3 ms·
"I think there's a huge opportunity for a startup that operates a social network that actually caters to people's desire for anonymity and privacy." The proble
by sixtofour 15y ago
"I think there's a huge opportunity for a startup that operates a social network that actually caters to people's desire for anonymity and privacy."
The problem here is that any successful endeavor will inevitably turn evil in some way. Once millions and billions are on the table, investors and shareholders will force the proprietors to scrape every bit of money possible off the table. That will usually involve some user-unfriendly behavior, like Google's real name, Facebook's ever changing privacy policy, and LinkedIn's recent emulation of FB.
The only way that social networks can be pro-user is if they are a standards-based commodity like email, where anyone is free to implement the standard and offer the service. You can get the same email service anywhere you want, interoperable with every standards based server on the planet. It should be the same with twitter-like message bursts, and social networks.
As it is now, bursting and socializing is built similar to early telephone networks, where many companies offered networks, they didn't connect to each other, and businesses had to have multiple network accounts to be able to interact with all their customers. That's just wrong.
fastmail.fm offers excellent email service. They should also (be able to) offer standards based bursting and socializing; so should anyone else who wants to.
- politician 15y ago"The only way that social networks can be pro-user is if they are a standards-based commodity like email, where anyone is free to implement the standard and offer the service." This already exists for social data, it's called The Locker Project. http://lockerproject.org/#vision http://lockerproject.org/#vision
- sixtofour 15y agoThis is great, a great start. But this is an implementation, not a standard. I think it remains to be seen whether github repos give us anything like the long term stability and stewardship of RFCs. Not to mention the longevity of github. Yeah I know, you can fork. Everyone can fork.