I'm building a project with a friend and we're going FB auth only for the initial closed beta. How long do you think that's a reasonable strategy before we have to build our own login system?
Just my personal opinion, but I would suggest having something like: "Don't want to sign in with facebook? We understand, here's why we're requiring it while we build: {...}, enter your e-mail here [email input box] and we'll notify you as soon as other login options become available". This would serve to explain why you are requiring it (which might be enough to persuade some folks) and would allow those who are interested but not enough to link their FB to have a way to keep in the loop.<p>As far as how long you can get away without having your own login system, it really would depend who your target market is - and what kind of community you are trying to build. I would recommend keeping track of # of sign-ups per week / month, and adjust the priority of the "create login system" task based on how that number changes.
I've been using FB auth only for a year or so on one of my Android apps. It's centered specifically around social sharing and I'm focusing on US based users at first.<p>Given the maintenance tradeoffs around other login methods, I'm pretty happy with my choice. I'm sure that I lose groups of users that are only on Twitter, or Orkut or choose your SN du jour, but in the spirit of keeping things lean I'm happy with my choice.<p>One can always add other auth methods when growth looks like a hockey stick. 500MM+ FB users should be enough to get started.
I have a similar inner struggle with a product I am working on. I personally cannot manage more logins for anymore sites. I am constantly resetting passwords. I wouldn't be against using a single login solution but faceboos kind of gives random people personal information about you (if you opt in) so it might be a little scary. I wonder what the general population thinks?