Once Facebook got wind of what he was doing it cancelled his account and then got to work removing his account from the service. I ran a search and it seems there are at least two accounts in his name still active although odds are those are not his accounts.
The feature he was using was being tested by a couple alpha testers who seem to have been subject to a non disclosure agreement and couldn’t disclose what they were using. Scoble was released from his NDA and gave some information about the feature on his blog. He said the new (and possibly never to see the light of day) feature works a little like the address book importers Facebook and other services use:
What does it collect?
Names and email address and birthday.
Why those? Because it’s trying to connect Facebook names with names in its database.
For instance, it learned that of the 5,000 people in my Facebook account about 1,800 were already on Plaxo.
It did NOT look at anything else. Just this stuff, no social graph data. No personal information.
There are a couple problems with this though. Facebook advised Scoble that running the script was a violation of its terms of service and possibly illegal in terms of US federal and state law. Another issue is that Plaxo knew this would not go down well with Facebook and apparently told its alpha testers and Mike Arrington that there was a chance Facebook would close off access to the service. If they knew this would be a problem and could compromise its users then why do it? I can understand the desire to facilitate portability between Facebook and Pulse but I think that should come from Facebookand not be forced by Plaxo. If Facebook wishes to remain relatively closed then let it. I use Plaxo Pulse, Jaiku and similar services because they are open, not because I am hoping that I can one day export my Facebook contacts to Plaxo.
This is one of those times where if what you are about to do doesn’t quite sit right with you then perhaps you shouldn’t do it. I am a little disappointed that Plaxo adopted this approach and would have preferred that they either work something out with Facebook in advance or not at all. That being said, I don’t share Paul Walsh’s sentiment and will continue using Plaxo’s core service and Pulse. The team there have been fantastic to me and I dig the services. While this feature may well facilitate a violation of Facebook users’ privacy (John McCrea and Joseph Smarr from Plaxo told Jeremiah Owyang that Facebook users explicitly consent to share their email addresses but I think there is more to it than that), Facebook is not quite as pure as the driven snow. I do agree with Jeremiah thoughts about exporting data and the social contract between friends on these services:
Anyways, here’s my opinion: Robert nor Plaxo don’t have the right to scrape these addresses without the explicit consent of the friends of Robert. I’m one of them, and I’d like to have an opt-in before Robert scrapes my data and exports them anywhere.
There’s a unsaid social contract that if you become my friend, I expect you to use my information that I share with you justly, and most of all, be considerate. On the other hand, what’s happening is just the taste of what’s to come, this is the year where the social graph will become portable, whether containers like Facebook or Plaxo allow it, or other tools come around to scrape it.
Plaxo and Dave Winer have commented that this issue is about portability of your data from one service to another and how Facebook’s reluctance to facilitate this more fully pretty much means that if others have the means to force this, then why not? My take on this is that if Facebook won’t allow me to export my friends list to something like Pulse or Orkut then I am less inclined to spend as much time on Facebook. That being said the way I reached out to my friends on Facebook is the same way I reached out to my friends on Pulse, Orkut and LinkedIn – my address book. In a way I actually don’t need Facebook to facilitate this sort of sharing but that is more because of the way I connect to my friends. My address book tends to mirror my friends lists. Perhaps Facebook could give users the option of publishing a vcard? That would feed into my address book and, at the same time, would suit Plaxo just fine given that I am a Plaxo user.
There is another little thing I haven’t seen anyone mention yet. Remember this piece? Perhaps Scoble and Smarr should eat a little of their own dogfood.
A Bill of Rights for Users of the Social Web
Authored by Joseph Smarr, Marc Canter, Robert Scoble, and Michael Arrington
September 4, 2007
We publicly assert that all users of the social web are entitled to certain fundamental rights, specifically:
Ownership of their own personal information, including:
- their own profile data
- the list of people they are connected to
- the activity stream of content they create;
- Control of whether and how such personal information is shared with others; and
- Freedom to grant persistent access to their personal information to trusted external sites.
Sites supporting these rights shall:
- Allow their users to syndicate their own profile data, their friends list, and the data that’s shared with them via the service, using a persistent URL or API token and open data formats;
- Allow their users to syndicate their own stream of activity outside the site;
- Allow their users to link from their profile pages to external identifiers in a public way; and
- Allow their users to discover who else they know is also on their site, using the same external identifiers made available for lookup within the service.
As far as Plaxo goes, I hope that this is the last time they try something like this. For one thing it does leave a bad taste in my mouth and it is also negative publicity they probably don’t need. Rather let Scoble become an evangelist for their service and attract users to Plaxo from services like Facebook than drag them across and hope they stick around. Just a thought.
Tags: scoble, scoblegate, plaxo, pulse, facebook, privacy, scraping, screen scraping, friends, friends lists
What do you think?