Tag Archives: FacebookConnect

8 Minute Abs? Try 8 Minute Facebook Connect!

When I wrote about the launch of Facebook Connect and Google Friend Connect last week, I tried to explain the similarities and the differences. One difference I suggested was ease of implementation. I knew that Google Friend Connect was really easy, that they had focused on “cut-and-paste a few lines of javascript”. But here’s what I said about Facebook Connect:

A second difference is target market. Facebook has clearly focused on major sites, like Digg, Hulu, and CitySearch, and while simple implementations can be done with very little coding, most will involve a bit more complex development.

Last evening, I learned otherwise, watching a video whose title says it all, “Add Facebook Connect to Your Blog in 8 Minutes“. Engineers from the Facebook Connect team, Luke Shepard and Wei Zhu, take you through the process with a great, straight-forward guide to implementing Facebook Connect. I’m not a developer, but I found the presentation really straight-forward. Great job, Luke and Wei! (Hey, Luke, you are welcome as a guest on The Social Web TV any time.)

8 Minute Facebook Connect

(I would have embedded the video here, but hosted WordPress has severe limitations around what can or cannot be embedded. Lame.)

So, it really will be interesting to watch adoption of Facebook Connect and Google Friend Connect in the “long tail” of the Web. I wonder if we’ll start to see directories that can let us know all the sites that implement either. Would be fun to check them out and to compare and contrast.

Anyway, I stand corrected. Congrats to the Facebook Connect team on a great rollout. I note that CNET is live today with their integration.

Tagged , ,

Birth of the Social Web: Facebook Connect and Google Friend Connect Now Available to All

December 4, 2008. Today may be remembered as the birth of the Social Web, as two major projects aimed at turning the Web social emerged from their restricted beta periods for general availability, Facebook Connect and Google Friend Connect. Together, these two major events sound the death knell for the walled garden phase of social networking. Early reactions to the news are quick to frame this as a head-to-head battle between Google and Facebook, but the truth requires a look at the details, and I think something much more profound is happening…

First, the similarities. Facebook Connect and Google Friend Connect share the same basic vision of the Social Web. Any site can become social, without having to build up its own social network. Users should be able to access those social features without having to experience the pain of usernames, passwords, uploading a photo, filling out a profile, importing an address book, and re-friending the people they’ve already connected with elsewhere. And, activity streams out to web-wide lifestream aggregators should become important engines of social discovery and growth for the site.

Now to the differences. One major difference between these two offerings is the technology under the hood. Google Friend Connect is built on the “open stack,” leveraging building blocks like OpenID, OAuth, and OpenSocial, whereas Facebook Connect is built on Facebook’s proprietary stack. A second difference is target market. Facebook has clearly focused on major sites, like Digg, Hulu, and CitySearch, and while simple implementations can be done with very little coding, most will involve a bit more complex development. Google, in contrast, has explicitly targeted the “long tail” of the web, sites that would never dream of writing their own social code; the focus of Google Friend Connect is to help these sites become social by cutting-and-pasting a few lines of javascript. The third major difference is one of strategy. Facebook Connect is all about making Facebook more useful to its users all over the Web. Google Friend Connect, on the other hand, is all about making the Web more social, with an approach that incorporates other social networks. For example, the current release integrates not only Orkut, but also Plaxo. (And recall that the earliest version also included Facebook, until Facebook shut that down.)

I’ve been playing around with Facebook Connect and Google Friend Connect all along the way while these services were being carefully tested and refined prior to today’s formal rollout. I like them both, but see lots of room for improvement. But that’s to be expected; this is a major shift in how the Web will work, and there’s a lot of complexity under the covers. Today marks the birth of the Social Web, and we should expect to see lots of rapid progress for this newborn.

For those who haven’t checked out Google Friend Connect yet, I’m including a few screenshots…

Signing up via Google Friend Connect

Turning on Sharing to Plaxo

Signed in with a single click

Activity shows up in Plaxo!
My activity showing up in Plaxo

Oh, and you can check out the “Dive Bomber” site I used for these screenshots here.

Update: I just realized that I can now declare victory on the prediction I made for 2008, a prediction I made on December 6, 2007!

Update: The new episode of The Social Web TV is now up, with Joseph Smarr and I addressing the question, “Facebook Connect vs. OpenID?”:

Tagged , , , , ,

Announcing Episode 15: “On Location at Yahoo!”

We’ve just posted this week’s episode of The Social Web TV, “On location at Yahoo!” I chat with special guest, Allen Tom, Architect, Yahoo! Membership, about last week’s Y!OS rollout and this week’s historic OpenID/OAuth UX Summit.

And, in case you missed it, I did a guest post on the topic at TechCrunchIT, entitled, Facebook Connect and OpenID Relationship Status: “It’s Complicated”.

Oh, and Dare Obasanjo wrote a response to my post, entitled Some Thoughts on OpenID vs. Facebook Connect

Tagged , , , , , , , ,

Live Blogging the OpenID/OAuth UX Summit

From the OpenID/OAuth UX Summit

I’m at Yahoo for the OpenID/OAuth UX Summit. The room is packed with 40 or so folks. Companies with representation include Yahoo, Google, Microsoft, Facebook, MySpace, Plaxo, AOL, SixApart, JanRain, Vidoop, Chi.mp, and Magnolia, and projects including Internet2 and DiSo. The Summit is a response to recent usability studies by Yahoo and Google that show the current state-of-affairs with OpenID and OAuth is quite poor, and we need together to find a user experience for the “open stack” that works for consumers.

I’ll be sharing observations over the course of the day.

First up: Facebook’s Julie Zhuo, sharing experience from Facebook Connect. Idea originated in 2006 with the Facebook API. Initial version didn’t have any flow back to Facebook. Clunkiness of UI. One question for the Facebook Connect UI: How much text is really needed? Showing evolution of the UI to address the fundamental question, “What is Facebook Connect?” Final version includes user’s profile photo (if user is logged in), and thumbnails for both Facebook and the site user wants to connect.

From the OpenID/OAuth UX Summit

Good discussion about what usability revealed, about informed consent and user confusion, and about whether this passes EU privacy laws. (Answer: yes.) Facebook research showed that users had little or no understanding or savvy about phishing and URLs.

By the way, I have to say it — great to see not only is Facebook attending this “open stack” summit, but that they’ve got four people here (including Dave Morin, Josh Elman, and Mike Vernal) and leading the opening session! That’s awesome.

From the OpenID/OAuth UX Summit

Now talking about the Connect Button. First version had tagline “Bring your friends,” but users didn’t know it was a button. Second version said “Register.” Third version said “Connect” and experimented with the user’s profile photo on the button. Final version is just the Facebook “f” and “Connect” or “Connect with Facebook”.

Discussing logout options: unified, per-site, hybrid. Unified is secure, but unintuitive. Per-site is intuitive, but not necessarily secure. Chose unified out of security. Question for the future, if Facebook Connect takes off, may be strange to log out from one site and be instantly logged out of Facebook and all other Connect sites. A good laugh, as Joseph Smarr suggests a slightly more complex alternative. Julie says, “But then you’d make the user have to think.” Joseph’s aside, “That’s spoken like a true mainstream consumer site.” Incredibly active session. Key takeways slide: streamlined login is important. Explain what is going on. Err on the side of security. Flexibility is important.

Next up: Max Engel of MySpace. “The Hybrid Login: OpenID and OAuth.” MySpace will support OpenID, OAuth, and a hybrid of the two. Will use a pop-up iframe. Allows the user to stay in context. Max is showing screens of the experience they are planning. Every MySpace user has a vanity URL, which will be their OpenID. Still trying to figure out whether to support logging with just “MySpace.com”. Key design elements will be similar to Facebook Connect.

Data types: content, address book, registration, profile, friends, activity. Big laugh as Max shows the original OAuth screen, that has so much fine print that it looks like it was designed by a lawyer! Lots of discussion about whether email address should be passed to the site. Why it matters: not just for communicating, but also to avoid duplicate account problem Plaxo has experienced as an OpenID Relying Party and Yahoo OpenID. Chris Messina advancing the idea of email address as OpenID, something under consideration for OpenID 2.1.

Max revisiting that MySpace Data Availability originally was to have zero cacheability of the data, which was not going to fly with anyone. Now planning a “portable profile” plus some cacheable MySpace-specific data. Allen Tom of Yahoo raises the point that the “cacheable” data is all on public pages already, so why not just mark it up with microformats and remove the caching restriction. “If Relying Parties don’t get the data they need, OpenID only creates complexity.” Max just mentioned Portable Contacts in his presentation. Drink!

So many tough questions about complexity and confusion vs. simplicity but lack of clear, informed consent. Good discussion about whether participating sites can use the profile data they pull in to do targeting (including ad targeting). Facebook team says that they allow the site to use the data for targeting on the site, but not to redistribute the data (to an ad network, for example). Makes sense.

Max says that the sell to major websites is much stronger for combination of OpenID, OAuth, XRDS-Simple, Portable Contacts, and OpenSocial. Question from the back of the room, “What do you call all of that?” Answer popping up from Max, Joseph, Chris Messina, and me, “The ‘Open Stack’!”

Rising chorus for coming together to develop a common UI spec for OpenID. A call for five volunteers. Hands raised include Chris Messina (Vidoop), Joseph Smarr (Plaxo), Eric Sachs (Google), Max Engel (MySpace), and, drumroll, Julie Zhuo (Facebook). That’s great!

LUNCH BREAK

Next up: Allen Tom of Yahoo. Over 300 million users have an OpenID from Yahoo. Question shouted, “How many have used it?” Answer: “It has exceeded our expectations.” 😉 But, yes, we’re all here because we know we need to improve the user exerience.

Launched BBAuth in 2006. Showing “Find Friends” on Facebook and LinkedIn, using BBAuth. BBAuth and OAuth is to grant long-lived credentials to third-party sites. “Cannot allow weaker credentials to be used to mint stronger credentials.” Talking about various security considerations. Login screen must never be framed. Anti-phishing sign-in seal must always be displayed.

Allen now showing the “scary screen” which users are shown to approve access via BBAuth. *Lots* of small print legaleze. “Based on the feedback on BBAuth, we changed our approach on OAuth, which is what we’ll be using going forward.” Now, been spending a lot of time looking at and talking through the OAuth permissions screen.

Allen now showing and talking about Yahoo’s implementation of OpenID. It is *much* improved over the version they went out the door with (shrinking 14 steps to two). Allen shares that “machine-generated” OpenID URLs have proven *way* more popular than user-selected. Surprised reactions.

Talking now about Plaxo’s experience as an OpenID Relying Party. The business rationale, the philosophical view, and the admission that OpenID experience is not yet today a clear net positive to the key metrics. But Plaxo remains optimistic that the situation can improve dramatically with what’s being discussed here today.

Next up, Magnolia’s Larry Halfft. They’ve used OpenID as a key part of their strategy to reduce spam accounts and have been generally pleased with the results.

Now, Eric Sachs of Google, who just showed what I think is the first public demo of Google as an OpenID Provider. Giving context: SaaS vendors get asked to be a SAML RP for enterprise IDPs. In parallel, Google Checkout folks had questions/issues with login. Giving examples of login on Buy.com and Amazon.com, as an inspiration for a new/better? login experience for OpenID/OAuth. Now the challenge of desktop apps and OAuth. Seems like “No, help me sign in” is the key verbiage of this new “LSO” login model Eric is advocating. Now Google Accounts vs. accounts for Google AppsForYourDomain. Downside to this LSO login approach is that it does not work well for IDPs who are not email providers.

Lots of good-natured joking as we try to do a demo, that requires a Windows computer with .Net and IE as the default browser. Not easy to find in this crowd!

It’s 3:00pm. We’ve now finished the formal agenda and are discussing how folks would like to organize the last two hours.

It’s almost 4:00pm. Joseph Smarr of Plaxo is demoing the “Open Stack” end-to-end stuff that was developed by JanRain for the Portable Contacts Summit. OpenID, OAuth, XRDS-Simple, and Portable Contacts working together to enable simple and secure sign-up with access to user’s profile and address book. Good discussion underway. Joseph now explaining XRDS-Simple and answering a lot of questions.

Joseph Smarr demoing the "Open Stack"

Chris Messina now leading a discussion about the proposal to extend the OpenID spec to allow email addresses as OpenIDs. Mike Jones of Microsoft asserts this creates a major security vulnerability. Discussion underway.

Some discussion of how to handle if the Provider site is down. Mike Vernal of Facebook responding to that question vis-a-vis Facebook Connect. Good response.

5:15. That’s a wrap. What a great day. The UX working group got a bit larger at the end, which is good. Eager to see what they come up with!

Tagged , , , , , , ,