• Print

OpenSocial: It's the data, stupid

While I’m a huge fan of the idea of an open social networking platform, I’m bemused by all the enthusiasm over Google OpenSocial. As I sit with what I learn, the mild skepticism I expressed the other day has turned into full blown disappointment. This is nothing like the social network operating system that I got so excited about when I first heard Brad Fitzpatrick and David Recordon’s thought’s on the subject.

My disappointment with OpenSocial was crystallized by an exchange between Patrick Chanezon, Google’s developer advocate for the program, and an audience member at the OpenSocial session at Web 2.0 Expo Berlin. The audience member asked something about building applications that can remix data from the participating social networking platforms. Patrick’s answer was along the lines of: “No, you only have access to the data of the individual platform or application.”

This is SO wrong. And it shows a fundamental failure to understand two key principles of Web 2.0:

  • It’s the data, stupid. (Formerly “Data is the Intel Inside”)

  • Small pieces loosely joined.

Let’s start with the first one. If all OpenSocial does is allow developers to port their applications more easily from one social network to another, that’s a big win for the developer, as they get to shop their application to users of every participating social network. But it provides little incremental value to the user, the real target. We don’t want to have the same application on multiple social networks. We want applications that can use data from multiple social networks.

And data mobility is a key to that. Syndication and mashups have been key elements of Web 2.0 — the ability to take data from one place, and re-use it in another. Heck, even Google’s core business depends on that ability — they take data from every site on the web (except those that ask them not to via robots.txt) and give it new utility by aggregating, indexing, and ranking it.

Imagine what would have happened to Google maps if instead of supporting mashups, they had built a framework that allowed developers to create mapping applications across Microsoft, Yahoo! and Google as a way of competing with MapQuest. Boring! That’s the equivalent of what they’ve announced here.

Would OpenSocial let developers build a personal CRM system, a console where I could manage my social network, exporting friends lists to various social networks? No. Would OpenSocial let developers build a social search application like the one that Mark Cuban was looking for? No.

Set the data free! Allow social data mashups. That’s what will be the trump card in building the winning social networking platform.

tags:
  • http://www.kensheppardson.com Ken Sheppardson

    I agree with all of the above, Tim, but I’d just point out that there’s a little ray of hope: the People Data API–a part of OpenSocial that’s yet to be released–would enable a single app to pull social graph data from each service that supports it, no? It not clear that this could be an app that lives as a widget in an OpenSocial container, but it could exist in some form.

  • Roupen N.

    I agree with the point but how to free all that data without getting tangled-up in the privacy policies and terms of use of the source sites? Also, the sites also know “its all about the data” so they likely would be less reluctant to participate and let go of all that GOLD (data).

  • http://radar.oreilly.com Tim O'Reilly

    I hadn’t seen that, Ken. Yes, that’s very good news, and hopefully there will be more.

    But it doesn’t hurt to keep the pressure on :-)

  • http://www.thoughtandtheory.com Dennis Eusebio

    Do you think this might be only the first step in reaching a point where data can be ported?

  • http://iss.im/ Nick Vidal

    Take a look at ISS. ISS is a set of open standards that help people publish their social graph. But differently from FOAF and XFN, it publishes what really matters: i.e. how are Alice and Bob connected semantically (and in real-time). Please check the screenshots and the two microformats:

    It follows the K-ISS philosophy and provides endless possibilities.

  • http://www.rooftopsolutions.nl Evert

    I had the same thoughts initially.. It is a good step in the right direction though, and since there’s now a common format for describing users and friends.. this already allows us to mix and match social networks and friends .. (just not on their platform)

  • http://voodoowarez.com rektide

    not to excuse such a piss poor excuse for a platform, but OpenSocial has massively been hampered by the total inability of the web browser to do the data access patterns you, me and every web developer worth a salt craves.

    currently the _only_ cross domain way of accessing data is appending tags that load JSON data. the same origin policy is a brutal beast and has put the web in retrograde, and hackers have gone to enourmous length to squeeze every ounce of leeway they can around it. its one of the greatest side effects of the webs “screen replacement” operation, that the web is comprised not of lots of iterconnected pieces, but a network of many many diverse standalone pieces.

    the reference documentation on this is
    crockford’s state of ajax
    and his JSONRequest cross domain ajax library.

  • http://20bits.com Jesse Farmer

    Arf! Well, I swear I didn’t read this before I wrote my post, “It’s about distribution, stupid!”: http://20bits.com/2007/11/07/memo-to-opensocial-its-about-distribution-stupid/

    I disagree that it’s about data portability. I honestly believe that’s a concern of (1) people interested in high-falutin’ ethical issues around Web 2.0 and (2) engineers.

    That’s not to dismiss these concerns, but that’s not what’s driving users or applications, IMO.

    It’s the fact that Facebook has an unparalleled distribution system. If OpenSocial can get that (and part of that surely means tying together the social networks) then we’ll have a horse-race.

  • http://20bits.com Jesse Farmer

    I should amend (2) from “engineers” to “a certain class of engineers.”

    As an engineer myself I don’t care if my social data is portable, I care if my application can get exposure and I can get users to monetize. Facebook makes that really, really easy right now. OpenSocial hasn’t proven itself in that regard.

  • http://epeus.blogspot.com Kevin Marks

    Tim, mapping identifiers between networks is a hard problem, not just because of technology but precisely because it is social norms that are involved. If the social networks are publicly articulated performative ones like MySpace or Twitter or Orkut, where the norm is that everyone in the network can see your profile and friends list, that is very different from the kinds of social networks where the list of friends is something only you know (like Instant Messaging networks).
    The problem of portability in a non-performative network is that it is too easy for you to inadvertently leak other people’s personal data; the difference between ‘me’ and ‘an application acting as if it were me’ is at the root of most of the security attacks we see on computers.
    We do want to find privacy-preserving ways to exchange social network data between sites, but we didn’t want to hold the release of OpenSocial for that to be completed, when there is lots of value with the initial, simpler per-domain identifier model.
    You can look at this as social data mobility in the opposite direction – the apps can come to the data, and they operate in the social context of the network they are running in.

  • http://webandlife.blogspot.com Andy Wong

    At the moment Google’s initiative of open social emphasizes on open API to build applications more portable, and more inter-operative, in order to, get eyeballs of developers, rather than end-users.

    I do think Google understands that the real purpose is to make data more open in order to aggregate data in different data islands while keeping needed privacy of end-users. And Google is in good position to aggregate data when such open api and related applications are getting popular. Of course, MS and yahoo etc. can join in, competing with better server farms, or other value-added services.

    After all, the so called social network operating system is to support user centric social network, and users’ data will effectively become users’ property.

  • http://www.krishworld.com/blog/ Krish

    Tim, you have hit the bull’s eye when you said “We don’t want to have the same application on multiple social networks. We want applications that can use data from multiple social networks.”.

    This is exactly the point. And, as Marc Canter likes to say, let us put the social into the applications.

  • http://www.techfornovices.com/ Tech For Novices

    Thats right. Ours is a small blog covering Open Social news called Open Social Monitor. We are not even remotely in your league. but thanks for putting up this comment

  • http://apps.facebook.com/iamhappy/iamx Karel

    I agree users don’t want to store application data on different networks, on which they are separately members. So the way I’m building my app platform is to store the application data in one model, and access it from each network, or from an ordinary website if users authenticate with my own or an openid user/pass.

    In that sense, SNs provide authentication, a friends graph, and distribution, while I keep all of my application data.

    On that score, any SN can provide the first two, while only myspace and facebook can provide the distribution (in different ways).

  • http://vanelsas.wordpress.com Alexander van Elsas

    It is actually really weird that we have to beg these service providers to open up our data isn’t it? It’s the world upside down. But as I said before, the current web 2.0 business model with free (ad-based) services is doomed to be replaced by something that provides the user value. It is not a user centric business model, but a network value model that enforces walled gardens instead of open space.

  • http://www.modadigital.es Adam

    Totally agree, data needs to be shared between platforms.

  • http://dannyayers.com Danny

    The technical side of data-sharing isn’t particularly difficult, given moderate conventions on data representations & protocols (OpenSocial is ugly, but does cover the important bits of the social space). If I have a widget that can see data in (sub)network A and (sub)network B then I can mashup. In practice you’d probably want to do this server-side, with a connector that behaved according to the OpenSocial interface.

    I disagree with Kevin a little – in 9/10 cases a person’s email address is a pretty good cross-system identifier, and there are other (inverse-functional) properties to help cement things.

    But Kevin’s main point – that social issues here are difficult, is absolutely on the nail. To be able to usefully share data without trampling on privacy, quite sophisticated access control mechanisms are needed.

    It’s here where Semantic Web technologies offer a way forward, because they allow rich, web-oriented descriptions of resources (and anything you can identify can be a resource). There’s quite a way to go with Web of Trust techniques, but they’ve received considerable attention over the years – such efforts are somewhat masked by the surface glitter of Web 2.0.

    It should be approaching common knowledge by now that RDF is also extremely useful for data integration, almost certainly the easiest way to hook OpenSocial-derived data to other arbitrary systems. (Ok, connecting between content-oriented systems may be easier at the Atom level, but the social domain isn’t fundamentally about content, it’s about people – which is why some of the Atom parts of OpenSocial are really ugly).

  • http://t3kbiz.com Mahesh Sharma

    Wow, i didn’t realise this was the case. This really takes the oomph out of Google’s announcement.

    From the business side i was thinking there would be so many possibilites in being able to port data from one source to another, and using it across a variety of platforms and applications.

    Imagine the depth of the collaboration that could go on!

    But by locking it out, it seems this to be just as bad as a walled garden social network..

    I mean at the end of the day people don’t want to have to continually load their information into a bunch of platforms, so the application will only have value where the most people, which is exactly the case now..

    Google has to unlock the information for this to have real value. Otherwise this announcement is just an attempt to undermine MySpace and co, rather than providing a truly free and open platform where developers and social networks can thrive.

  • Hisso

    Tim, you’re not the only one: Why I won’t use Facebook (not by me).

    What about Six Apart’s plans: We Are Opening the Social Graph?

  • http://www.itbreaks.net riper

    I totally agree with you but maybe it’s just a fisrt step…

  • http://www.fledermaus.co.uk Ian Batty

    At last, some kindred spirits :-)

    Yes, the People Data API is a step in the right direction (though all we can see now is the preview of the developers guide) but the key thing here is:
    Who do I trust with my personal data?
    Probably the answer is that I only trust one person; me (I am pretty sure that I don’t want every application going to authenticate me at Google, that smacks back to Microsoft Passport)

    So, now I need to use CardSpace, Liberty Alliance or some other Identity Management system to verify me, and allow me to control what information is passed. As well as doing the little bit of actually authenticating me.

    When identity and the social worlds collide we truly have a social web 2.0

  • http://weblogs.macromedia.com/jd John Dowdell

    If Dick Cheney and NSA were pushing the same statements, would you feel so sanguine…?

  • bob

    Yes to “It’s the data stupid”. No to “Small pieces loosely joined”. We advise other countries to be democracies, but we don’t advise them to have two legislatures, states, a constitution, etc. Tim crosses a line when he prescribes technical approaches to the goal.

  • http://www.loudmouthman.com Nik Butler

    Of course its about the data. Data is the next dot Oh step were all about to take to shift the markets forward. But in a shameless plug for another oReilly conversation see here for how Data is next , again,

    http://www.oreillygmt.co.uk/2007/07/the-new-thing-i.html

  • http://billy-girlardo.com/ BillyG

    After reading Danny’s Open? Social? story the other day, I could only leave for myself, “The People Data API is cruel and unusual. The terms of service seem social like bubonic fever at the orphanage. Google transforms into Microsoft of Old. I couldn’t agree more, they get on my nerves more and more everyday”.

    I’m sure G will do fine though, whether I drink the Kool-Aid or not. After all, I don’t work for them, and I’m sure not gonna spend time making another version of all my sites so they’re compliant with the OpenSocial API.

  • http://manwithnoblog.com Gary Barber

    Have you considered OpenSocial as a stepping stone. Yes its not taking the right approach. Yes it walls the data in and doesn’t allow information to joined for various sources to a standard for on ultimate API mashup.

    But its a start. It’s a direction roughy in the direction. We just need to make it more open and usable outside the application. So why can’t we (as a web community) take OpenSocial as a part example and extend it.

  • http://radar.oreilly.com Tim O'Reilly

    Kevin and Danny –

    As to the subject of this being difficult, it seems to me that it’s possible to cut the Gordian knot simply by letting the USER own and control all that data. All of a sudden it gets a lot easier if instead of thinking that it’s these companies sharing THEIR data with each other, we start thinking about them giving me back MY data, with a management console that lets me then re-export that data to any of the social networks.

    Put the user back in the middle, and all of a sudden, this becomes a much more manageable problem.

    That’s why I keep talking about Address Book 2.0. Give me a management console for all my social data, and tools for figuring out what to display where, and this all becomes much easier.

    This is where the iTunes model really works. It doesn’t all have to happen in the cloud. Give me a local management console.

  • Paul Lindner

    Hi folks,

    Good comments all around. However I’d like to posit that data access is _not_ the problem. We’ve had universal standards for years now with little uptake. Tribe.net, Typepad, LiveJournal and others have supported FOAF for many, many years, which encompasses the OpenSocial Person and Friends APIs. Not much has come of that — there isn’t a large enough base there to get people interested.

    Now you have a broad industry consensus on a single way to provide all of the above plus activity stream data. You have a rich client platform that allows you to crack open that data and use it in interesting ways, and finally you have a common standard for social networks to interact with each other based on the REST api.

    So Patrick’s statement at the Web 2.0 Expo is correct, a app running inside a container only allows you to see what that container shows you. However that does not mean that a container could not contain friend references to external social networks via it’s own federation mechanism. Movable Type 4.0 has shown that you can support any OpenID login in a single system, there’s no reason to believe that social networks could not leverage OAuth to do the same.

    And here’s a final point to consider — you have Myspace opening up to developers. That’s huge. That alone is going to draw more developer attention to this problem than much of the oh-so academic discussions of the past few years.

    I suggest people that _want_ OpenSocial to solve all the social graph ills get involved on the API mailing list and make sure that those elements are addressed as OpenSocial evolves.

    There’s a tremendous amount of momentum. Let’s not waste this chance.

  • http://www.mikearauz.com Mike Arauz

    I completely agree. Letting the data out is the fundamental game changer.

    And I believe it’s only a matter of time.

  • http://doncam.livejournal.com/tag/opensocial Don Park

    I too was disappointed when i realized the limited scope of the javascript API. As the first comment points out, only a piece of OpenSocial has been released. The “Data APIs” which are RESTful APIs will open the floodgates to accessing any of the social networks. OpenSocial may not address the idea of consolidating profiles across networks but it opens up each and every network to that possibility!

    There will be another layer, the Address Book 2.0 as a local application is one way, that aggregates data from the social networks. OpenSocial is what’s going to enable that aggregation app to read/write data out of MySpace and Orkut. I believe the Data APIs are not out yet because they haven’t solved how authentication is going to happen – OAuth is the obvious choice here and I hope they use it.

    Where does AB2.0 store its information, especially the meta information that connects my myspace profile and my orkut profile under a higher-level abstraction? Id like to think that gets stored in a FOAF document.

  • http://theprogrammersparadox.blogspot.com/ Paul W. Homer

    Its always been the data, stupid :-)

    After all, software has only ever been just a tool to manipulate data. All of our functionality boils down to just ways to collect and manage data. The success of the World Wide Web wasn’t about really cool distributed software, it was about mass access to mass data. Data, data, data.

    We’re so focused on functionality that we forget to see the data, but the only thing a computer really does is help us make big piles of data. Beyond that, its useless (unless you turn it into a flower garden :-)

  • http://mywhurls.com/ Nick Zadrozny

    Hear, hear. Keep the pressure on.

    I am actually the lead developer for an app with exactly this vision: to open up one’s entire social graph. So even if OpenSocial only allows API access to one site at a time, this is a huge time saver for us because it defines a common API to pull out that social graph information.

    Because we’re deliberately using semantic technologies to store and map this stuff, we can in turn provide RESTful API access to pure RDF or Atom or whatever. We want a use to be able to map out their entire social graph, and have the ability to reuse that data wherever and however they want.

    We’re in very early alpha, but we’re all about being a part of the conversation, so check it out and give us some feedback. http://mywhurls.com/

  • Bud Vieira

    As a person who builds applications in the social networking space, it seems to me that OpenSocial would allow my members to do something like update their friends list on Orkut and MySpace when they added a friend on my service. If that’s possible, then it seems like the only people (as developers and entrepreneurs) that benefit from a more radically open data model would be those who hadn’t done the hard work to get an audience of their own, who just wanted to use other people’s data in a mashup. Letting them do that without getting the permission of the end users of the different services doesn’t seem like a “power to the people” move to me.

  • http://burningbird.net Shelley

    “As to the subject of this being difficult, it seems to me that it’s possible to cut the Gordian knot simply by letting the USER own and control all that data. All of a sudden it gets a lot easier if instead of thinking that it’s these companies sharing THEIR data with each other, we start thinking about them giving me back MY data, with a management console that lets me then re-export that data to any of the social networks.”

    Are you saying you want something like an independent application that allows you to add a friend, and then click on which social networks you want this friend added to?

    This strikes me as one thing that will be allowable via OpenSocial, once the APIs are available, as Ken mentioned above.

    Danny and Kevin are talking about data problems within widgets within one of the social networking applications. (Well, I think that’s what they’re talking about). The only way to break the JS sandbox is with the JSON kludge, which opens up massive security problems.

    Like Danny, though, I’m not concerned about universal identifier: email address, plain and simple.

    Ideally your console would be able to access your friend’s console and tell you if they’re amenable to being added to Facebook, LinkedIn, and that sort of thing, but that’s just data that can be represented as a FOAF file or something of that nature. As Danny also mentioned, RDF is quite good at this. Ahem.

    Having said all this, and maybe I’m tired and fuzzy, but I’m not sure I completely understand what you’re hoping for, if it’s not what I and others have mentioned.

    Perhaps if you can walk through a scenario of what you’d really like to see happen — a real world use of Address 2.0? Or what you expect to be able to do with various of the participants if the data was free enough?

  • Paul Prescod

    Open data is important but what about connectivity between the apps? What about Orkut users seeing status updates and MySpace users subscribing to private Flickr photo sets.

    Users don’t want to synchronize a bunch of different accounts. They want a single account that acts as if it is part of every network.

    An analogy: You’re talking about moving mailboxes between Outlook, Thunderbird and Gmail because its my data. That’s great. But isn’t it more important that I can send email to people no matter what mail client library they’ve selected?

  • http://www.bluepal.com Raju Manthena

    I agree with Tim. We just don’t need a common API. We need a common CHANNEL. Similar to whatever Google is doing for the web-content, would be nice to have something for social-content. Of course, this is all good for an Application User. What about Network Provider? Practically speaking they wanted to compete with each other and not to partner with.

  • http://www.ecademy.com Julian Bond

    Part of the problem here is that there’s multiple things going on under the OpenSocial banner. At the minimum there’s:-

    - Write Once – Run Many gadgets. This should appeal to all the Facebook gadget developers that they can leverage their experience on many other platforms.

    - Rich profile pages for social networks that can’t afford to reverse engineer Facebook.

    - Common APIs for social data access. Look at the Twitter app and mashup ecosystem and now imagine it driven by many sites having the same APIs instead of just Twitter.

    It’s really important that the commentators don’t dismiss the whole thing because they can’t see the utility in the first one while ignoring the others.

    And of course one winner in all this is Google’s Orkut. They’ve managed to get the whole industry behind improving Orkut. ;)

  • http://boinkinchipmunks.com thacker

    Has anyone given any serious thought to OpenSocial and mashup capability and its downside impact? Mashup of products and services is one thing, but people?

    To complement the flippant one-liner of “It is the data, stupid.”, possibly give thought to these flippant one-liners: “It is in its use and application, idiot.” or “Think beyond your own narrow parameters, pinhead.”

    O’Reilly, to cover just one small area of probable abuse, do you have any children? Never mind. ::Fervently coding a mashup to find out all about them::

    Ladies/Gentlemen, please give thought before you propose and give thought before you execute. As far as the social network phenomenon, the best social network is to turn off the computer and step outside.

  • http://www.bluepal.com Raju Manthena

    I like 2 statements from ‘thacker’:
    (1) “It is in its use and application”. It is true that whatever we see now was NOT even thought through by the original inventors. Could be even possible to fill the gaps (with the given initiation of OpenSocial) as it might lead to someone to build the needful and address the issues. Yes, THOUGHT it is needed!

    (2) “the best social network is to turn off the computer and step outside”. At some point, everyone should realize what is it really and what is needed personally too :)

  • http://tagschema.com Nitin Borwankar

    Hi Tim,

    The issue in my mind isn’t just about user generated social network data, it’s about any and all personal data on the Internet.

    IMHO, it’s not possible to be consistent while complaining about social network information while ignoring the bigger issue of data privacy on the ‘net.

    Again, IMHO, we need to rethink data on the Internet from the ground up.

    I don’t see how one can be accomodating about some personally identifiable data being public while objecting to other such data.

    The choice of what data to expose must remain only with the individual who the data refers to.
    I don’t think there can be any grey areas here, else the discussion doesn’t meaningfully converge.

    Some thoughts I penned a while ago are at

    http://tagschema.com/blogs/tagschema/2007/07/some-thoughts-on-data-rights.html

    Thanks for keeping the issue in the public eye.

  • http://tim.oreilly.com Tim O'Reilly

    thacker –

    Yes. I do have children.

    And that’s why I say that the solution is not to aggregate social network data across platforms, but to give an API that allows me, the user, to access my data from all applications and platforms, and then syndicate it out to others. No one is suggesting that this data be made completely open.

    But of course, if it’s already on the internet and available without restriction, yes, mashups would be good.

    As far as stepping outside: that’s an answer to virtually any computer application, and I try to do as much of it as possible.

    But people like being connected. That’s why you see people outside listening to their ipod rather than to the birds, talking on their cellphone rather than to the person next door.

    But the pendulum swings. I do think that we will come to an end of the infatuation with online, and in fact, we already see signs of that with the hackers we profile all the time in Make magazine.

  • http://tim.oreilly.com Tim O'Reilly

    FWIW, there’s lots of backchannel telling me that Google is far from hostile to the points of view I express here, and in fact, is thinking very much along the lines that I am hoping for. So if this posting is being taken as negative, take it as negative as to this first step being insufficient, not to the ultimate direction. I’m looking forward to seeing what comes next.

  • http://boinkinchipmunks.com thacker

    O’Reilly–

    What may come next, along with an ultimate objective, is worrisome. It is not a negative reaction. It is a concern in hopes that it is addressed before a negative impact occurs. Social networks, their owners and developers do not historically have a good track record in forethought to processes and execution, let alone the ability to voluntarily take proactive steps when failure has seeped into the process, e.g. the necessary action that has been taken by the National Association of Attorneys General [NAAG] to affect certain change within social networks.

    The hard questions need to be asked. It is a responsibility of everyone within the Internet communication chain, more so for those with substantial influence.

  • http://dasht-exp-1a.com Thomas Lord

    Thacker:

    You smell “attractive nuisance” too?

    -t

  • http://blog.vortexdna.com Kaila Colbin

    I have to agree with Tim here that the key issue is putting the user back in the middle of the equation. We all know that requiring customers to stay with us is an old school model, and we mock entities like the RIAA for attempting to continue the tradition. Even so, the impulse to try to force our customers to stick around, through golden handcuffs or locked-down data or whatever, is still powerful and pervasive, as evidenced by the second comment on this thread:

    …the sites also know “its all about the data” so they likely would be less reluctant to participate and let go of all that GOLD…

    We need to understand and accept that the only sustainable way to keep customers in the future is by providing what they want and continuing to do so. It becomes a question of self-confidence. If I’m confident in my relationship, I don’t worry about my partner speaking to other women. If I’m confident about the value I give my customers, I don’t worry about other companies having access to my customers’ data (which, as Tim also points out, doesn’t belong to me anyway; it belongs to them).

    The lion’s share of any company’s focus should be how to continually increase the value it provides its customers. If OpenSocial doesn’t evolve to support that, I’m quite sure another standard will.

  • http://gowrisivaprasad.wordpress.com/ Gowri Sivaprasad

    Agree with you Tim. This is what I was referring to in the 2nd November entry in my blog.

    Setting the data free will allow the users and the developers to “automate” various activities and mashups across social networks. But will the networks set their most valuable asset free? Whats the business value if you are MySpace?

  • http://htp://blog.debreuil.com Robin Debreuil

    I sense the end of email, once social network data is open, my interests are open, my friends are known — and now my spam comes (seemingly) from a friend, mentions my work in the subject, and my interests in the body. Even people will have a hard time sorting that, it will become too painful, and end.

    Careful what you wish for, the first rule of making things better is to not make them worse. There is no way to secure that data once it is opened up. So while I can always change my email, it is harder to change my friends and preferences in life, so dumb mistakes now will follow us around for a long time.

  • http://tim.oreilly.com Tim O'Reilly

    Robin –

    I fail to see how more information leads to more spam. Once your email is out there, you get spam, tons of it, more than you can imagine, without any targeting at all. The essence of spam is that it doesn’t know anything about you except about how to reach you. If it were more targeted, it might occasionally be “not spam,” (i.e. interesting to you) but the amount has nothing to do with the relevance. Because spam costs nothing to send, there’s no need for relevance.

  • ahoving

    best line i’ve read in all this so far:
    “the meta social platform is the web”
    worst line: step outside.
    when all the networks are eventually interconnected and don’t have to be doing alot of duplicative signing up, signing in and profiling, then we’ll have more time to spend outside. ;-)

  • Rick Ringel

    It -is- about the data, which leads to the question of cashing in on the value it provides.

    From a SN provider perspective, wouldn’t open (but secure) access to the data they host bifurcate their business into a GUI component, and a social database component? I think we have a general agreement that the value is in the social database. But, the question of how to monetize that end of the business remains.

    Google’s suite of APIs indirectly drive clicks to their advertising services. It is clear they have a unique position in the ecosystem: one that turns their free backend data into revenue without ownership of the GUI. Isn’t that a revenue model that is generally unavailable to social database providers?

    Perhaps in the long term, the industry will cut out the middle man. If I publish my personal information from the context of my personal social node, then the need for central social databases goes away. Instead, SN GUI/mashup services will depend on interfaces to this social node server, and the social node can be in an unlimited number of social networks. I think this architecture would create incentive to build secure publication APIs around a social node, but frees the soical node from generating revenue. Google’s OS is applicable to social nodes.

    Tim, you are absolutly correct about who own’s the rights to use personal data. The creative energy I put into my network persona is my claim to related property rights. But, how do we create ‘property lines’ in web 2.0?

  • http://theprogrammersparadox.blogspot.com/ Paul W. Homer

    Sometimes techies loose sight of what it is that they are really trying to accomplish.

    I find facebook entertaining, but in truth it is not really important to my life in any significant way. Most of the Web 2.0 is similar, just another channel to play around in and interconnect with other people. If anything it is more a replacement for wasting time watching TV, than anything else. Pure entertainment.

    Besides goofing around, what is it that people expect these Web 2.0 tools to do for them? Open Social is only really useful if it lets the user do something useful. Personally, the only thing I really want to be able to do as a user is ‘exist’ in multiple systems at the same time, with the least amount of effort. In truth, I’m only going where my friends go, something that changes as each new fad comes and goes.

    Oddly, an interface that allows the users to easily move from one social system to another, or keep all of their data reflected in multiple systems works well for the users, but not for the vendors. A captive audience is a good thing. That’s where the money comes from.

    This seems to me to be one of those circumstances where the ‘right’ technology will never exist because of the underlying complexity of human organization. If it wasn’t for people, our technology might actually get built properly…

  • http://www.centiare.com Snerfling

    So now we’re finally getting to the crux of the issue: What are the objectives/motivations for individual & business users to list their personal/company data on a social network?

    While we can surmise that SNs provide interesting entertainment for individuals, businesses have only one goal – to increase EPS, regardless of the specific form of business.

    That means, in essence, two things: driving higher revenues and/or achieving greater efficiencies (lower costs per productive unit). Let’s focus on the sales side for now.

    The key to most sales (other than retail) is not to make an actual on-line sale, but to provide greater information about the product/service provider to the eventual consumer(s).

    3rd party mash-ups (Web 3.0?) are the key to having certain components used in a recombinant fashion so that data becomes useful information. There is absolutely no limit to the potential creativity, and number/variation, of mash-ups.

    If this is the future, then why would any business submit *their* data to a proprietary SN rather than utilize a standard mechanism in which to expose (modularize/atomize) it to the Web in its entirety?

    What this really suggests is that there isn’t a viable long-term business model for SNs. The real play is a common Web standard by which data is published specifically by *owners* to expose it to search engines and 3rd party mash-up developers in order to achieve greater visibility.

  • Ritesh Bawri

    Information is gold. The gold belongs to the individual. Applications must be created that allow individuals to own this information and share it with sites as they deem fit when they deem fit.

  • http://www.betterlabs.net chetan

    i need to get ssh login db , server directory
    on my own machine & publish to opensocial same like facebook-joynet one year free account

    should is there any provision with google?
    should i develop rails web application ?

  • http://leagueoftechnicalvoters.org Silona

    let not forget people’s rights in regards to that data and it’s usage as banks and credit cards start participating in the commoditization (if that is a word) of reputation…

  • Lars

    just read about Google’s Friend Connect, that seems to promote the cross – social network data exchange.
    Is that right? If so why can they do so?

  • Sebastian

    Hi Tim,

    from the user’s point of view I can totally understand your disappointment. But as we both know there are also other forces which want to be respected such as the social networks themselves that do not really want to share data and the law which maybe has a more critical view about sharing user data. I just want to mention that this is not only a question of implementation which would not be that complicated. Just adding some more API calls and it would work.

  • http://tim.oreilly.com Tim O'Reilly

    Sebastian,

    I hear you about the complexities. But imagine how many of those complexities go away if you give the user complete control over the data.

  • http://www.graphictruth.com Bob King

    I got here due to social networking, because the conversation was interesting to me – not the means, but the ends.

    I’m an ethics blogger, and the more I blog about ethics, and the more problems I see that could have been easily avoided simply by the choice to not misuse people (or their data) the more I realize that a major key to elegant systems on all levels is simple. “It’s the ethics, stupid.”

    I tend to write essays – here’s the conclusion to one. It has some other relevant bits, (one cited below), but in essence, it’s about the ethics of social interactions.

    But the point that’s assumed in this essay, I now realize (and thanks for that insight, y’all) is that an ethical solution also tends to be an elegant solution.

    Of course, this is not a new idea:

    …consequences of unethical decision making tend to be diffuse and unpredictable in time and space. It’s far, far more difficult to protect your interests from the blow-back of an unethical decision than to do the right thing in the first place.

    Furthermore, though the blow-back is unpredictable in time, the general “progress curve,” the rate of social and technological change we currently experience, tends to make all consequences more dire. The up-side is that ethical decision making will tend to reap far greater rewards far sooner than conventional wisdom would suggest. Still, it’s an obvious idea, one that our nation was actually based on, the idea that the fundamental unit of society was the individual.

    I’m a Constitutional conservative. Not in the Alito sense, but in the sense of someone who has a decent, if casual background in the writings of the founders and their intents – which was primarily the intent for the Government to secure the rights and liberties of individuals, in the belief that, aside from being intuitively right, ethical and morally proper, it was also the simplest way to ensure a minimax outcome – the most possible approaches to social, cultural and structural issues at the minimum global cost.

    My view of the last thirty years – which has been a view from the bottom tier of the pyramid, or close to it most of that time – has been that the presumptions of the Regan Revolution have proved to be deeply and irredeemably flawed. For, if you strip away the rhetoric, it amounted to this, “Look good, smell good, talk a good line and put your conscience in your pocket.” And we have allowed a government to grow up with those very same values – the idea that “money has no smell” and that the worth of a constituent is measured in their ability to generate campaign funds. Furthermore, we have abandoned ourselves to the tender mercies of centralized planners and social experimenters who are neither persons of good will or even of satisfactory qualification – were there any satisfactory qualification for such a thing.

    Meanwhile, back in the real world, none of that is true. In Middle America, ethics still work and honor and one’s good name matter – because if nothing else, if you cannot be trusted, there’s someone else who can be. And if you are truly criminal – well, even if the law might shield you, it may come down to an axe-handle to the kneecap.

    Or to boil it right down – ethics is the art of making correct decisions that bring the maximum potential benefit at the least possible potential cost to those impacted by the transactional costs.

    An electronic social network is a means to an end; the rules, principles and core issues are the same whether the means of networking are individuals gathering in churches, taverns and coffee-houses or interacting on “social networking sites” with various “flavors.”

    Me, I like to be able to smoke, drink and use rough language – so I much prefer the rowdy context and company found there to the rather prissily efficient facebook.

    Do you prefer to shop at Costco or Nieman-Marcus? Why? (And is it any of my business to know, other than observing that you clearly do?)

    We must generally accept that, first, much such information really IS “none of our business.”

    Furthermore, we should rejoice in that fact, for in that phrase, we realize that we really do not need to know what any particular person generally does – what we want to know is the probability of many people “beating a path” between here and there, from one node to another.

    And generally, we can infer why well enough.

    This tells us exactly how to monetize any particular “node” – it’s by providing an environment where particular transactions between persons will be advantaged, whether they be social or commercial, and by presenting them with things we know to be fairly probably of interest to them, even when they did not specifically come there for that.

    And of course, the person who designs and maintains that node is entitled to a return on that investment – which needs to be as automatic, painless and with as mandated decision trees as possible.