Twitter as a protocol
I spend a lot of my life on twitter. I get a lot of my news from it. It helps communities form and react to events quicker than ever. It is unique in it's reach.
It's also unique for another reason. It is all a single company. I find this deeply disturbing. Unless you are willing to engage with this one company, you are shut out of a vastly important swathe of online communication. Imagine if email or web pages were owned in the same way - everyone around the world with a hotmail account an a google-plus page. Layers over the top providing some distinctiveness, but the whole edifice vulnerable to the whims of one corporation.
So. What could be done about this? Well, way back when, twitter was referred to as microblogging. Blogs are not homogenous(1). Why not? Well, there is a vague consensus about structure, and a simple syndication/subscription process allowing people to follow lots of them. Couldn't microblogging go the same way? I think it could - and probably should. This has two prongs: the tech and the company. Tech
Any micro blogging platform needs to supply two things, a publishing platform and a subscription ability.
The publishing is, I think, the easy bit. A microblog-stream is essentially an XML stream, with content surrounded by a cloud of metadata, detailing who posted it, where, and a direct link to the specific microblog. It would be quite straightforward to turn a well formed XML stream of this type into a web page, or use it to drive an app.
Subscription is slightly harder. You don't have a single database of every microblog stream being published. You'd need to spider the web to find them and add streams to your list of streams to merge and display. I'd also expect display rules to be part of the subscription - character limits, picture harvesting, short-link resolving etc.
With the right open source packages, anyone could host this on their own site; allowing the hosted stream to be spidered and pulling in and displaying the requested streams in a channel alongside. Which people with their own website or a dedicated tech department could do. Hooking this up to third party clients and authenticating posts/changes is eminently possible. The company
I've just described a ton of tech. Where does it come from? Well, the hosting process I've just described above has non-zero technical requirements. What about a hosting company? I admit I'm cribbing wildly from the WordPress business model here, but I'd expect the company that built this software to have a massive advantage in hosting it.
Fundamentally, as open as you make this, not everyone will want to host this themselves. I'd anticipate two tiers of users. On the one hand you have people required to use default settings and the ad-supported website, on the other you have people willing to pay £Xpcm for the right to use third party apps, have an ad-free experience and tweak their settings more.
Also, there is a ton of b2b potential here: Have a giant database of spidered microblog streams? Sell access to people who want to also provide a hosting experience who can't afford to spider the web. Have a ton of data about how people connect and communicate? Don't just sell the data, sell insight (selling raw material is rarely as lucrative as selling a manufactured product). Have a dev-team with bone-deep knowledge of the micro-blogging standards? Sell custom clients to companies with specific customer-service requirements. Risks
Of course, it does need to be interoperable with Twitter - at least initially. It's the only way to build a user base. This will be tricky. I would say that it needs to be done in an if-this-then-that(2) way, except twitter recently deliberately stopped ittt from working with Twitter. It will also be messier, more asynchronous and more varied than twitter. Personally I think this is a good thing, but it may limit adoption rates. It also raises the issue of abuse. How to spot it and how to stop it.
In some ways having this be open, paradoxically makes it easier. If companies are just hosts, they can have policies, which when breached just get you kicked off (since they aren't denying your right to participate, they just don't wish to help). I can imagine trolls migrating to "@subscriber.trolls.com
" or "@troll.mydomain.com
", which should make them easier to manage. Conclusion
So, why aren't I working on this in coffee shops on my laptop rather than write this? I'm not a web-developer. I'm a kinda-developer in a ludicrously high level language, with a tendency towards requirement analysis and design. Also, while I think this is important and should be done, honestly, it's not something I feel so strongly about I'm willing to spend all my free time for the foreseeable pursuing it - which is what it would need. However, if you see anyone trying something like this, point me at it and I'll contribute where I can.
(1) Ignoring the recent trend toward everyone and their dog having a tumblr instead of a blog.
(2) A web-service that monitors other web services, when they update/report a specific result, it performs some activity you describe - emailing you or cross-posting to another service etc.