Log in

No account? Create an account

Previous Entry | Next Entry

OpenSocial Isn't the Entire Answer

Tim O'Reilly writes on O'Reilly Radar:
While I like the direction of Google OpenSocial, not only may Google be too late, as Mark argues, I don't think they go far enough. A framework and a set of Google Gadgets for building "social applications" misses the point. We don't want to build more applications that look like Facebook applications. It isn't about a social UI. It's about deeper re-use of social data to enliven any application. Some of those applications may have a minimal UI, like Google's breakthrough search app. OpenSocial doesn't give us any of that. Ajax widgets are a halfway house, an attempt to sandbox the kinds of applications that can be created.

I completely agree. Don't get me wrong, I think Google OpenSocial is a great step forward to allowing gadget style applications (ala Facebook apps) to be run in a distributed environment. At the end of the day I don't want every application I interact with on the web to be via a mammoth social network. I do see OpenSocial adding value of being able to integrate small common applications across social networks. It scares me though as it seems like a slippery slope if web entrepreneurs no longer think about building standalone services, but only those that ride on the back of these networks. If the Web is the platform, then how is moving from one large silo to a few large silos that much better?

As I'll be discussing in one of my talks at Web 2.0 Expo Berlin this week, OpenSocial is a great step forward, but as Tim says it doesn't go nearly far enough. I want APIs powered by OpenID and OAuth that let me control what services see which of my friends. I want to be able to interact with my friends all over the web; I want distributed applications (Dopplr, LiveJournal, Twitter, etc) to have controlled access to my friends from the mammoth social networks.


( 6 comments — Leave a comment )
Nov. 5th, 2007 07:33 pm (UTC)
Thanks! That´s exactly why I´m disappointed with OpenSocial so I hope that something more will come in near future.
Nov. 5th, 2007 07:52 pm (UTC)
URL as identifier
I have to agree. While it's a step forward, I don't think the direction is correct. Some remarks:

1. Person ID (as I understand it) is an URL of user profile in SN. That's great, but OpenID strength is not that URL's were chosen as user ID's, but in user control upon URLs. It would be great id MySpace and other "launch partners" were forced to give people more control, to allow them to delegate an OpenID (using meta tag or XRDS file).

2. OpenSocial divides whole space in two area: "containers" and "application" (much like "operating system" and "application". I'm very interested if "containers" won't be perfectly interchangeable thus obsolete. What if I could use the same apps in MySpace and LinkedIn? Where the difference would be? In "branding?" So I expect much tension between "containers" and "application" providers (we saw it before in OS space, the MS trial on embedding an internet browser is a good example when Netscape (app provider) fought with MS ("container" provider). OS serves many purposes unlike "container". What's the minimal container functionality? Add/delete relationships and let users embed apps. That's pretty simple, there's not much area for improvement. And perhaps more technically inclined people will install their own SN nodes (like some of them install standalone weblogs on Wordpress and Movable Type).

That's my thoughts :) Pls, pass my regards to all OpenID Europe Foundation ppl in Berlin, especially to Snorri :)

Marcin Jagodziński
Nov. 5th, 2007 08:41 pm (UTC)
API Standards are not enough
Thanks for this post David - I've been surprised by the lack of comments around both data ownership/control and data standards. As I understand OpenSocial it just standardizes the calls, not the format of the data itself so there is still a lot of clunkiness and friction that OpenSocial does not seem to address. On top of that, as a user will all of these social networks allow me to define which of their partners get my data? This could cause huge issues if not transparently implemented.

It is definitely a step in the right direction but as a user it makes me nervous because there is so much that is not defined around the protection and storage of my data.

Rachel Happe, IDC
Nov. 6th, 2007 03:35 pm (UTC)
With the OAuth enabled RESTful API you can start to see cross container communications finally work. Right now it's just a series of email scraping toolkits that try to pull you into one network or another.

Is it the best way? Very likely not, however if implemented by all parties it will go a long way to making sites interoperate well while giving the user the necessary controls over where and how their information is used.

Additionally with Open Social you can finally get the guy in the garage running applications on their own open source social network container. Try running your FB apps on anything approaching that!

I don't think everyone grasps what this can enable. There is already over 850 people signed up for the open social dev group in a matter of days. There's momentum and a lot of developer time that can be put to good use to make some of these dreams a reality for real world users, not just the geeks that care about this stuff.
Nov. 6th, 2007 04:27 pm (UTC)
Re: Disagree
I don't think we're actually disagreeing. Today OpenSocial doesn't give standard OAuth enabled RESTful APIs. I know there is talk of it and it will be really great if it does.

What scares me is if everything just becomes a gadget atop existing social networks. I rather see the model as distributed applications using APIs, data formats, etc to mash data together.
Nov. 9th, 2007 12:49 am (UTC)
I have a dream... ;-)
that one day nearly ever website is sort of a social network in itself, I can see whom of my friends are on there, I can provide access to certain groups of friends for that site (or use a default), I am logged in with a central id and I can extract any data I leave there again. All of this with as less interaction with the actual system as possible.

And I agree that OpenSocial is really addressing a sort of different problem and I see the step forward mostly in the political area in that companies start to talk to each other regarding standards.

( 6 comments — Leave a comment )