Google+: Not a Product In Isolation

Google+ may look a lot like other social networks where we've participated before, but it's not a product in isolation. Every piece has been designed to help push the industry in a new direction of interoperability and selective sharing. Edd Dumbill of O'Reilly gets it and breaks it down.

Much of the power of Google+ becoming the social backbone comes from intelligent focus on standards, from people who have pushed for standards long before they joined Google - helping move the industry from the inside out. Read the article.

/via My Google+ Profile.

Google's Data Liberation Front Frees Your +1s

In a new and innovative way to leverage the company's new Hangout feature as part of Google+, Google's Data Liberation Front held a small press briefing with a handful of tech journalists today, walking them through why the company is focused on leveraging open standards and helping users get their data out. Alongside the discussion, engineering manager Brian Fitzpatrick said the company has extended its exports to include the +1s you've made for Web sites - a small bump obviously, but one that demonstrates their seriousness about making getting your data out of Google as easy (if not easier) than it is to get it in.

Without speculating on other company's practices (namely Facebook), Fitzpatrick asked those of us participating if we would recommend a restaurant which locked its doors to prevent us from leaving after we sat down for a meal, or if we would recommend people rent an apartment that demanded it keep your furniture and family photos once you moved. The obvious answer is of course not... and Fitzpatrick said the same should be true for your online content. He recounted how when the Data Liberation Front first started with Blogger, there were some internal concerns at Google that users would leave the platform en masse for WordPress or other solutions, but in fact, they instead regularly downloaded content but kept posting - using the exports as local backup. (This is what I do as well)

Google Takeout Liberates Your Content from Multiple Services

Fitzpatrick pointed to Google Search as an example of not locking in one's data, as users, with many choices, will use the engine and can go to any other when they like. But with many online services, content goes in and doesn't come out. As I've demonstrated with my own personal backup of my Facebook wall and photos, you can get your data out of the social network, but it's possibly not configured very simply to move to another platform altogether. This is a main focus for the Data Liberation Front team, said Fitzpatrick, who said the effort is made to point to XML, Activity Streams and Microformats wherever possible, letting your data be interchangeable within services.

Downloading My +1s is a Mere 25KB.

Now that My +1s are Downloaded, I Can Move them Elsewhere

Lost in much of the coverage of Google+ a few weeks ago, the Data Liberation Front introduced its Google Takeout policy on the same day, making it easy to download data from multiple services and take them elsewhere. This list includes your +1s now, as well as Google Buzz, your Circles and Streams from Google+, and Picasa Web albums. As they mentioned at the time of that launch, if they make it simple for you to get your data out of the company, they have to work harder to keep you in.

Echo Prepares World for Migration to a Web of Streams

The foundation of the Web is moving from one of unchanging static pages to continually updated streams fed by external services. Gone is the time when publishing an article or a site was the final step in creating content. Instead, this can often signify the beginning of the content's life, as it is shared, retweeted, liked, commented on, and further distributed. Echo, a company that first made its name in the comments space, explained to a full hall of industry onlookers Tuesday morning that dynamic social experiences are going to rule the future of the Internet, further arguing they have placed themselves at the center of this trend, having built a flexible environment for people to tap into the activity streams and benefit from the Web's continued evolution, instead of seeing visitors capitulate to household name social networks.

Companies like Facebook, Twitter and LinkedIn have unlocked billions of dollars in revenue around user streams of content, while traditional publishers have often held to traditional business models, unable to keep pace, said Echo's CEO, Khris Loux. "Today's publishers are ill-served to serve even their own audiences," he said. "The conversations, traffic flow and revenues have moved on."

What Echo recommended was to give publishers, ecommerce sites and startups the capability to craft user-centric, stream-centric experiences themselves, tapping into activity stream data as the unit of currency to power interactive next generation sites. To accomplish this work, the Echo team has created a stream server to deliver real-time experiences that support the existing brand and content without traditional costs of real-time. It was explained the company's new stream server supports thousands of queries per second, with live updates, based on open standards, with each item passing through the stream having the opportunity to be touched with the appropriate action (such as likes, replies, flags, votes, promotes, etc.)

But rather than building out a complex product portfolio to white label or charge premium fees to customers, Echo told attendees yesterday that they work "up to the API" and partner with other startups to connect their services to end users.

"We have coded up to the API and no further," said Loux. "It's scary as we can't solve anyone's problem ourselves. But that also means our partners have as much power as we do."

Echo, since its product shift in 2009, has signed up many top name brands to its platform, some of whom presented Tuesday, including Sports Illustrated. Displayed on SI.com and other partners' sites were real-time comments, user reactions, badges and achievements, and a visually appealing image gallery for hosted graphics. Partners supplying many of these things included Realtidbits, Janrain, Klout, Badgeville, Branchnext, Clicky and Arktan, who claim to have connected to Echo through open standards, enabling customers to switch providers in the future if desired.

"It's the beginning of a new way to look at software," Loux said. The company hopes to also offer a custom application store that enables partners to connect hours of effort to dollars through showcasing their apps on the market. Executing on a planned feature that is stream-centric looks right. Echo's demos from Tuesday and well before are showing real promise about the future of the Web. Rather than shying away from competing with the Web giants, Echo and its partners are set to offer a serious alternative - keeping ownership of the entire social experience on the corporate site.

Disclosure: Chris Saad, an Echo employee, and I occasionally collaborate on the Edge Theory conversations series, and I consider him a good friend.

OpenID Foundation Finally Grabs @openid Twitter Account

The OpenID protocol, aimed to authenticate users in a decentralized way, helping individuals own their own identities instead of requiring unique accounts for every service, has had something of an ironic issue for some time - given that the OpenID Foundation didn't actually own their own ID for one of the more popular discussion forums out there - Twitter. Today, this issue can be put to rest as Facebook employee and open source advocate David Recordon announced this afternoon they had gained access to the @openid account - giving folks a central place to keep tabs on updates to the protocol.

While OpenID at times has been pilloried for being obtuse, making logins more complicated rather than less, the project has helped push the concept of decentralized identities away from individual domains, including last week's news that Flickr had adopted the standard. The protocol, developed by now Google employee Brad Fitzpatrick in 2005, while at Six Apart, has seen deployment at a significant and influential number of Web services, and has corporate members including Facebook, Google, IBM, Microsoft, PayPal and Yahoo! on the foundation's board of directors. Ironically, nobody from Twitter is involved on the board.

The release of the account may or may not have been related to yesterday's OpenID OAuth Summit, held at Facebook headquarters, which saw stakeholders from a variety of tech companies, including Kevin Marks of BT, Chris Messina and Joseph Smarr of Google and Tantek Çelik, computer scientist at Microformats.org, participate to potentially discuss the differences between the at times complimentary but competing formats. Also, the Internet Identity Workshop (IIW) kicks off today in Mountain View, where many of these same folks will be participating.

Debate on OpenID versus OAuth continues to rage on the Web, of course. Recent entries include the Halloween post on Codebase: OpenID OR OAuth – That is the Question. Regardless of the eventual resolution, proposed as a new agreement called OpenID Connect, adoption of open standards and reduced complexity is a good thing. You can now follow OpenID's happenings at @openid.

Google Buzz API Debuts, Fueling Potential Network Growth

Google Buzz celebrated its 100th day of public existence at Google I/O Wednesday with the introduction of its much-awaited API, and the immediate adoption of that API from some interesting showcase partners, including Seesmic, TweetDeck and SocialWok. In the last three months of Buzz's availability, I've discussed how Google had aimed to avoid offering proprietary standards for developers to point their applications to, choosing instead to work with open tools including OAuth, AtomPub, Activity Streams, Pubsubhubbub, RSS and WebFinger. On Google I/O's first day, the company showed the first fruits of that labor, giving the fledgling network the opportunity to tap into the same developer community that other services, including Twitter, Facebook and LinkedIn, have enjoyed.

Chris Chabot (@chabotc), a developer advocate for Google, explained the role of Buzz at Google Wednesday, saying that "The Web is better when it is social," chronicling the rise of social activity to the point where having a social profile online is as common as having a mobile phone. "The Web is becoming a social place. It is moving from a static place to one about people sharing experiences, playing, working and talking together," he said.

Initial Partners Using the Google Buzz API

But as Google's social team has chronicled several times, including Chris Messina's talk on Activity Streams at SXSW this Spring, many of the different destinations on the social Web speak different languages. Rather than inventing a new language, Google wanted to utilize existing standards to avoid such headaches, and make Buzz part of the Web's infrastructure, "not a special destination."

In April, after one particularly nutty report was issued on Buzz's aggregation statistics, we proved the majority of Buzz's active content originates on that network. The new Buzz API makes it easier for third party clients, such as Seesmic and TweetDeck to push comments, likes, rich media and other updates to the network, but also to consume content from Buzz in users' favorite application. This seemingly minor move - one already seen by Facebook and LinkedIn, puts Buzz on par with those historically larger networks, and ahead of scrappy startups like FriendFeed, which never gained such prominence.

Google Buzz Displayed In Seesmic (Next to Foursquare)

Marco Kaiser (@marco) of Seesmic, who demonstrated his company's application pulling and pushing data from Buzz, said Seesmic wants "to support as many services as possible to as many products as possible", and as of today, all of their products, including the Web version, Desktop and Android clients, let you add Buzz as an account and see Buzz content alongside Twitter and other updates. This caps a busy week for Seesmic, following their Monday addition of Foursquare and other updates.

While Twitter has had an off and on relationship with its own development community, the ease of developers creating applications and Web services has led to much of the service's growth. Buzz, assuming expanded adoption of their API by other partners, could receive the same kind of boost that so far has been lacking in the service - seemingly siloed despite so many promises to connect with the outside world.

In addition to the announcements by Seesmic and other partners, Google introduced a testbed for developers to try out the API, which you can find at http://www.buzz-demos.com/. The test site lets you display your stream, add, edit and delete content, follow and unfollow others. It surely won't be all that long until many of the automation tools that have existed in Twitter's ecosystem make their way to Buzz.

As someone who long begged for services like FriendFeed to be supported by major clients like TweetDeck and Seesmic, but didn't see it happen, the rapid adoption and seeming ease of the Buzz API is a huge endorsement for the future of the network. It could have something to do with Google's name being attached, or the ease of the standards, but the potential is there nonetheless.

Google's official post on the new Buzz API is here: Introducing the Google Buzz API. You can find me on Buzz here: http://www.google.com/profiles/louisgray.

Google Buzz Team Tweaks Tweet Imports, Hints at Filtering

As Google Buzz users discover their own approach to using the service, and being selective as to which programs they pull into the site, many have been dissatisfied with the way Buzz has handled Twitter. Unlike other aggregation services, including Cliqset and FriendFeed, Buzz's imported tweets are not imported in real-time, and can seem very disjointed - made worse by their being imported in clumps. Today, one Googler explained they were going to solve the clumping issue in what was termed a public experiment, making things "less awful" in the interim before more capabilities are released, including filtering of imports by source - a popular option seen on sites like FriendFeed and Facebook.

Why Buzz does not pull in tweets in real time is extremely curious, given Google's much-publicized relationship with Twitter, their focus of late on recency in search, and of course, Buzz's centering around standards, such as PubSubHubbub, a big player in the real-time world. That Twitter does not support PubSubHubbub is obviously one contributor, but other services have managed to find a way to get Twitter updates with practically zero delay, through leveraging Twitter's APIs. The result is a truly unsatisfactory experience, contributing to many people removing Twitter as a service feeding Buzz.

The way Buzz polls Twitter for updates now can mean older updates get pulled in as new along with more recent items. To date all "new" updates were added concurrently, but now they will be redistributed using the timestamp of the tweet from Twitter, in effect shuffling those updates throughout the network. This will make Buzz's mysterious "unread" counts even less relevant, and should be seen as a band-aid.

Updates from Twitter Now Shuffle In By Their Official Timestamp

Josh Wills, an engineer on the Buzz team, explained:

"Personally, I really like Twitter, and I want better integration between Twitter and Buzz. My hope is that this is a temporary change until we work out a way to play nicely together that is beneficial to everyone."

Josh went on to say that the changes were intended to "make things less awful" while things get "worked out" to make the tweets appear in real-time. Whether that's a business development decision or an engineering one, to make things "play nicely together" was not explained, but he added that he believed "real-time twitter updates don't make sense w/o the highly-anticipated source filtering."

The ability to hide sources, such as Twitter, from one's feed in Buzz will be as important as it has become to hide services in FriendFeed, or to stop seeing Farmville updates in Facebook, should you not be interested. While Buzz has been fairly close-lipped in terms of talking about future features, Josh's comment referring to the filtering shows it's on their radar, and no doubt coming, rather than being speculated about endlessly by hopeful users.

Older Tweets that do enter Buzz can of course be made more prominent if they gain conversation, and therefore, get bumped to the top of your feed - but from what I've seen, most engagement comes on native items on the network.

EdgeTheory: Is iPad Mania Fully Justified?

EdgeTheory Conversation 7 is about the big news of the last week – Apple’s latest digital device, the iPad. While millions of words and thousands of articles have already been written about the iPad, there’s always more to find, especially as the product spends more time in the hands of early adopters. This week, Chris Saad and I talk about whether the closed nature of the iPad (when contrasted with open standard solutions) is something we should be avoiding, and where the iPad fits in a world of laptops and the iPhone.

Are users getting screwed by needing to buy multiple versions of applications with the iPad’s debut? And should it be a mobile device, or a fully functional computer? We talk about this and much more in another EdgeTheory Conversation.

This and all other EdgeTheory conversations can be found on the dedicated ET Conversations site.

Listen in below:

Cliqset, Status.net Support Salmon for Comments Migration

As Marshall Kirkpatrick noted in a post on ReadWriteWeb this morning, the Salmon Protocol project we first introduced back in October of 2009 looks like it is progressing beyond the planning stages, as it has been integrated in two small, but influential, social networking sites, Status.net and Cliqset - both of whom are strong open standards advocates laboring away in the shadows of larger communities.

The Salmon Protocol, which aims to define a standard protocol for comments and gestures (such as likes) to swim upstream to the originating blog, hopes to unify conversations across diverse locations. It's planned for adoption by Google Buzz (See: Designing Buzz for a Google-Free World) and solves the problem that first blew up back in 2008 around fractured conversations.

While many people, including myself, have adapted to a new world whereby conversations take place in a wide array of communities, it would be nice for the content originator to have one centralized location to see all downstream activity.

Evan Prodromou of StatusNet mentions integration of Salmon's digital signatures in a blog post from Sunday, where he also notes support for Activity Streams encoded in RSS 2.0 and Atom. The move, paralleled by deployment at Cliqset, should be what I hope is the start of a lot more announcements, from small and big companies alike, to make Salmon a reality. The next big target, which I am biased in favor of, obviously, would be for Disqus to integrate with Salmon and pull comments upstream to this blog.

I managed to find time to talk with Darren Bounds of Cliqset at SXSW earlier this month. In our quick discussion, recorded on CinchCast, you can find his comments on their support of open standards, and that network's direction - which could hint at why they're an early adopter of Salmon.

Open Identity Exchange Proposes Identity Trust Framework

Today, at the RSA conference, the Open Identity Exchange (OIX), aimed to increase trust in online identities, and backed by the OpenID and Information Card Foundations, announced its inception. In parallel, the U.S. Government is recognizing multiple technology companies as meeting federal standards for identity assurance, including Google, PayPal and Equifax, essentially securing users' ability to register and log in at federal Web sites with credentials from each of those services.

Goals of the Open Identity Exchange include building online users' trust and confidence in the exchange of identity credentials, standardizing these interactions and reducing hassle with online logins, registrations and purchases. As practically any Web user knows, frustrations with remembering scads of online user names and passwords, each corresponding with different sites with varying trust levels, can be a complete pain - no matter how much effort is taken to standardize, and the alternative, keeping one password for multiple services, which many do, has many more problems of its own.

OIX and its members are looking to reduce the problems with today's Web and move toward further highlighting open standards. Founding members of OIX, a non-profit corporation, include Booz Allen Hamilton, CA, Equifax, Google, PayPal, Verisign and Verizon.

The Often Complicated Process of Assessing Trusted Identity Online

Google's participation in the exchange follows the company's hirings of some of the more vocal advocates of OpenID and the open movement in general, including Chris Messina and Joseph Smarr. Earlier this week, a Google spokesperson wrote by e-mail that the inclusion of the company as part of OIX's launch should not come as much of a surprise.

"As you probably know, Google has long supported and contributed to the development of identity standards such as OpenID and OAuth, largely in order to increase online security by reducing the reliance on password use across websites." they wrote.

A white paper on the new OIX Web site, entitled "An Open Market Solution for Online Identity Assurance", explains how open identity technologies, including OpenID and Information Cards, serve to take closed user name and password systems deployed by most Web sites and expand them to accept identities issued by other parties, such as Google, PayPal and Equifax. Much of the paper, and OIX's mission, centers around the issues surrounding identity, including social, business, legal and emotional, such as trust.

This model of trust is explained in a second piece which defines a new "Open Identity Trust Framework (OITF)". The OITF paper shows holes in today's trust frameworks, and questions how people passing along personally identifiable information can be sure their data is protected with acceptable technical, operational and legal safeguards, while proposing a structured role for policymakers, providers, assessors, auditors, and dispute resolvers, to be sure that all participants are acting in a trusted manner. It may seem overly bureaucratic, but considering the Federal government needs to accept its findings, process is a good thing.

Lest you think this just yet another association or bureaucracy with talking heads looking to grease the skids of online growth, see the conclusion of the OITF model paper, where the authors explain a data utopia: "
Imagine 
that 
the 
OITF 
model
 takes
 off
 and
 identity 
aspects
 of 
all 
digital 
communications 
become
 reliant 
on 
this 
new
 layer 
of 
the
 Internet. 
Society 
could 
become
 dependent 
on 
this 
type 
of 
infrastructure 
for
collective 
action. 
The 
authors
 want
 to
 make
 it 
clear
 that 
trust 
frameworks
 for 
identity 
information 
portend 
to 
be
 so
 important 
for 
the
 future 
information 
society 
that
 they
 warrant 
extensive 
scrutiny, 
participation, 
and
feedback
 from
 a
 wide
 representation 
of 
stakeholders.
"

You can find out more on this new exchange at http://openidentityexchange.org. In addition, Google posted on the announcement on the company's online security blog: Federal Support for Federated Login

The Future: Operating System And Application-Neutral Data

We are now growing accustomed to the concept of the "cloud", where our data will be increasingly stored in Web services, not on local disk, accessible from any computer, operating system or browser. But, despite the adoption of standards from major players storing our personal data, the choice of services causes serious vendor lock-in, as the data suite, be it from Microsoft, Google, Apple or other providers, is not only interpreted by their offerings, but stored there as well. This storage and management of our data makes migration between services incredibly difficult, and still leaves us at the mercy of a large company, whose priorities may not be the same as our own.

The time has come to start on a path to true ownership of data by the individual, reducing applications and Web services to the role of filters and containers, rather than hosts, who can propagate lock-in as these services spread to mobile devices and tablets from their desktop roots.

The Potential For Owning Your Data Outright

What makes a digital device mine, be it a laptop or a cellphone or an music player, is the personal content that is stored, and how that data is translated, stored, presented and categorized. Similarly, when we make a choice as to our preferred Web services or technology providers, we are, for the most part, passing our content to them exclusively. While we may have made the data location independent, it is far from being service independent, and any potential future switching will have dramatic impacts on time and productivity, including:

• Complicated export and import of personal data
  
• Differences in the interpretation of data between similar applications
  
• Potential loss of metadata between services
  
• Reduced backup stability as differing instances of our data is housed at differing services

This headache is a major part of vendor lock-in. Today, when I make a choice as to what brand computer to buy, or what phone to purchase, while I may be committing to a brand or a suite of applications, all I am really doing is asking this product to provide its own interpretation of my data, including:

• My Contacts and Relationships
  
• My Music Files
  
• My Videos
  
• My Photos
  
• My E-mails and Hierarchy
  
• My Documents
  
• My Bookmarks and Hierarchy

Increasingly more important than the actual data itself is its metadata, or data around data. How is the data structured, meaning... Do I have my e-mails in folders with subfolders and rules? Do I have photos in specific albums? Do I know how often I have played a specific song or genre? When were documents created or last edited?

Today, for the most part, we are choosing from three major service providers, although there are alternatives. We can select Google, who offers Google Contacts, Calendar, YouTube, Picasa, GMail, Google Apps and Google Chrome for the majority of our needs. We can, instead, select Apple and use Address Book, iTunes, iPhoto, Mail, and Safari. Or, we can stick with Microsoft, and leverage Outlook, Windows Media Player, and Internet Explorer. (Or their online equivalents)

Despite standards adoption, not all programs interoperate well. No doubt there remain issues with meetings from Microsoft Exchange being received by Apple Mail, and the integration of Web browser bookmarks and Web history is not shared between Google Chrome and Safari. These minor problems are greatly magnified when you consider the potential for future switching, as you migrate from one platform to another or one computer to another - largely because in every case, the applications themselves, even those that are Web services, are storing our data for us, and interpreting it in their own way.

I think it is time for a change, that lets us own our own data, turning the situation on its head.

Instead of hosting our own data with the service provider of the day, we should host our own data in a standard format, which will be adopted by the leading providers, whose applications will tap into us directly, and pull down our data and its metadata. If I chose to log in with GMail one day, I would authenticate who I was, and GMail would pull down my e-mail stream, complete with e-mail activity history (such as replies and forwards). The data would not be stored on Gmail, but instead be more like a read-only process, whereby changes to data, including sent items, would not be stored in GMail, but written back to my personal "cloud", if you will. Similarly, if I opted to log in to Microsoft Outlook, tapping in to my own, authenticated, account, I could browse my contacts in their application, through their filter, but the data would reside with me.

Hosting one's own personal cloud with our own data is not an end run around large corporations in fear of Big Brother, but instead, for real, true, portability. In this situation, a longtime iPhone user could pick up an Android phone, enter my own personal ID (be it through OpenID or some other standard), and pull down my details into all of Google's native applications. Similarly, I could log in to any Microsoft, Apple or Google powered device and become me, not with my data hosted on the new machine, but with my data being read, like a Web page, on that device, in their own lens.

Even as we on the Web are rallying around these concepts of standards, and the cloud, we are seeing the concept of vendor lock-in be as true as ever. The switching costs from hardware device to hardware device, OS to OS and Web service to Web service remain completely too high, and the way around this problem is to take back our data, make it personal, and enforce standards that get the major players to come aboard. While we may not all have all the broadband access necessary to make this a solution today, it's 2010, and we should be well beyond the same issues we have been facing in computing for the last 20 years.

So how do we make this happen?

Is There a Looming Battle Over OAuth's Successor?

The OAuth protocol, used on many popular Web sites and applications to pass your credentials between sites without requiring the entry of your user name and password, including Twitter, is potentially under pressure from a team of techies representing Microsoft, Google and Yahoo!, who have introduced a competing specification interpreted as being aimed to succeed OAuth, called Web Resource Authorization Protocol, or WRAP. Eran Hammer-Lahav, the Director of Standards Development at Yahoo!, who helped coordinate many OAuth contributions, and created a formal specification for the initial OAuth standard, recently panned the move, saying, "The road to hell is paved with good intentions," adding his own proposal for OAuth 2.0, which he hopes will better separate between authentication and authorization.

Today's OAuth standard is known to have its imperfections. Hammer-Lahav notes in his 2.0 proposal that OAuth is essentially "unusable" for mobile devices or installed apps, and also suggests that OAuth "does not adequately support large providers". But he says the move to create WRAP has confused developers' focus, and diverted resources, calling it "just one illustration of the demise of the OAuth community".

But his opinion, unsurprisingly, is not universally accepted. David Recordon of Facebook, also on the boards of the OpenID and Open Web Foundations, states in the comments of the post that Facebook is not supporting OAuth 1.0 as it is simply too heavy - requiring a massive increase in HTTP requests, also adding that other developers find OAuth "too difficult to correctly implement".

David followed on to his initial comments with a post to the IETF mailing list, which you can see here: Facebook, OAuth, and WRAP. In the note, he highlights the belief that the proposed WRAP alternative maps well to the company's current authentication process, adding WRAP simplifies the development community's learning curve.

The discussion, which is ongoing, may end up splintering development communities between sticking with the current version of OAuth 1.0, looking at WRAP as an alternative, or trying to support a new OAuth 2.0, as specified by Hammer-Lahav. But if you have wondered why Facebook Connect acts one way and Twitter OAuth acts another way, it's because they are different approaches entirely. If this discussion is any indication, one can expect there to be continued divergence, rather than a single way to deliver user authentication and authority between sites and applications in the future on the Web.

For another viewpoint on this broad topic, see Jesse Stay's post: The Future Has No Log In Button. Also, DeWitt Clinton of Google, on FriendFeed, says the open discussion "is good".