Open Social compatibility: social networks and applications vendors working together
Now that MySpace, SixApart and Bebo have joined Orkut, Friendster, Ning, Xing, Slide, RockYou, iLike, hi5, LinkedIn, Tianji, Engage, Flixter, Plaxo, Oracle, Viadeo and Salesforce - among others - it looks like the OpenSocial juggernaut is steamrolling along. Now we just gotta hear from Yahoo, Microsoft and - Facebook.
Facebook claims Google hasn’t even asked them to join the alliance and folks are already arguing whether Facebook will support both models or change and just support the open OpenSocial approach. VentureBeat got it wrong - assuming this has anything to do with Google gettings its hands on user’s data. Flixter has shown how easy it is to create an OpenSocial application and Friendster founder Jonathan Abrams even had the audacity to open his mouth to Wired.
Me - I’m just sitting here, smiling and wondering about interop and whether all these platforms are really gonna open up their social graphs with unique identifiers. After waiting four years - who’s in a hurry?
Sure we’re seeing all sorts of discussion over whether OpenSocial competes with Facebook or will Facebook join the OpenSocial alliance but the real issue is - will this alliance work - at all?
Needless to say OpenSocial is gonna affect and effect allot of companies, markets and vendors. There will be startups who nothing other than ‘cross-social network’ applications. There will be VC funds who do nothing but fund ‘OpenSocial apps’. Like any other stampede - they’ll be on-going competition to be ‘first to market’ and contests and giant directories will be born overnight as well.
I give my own opinions on this Kara Swisher interview - as I’m hoping Mark Zuckerberg can swallow some humble pie and go that final 2% of the way - Facebook hasn’t already traveled. I just finished another interview with Susan Bratton and something tells me this ain’t ending here - today.
Plaxo is gonna throw a party tomorrow (Friday 4-6 PM at their offices in Mountain View) - and I’m gonna try my dammed-ist to connect all this up to our upcoming launch of MyKingsWorld and the Bell Video Store.
But let’s make sure that the implementation of OpenSocial is done PROPERLY first!
Lets make sure that:
- the user’s profile data can be accessed - both exported and imported - and that the user gets full control over what gets moved around. Ideally users would also be given ‘opt-in’ control over whether or not their data can be moved in their friend’s social graphs.
- user’s social graphs also have to be under their control and a full disclosure of that social graph’s unique identifiers (eg. their email address) is crucial if any developer is able to do anything with this social graph - that is not inside the firewalls of a social network (SNS.)
- applications and services have to be free to display any kind of advertisement or monetization model they wish - without demands for kickbacks, royalties or pieces of the action.
Add to these existing OpenSocial specs - assumptions you can apply to:
- communications, IM, presense and VoIP also need come APIs - so Twitter customers can move their list of followers to Sessmic and Meebo can act like a gateway to Skype or vice versa.
- reviews, events and databases of people - must be shared and made available (except to spammers - of course!)
- listings marketplaces and ad networks must be open - to allow any vendor to participate in - at any level
Well guess what?
Just because MySpace and Friendster say they’re gonna support OpenSocial - is completely different from them actually allowing a user to export their list of friends - with unique emails for each friend. This I gotta see.
As Saul Hansell points out - OpenSocial is not something that competes with Facebook, it is a technology - a feature. But to make interop work well - you have to have compatibility between different implementations of the spec.
Will Google provide a validator and compatibility testing labs? What happens if something doesn’t work?
What happens when some of the alliance members don’t adhere to ALL of the APIs and thwart or warp the APIs? What happens if they DON’T let their users more their data around - but only SUCK IN external data?
What happens when OpenSocial social networks don’t give us access to email addresses for all those social graphs? Who do we complain to? Google?
Details are still lacking, but I’m sure this is something Google just WISHES they could do - run a compatibility test labs - right?
November 1st, 2007 at 5:28 pm
Wow, it really sounds like you don’t get it at all.
November 1st, 2007 at 6:35 pm
[…] The total named partners comes out at 20 if you include Newsgator that is not included on the list Marc Canter rolled […]
November 1st, 2007 at 10:36 pm
[…] Dodgeblogium wrote an interesting post today onHere’s a quick excerptOpen Social compatibility: social networks and applications vendors working together Now that MySpace, SixApart and Bebo have joined Orkut, Friendster, Ning, Xing, Slide, RockYou, iLike, hi5, LinkedIn, Tianji, Engage, Flixter, Plaxo, Oracle, Viadeo and Salesforce - among others - it looks like the OpenSocial juggernaut is steamrolling along. Now we just gotta hear from Yahoo, Micros… Read the full post from Marc’s Voice Tags: Uncategorized via Blogdigger blog search for action. […]
November 1st, 2007 at 10:58 pm
user’s social graphs also have to be under their control and a full disclosure of that social graph’s unique identifiers (eg. their email address) is crucial if any developer is able to do anything with this social graph - that is not inside the firewalls of a social network (SNS.)
I haven’t seen the opensocial specs in detail yet, but the graph really shouldn’t use the email address as the unqiue identifier. That’s a really bad idea because it’s incredibly difficult to automate user behavior using emails. Make it URLs so that people can then support OpenID and even more importantly, you can get contextual data to act on the user’s behalf. The end result of open social data is automating human behavior so you don’t have to worry about day-to-day crap. For instance, it’s impossible to me to automate arranging calendar synchronization over email — whereas if you had an HTTP endpoint, you can hit the URL, check if you both had the calendar protocol/endpoint set up (whether it be via AX, uFormats, whatever). I know email is presently the unique identifier on the web, but I think the identity community should be looking forward, not backwards.
November 2nd, 2007 at 2:46 am
Got what wrong?
November 2nd, 2007 at 4:15 am
Marc: you’ve been beating the drum for years for social interconnect. When I was briefed on OpenSocial a couple of weeks ago I was smiling and thinking of you. Continue to push the envelope on the implementation issues and keep your eye on the efforts of this ‘open’ alliance. Lets hope that its open. The world is going to turn out just the way that you envisioned it years ago. Congrats.
November 2nd, 2007 at 5:58 am
Marc
There’s a lot of geek speak on the Open Social announcement. If you need to explain it to executives, I’ve tried to make it very clear and concise from this post, weighing the pros and cons.
http://tinyurl.com/3dtqs6
Love to hear any feedback, even if you don’t agree.
November 2nd, 2007 at 11:30 am
[…] Marc Canter - Open Social compatibility: social networks and applications vendors working together […]
November 2nd, 2007 at 1:40 pm
[…] and still preserve all the interactions with their friends (the bit that really matters).” Marc Canter says: “Me - I’m just sitting here, smiling and wondering about interop and whether all […]
November 3rd, 2007 at 7:08 am
[…] See also: Marc Canter on OpenSocial compatibility–social networks and applications vendors working together […]
November 5th, 2007 at 11:20 pm
[…] Canter doesn’t care if it’s a rumble or not, as long as it moves the needle forward on opening up […]
November 6th, 2007 at 2:30 pm
While i agree that it would be great if openSocial was what you described, it is not the case. openSocial is at it’s heart about app/widget portability, and has nothing to do with portabilitly of user data, including social graphs. the only kind of cross network portablitity would be provided by the app developers, for example, if i make an app that stores a users favorite movies, i would then store a list of movies associted with a unique network/user id that is given to me by the social network, when that users profile loads, some javascript sends the id of that profile to me, and i return the movie list. i can also do things like push an feed item saying this id added this movie to its list and the network can do what it pleases with it. the only possiblity for portablility here is if i, the app developer, allow users to link multiple id to a single movie list in my database. the networks themselves share no data.
with that said, i’m sure some interesting grey box techniques could be used to scrape data from multiple networks, which could make it possible to have some sort of portable profile that just exists in the cloud and that be exported.
November 17th, 2007 at 1:18 am
haven’t seen the opensocial specs in detail yet, but the graph really shouldn’t use the email address as the unqiue identifier.