Andreas Gal and I were among those calling for Persona to be integrated into Firefox ASAP, for scaling leverage against Metcalfe's Law. We had frustrating, protracted arguments about it with Ben Adida. I found resistance to the idea to be based on ill-concealed fear and loathing of dealing with the Firefox codebase, and (possibly as a consequence, not cause) explicit preference for doing a JS "shim" library and promoting it to web developers in competition with FBConnect.
That worked about as well as you would expect.
Eventually, Mark Mayo got Firefox Accounts going, but it was non-federated. In truth so was Persona: Mozilla ran the only IdP of note. Also, prior to Accounts, the protocol seemed to fork in anti-federated ways, but to me that was just teething pain, to be overcome by further evolution.
The fatal problems were threefold:
1. Facebook had huge scale and even in 2011 (browserid days) it had already won.
2. The Persona team was averse to integrating into Firefox, for whatever client population "interop readiness" pressure that might have put on servers (Metcalfe's Law is a barrier to new protocol adoption).
3. Users don't grok federated identity. Relying party? (That's the first party, the site to which you're browsing with clear intent and understanding of its identity -- assuming you haven't been phished.) Identity provider? (What's this sketchy popup I get every week or so asking me to re-login to some third party?) The whole federated Rp/Idp/browser three-body problem is confusing and looks like some kind of hack, not just phishing but popup malware.
The initial centralized or under-federated situation to me was not fatal, but could have become so if problems 1-3 didn't doom the whole effort.
Firefox OS indeed suffered from slow and half-hearted commitment from July 2011 on. Not even half-hearted: at first, it was a pirate ship. The CEO told another exec that in previous jobs, someone would have been fired for launching it via a post to mozilla.dev.platform (even though drafts of that post had been discussed and vetted by all execs who were paying attention).
Don't get me wrong, even with aggressive resourcing from mid-2011, Firefox OS might not have made it. But half-hearted, slow-rolled "investment" was worse than either "do" or "do not". No half measures, as Mike in "Breaking Bad" taught.
None of my employees called for me to be fired. You're confusing six Mozilla Foundation employees with people who worked for me in the (arm's length, for profit subsidiary) Mozilla Corporation. See http://arstechnica.com/business/2014/03/mozilla-employees-to... (which, typical of coverage at the time, fails to note those employees worked for an entirely separate org from the one I was CEO of).
That worked about as well as you would expect.
Eventually, Mark Mayo got Firefox Accounts going, but it was non-federated. In truth so was Persona: Mozilla ran the only IdP of note. Also, prior to Accounts, the protocol seemed to fork in anti-federated ways, but to me that was just teething pain, to be overcome by further evolution.
The fatal problems were threefold:
1. Facebook had huge scale and even in 2011 (browserid days) it had already won.
2. The Persona team was averse to integrating into Firefox, for whatever client population "interop readiness" pressure that might have put on servers (Metcalfe's Law is a barrier to new protocol adoption).
3. Users don't grok federated identity. Relying party? (That's the first party, the site to which you're browsing with clear intent and understanding of its identity -- assuming you haven't been phished.) Identity provider? (What's this sketchy popup I get every week or so asking me to re-login to some third party?) The whole federated Rp/Idp/browser three-body problem is confusing and looks like some kind of hack, not just phishing but popup malware.
The initial centralized or under-federated situation to me was not fatal, but could have become so if problems 1-3 didn't doom the whole effort.
Firefox OS indeed suffered from slow and half-hearted commitment from July 2011 on. Not even half-hearted: at first, it was a pirate ship. The CEO told another exec that in previous jobs, someone would have been fired for launching it via a post to mozilla.dev.platform (even though drafts of that post had been discussed and vetted by all execs who were paying attention).
Don't get me wrong, even with aggressive resourcing from mid-2011, Firefox OS might not have made it. But half-hearted, slow-rolled "investment" was worse than either "do" or "do not". No half measures, as Mike in "Breaking Bad" taught.
None of my employees called for me to be fired. You're confusing six Mozilla Foundation employees with people who worked for me in the (arm's length, for profit subsidiary) Mozilla Corporation. See http://arstechnica.com/business/2014/03/mozilla-employees-to... (which, typical of coverage at the time, fails to note those employees worked for an entirely separate org from the one I was CEO of).