> If they wish to use Apple then that is their own choice, but on Android it's quite trivial to download Red Hat's open source authenticator app[1] from f-droid

> If they wish to use Apple then that is their own choice, but on Android it's quite trivial to download Red Hat's open source authenticator app[1] from f-droid (the website, you don't even have to install the store if you don't want that). That's even less intuitive than using the default app store. That's a whole new slew of concepts you need to grok (you can download an app from the web and install it without an app store; what is this fdroid thing? is this a virus? what do these dialogs mean?), plus training people to do this without also giving them the knowledge when and why this is safe isn't exactly helpful, but that's a lot to ask from a simple sign-up flow for a hypothetical niche app built by a hypothetical two-person team. > And if people don't have a phone with support for apps, then you can still fall back to SMS. You can, but that adds to the complexity and support burden and probably also costs you users due to sign up friction. > Fun fact: my grandpa can't use SMS either, your solution is not as universal as you make it seem. He never has been able to due to sight issues (it's not an age thing, though it doesn't help if you're close to illiterate and now need to start to learn how to use solutions for sight-impaired people due to this information age having onset). That's an interesting case. I'd like to think there would be a fallback for people like him, but I guess, in the vast majority of cases, he'd just be left out. The current state of inclusivity in tech is abysmal, though I've seen vision-impaired and deaf people get around their devices surprisingly well; it's still an embarrassment that this industry won't do better. It's hard to get this right when it should be hard to break this, but current frameworks and paradigms don't prioritize this. It's shameful IMO. I do think SMS is a lot more accessible than authenticator apps and the like, even though that still will not work for everyone. > Can't we have the better solution as well as the accessible one? I'm not saying you can't or shouldn't offer the best solution you can. By all means give me Yubikey support and several fallbacks. But I can see quite well how not everyone might want or be able to. > No no, you got that backwards. All Facebook needs is your phone number, or whoever it is that pinky promises to only use your phone number for security. Facebook definitely should get rid of SMS factors. If anyone has the resources to do much better, it would be them and the other giants. Not sure how they handle that, though. They'd still collect phone numbers in any case, but they'd happily image people's internal organs if they could, so that is a separate issue. Apart from that people seem to be quite happy to use their phone number for signup if it makes signup quicker and less annoying. Even if the primary flow is email and alternatives are hidden in another tab, phone number still tends to get used a lot, in my limited experience. Same with Facebook/Google login. Plus, for most people ai guess it isn't as black and white; in quite a few cases I've given my phone number even if signing up via email, because it helps a lot if people can just call me in case of issues (e.g. the restaurant is out of my extra topping). That said, I'm all for offering as much choice as possible, and I'm not happy with the inflationary use of phone numbers as the only way to sign up, and I'm all for Yubikey support in every app, and it's disappointing that OS/browser vendors don't make this easier and more convenient, and if anyone wants to let me have as many anonymous phone numbers as I need, I'm very interested. But, still, I can totally see why a resource-strapped product/dev team might come to the conclusion that SMS second factors are sufficient for now.