3 ms·
I was an early adopter of Persona (my site is billed as example in the docs) and was similarly disappointed when Mozilla gave up on it. In retrospect, I think
by stickfigure 11y ago
I was an early adopter of Persona (my site is billed as example in the docs) and was similarly disappointed when Mozilla gave up on it.
In retrospect, I think that Mozilla made a mistake by centralizing the fallback identity provider. The fallback provider was just a temporary edifice to bootstrap the protocols; it didn't have to be run by Mozilla. Every website could have run a small stack which remained fully self-branded and would only verify email addresses for its own purposes.
I understand why Mozilla took the route they did - centralizing the fallback provider eased the RP implementation, made it possible to rapidly rev the protocols, and in theory made it more convenient for users since one Persona password would work across multiple sites. In practice, however, users were confused about the extra branding and the vague sense of logging into Persona so you can log into a site. More critically, it made the whole project depend on the whims of Mozilla - it's not just software we depend on, but infrastructure, and without Mozilla's support the infrastructure will eventually die.
If Persona is revived, I hope it becomes a complete software stack that every RP can run independently. The fallback IdP should be 100% branded by the RP so users are never confused about what site they are logging into. And as non-hosted software, it should be able to live and evolve as open source software without fear that some tepidly supported server will go down (or simply fail to evolve). IMHO, this is the only way that Persona (or at least the auth standard, which is what we care about) can survive long-term as a community project.