3 ms·

The API/protocol is a good way of implementing the idea in a privacy-friendly way although I think everyone should have been clear that it is not as private as

The API/protocol is a good way of implementing the idea in a privacy-friendly way although I think everyone should have been clear that it is not as private as simply disabling Bluetooth or not carrying a phone. It is pretty benign. But being honest about that is important if for no other reason than to build trust. However, the apps built on top of the API have (at least in some instances) clearly fallen victim of scope creep - eg the UK app came with venue check-in functionality (which then suddenly became legally mandated and there was no separate check-in app so you were out of luck if your phone didn't support the API or you didn't want to enable Bluetooth) and a function which tells you how "risky" your postcode is which you can't switch off. There was even brief talk about it being extended to include "vaccine passport" type functionality. This was, sadly, predictable, and is why many people simply won't trust it. The other issue with the app-based approach is that it sadly appears to be pretty useless. At least the UK app appears to notify for far too many contacts, large numbers of which are false alarms of one sort or another (neighbours, people sat in different offices, etc). While using Bluetooth proximity was a good idea, I think it's fair to say at this point that in the real world it just hasn't worked.

fossuser 6y ago
I agree with you that relying on state/country cooperation to build the apps was part of the problem. I don’t agree that it’s useless, 15min of close proximity to a confirmed positive COVID patient within the timeframe is the best possible determination of exposure and people should quarantine if they get it. I could be persuaded that too many are notified if there’s decent evidence of it, but with the massive case load/spread in the UK right now - I’m more likely to suspect that amount of people exposed is massive.
- tomatocracy 6y ago
  I'm not sure current case numbers really tell you much either way - you could equally argue that it's evidence that the app didn't work of course. The issue in particular is that Bluetooth proximity isn't reliably able to distinguish between "next to but with a meaningful barrier in between" and "actually next to". In my office, for example, we have areas adjacent to one another separated by floor to ceiling glass walls. People working on one side of the wall don't work with people on the other side and are not allowed to access the other area. There are quite a few scenarios like this (apartment blocks with thin partition walls between apartments, sat in a traffic jam next to another car, etc.

fossuser 6y ago

I agree with you that relying on state/country cooperation to build the apps was part of the problem. I don’t agree that it’s useless, 15min of close proximity to a confirmed positive COVID patient within the timeframe is the best possible determination of exposure and people should quarantine if they get it. I could be persuaded that too many are notified if there’s decent evidence of it, but with the massive case load/spread in the UK right now - I’m more likely to suspect that amount of people exposed is massive.

tomatocracy 6y ago
I'm not sure current case numbers really tell you much either way - you could equally argue that it's evidence that the app didn't work of course. The issue in particular is that Bluetooth proximity isn't reliably able to distinguish between "next to but with a meaningful barrier in between" and "actually next to". In my office, for example, we have areas adjacent to one another separated by floor to ceiling glass walls. People working on one side of the wall don't work with people on the other side and are not allowed to access the other area. There are quite a few scenarios like this (apartment blocks with thin partition walls between apartments, sat in a traffic jam next to another car, etc.