Motivating vendors to embrace DataPortability

[dan.col]

The DP website lists 'not wanting to share data for competitive
reasons' as a challenge to getting vendors on board with DP. This may
be true to some extent, but another factor is dominant: data needs
are driven from functional requirements, so managing different
functional requirements is going to be a key challenge.

Example: let's say one insurance company uses 'job type' as a required
input variable for quoting for a certain type of insurance. If
another insurance company does not use this input variable at all,
will complete DataPortability* ever be possible ? Or do we settle for
partial DataPortability ?

* As you can see, my take on DP is that it facilitates seamless and
automated transactions between organisations through the use of common
data definitions (and possibly data models). The organisations can be
business or individuals.

Furthermore, as innovation drives ongoing changes to functional
requirements, and thus data needs, a strong commitment is needed by
all participants to maintain DataPortability. Unfortunately, very few
businesses understand, and thus value, data portability, so
maintaining this ongoing commitment will be very challenging.

How then do we motivate vendors to make the commitment to DP ?

I note that the DP website and many of the forum posts focus on social
networking websites, such as YouTube etc.

Would the DP community be more likely to convince vendors if they can
demonstrate the values of DP through an industry by industry focus ?
This will allow each industry to reach a critical DP mass, and then
act as a role model for other industries. After all, it is a lot
easier for an executive to agree to something that is already
delivering value for another industry, than to accept the DP concept
as a good, but unproven, idea.

Focusing on social networking websites could work... it would raise
overall awareness, and, best case, users could start to 'demand' it
from the websites they use/businesses that they buy from. But whether
the executives who approve software development would value these
demands is another question.

While DP for social networking is a nice to have, personally, I will
see DP as successful the day I don't have to type near identical
details into multiple business websites to choose which one of them to
buy from.

[Elias Bizannes]

DP has been raised for other industries, such as healthcare, and it
certainly is welcoming. Since January, this project has expanded way
beyond social networking.

What I like about your idea though, is we take an industry by industry
approach. Analysing the issues for example by industry, gives us a
better understanding of how we can pull this off. Would love to see
you move this forward - join the Steering action group chat and we can
develop it.

[Christian Scholz / Tao Takashi (SL)]

This is maybe another point for going field by field, seeing what the problems are in this specific field, how to solve them, how to argue to get businesses on board and so on. I think the only way to get going is by going industry by industry anyway, it's very unlikely that everybody will jump onboard at the same time ;-)

I also wonder if it makes sense here to really allow specific working groups to be built for such a specific topic. But it probably all depends on how different the problem space is from each other.

-- Christian

--
Christian Scholz
Tao Takashi (Second Life name)
taota...@gmail.com
Blog/Podcast: http://mrtopf.de/blog
Planet: http://worldofsl.com

Company: http://comlounge.net
Tech Video Blog: http://comlounge.tv
IRC: MrTopf/Tao_T

[Julian Bond]

"Christian Scholz / Tao Takashi (SL)" <tao.t...@googlemail.com> Thu,
13 Mar 2008 13:59:04

>This is maybe another point for going field by field, seeing what the
>problems are in this specific field, how to solve them, how to argue to
>get businesses on board and so on. I think the only way to get going is
>by going industry by industry anyway, it's very unlikely that everybody
>will jump onboard at the same time ;-)
>
>I also wonder if it makes sense here to really allow specific working
>groups to be built for such a specific topic. But it probably all
>depends on how different the problem space is from each other.

I had a thought one morning this week that one aspect of DP is a
clearing house for Website and Standards body annoyances and a bunch of
noisy individuals who go and engage with the website owners or standards
community to try and effect change.

This suggests having a database something like the WINE applications
database. Where there's an entry for each group we want to target, a
list of comments or discussions about changes needed and one or more
people who commit to trying to get those changes made. eg

Twitter
- Support oAuth
- Export Contacts as FOAF
Led by: A.N.Other

Google Contacts API
- Use a standard profile Schema instead of creating another one
- Use oAuth instead of AuthSub
Led by: John Doe

Yahoo OpenID
- Support Sreg and AX
- Improve Yadis discovery
Led by: Jane Doe

IDCommons
- Encourage the Schema WG to maintain a list of schemas and push the
industry to converge them
Led by: Mr Identity Person

XRDS group
- Build and maintain a list of service types
Led By: Miss Auto-Discovery

And so on.

This could easily be extended to specific industries or non-consumer
areas.

--
Julian Bond E&MSN: julian_bond at voidstar.com M: +44 (0)77 5907 2173
Webmaster: http://www.ecademy.com/ T: +44 (0)192 0412 433
Personal WebLog: http://www.voidstar.com/ skype:julian.bond?chat
Adult Friendly Design

[Phil Wolff]

This is the reconciliation part of the Sync Layer of the Social Stack [1].

Since there are an infinite number of world views, let's not strive for the impossible universal dictionary.

What is feasible?

Learning. Having DP-consuming sites learn from DP-providers.

The providing service must provide enough meta-knowledge that the consuming/syncing service can learn to import/adopt an alien set of data models and policies.

I think the use case looks a little like this...

I operate a site, say Chocolate Heaven, a cuisine community.

• Andy comes from facebook.  Basic profile info is understood but I cannot find Andy's favorite chocolates. - missing data
  
• Betty comes from Chocolatier. I understand most of it, but i need to learn some new relationship attributes (A told B about chocolate C). - new+relevant strucs
  
• Chuck comes from HotJobs. I get some of the profile information, and much of the community data, but don't know what to do with the massive and sophisticated career data (and could not care less). - new+irrelevant strucs
• Daisy comes from facebook two hours after Andy. And facebook's data model changed.
  

This happens thousands of times per hour, so the technical part of this must scale.

Distinguishing between known/unknown and relevant/irrelevant will be hard, but it's crucial. Niche standards may help, even a library of standards. But we may need simplifying abstractions or technologies to make this easier, to tolerate a zillion ever-changing variations.

Or we can punt (American football metaphor for avoiding the hard problem and taking your chances on a lower value, lower risk strategy). What do you think?

- Phil

[1] http://skypejournal.com/blog/2008/03/skype_and_the_social_network_s.html