the report and issues a citation to anyone where there is no route between those two points that allows the calculated average speed without exceeding the speed limit. The report isn't automatically generated if the average speed was calculated to be 490MPH because cars don't go that fast so issuing a citation would trivially allow it to be challenged as a data irregularity, and the report is requested by the department issuing speeding citations rather than the one (if any) investigating data irregularities. Now you get a ticket for doing an average speed of 87MPH along a route where the speed limit never exceeded 55MPH. You claim it's a data irregularity (cloned plate), but that's a possible speed, because the automated report didn't include the impossible speeds. There may have also been sightings that would have implied an average speed of 490MPH, but that report was never requested so it doesn't exist. The bureaucracy comes up with some excuse for why you can't run your own reports that don't already exist. You could request the raw data and then do the calculations yourself, but then you have to make a thousand requests to each individual locality, which is purposely too much trouble for most people to bother, and even then there is no guarantee that the person who cloned your plate was spotted in a location that would have produced an impossibly high speed.
  - multjoy 2y ago
    ><The department issuing traffic citations requests an automated daily report showing every vehicle observed at multiple locations where the two sightings imply an average speed between 55MPH and 170MPH. Then the citations department takes the report and issues a citation to anyone where there is no route between those two points that allows the calculated average speed without exceeding the speed limit. The problem with that is that isn't how average speed cameras work, and where speeds are ludicrously high then there is a human operator looking at it.