FOSDEM '10 is a free and non-commercial event organized by the community, for the community. Its goal is to provide Free and Open Source developers a place to meet. No registration necessary.

   

Interview: Evan Prodromou

Evan Prodromou will give a talk about StatusNet at FOSDEM 2010.

Could you briefly introduce yourself?

My name is Evan Prodromou. I'm originally from San Francisco, but I've lived in Montreal, Quebec since 2002, where I have a wife and two small children.

I've been a web developer and open source enthusiast for 15 years. I have been involved with various projects like Debian and Freenet in the past, but most people know me as the founder of Wikitravel, the open content travel guide project. I also founded the StatusNet project in 2008, and am the CEO of StatusNet Inc.

What will your talk be about, exactly?

I'll be discussing our software, StatusNet, with an emphasis on scaling the software to meet the needs of the Identi.ca web site.

What do you hope to accomplish by giving this talk? What do you expect?

I'd like to accomplish two things. First, I'd like to give people an idea of what StatusNet is, what it does, and why open source and distributed social networking software is important. Second, I'd like to show a real-life example of how you can scale up an open source LAMP stack tool while still keeping it valuable for small and single-user sites.

What's the history of the StatusNet project? How did it evolve?

I started StatusNet in May of 2008. I'd been thinking of having an open source Twitter clone for about a year, and since no one else had made one, I decided to give it a shot. I launched Identi.ca in July of 2008 and got a huge amount of attention, which I think is what gave me the incentive to keep going.

In the last 18 months my company, StatusNet Inc, has been extremely successful in developing the software and the project. We have almost hundred people who've contributed code, and we've attracted great developers like Brion Vibber (of MediaWiki) and James Walker (of Drupal) to work on our team.

We've launched the private beta of our cloud service, status.net, and we're adding thousands of new sites per day to it.

Identi.ca is the most prominent StatusNet website. Can you name a few other prominent sites that are using it?

There are about 700 sites on the open web that use our software; sites like bleeper.de, army.twit.tv, todaysmama.com, phitter.com, and chum.ly. However, the software is used much more in the enterprise, where companies like SAP, Sun, Motorola and Intuit use it to keep employees informed about what's going on.

What are the biggest scaling problems behind Identi.ca and how have they been solved?

Probably the biggest scaling problem with Identi.ca has been supporting a messaging inbox format. We've gone through four different data structures for our inboxes over the last 18 months, as well as using some other scaling techniques like master-slave database replication and memcached.

It's a fascinating issue -- keeping information on the tip of the moment easy to receive while older information is still accessible and manageable.

Why is that subscribing to an account works federated among StatusNet servers, but groups and direct notices don't? Is this something that will be implemented in the future or is it a technical impossibility?

The main reason is that the StatusNet software has evolved faster than the OpenMicroblogging standard has. We're working hard to fix that, and our upcoming 0.9.0 release will include support for the new OStatus remote subscription protocol. It does support groups; direct messages will probably come later. Adding private data to public streams is incredibly hard!

How big is the StatusNet developer community?

We have about 75 people who've made contributions to the core code. I consider our developer community to extend beyond that group, though. We also have more than 300 web, desktop, and mobile applications that support StatusNet through our Web API. On top of that, we have maybe a couple of dozen "themers". And we have another few dozen people providing peer support, doing documentation, and doing advocacy.

What features can we expect in StatusNet in 2010?

I don't know! That's part of the fun. We're going to be releasing a 1.0 version in March 2010 which is considerably slimmed down. We're keeping a small core of web-based microblogging functionality and breaking out most of the code that does external interactions (like the Twitter and Facebook plugins) or optional functionality (like OpenID login) to dedicated plugins that will evolve on their own.

My goal is to see this plugin ecology grow rapidly. Personally, I'd love to see more syntax extensions (I like what's going on with microsyntax.org, for example), connections to other Web-based services, cross-site subscription tools like FETHR and RSSCloud, vertical integration using OpenSocial and other Web APIs, and integration with IM systems like IRC, AIM, MSN and Yahoo. We really want to bust it wide open and see what people do.

StatusNet is rooted in the same distributed and open philosophy as Jabber, RDFa, etc. How do you foresee the future of this kind of "open web"? Won't people keep using the closed counterparts like Twitter, MSN, etc., simply because they are easier to use?

I think that in 2010 we don't have the same problem with ease-of-use that people had "back in the old days" with open source software on the web. Cloud services like wordpress.com make it as easy to set up a WordPress blog as it is to set up one on Blogger or TypePad or whatever. Our cloud service, status.net, is in private beta and will be available for public signups this month.

Even if you're not going to use a cloud service, installing open source web software is much easier than it used to be. WordPress has really set the standard with its 5-minute install process; other web software needs to meet or beat that. We have a very similar install program that should get you working with basic Web microblogging very quickly (or tell you why you can't).

There are two things that I think keep people from changing to Open Web technologies from closed services. First is the network effect: that is, the value of a communications medium goes up by the number of participants. So, large social Web sites that are closed have great 'stickiness' for people.

We can counteract these effects by giving people a way to stay in touch with their existing networks using open web software. You can tap into Twitter and Facebook using StatusNet; hopefully we'll support other services like LinkedIn, MySpace and Ning later this year. Giving people access to their existing networks makes it easier to try out new tools.

The second thing is price. If a proprietary, closed web service is the same price as an open web one, why switch? I think as people begin to understand that their social network presence is important, they're going to want to take more control over them -- something proprietary services won't let you do.

All of which is to say: I think open web tools have great competitive advantage over closed tools, but it's an incremental process getting them implemented.

Creative Commons License
This interview is licensed under a Creative Commons Attribution 2.0 Belgium License.