If you have a web site that has about 60,000 users and maybe about 500 or so active daily, and you want those users to be able to sign into another site you own, you might consider doing single sign on.
Of course you can just copy the DB over, or write something that syncs across two databases so they share the same user table.
What do you think would be the acceptable loss? A few people with duplicate accounts? How many problems do you think you would encounter? If this was a business decision, what cost would you incur to build single sign on versus a hack?
These are all valid questions, but the biggest question that site owners do not ask themselves is this: Would the user give a shit?
Ask yourself why a user wants single sign on. Is it because they could share the same profile? Well, that doesnt seem to be the case for Open Social, Open ID or Facebook connect. So why does a user want single sign on?
I know why I want single sign on... no more accounts! I hate signing up for more accounts. Every site has a different requirement. 4 characters, 6 characters, characters and numbers, no real words, don't use your name, your initial, common names, etc. So when a site asks me to create an account, I bail. Yup, I bail. This is why single sign on is so important. It's to keep people like me from bailing.
Before you commit any capital or resource in creating another closed single sign on, ask yourself if your users would rather be able to use your universe of sites, or use their facebook, yahoo, google, aim, or open ID account to log in.
Don't worry, your ego will recover.
