I really, really wish an Open Source alternative to Facebook and Twitter would happen so we can put the proprietary social era behind us.
Define an open standard for a message; a photo-share; tagging; etc etc. Include stuff that clients and servers MUST do, or MUST NOT do. Write a nice server, and host it. Write a nice client (Browser extensions? Website?)
Now you have an open ecosystem that is as flexible as people want it to be. There's a reason that all discussion sites end up being a worse implementation of Usenet.
Exactly. The problem isn't that we don't have useful social networks. The problem is that we have too many of them, with varying heights of wall between their respective gardens that make interoperability and data migration absurdly difficult in many cases.
I suspect the roadblock is that these networks are tied together by the underlying concepts of identities and relationships, but for many valid reasons a lot of people would like to present different identities in different contexts or treat some relationships differently to others. Until there is some reasonably standardised approach (or set of compatible approaches) to representing these ideas, there's no robust foundation so you can build different kinds of messaging and content sharings systems on top.
That in turn leads to practical issues of authentication and trust networks, and we haven't completely solved those problems in a way that Just Works for non-technical users yet.
Facebook lets you talk to your old friends, twitter lets you meet new ones, and google plus is there if you want to get serious. Tumblr for a more expressive context, and wordpress for the rest.
The problem with creating a new social network is that you'd have to replicate a dozen companies with varying degrees of porosity in their services.
Add to that my second point: we don't know what we want yet. Things are still evolving and developing, in a high pressure cauldron of talent and money. Once things cool down and take shape we'll see what we like and someone will spend time making open source plugins. It's a really fun time to play on computers.
Not quite, because everything is so heterogeneous that there's no unifying, consolidated view to help me use it.
I want a single private, secure, easy to use platform to look up my friends and colleagues on-line, and on top of that I want to have convenient plug-and-play functionality for everything from IM to sharing photos to collaborating on a coding project to scheduling a business meeting.
Right now, we have the Internet infrastructure sitting a level below all of this, and numerous web sites (not to mention e-mail, IM tools, Usenet, distributed version control systems, etc.) trying to be both the platform and one or more of the plug-ins. This is not what I want as a user, i.e., as a person who wants to get stuff done and doesn't really care whose badge is in the corner of the screen as long as it all works properly.
Because at the end of the day user experience is order of magnitudes more important than some arbitrary geek obsession with being open. Which isn't important for most people since they are used to having their personal information owned by third parties e.g. banks, employers, insurance companies, government.
You hit a really important point here: 3rd parties have held personal information since the dawn of civilization: messengers, small-town post offices, the US Mail, Ma Bell, stores that tracked purchases against your line of credit in a ledger, etc.
A computer using keywords in my emails to display relevant advertising is orders of magnitude less scary than a town gossip who runs the post office and tells my extended family and neighbors anytime I receive a letter that might indicate I'm deviating from social norms or religious morals.
The code running Amazon doesn't judge people for buying erotica; the town bookseller does. The card catalog won't out you to your friends for looking up LGBT reading material; the librarian might. The code running Google Maps doesn't care and won't tell anyone that you're driving to a woman's house and then a restaurant or bar most Saturdays; the waitress/bartender will almost certainly tell a friend/stalker/jilted lover/cop who you were with.
Big Data is far less threatening than Little Data.
I don't get to commit murder even if I feel it's justified; Amazon doesn't get to refuse to comply with a court order even if it feels it's unjustified. Nobody, individual or corporation, is going to sacrifice their well being or sit in jail indefinitely for contempt of court to protect the privacy of your purchase history.
The only people who really pull that stunt are journalists.
The threshold for obtaining a subpoena is remarkably lower than that. NSLs -- I wish I could tell you, but the whole program is classified.
The EFF reports that mobile providers received 8 million law enforcement requests for GPS data: https://www.eff.org/deeplinks/2009/12/surveillance-shocker-s...
And ... due process like subpoenas and court orders? Not so much:
AT&T said it now responds to more than 700 request a day, about a third of which do not require court orders or subpoenas.
http://news.cnet.com/8301-1035_3-57468316-94/cell-carriers-s...
I'm all for rule of law. This isn't.
Government is somewhat closed, but it also has an army and isn't going anywhere soon. Facebook is not trustworthy as a single source of truth.
Switching banks is inconvenient too, especially if you have outstanding loans.
If you have a Facebook group where personally identifying participants isn't desirable (i.e. you would want non-FB users participating) than you're probably looking for a wiki or Google Sites instead. Facebook has every right not to be a web publishing tool.
And my point still stands that the overwhelming majority of people are comfortable with their most private data being in the hands of third parties. And Facebook has long reached that point where people can trust it for managing their semi-private data.
Personally, I think the future of "social networks" will, in a sense, go back in time once again where it will be peer to peer, encrypted end to end.
I think the guys working on the Freedom Box are trying to crack the riddle.
Except, people like to use Facebook as a broadcast medium, and the message size is likely to get prohibitively large if you're encrypting to a lot of people.
Do you have any evidence of this strange claim?
Encrypting a message to lots of recipients in an efficient way has been solved for email (via OpenPGP / GPG) a long time ago.
It is a positive side effect of the hybrid algorithms which combine symmetric with asymmetric algorithms, originally introduced to reduce computation time for classic (single-recipient) asymmetric cryptography.
I guess I was thinking more along the lines of the actual network it sits on will help, too. Like CJDNS, an added protective layer for privacy, at least.
Public key encryption generated automatically for each user using the freedom box (I think I watched a video of someone explaining their concept of this for the freedom box).
In the end, I don't know. I'm only starting to get involved.
Actually, to make that work I think you'd need to add three things:
1. An extension to the vCard format to represent relationships (maybe this already exists? can you embed RDF triplets in vCard?)
2. A plugin for the popular mail servers that displays vCard info of its users as html in response to web requests (with appropriate RDFa metadata).
3. Some way to make an http request that proves "I am johndoe@example.com" or "I am the mail server responsible for johndoe@example.com" so that servers could make a distinction between private data, public data, and "friends only" data. Maybe have a chain of trust that starts with DNSSEC keys?
There is FOAF ("friend-of-a-friend", based on RDF). At the moment, this is mostly provided by websites (such as identi.ca), but could also be transferred via E-Mail.
There might be some way to embed FOAF into vCard, but why should be care? E-Mail allows for more than one attachment ...
The nut to crack is to allow a user to transfer the network effects he currently has on Facebook to a new network.
Obviously the problem of porting the data is 100% up to Facebook to allow, which I strongly doubt it will ever do (what is its commercial interest in allowing that?)
I didn't say that many open source implementations exists, just that many implementations exists. Diaspora might be a good candidate for a software stack, but the point is that having some social networking software is utterly pointless if you don't have a way of getting users onto it.
The much derided Diaspora (sign up at http://diasp.org) works perfectly well for Facebook's core functionality, and is fairly pleasant to use.
Of course, the real problem with that would be paying for it. Hardware and bandwidth isn't free and it seems like most people are unwilling to pay anything for social networking.
Why don't people just use preexisting technologies like email? What's the obsession with social networks anyway? Let's rethink this and come up with something radically different.
There is a alternative, all inclusive solution right under our noses...
I agree, there should be something created completely different. However, I don't think that can really be achieved. I think it can be made to look different but in the end it will just be a different implementation of the same thing.
I think the next big thing won't be necessarily social networking. Social is basically an expected feature of all new software. There is nothing new to be had there. People like to share things and talk with friends, meet new people, etc. That won't change.
The appearance will change and the technology that becomes big will have all of that stuff in it. I just hope it's free software and disconnected from a centralized source.
Personally, I think the next big thing will be accidentally created by the "the people" and not some corporation.
I'm sorry, I didn't understand that part. Is that another language?