If something Twitter like is ever going to make it’s way into the world’s schools in any big way, the core architecture or protocol of microblogging which today we see as Twitter, needs to be made decentralized and into an open standard. This process would allow Twitter to move from a consumer space to an enterprise solution.
There are many high scale school social mapping based applications that potentially will only evolve if an enterprise class microblogging standardized architecture exists. This would create high and reliable scalibility – the potential to scale beyond just one school … a whole district at the very least and the ability to scale higher – all in realtime – applications such as: all call announcement system, all call emergency system and many others based around staff to staff, administration to staff, admin to parent, parent to staff, student to student, student to staff etc. communications systems.
Dave Winer wrote a post yesterday about decentralization microblogging that he says was driven by Scott Hanselman post, Why is Microblogging centralized? If you have been following Dave’s blog of late you have been reading his posts about the problems with Twitter scaling and how they are related around the centralized design of the system. Dave’s solutions to the problem always seem to be based around RSS at the core, technology that he has mastered and been the core developer of.
I left a comment at Dave’s post that has really nothing to do with what Dave is proposing which is backing up my Twitterstream as RSS on another server somewhere or allowing my Twitter client to poll RSS feeds when Twitter is down. The comments I left is more in context with Scott’s original post and proposal to distribute Twitter into a generalized spec for microblogging.
Don’t get me wrong, I respect Dave and all the work his has done in RSS and I understand what he states in is post is not a full solution. I just wish he would start thinking and pointing people beyond his RSS forest. Yes RSS or something XML will be woven into this solution on the backend as transport language but the solution at the core is bigger then RSS alone.
What the social tech industry needs to do right now is get past it’s amazement of Twitter and any thoughts of bolting band-aids to the present scaling non-distributed mircoblogging architecture of Twitter and to start building something open and scaliable for the opportunities that are up ahead.
Dave went for a walk and came back with this thought “If Twitter were to go down, then the desktop client would fall back to polling the feeds. It would probably be slower, but it would work.”
Sorry Dave I don’t want Twirl probing 100+ feeds when Twitter goes down. This is not a solution or option for me and is the reason static RSS pages sitting behind RSS never became the achitecture and won’t be the future realtime microblogging solution. The fact is the Twitter architecture, with all it’s problems scales by orders of magnitude over static RSS files sitting behind http servers. The answer up ahead is to move the microblogging load off of the Twitter sphere and off of the main Twitter servers and onto a decentralized “microblogging cloud” architecture that is deployed around the world and most importantly “owned by no one”. That is the proposal and conversation you want to have with the young microblogging developer community.
The present conversation really needs to go straight to the development of the Microblogging architecture. Wasn’t that the answer with the web? ie. http, html, dns. If the Web was designed around a Twitter like architecture we all would be serving our web pages and blogs served off Tim Berners-Lee’s server farm at CERN right now.
Someone or some group needs to start standardizing a microblogging protocol that would be supported by all present and future micro blogging clients/catchers/applications. The investment into the future of microblogging by education is dependent on it.