Twitter is getting crowded with Data Portability talk. Let's discuss it here.

Mick's thoughts in no particular order.

This is all new. Facebook started in a controlled way (students only) and has opened the doors a little with the App platform.  It's early days and us in the echo chamber are rightfully asking lots of questions. It's good to ask but don't expect easy solutions.

 

Devils Advocate - are we all caring about this a lot out of fear of Facebook Monopoly? I'm an open source guy, but it mainly shits me that they might be the social operating system limiting my next business.

I am not sure why answers need to be easy. Also this is not new. The problem is our definition of 'social network' is too narrow. Tangler is a social network, gmail, flickr, upcoming, digg and all of them - I want ONE buddy list - the rest are just applications inside which i can interact with my friends

I think the One Buddy List will come but not this way.

How then?

Creative commons and openID mash ups?

I think we are on the right track, but I'm just not clear on how something like this can be open source.

Also, are we creating tools for us geeks, or for the common people?

The average dude just wants to use stuff safely.

And actually, if you asked the average duddette whether they want a single system with all their friends, that's lots of fun, easily to use and safe, or if they want complete freedom to easily move things to new systems, they'd go with the former.

If I collect my friends information together in Outlook, it is my data. I move it to my mobile PC, my desktop, to plaxo. I also have a social contract with the people in that list not to send it to spammers etc.

If I build a Facebook profile with lots of friends then that collection is also my data and I still have a social contract not to share my friends email addresses etc with others. I have put time and effort into collecting that data. That data is my digital world. I want to move it to other places. This is no different to what I do on my local PC today, but is more tricky because it is on the internet. Thats why the time is now to think about that properly.

This debate is excellent because it brings the issue into relief with an actual use case.

The data portability proposition is to find a way to give me the control over my own data that does not require scraping, nor exposing my data (including your email address) to other people.

Perhaps the answer is that I should build and maintain that list on my own service somewhere and then future Facebooks, Tanglers, etc let me connect to that to bootstrap my social net...?

 

Guys like it or not, geeks solve things first, then the mainstream just use it

Making it open source is SUPER simple - you define the end-points and use standards to encapsulate the data

90% of the solution is already designed (on the DP google group) and implemented (in the faraday media labs)

It's up to vendors to make it easy - DP is about pushing vendors. The only that DP does to help users is help them identify the vendors that join in by giving them a common Badge/brand to look for

That is true Chris BUT...  Twitter is full of geeks this morning saying they don't get it... so they are not going to start work....  but I reckon that the issues before us today are EXACTLY the right fuel to engage real development.

yeah so the problem is helping geeks get it - not end-users - YET

Users need to drive vendors... vendors will always choose lock in.

Again - consider what happened with DRM.

Yeah I agree but there are two different conversations going on here... to IMPLEMENT DP needs a geek story. To evangalise DP needs scoble and other high-profile cases, a common brand and user-facing materials - but there is no point promoting to users heavily until we have the reference design in place and some pilot sites from smaller, more nimble vendors

The story needs to evolve as the audience changes

Early Adopter Developers -> Developers -> Early adopters -> Mainstream

Mainstream may not even know about it in the end - they may just magically see the same friends list everywhere they go and authorize

I think you are right with the adoption flow... but developers still need to understand how they are going to make people's lives better... so it still needs to be expressed in terms of users/people in my view.  'Implement hCard and Open ID on your site' doesn't do it for me.

just finish the sentence "... to get complete and up-to-date, and constantly refreshing data about users as soon as they join"

precisely

 

but that's not USER value - that is developer value

which is fine - that's the story they want

the end-user campaign is different

'look for the DataPortability badge so you know you don't have to fill in your profile again'

But is there a central source still, like openID? Or is it just a network layer type function?

I need to make a diagram maybe

this is the text version http://groups.google.com/group/dataportability-public/web/reference-design

OpenID -> YADIS (a list of services you use) -> oAuth/oAuth Discovery on given services -> Retrieve data using common API and Standards

Totally distributed, the only central point is the OpenID provider you trust which returns your YADIS file

Yeah, that is the golden rainbow, but I think it's still ten years off.

aaas if

and just cause something is 10 years off why is it not worth starting/doing?

Tangler could implement this in a week

and would win friends and influence people if it did

the tech is here, but the actual real world solution is a long way off.

Lots of people are using FB and other SN's but that doesn't mean they care or get it.  The concept of RSS is still boring and painful to most. OpenID is just gaga.

I know it will happen and I know it's worth to work on it now, I'm just wary of panaceas
 

People don't care about HTML - but they care about the Web

they don't care about RSS but they care about the tangler notifier

the technology does not matter - the result matters - so we just need to show them results - which means we need to just implement it

yep. I just struggle to see the structure working in such a subjective, vague and changing arena like my social graph.

Social graph is just data

and its just a fancy name for 'buddy list'

And it can be more than adequately described in XFN/FOAF/APML - so we simply need to help developers find and sync a user's XFN/FOAF/APML files

MSN Buddy List is simple.

Global Buddy List is messy and changing and vague.

I question the marginal value of managing it. It's like Open ID for me. Sounds great, but it really doesn't save me that much time.
 

BRB

Well I think its a mistake to think of it as a global graph probably - it should be a personal graph - a Personal Social Network rather than a Portable one.

And separating OpenID from the issue is also a mistake - OpenID is a platform on top of which this is all built. DP is what gives OpenID Value - that's the point - it is the next step after OpenID so that more value is realized by using an OpenID login

not separate things

I know they are two layers of the same issue. I don't know enough about it yet to see how OpenID is a trusted easy to access, versatile central source that easily manages all my many many social apps (as you said, nearly everything online is)

Mick, I think you are wrong in your 10 years away estimate. Users will need this very soon and Chris is right that it can be implemented very quickly.  The value it adds is immense.

You say "it really doesn't save me that much time." - I think we will look back and think how silly it used to be that. - just to have a conversation - we needed to manually input all our friends details again into Tangler.

Would you argue that Nokia should not allow exporting address book data to people that move to HTC?
 

Mick then u need to learn more about it

Phil - Agreed - I'd say 5 years max till it's just 'expected'

The time to begin, however, is now

I totally agree that it should be allowed and I know it will happen, and I'm behind it, I just think it's a big, complex task that a single vendor can't win due to trust and that open source can't win quickly because it needs vendor support.

Yep, i'm catching up.........

Of course when I say 10 years, I mean 10 Mick years which is about 4.6 normal years.
 

haha - you are totally right there - 1 vendor can't win - that's why it needs to be done in this common/shared space

i think in the end it will look like the banking system. Our currency is our data, trusted identity providers are the banks, and dataportability is the banking network/eftpos/atms etc

I think there's a cause and effect at work here. Right now the value is low, but I think once it catches fire the sophistication and tools that support and manage that centralized data will become very powerful and very useful. That will quickly snowball.

Not sure when that tipping point is though. Still a little way off for large numbers of users.

A key issue will be privacy.
 

As i say over and over... The issue of privacy is just subset of the issue of Control.

And to Quote Chris Messina "

Our notions of privacy must adapt. That starts by developing the language to discuss privacy in a way tha's obvious and salient. Simply demanding the protection of one's privacy is now a hollow and unrealistic demand. Our systems should protect our privacy by allowing us to fine-tune the trust we put in our friends"

So nothing can help the cause more than true forms of user-centric data control

All true Chris.

One thing about privacy is it's a burden. It's hard to control, so most people just default to the default. I think that will turn ugly.