• Print

Scaling Google+

Joseph Smarr of Google+ on early lessons, an API, and pseudonyms.

Last week at OSCON, Google social web engineer Joseph Smarr sat down for an interview with me about Google+, the long-awaited social network that the search engine giant launched earlier this summer.

We covered a lot of ground during the interview. Smarr connected what he and others learned from Plaxo Pulse, where he was the CTO, to how the Circles tool in Google+ builds granular control into public and private sharing. He also said scalability comes in different flavors — it’s not just about infrastructure, but rapidly scaling the user interface with feedback. Finally, we talked about the future of the Google+ platform and the possibilities of an API (more on that below).

When asked about what surprised him the most, Smarr pointed to the high rate of public sharing on Google+, versus how the social network had been used internally at Google before launch. “People are getting these incredibly high engagement discussions,” he said.

The Google+ API

Smarr said the Google+ team is spending a lot of time thinking about an API, drawing from what they learned from Google Buzz and the experiences of other Internet companies building platforms.

“On the one hand, clearly the goal is to not just have another social network, to really help not only Google’s products but to make the web in general more social, more open, more connected, and APIs are a crucial piece of that,” said Smarr. “We actually got a far way along the road with the Buzz APIs, and not only having a lot of access to the activities and the graph and so forth, but with a lot of these modern standards, like pubsubhubbub and Webfinger and so forth. That’s the style of thing we’d like to bring. “

The challenge for the Google+ team working on the API, as Smarr explained, is that the devil is in the details.

“One of things people seem to really like about Google+ right now is it’s 100% authentic,” Smarr said. “Every piece of content was created by a real person sitting in front of Google+ and deciding who to share with. Balancing the obvious need to get more content in and out from more sources while maintaining that authenticity is something that we’re spending a lot of time playing and iterating and coming up with. I think you’ll see things trickle out over time as we get bits and pieces we’re happy with.”

Web 2.0 Summit, being held October 17-19 in San Francisco, will examine “The Data Frame” — focusing on the impact of data in today’s networked economy.

Save $300 on registration with the code RADAR

Google+ and the identity issue

Smarr also talked about the nature of identity on Google+. Those following the launch of the service know that Google+ and pseudonymity is a hot-button issue. As the Electronic Fronter Foundation’s Jillian C. York noted in an essay making a case for pseudonyms, Google+ has changed some of its processes, moving from immediate account deactivation to warning users about the issue and giving them an opportunity to align their Google+ username with its “real name” policy. This week, Kaliya “@IdentityWoman” Hamlin became the most recent person to have her Google+ account suspended. (Given her work and role in the digital identity space, Hamlin’s use case is likely to be an interesting one.)

“There’s cases where that authenticity and knowing that this is a real person with whatever name they tend to be called in the real world is really a feature,” Smarr said during our discussion. “It changes the tone of discussions, it helps you find people you know in the real world. And so, wanting to make sure that there’s a space that is preserved and promoted is really important. On any of these social networks, it’s not enough to write the code, you have to make the right community. Lots of networks choose different approaches to how they do that, and they all have different consequences. It’s not that one is inherently better or or more valid than the others, it’s just that if you don’t do anything about it, it will kind of take its own course.”

There are clearly some gray areas here, particularly given Google’s global reach into parts of the world where using your real identity to share content could literally be life-threatening. “Obviously there are a lot of cases where being able to share things not under your real identity is valuable and necessary, and Google has a lot of products like this today, like YouTube,” said Smarr. “If you’re posting videos of authoritarian governments during a revolution, you may not want to use your real name, and that seems pretty valid. Whether or not that type of use case will be supported in the Google+ as you know it today is something that we’re all thinking through and figuring out, but it’s not meant to stop you from doing that in other products.”

Smarr had one other comment on identity that goes to the difficultly of creating social networks in domains that may be hostile to free expression: “It’s not just enough to offer the ability to post under a pseudonymous identifier. If you’re going to make the commitment that we’re not going to out your real identity, that actually takes a lot of work, especially if you’re using your real account to log in and then posting under a pseudonym. We feel a real responsibility that if we’re going to make the claim to people ‘it’s safe, you’re not going to get outed,’ then we really need to think through the architecture and make sure there aren’t any loopholes where all of a sudden you get outed. That’s actually a hard thing to do in software … we don’t want to do it wrong, and so we’d rather wait until we get it right.”

As I said at the end of the interview, if anyone is going to solve the engineering challenge of enabling its users to securely and anonymously connect to its social network, Google would have to be near the top of the list. One potential direction might be further integrating the Tor Project and the Android operating system in the context of a Google+ API. What’s clear now, however, is that if Google+ looks like a social backbone for the Internet, there’s still a lot of growth ahead.

Related:

tags: , , , , ,
  • http://twitter.com/jmcesteves J M C Esteves

    “a hard thing to do in software”, “real account” ← still expecting users to trust Google. Water wets, data ultimately leaks. Beyond that, why trust Google itself? Why trust others who have the power to demand our data from Google? Go on doing it “in software” for users feeling good about it, but it could be done in policy immediately — for those who would prefer to open accounts providing no name data beyond their pseudonyms. Unfortunately, with no signs of such a move, ‘real names’ seem of the essence for the ways in which Google will use us as product.

    “parts of the world were using your real identity to share content could literally be life threatening” ← still a restricted view of risks. Abusive government behavior is not restricted to ‘parts of the world’ where such abuses are lawful. And there are more sources of threat to consider. Some have to hide from their own family.

  • http://www.espy.cc Graham

    The really annoying misspelling of “where/were and there/their” makes many of these articles suspect. Bad writing, while it does not = bad logic, it makes such articles ‘suspect’. However, I expunged my facebook account because privacy is important to me. I no longer have a twitter account for the same reason. Nor will I have a Google+ account. It is nobody’s business but my own what I call myself, where I am located, or what my preferences regarding anything are. Sadly millions in the gutless herd would give up their virginity if it meant they could be accepted as a part of something! Sad, sad sad. I may even be one of the first to totally disconnect myself from the internet soon. It was a brave experiment. Now it sucks radically.