Start with protocols for handling the data well. Then figure out what flows work. Start at the bottom with what already is, with what is concrete & true, and b

Start with protocols for handling the data well. Then figure out what flows work. Start at the bottom with what already is, with what is concrete & true, and build forward from that. I both want us to be careful & considerate, but also, I find the assaults against progress like this to be impossible. We can keep tattering efforts to advance forever, create ever more gates, raise ever more concerns about any topic forever. I used to be able. To process this kind of cross examination as a tempered concern & moderation, but increasingly I feel like grace is lost, that- whatever the intent- we are suffering from a large amassed anti-will, a will against. Having the technical standards & options to do these capabilities doesn't imply that every system has to offer them. Most should, & there should be good options for managing them (and the concern here is valid that the options indeed be good & usable) . But security keys have been a miserable failure, in huge part because users have never been given options to manage their contents, because the device has been the sole authority & holder. It's non negotiable in my mind that exporting importing files needs to happen ASAP, with low barriers. We keep roadblocking progress with concern that some users will make awful mistakes. If we had forever we could perhaps make your perfectly safe & secure utopia, where power users and the lowliest end user both happily safely operate, but we do not have forever. Given this, we ought create powerful technical underpinnings that make user capabilities possible & evolvable, then let product iterate & hone on how they can best marshal & expose subsets of those capabilities in ways catered to their users. What we cannot do is gate progress on imagined self-threatening users, what we cannot do is to forever pick the lowest common denominator of options, out of fear.