I can kind of understand the vilification of this feature, and the reasons for that are explained in other comments on this post, but I think it's a major improvement to how it would have been before, and on other current platforms as well. Unlike other ways to keep in touch with friends, like WhatsApp, WeChat, Kakao, Facebook isn't just a messaging platform (although that's how I primarily use it) and so it's actually possible to see someone marked as safe without having to directly message them about it. This is a direct improvement over other platforms where the only way you know if someone is safe is to actually message them.
I don't really buy the argument that you would just assume someone is safe before; you would absolutely have thw worry that if there was a disaster in London, you would want to know if your friend is safe. Previously you couldn't easily contact them, though: if you even had their phone number, calling them internationally wouldn't be easy or sometimes even possible, but often you'd have their address, and hope they would respond to you. Now it's much easier to keep in touch.
I also feel like people are making controversy over nothing when they think that asking if they're safe when they're in London during that fire is too much if they're not in the vicinity. Facebook is in a catch-22 here; Facebook either knows your (roughly) exact location and knows if you were in or near the apartment building at the time, which would make people cry about Facebook tracking you everywhere, or it doesn't and it asks if you're safe if you're in London. Even in the image from the tweet that this article references there is a "Not in the area" button you can press. There's really no way to correctly do this without having really accurate and very up to date information about the people using Facebook, which isn't always possible.
Could Facebook improve the ways it determines if a user is in the area? Yes, of course; a simple way would be to look up IP address block(s) and see if the user is in a block they look up, then prompt them, although it's not really that simple. I also run into issues with Facebook thinking I'm in Japan when I'm not, even though I left nearly a month ago. Facebook really could also improve the UI around it; the point at the bottom of the article when it says that the writer has 100 (probably) London based friends, 97 of which are not "marked as safe", which is terrible UI. But I do absolutely disagree that this feature is worth removing based on the arguments presented in this article.
> I don't really buy the argument that you would just assume someone is safe before
The problem is that it doesn't matter how you react to the feature. The problem is how groups of people react to it.
This isn't much different than saying you'd calmly walk to the exit if caught in a fire in a club. Even assuming you evaluate your own behavior in that situation correctly, it has zero effect on everyone else.
> There's really no way to correctly do this without having really accurate and very up to date information about the people using Facebook, which isn't always possible.
So maybe if you can't do something reliably, and there are risks to getting it wrong, the proper response isn't 'hold my beer?'
Let me put it another way: I don't understand your argument, so I clarified and asked if your argument was X. You said it's not, so what is your argument?
The argument is that, based on the news reports, the system breaks often enough to risk causing more panic and alarm than it solves, so instead of doubling down on a bad idea, perhaps the right course of action is to reconsider the feature.
You seemed to want my argument to be some absurdly rigid form of "ship only on perfect performance". It isn't - it is more that treating misfires and the predictable, if irrational, responses something like this will generate in actual emergencies as if they were as trivial as, say, bad ad targeting is irresponsible and should be given substantially more thought.
I don't really buy the argument that you would just assume someone is safe before; you would absolutely have thw worry that if there was a disaster in London, you would want to know if your friend is safe. Previously you couldn't easily contact them, though: if you even had their phone number, calling them internationally wouldn't be easy or sometimes even possible, but often you'd have their address, and hope they would respond to you. Now it's much easier to keep in touch.
I also feel like people are making controversy over nothing when they think that asking if they're safe when they're in London during that fire is too much if they're not in the vicinity. Facebook is in a catch-22 here; Facebook either knows your (roughly) exact location and knows if you were in or near the apartment building at the time, which would make people cry about Facebook tracking you everywhere, or it doesn't and it asks if you're safe if you're in London. Even in the image from the tweet that this article references there is a "Not in the area" button you can press. There's really no way to correctly do this without having really accurate and very up to date information about the people using Facebook, which isn't always possible.
Could Facebook improve the ways it determines if a user is in the area? Yes, of course; a simple way would be to look up IP address block(s) and see if the user is in a block they look up, then prompt them, although it's not really that simple. I also run into issues with Facebook thinking I'm in Japan when I'm not, even though I left nearly a month ago. Facebook really could also improve the UI around it; the point at the bottom of the article when it says that the writer has 100 (probably) London based friends, 97 of which are not "marked as safe", which is terrible UI. But I do absolutely disagree that this feature is worth removing based on the arguments presented in this article.