The bright future of the SUMO platform

When the SUMO project started back in 2007, one of the first big questions was what web platform to use as a foundation for the upcoming Firefox support website. The Mozilla web development team started by doing a thorough analysis of the available content management systems (CMS) currently available in order to reach a conclusion. The outcome of that analysis was that TikiWiki was most suitable for our needs. Among other reasons, Tiki was best because of its many bundled features, strong multilingual features and powerful wiki and forum integration.

What followed was some rapid development to get Tiki deployed on our server and adapted to our needs. We received great help from the Tiki community, ensuring that the initial launch of Firefox Support was smooth and successful. Unfortunately, we didn’t really have a plan for how to ensure we remain synchronized with the continuous development of the Tiki platform upstream, so we just kept developing on our local codebase.

Because Firefox is such a popular web browser, we get a lot of visitors on the Firefox Support website. And when we say a lot of visitors, we mean a lot: over 16 million page views per week! www.mozilla.com is in the top 150 sites in the world, and SUMO naturally gets a good portion of the traffic.

As a result, a significant part of our development focus has been on increasing performance. About two years after the site officially launched, September 2009, we had reached a point where we would be forced to rewrite parts of the underlying infrastructure of Tiki in order to keep up with the increasing traffic to the site.

We had reached a crossroad when we had to decide on what do with the SUMO platform:

  • Should we upgrade to the latest version of Tiki? Remember, we had over two years of local development that hadn’t been upstreamed.
  • Should we continue to patch the Tiki platform locally, i.e. continue with what was essentially a fork of Tiki?
  • Or should we switch to another platform that was better suited to our particular needs?

Abort, Retry, Fail?

Interestingly, in our second analysis of the available platforms in 2009, Tiki still came out as the winner! There simply is no existing CMS that is more suitable for us.

However, that doesn’t change the fact that Tiki is not entirely optimized for our needs. In order to change this, we would have to invest a huge amount of time to rewrite structures and code in Tiki relevant to SUMO, and we would have to take great care to not mess up with all the thousands of other Tiki-powered websites out there.

This led us to the realization that the right path for SUMO is neither of the above options, but to instead do a clean break and work together with the AMO team to develop something new that is optimized exactly for our specific needs: an excellent open source support platform that can handle over 350 million Firefox users.

TikiWiki really is a fantastic CMS — the fact that it still is the best fit for Mozilla among the options available is amazing.

Now our challenge is to build the next generation of the SUMO platform: SUMO 2.0, codenamed Kitsune. More on that soon!