This a project designed to help people find alternatives to walled garden social media sites, and to delete their accounts and move off such sites, in protest at their ongoing disrespect of user rights, and their refusal to interoperate with other similar sites by implementing open standards. Such sites use people's friends to suck them in, and them invade their privacy (FaeceBook) or decide to charge for their service out of the blue (want Ning? give Bling). The idea for Fyre Exyt came out of Quit FaceBook Day, which Strypey joined. After deleting the profile under his own name, he figured it would still be good to have a subversive presence within FarceBook, so he decided to set up a new profile under the name Fyre Exyt, with the express purpose of continuing to encourage people to leave FaeceBook and agitate against walled garden sites in general. For criticisms of walled gardens check out the blog post Transition To Web Free.
General update (2017-04-30)
Strypey completed a blog post giving a brief history of the GNU Social branch of the family of federated social networking apps. More blog posts are planned on the history of the Diaspora and Friendica/ Hubzilla branch of the family, and on the road towards consensus on standards. At some point all this research might be gathered into a detailed paper on the history, present, and potential future of federation. More projects relevant to the contents of this page include the resurgence of GNU Social, Hubzilla (mentioned here as Red Matrix), Mastodon (another OStatus client like GNU Social), Mumble (like IRC plus voice chat), Matrix (protocol), Riot (chat client for Matrix), RocketChat, MatterMost, and the demise of Tent and Lorea.
Another page has been in the works on Strypey's decision to delete his Google account,
Leaving the Googleverse
. The information on software that can be self-hosted from that page and this one is being compiled in a page on Free Code Self-Hosting Packages on the P2P Foundation wiki. Information on standards and protocols from this page will be added to the relevant P2PF page, or a new page there, ASAP.
General Update (03/04/2016)
There seems to be a new lease of life in the federated social network space and I'm working on a blog post outlining some of the developments over the last few years that might account for that. In the post-Snowden era, there is growing interest in new social network platforms. Since the last update another wave of new projects have popped up including text-based microblogging sites like Twister, Minds, Ello, and Synereo, and voice/ video chat projects including Unseen.is, Ring, Signal, Tox, and a plethora of WebRTC-based sites (eg Palava.tv and meet.jit.si).
The more I look into the whole "social media" thing, the more difficult the whole thing become to define. Obviously it's more than just 'software that moves blocks of text from user A to user B' because that describes almost everything on the net (except for voice and video chat, and music and video sharing sites). There are three functions that are often included in lists of social media platforms (this is a first attempt at clarifying all this, and will need more work, comments welcome):
user-generated content hosts
- blogs (eg Blogger, WordPress)
- forums (eg Reddit, PHPBB, Discourse)
- microblogging (eg Twitter, GNU Social)
- photosharing (eg Flickr, InstaGram, SnapChat)
- events promotion (eg Google Calendar, FaceBook Event pages)
- video sharing (eg YouTube, Vimeo, EngageMedia)
- public like newspapers and televisions stations, and persistent like the newspaper archives in libraries
- many-to-many, like the web and email
- based on publication by individuals to a self-selected swarm, none of whom may be known to the publisher
- based on peer-to-peer networks managed by individuals users (users choose who to 'follow' or 'friend' or 'subscribe to' and which of their posts to read and reply to/ comment on)
- chat (eg IRC)
- private, like telephone calls and letters, and only persistent if participants keep/ record the exchange
- one-to-one or many-to-many (emails, chat, IRC)
- based on messages written or spoken by individuals and addressed specifically to another person or group of people
- decision-making (eg LiquidFeedback, Loomio)
- messaging (eg Apache Wave, Slack, MatterMost, CrabGrass, Rocketchat)
- video conference (eg Skype, Hangouts, Jitsi Meet, Palava.tv)
- code repository (eg SourceForge, GITHub, GNU Svannah, GITLabs)
- private, like conference calls or internal memos
- one-to-many or many-to-many (although one-to-one private messages may also be supported)
- based on group space into which group members can publish messages
Both proprietary and federated versions of services in all three categories usually involve a server on which the user sets up an account via a web portal. Theoretically, it's possible to implement any of these as a P2P system (eg Twister). The challenges to this in practice include:
- the need to be always-on and always-connected, so messages and other content can successfully transfer from user to user
- IPv4 work-arounds like NAT (see The Digital Imprimateur)
- the asymmetry between the low-upload internet connections offered to ordinary customers, and the high-upload internet connections offered to hosting companies
- the battery drain, network access challenges, and bandwidth costs of operating P2P software on mobile devices, particularly tablets and "smartphones", but already a problem for users whose primary computing device is a laptop
- the difficulties in funding development when there is no central portal on which to sell advertising or seek donations (except the website for information about and downloads of the P2P client), and really no business model in the traditional sense
General Update (18/082013)
Time for another great escape, this time from Google. Our Google have been undergoing rapid Facebook-ization, abandoning many of their services (eg Wave, Buzz, Reader, and now Voice and XMPP chat), and integrating what's left into their G+ microblogging engine. The same issues apply, but sadly, a lot of the alternatives I listed when I left FB were Google projects, so they're now off the menu. What I'm really keen to explore is what social software is available which works peer-to-peer, relying on no third party whether corporate or not-for-profit.
I realised today that the Core-Us page which documents resources relating to realtime communications (instant messages with text, voice, and video), is an example of where this research project needs to go. I'm going to not just update this page - which has become a lengthy mess, despite the organised way it began - but make it an entire subwiki, with a page for each of the topic headings here.
Projects I've discovered recently are the FreeYourSpeech blog, which mainly discusses WebRTC and DruCall, a P2P social network client called Nightweb with both desktop and handheld versions, WASTE again, a fork of the defunct WASTE, another FreeNet style distributed BBS style system, the new BitMessage P2P email system which evades metadata collection, and CyptoCat, a browser-based chat client providing end-to-end encryption. Also as of June with year, StatusNet, GNU Social, and FreeSocial merged under name GNU Social, while Evan .
General Update (30/08/2012)
At some point I must rewrite this page to bring it up-to-date. A lot has changed in the ad-hoc evolution of the open social web in the last couple of years. A number of projects which looked promising have fallen over (eg Wave, Buzz, Social Graph API), or turned out to be underwhelming (Identi.ca, Diaspora), . New projects, services, protocols and organisations have emerged (SafeBook. RetroShare, Google+, WebFinger, PortableContacts). Some new projects have emerged *and* fallen over (eg GNU Social, DiSo, MugShot, Mozilla Contacts, PlaxoPulse). The social networking battlefield is now littered with corpses, but many may be worth searching for weapons and ammunition (architecture, protocol implementations, and source code) to carry on the fight for open federation, and semantic tagging.
In 2009, articles by Andy Oram, and Chris Messina and Jyri Engeström, discussed some of the engineering problems with decentralised and distributed networks in a real social situation (one with a full range of internet-capable humans, not just idealistic geeks). There have been huge leaps forward in the development of the open web protocols that we actually need to allow a federation of independent social hosts to share data with each other in a way which is secure, yet user-friendly. Building on the experimental work of the Free Protocols Foundation, with their Libre Services Manifesto, the Open Web Foundation has emerged as an incubator for emerging web protocols, on their way to maybe becoming standards (if they take off). A number of new protocols have been created (eg PubSubHubbub, rssCloud, ActivityStreams, WebRTC for voice and video). Some existing ones have gained in popularity (eg OpenID) while others have languished in obscurity (eg Mozilla Persona/BrowserID).
There have also been huge leaps forward have been made with web the emergence of semi-professional web tv sites like Blip TV, ongoing development of video-hosting projects like Plumi, and P2P video-streaming with BitTorrent Live. Flash Player spyware, following the "smartphone" trend, has started hijacking users' cameras and microphones, with a trip to Adobe's site required to switch the spying off (for each browser, on each computer), meanwhile Adobe has stopped developing its player for GNU/Linux. Gnash is a vaguely-usable alternative in the short term, but I can't wait for HTML5 video tags, and open audio-video formats (eg WebM) to replace flash for web video. No idea what kind of standards could replace ShockWave games and animations, but no doubt they're being cooked up as you read this.
Update 10/08/201: Discoveries from the Appropedia IRC chat this morning. Patrick of Mutual Gift wrote a security comparison of various social media platforms. There is a comic on distributed social networks. You can Try Friendica here. Friendica's Mike Macgirvin is proposing a new system called Red Matrix, which uses the Zot Protocol. SecuShare is built on top of GNUNet, and aims to be user-friendly.
Update: 23/10/2012: Rysiek restated most of the assumptions here with less detail in his post Breaking the Garden Walls.
Update: 29/04/2013: Lots of exciting new stuff is happening. Identi.ca is changing out its StatusNet engine for the new pump.io. Thanks to the discussion thread on that announcement, I learned about Tent.io, another federated social media platform. Also MediaGoblin, a GNU Project sponsored video CMS which seems to be setting out to use PHP to do the same thing Plumi has been doing on Plone for years (see EngageMedia for a working example of Plumi). Great to see Autonomo.us active again. Through their blog I've found a new sister project to Disintermedia, CitizenWeb, run by Jacob Cook. Through them I've checked out BitMessage, an encrypted P2P message system inspired by BitCoin, and ArkOS, Cook's personal server OS project designed to turn a Rasberry Pi into a variant of the Freedom Box.
Features and Alternatives
So what exactly are the services that make FB such a hot property, and what alternatives exist?
Think of this either as our online CV, or business card, or just a homepage where we can list stuff we like, next to a goofy photo of us. There are a number of places we can put up a basic profile - mine currently uses the personal wiki provided with accounts on CoActivate.org, which runs on the free code OpenCore software. A TransitionTowns team has recently put together a new networking site which can also host profiles, as well as news streams, and basic pages for TT initiatives.
Almost every online service we can sign up for has a profile. Wouldn't it be great if we could keep one profile up-to-date on one service, and get all the other services we use to just link to it, or pull the data on it into their page through an open standard like RSS? One experiment in this 'hosted somewhere, available everywhere' model is Gravatar (Globally Recognised Avatars), a service which allows users to upload avatar photos/images to be associated with an email address, and shared automatically with other sites where the user has entered that email address. Gravatar is now owned by Automattic, owners of the free code blog engine Wordpress.
This is one of the goals of
, to provide a standard for the sharing of profile data between sites, while giving the user total control over what is shared. Other implementation of this idea include the Social Graph API, created by Google. In fact 'Social Graph' seems to be replacing 'Web 2.0' and 'Semantic Web' as the new buzzword for online development.
This is basically a glorified address book, although the beauty of the FB friends list is that rather than having to keep our own address book up-to-date, it gets updated automatically every time someone changes their profile. Another site called Plaxo have been providing that service since 2002. I signed up, as did a couple of friends, but it's been an age since I updated my profile. Although I could see the usefulness, I didn't want to encourage my friends to trust their personal contact information to a faceless internet business, especially one with fairly aggressive Bebo/ Hi5 style marketing techniques. The same is true of FB.
This basically the collective aspect of the same problem described for individual profiles above, and various protocols such as FOAF (Friend of a Friend) and XFN (XHTML Friends Network) have attempted to define open methods for graphing our relationships online.
This one is a no-brainer. Um, this is called email. Most of us are familiar with it. It works on a bunch of well-established open standards, and there is free code software for every part of it. Most email clients (whether webmail or desktop based) are more intuitive and useful than FB messages.
The message inbox is the most irritating part of FB. There's no way to archive or file messages into categories, or even separate personal messages from messages to groups of friends, or from the tofu messages that come from the hundreds of groups friend want you to join (no folders, no labels, nothing, zip, zero). You can have FB email you when you get a message, but it won't send you the message, no, you have to login to FB and use their cludgy message inbox to read it. You can group your friends into lists, eg according to what area they live in, or what activity you know them through, but you can't send them all a message if the list has more than 20 people on it. Instead you have to make your message an event (see below).
Update: Turns out I was wrong about emails from FB not containing the message text, but did you know all the messages you've sent people get deleted if you delete your FB account? Also, did you know FaeceBook have been content-scanning and censoring your private messages to your FB friends?
On this front, FB is no different from the plethora of proprietary, non-interoperable instant messaging networks like ICQ/ AOL, MSN, Yahoo Messenger, Skype etc. The main innovation of FB being the way the chat client is built into the website, rather than relying on users downloading and installing a separate client program. Actually Google did essentially the same thing by incorporating GTalk into GMail, but instead of creating another closed network, they used Jabber (now an IETF standard called XMPP but I like the informal nickname better), an open standard for instant messaging, 'presence' ( whether your friends are available to chat or not), and offline messaging. Once the GTalk system was connected to the Jabber cloud, users could not only connect with each other, but with anyone using any Jabber client and server.
From the 'Credit where credit is due' files, FaceBook Chat has begins to move towards Jabber compatibility.
Groups and Discussions
This is one area where the traditional tools of email lists and web forums have proved less than ideal. Being able to form a new group out of a network of like-minded contacts, hold discussions, and organize events at a distance, are definitely useful. One emerging alternative aimed at activists, who increasingly use online tools to network, is Crabgrass, a free code software that can be used by anyone to set up their own social networking site. A working example can be found at We.Riseup.net.
Status Messages and Link Sharing
In other words micro-blogging, the service made famous by Twitter, and also offered by Identi.ca who use the free code StatusNet software. More recently there is Buzz is another service that allows users to share short messages (as well as links, photos, videos etc) with friends in a network, in this case your GMail contacts. Unlike FB, Google have demonstrated their committment to Data Liberation by making it easy to export data stored in your Buzz into another service.
One Rule To Ring Them All
Some people like the fact you can use your FB login on other sites. The same people were probably excited about signing up their email address as MSN Passport 5 years ago. Give me an OpenID any day.
Update: Another way to manage your online identity is TrustFabric, a free code p2p client which claims to securely store your private data and help you manage your relationships with services.
Photo and Video Sharing
There are numerous sites for this. I'm ambivalent about FlickR, owned as it is by Yahoo, but as mentioned above they have been strong supporters of CreativeCommons, and as of 2005, most of the software they used was free code. YouTube hosts most of the videos shared on FB anyway (update 2011: they are also supporting CC and will be making CC-licensed videos downloadable). Google have just bought up On2 and announced their intention to release their VP8 video format as an open standard, so they can use it with the video tags in HTML5 to make YouTube player nicer with free code operating systems like GNU/ Linux. The WebM projectM was recently set up to offer an open standard for web video, incorporating VP8 for video, and Ogg Vorbis for audio. We just don't need FB for this.
I've got to say this is the killer app of FB for me. The ability to create an event page, and invite people to it by selecting from a list of people I know, knowing that they will be reminded of my event every time they look at their profile page. Also the ability to have an online calendar, one that invites me to events I'm likely to be interested in (because my contacts are inviting me). However, FB events has its flaws. For a start, I've made groups of contacts (say, everyone I know in a particular city), and tried to send out an invite to that list. FB throws a spew because I'm inviting too many people at once. Grr. Events stay in my list, even when I RSVP that I don't intend to go. Why? There is no way to import from events calendars on other websites, eg Aotearoa.Indymedia.org, or TransitionTowns.org.nz, or Permaculture.org.nz, or export events to them. I could definitely imagine a system that improves on this.
OneSocialWeb are proposing a decentralized events system based on the Personal Eventing Protocol, a subprotocol of Jabber.
Decentralized Like the Internet
There are various projects developing free code social networking CMS software which includes social networking features; eg Elgg, Pligg, and BuddyPress. In theory, any CMS could attempt to position itself as the new FB, but they've been around at least since the now defunct Active.org.au, and the sites Indymedia built on their technology. The real problem now isn't the lack of publishing platforms - but rather an agreed means of stringing them all together.
Federated Social Networking
Update 11/07/2011: Diaspora has now been reimagined as a way of federating social sites together, more like Appleseed, or OneSocialWeb
Probably the most exciting FB alternative idea I've seen so far is Diaspora. It is intended to be the social networking equivalent of BitTorrent, a peer-to-peer network with GnuPG encryption built in to protect user privacy. It will conform to a range of open standards, including OpenID, VoIP, and probably XMPP. It will interoperate with other social media services like some of the ones mentioned above. The only major downside I can see is that our profiles and information are only shared when we're online, which is a downside shared with any server-less system. After getting listed on KickStarter, a site which helps people with creative ideas connect with angel investors who can donate funds to make them happen, Diaspora raised pledges of US$10,000 in just 12 days. Watch this space!
Update: in the wake of Diaspora, the developers of free code P2P social networking software Appleseed have come out of the woodwork with a fresh build of working code written over the last few years, and a test site up and running. OneSocialWeb is a few steps ahead of either, with mature code, and multiple servers already connected to 'the federation'. They list their inspirations as "activitystrea.ms, portablecontacts, OAuth, OpenSocial, FOAF, XRDS, OpenID". Lorea is another distributed free code offering inspired more by the semantic web than Web 2.0 hype.It was heavily used in Spain in the lead-up to the May 15 Uprising.
Another approach comes from the GNU Project. GNU Social is intended as a protocol for linking social networking servers in a peer-to-peer network, with the development of peer-to-peer clients as longer term goal. The initial plan is to develop an implementation of the proposed protocol in co-operation with Status.net.
Update 11/07/2011: Google have abandoned development of wave, but the open protocol and the free code they wrote for their implementation are still available to the community. Maybe its time will come. Maybe not ;)
Another decentralized client has been with us for some time. Google Wave hasn't made much of a splash yet, but I'm intending to get a few people together to test it out. Google has released the standard behind wave as an open protocol, meaning as with GTalk and XMPP, anyone with a server can run an implementation of Wave, and connect to the Wave cloud. Google have even released a bunch of the software they've written to implement the wave standard as free code, to help get you started.
Corporations generally suck, but Google don't suck often, and I continue to be exponentially more impressed with the IT innovations they give away than I am with anything that "Intellectual Property"-loving companies like Microsoft, Apple and Adobe patent and copyright. Google were the first major player talking about opening up social networking to interoperability with their OpenSocial framework, and although there's been limited buy-in from the other big players, Wave might get some traction one people start getting their heads around everything it can do.