Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Not parent but...

> There is a difference between "track you and show you ads" and just "show you ads." A print newspaper can show you ads.

And a federation of networks can get together and decide to share what data they have about you with each other.

> Not really. If Google or Facebook has your data but your "friends" have access to it, your friends also have your data. They can copy it and keep it indefinitely. Removing the middle man doesn't reduce your own control, it only removes the middle man's control.

Could you clarify this? I'm really not sure what you mean, sorry.



> And a federation of networks can get together and decide to share what data they have about you with each other.

Federation means competition. You can choose providers you trust and who agree not to do that. Or you can run your own, which in a properly designed system shouldn't require more than some free software, a domain name and a machine to point it at. Which ordinary people might buy off the shelf as an appliance that plugs into your modem.

> Could you clarify this? I'm really not sure what you mean, sorry.

OK, so right now you have a funny cat video, so you upload it to YouTube and then post the link on Facebook. So Facebook can tell who you've shared the link with and who has clicked on it, YouTube can tell what you've uploaded and who has watched it, and anyone you've shared it with (or, in the absence of authentication, anyone in the world) can download it and save a copy on their hard drive.

In the alternative, you upload the funny cat video to some distributed CDN thing like Freenet or BitTorrent, and then your friends get the link in something that resembles an RSS reader that gets its feeds from something that distributes messages how email does.

So anyone you've shared the video with can still download it and keep it but there is nobody who has the complete list of everyone you've shared it with and there is no third party middle man who can keep track of everyone who is looking at it.

Whoever is doing the uploading can still tell who is downloading but presumably these are your friends who are not keeping permanent records rather than some corporation bent on converting every aspect of your personality into a data set for selling hotdogs and health insurance.


On the other hand, assuming you can't convince your friends to all choose the same provider, almost every provider will know everything you post.


You're assuming the providers are allowed to see what you post. They don't even need to see the destination. The sensible way to do message routing is: You encrypt a message to your provider which contains the destination's provider and a message encrypted to the destination's provider, which itself contains the destination and a message encrypted to the destination. If the providers aren't colluding they won't even be able to tell who is communicating.


Who is creating and maintaining the protocol with which the members of the federation speak to each other?


Protocols are created by the first implementer, as usual.

Should it ever grow big enough IETF, W3C, ECMA, ITU, ISO and the likes might be appropriate maintainers (depending on the details).

See SMTP, XMPP which are both federated.


Yeah, as troubling as Facebook is, are you literally suggesting the W3C could do a better job running a social network via a federated protocol than Facebook with a centralized one?

That SMTP and XMPP exist are just more proof for the value of Facebook (and to be clear: I hate FB and don't have an account, but I still see that it has value). SMTP has massive amounts of users, likely more than FB, and yet FB still exists because SMTP . XMPP is a good protocol, and yet no one is using it because federation makes it too hard for these things to work in the market.


IETF maintains the SMTP specs, yet their mail server operation is one tiny (set of) server(s) among many.

The idea with federated systems is that the standard body isn't _running_ the network. So no, I wouldn't expect W3C to run a social network, federated or not. Their own node, maybe.


That's kind of a matter of semantics, I wasn't sure how to say it properly. "Directing" a social network? I understand they aren't literally paying for the nodes, but defining the protocol gives you quite a lot of power and centralization.


It's a multi-stakeholder model. In the end a "social networking protocols working group" would primarily consist of delegates from all social network operators that implement the standards.


IETF? Mozilla? Who is creating and maintaining TCP/IP or HTTP or DNS?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: