Hi! Thanks for the TownHall, loved it!
I want to comment on the second episode: https://twitter.com/i/spaces/1yNGaYZnqmdGj.
There was a lot of critique of the Internet Identity, most of it well deserved.
As a person intimately familiar with the conception of the II service, I’d like to provide some historical background.
- As many developers learned the hard way, the II derives a unique, unpredictable principal for each domain. That was a conscious decision. We considered having user-picked profiles modeled after browser profiles (I wanted this design particularly badly). Still, we decided to go for maximum privacy by default. Some people love this feature, but I’m not one of them.
- The UX of the service is not great indeed. Unfortunately, we didn’t have time for UX studies at the time: we had only two weeks to implement the first version of the II service, from the conception of the protocol to the functioning service! The II team improved the UX substantially over the last year, but some of our early decisions still bind them.
- It’s hard to see with a naked eye, but we invested a lot of effort to make II not unique in any way: anyone can brew their identity provider with different privacy trade-offs. If you crave diversity and don’t like II — stop complaining and build your service! The only problem is that it’s impossible to transfer an existing II anchor to another provider, but there are plenty of new users to attract.
Finally, the recent DIP-20 proposal for subaccounts (DIP20 Community Proposal (PsychedelicDAO)) goes against the Web3 values you admire so much: the transparency and ability to link information freely. Allowing canisters to masquerade their identity based on who they talk to looks suspiciously similar to the Internet Identity design.