ExpressionEngine Goes Open Source after 16 Years

In a post titled “Open Source Has Won,” EllisLab founder Rick Ellis explained why ExpressionEngine is going open source after 16 years. The content management system is an evolution of the pMachine blogging software first released in early 2002. EllisLab previously required a license fee to use the full version of ExpressionEngine, which is built on object-oriented PHP and uses MySQL for storage.

“Although open source was a viable licensing model when we launched our first CMS back in 2002, it was not apparent then just how dominant open source would become on the web,” Ellis said. “It wasn’t until Eric Raymond wrote The Cathedral & The Bazaar that open source would even begin to enter the general public’s consciousness. Since then we’ve watched the open source market grow rapidly and continuously.

“Today, over 90% of the CMS market is open source. In fact, it’s nearly the de-facto license model for all-things web. Stunningly, the market is expected to triple in revenue within the next five to ten years, and it’s estimated that over 70% of businesses worldwide rely on open-source software. To say that the internet is open source would not be an exaggeration. It’s that dominant.”

Ellis said he had wanted to migrate to an open source license for a long time but had not yet found “the right strategic and financial partner to enable the full vision of what we hope to achieve.” The first part of EllisLab’s business plan is to build a successful services model and then branch out from there.

Prior to licensing ExpressionEngine under the Apache License, Version 2.0, EllisLab’s commercial license imposed severe restrictions on what users could do with the software. Users were not permitted to do any of the following:

  • Use the Core License (free) for any client or contract work.
  • Use the Software as the basis of a hosted blogging service, or to provide hosting services to others.
  • Reproduce, distribute, or transfer the Software, or portions thereof, to any third party.
  • Modify, tamper with, bypass, or in any way impede license registration routines in the Software.
  • Sell, rent, lease, assign, or sublet the Software or portions thereof, including sites in your multi-site license.
  • Grant rights to any other person.
  • Use the Software in violation of any U.S. or international law or regulation.

Additional stipulations encouraged users not to share code by keeping their repositories private, and to make sure they were paying for commercial licenses if they were being paid for their work.

There was simply no way ExpressionEngine could capture any significant amount of market share with this kind of restrictive licensing and its usage has steadily declined over the years. It is currently used by 0.3% of all the websites whose content management system w3techs can detect. By this or any other measure of market share, ExpressionEngine stands as a sobering monument to the importance of giving a project a license that empowers its community to continue adding wood to the fire.

“The community is mostly gone at this point and I don’t even think its related to them charging for the software but they just stopped responding people and helping them in their forums,” reddit user @netzvolk commented on the news.

“I have paid EE multiple times in the past but considered NOT paying anymore because third party developers are gone, the community members are gone, the tutorials and books are gone….EE 2 was the best version so far. Moving to yearly releases also caused more harm than good in terms of building a stable ecosystem around the product.”

ExpressionEngine’s new open source licensing is a major win for its remaining users. How much further down the road would the software be if the decision was made years ago? There’s no way to know, but moving forward users will have more input and influence over the future of the software.

“I suspect open sourcing EE is an approach to get that community and developers back,” @netzvolk said. “EllisLab can still make money with consulting, support and add-ons.

“But all those suffer if nobody is using the product anymore. This is more about expanding reach to stay afloat than anything else because some of their past bad decisions are what created alternatives like Craft. EllisLab turned an amazing product into a forgotten one in just a few years. I hope this means some change, and maybe, maybe one day the old developers and hard core EE community members come back.”

Users can only speculate on why EllisLab is making this move after 16 years of keeping its software locked down under restrictive licensing, but Ellis makes it clear in his post that the market decided long ago.

“Open source has won,” Ellis said. “It’s not even a contest anymore.”

12 Comments


  1. See what crazy licensing does!

    Anyone remembers Movable Type from SixApart? I started using that back when WordPress was new and WP did the smart thing while Movable Type was charging insane prices for their licensing and had so many restrictions.

    Look where Movable Type is now. It’s a shame because I found MT was actually better than WordPress. But, look what happened to them in the end.

    As for ExpressionEngine, will have to see if this move gives potential.

    Report


  2. Hi Sarah, interesting info. However, EllisLab founder Rick Ellis is quoted as saying that “over 90% of the CMS market is open source.” And according to w3techs, WordPress has a CMS market share of 59.5%. So how much of the CMS market would be open source w/o WordPress?

    The somewhat successful Concrete5 has been open source since 2008 but its CMS market share is still a paltry 0.2% according to w3techs – even worse than ExpressionEngine’s 0.3% market share.

    Plus several non open source CMS options have much higher market share than open source Concrete5 (Squarespace, Weebly, Wix, etc).

    So being open source is only one of several factors to consider. For example, according to techspot.com, open source Android has 61% of the US smartphone market, and iOS has 38.7%. But I doubt many people believe iOS is a failure and Apple should make iOS open source (plus most people seem to agree that iOS provides a better and safer user experience).

    In other words, ExpressionEngine apparently failed at competing with non open source CMS competitors. So now it’s so desperate it’s going open source thinking it can take on the biggest CMS in the world? Do they know something others such as Joomla, Drupal and Concrete5 don’t know? I wish ExpressionEngine great success, but they’re definitely facing long odds.

    Report


    1. “over 90% of the CMS market is open source.”

      Maybe it was meant that 90% of CMS systems that are out there are open source, not evaluating their market share? Just that 90% of CMS systems that exist are open source. Maybe I misunderstood that.

      Report


  3. There is a lot you can do with EE that you can’t with WP. It’s a great tool for a mid range custom site. I hope this is a successful move for them.

    Report


    1. Can I request an example or two of things you can’t do with WP and can with EE. Seriously, asking for information, not to be annoying.

      Report


      1. It’s been years since I used EE but I remember it being a lot more versatile for creating any content you wanted. WP started as a blogging platform which became a CMS through Advanced Custom Fields. EE was a CMS that you could use to make a blogging site or any other type of content site. The biggest difference now will be in the database where EE is going to make tables for content types and WP is going to save everything as a post type. EE will probably be more efficient on the backend. For 99% of sites, it doesn’t matter.

        EE lost out because they started charging too much and WP became more attractive with the giant community of plugins. EE had a free version and they took it away over time and became more restrictive. They pushed the small time web people to leave. The plugins for it became expensive and they still are even if they are going to open source the main software. All the plugin developers wanted a piece of the pie. They have made their model so inclusive that Craft is now overtaking them by basicly being the cheaper version of EE.

        Report


  4. It’s funny how a software that most have never heard of wants more market share and only now in 2018 has decided to go open source. This doesn’t seem like it was ever deemed ‘the right thing to do‘, more of a move to grab more money wherever possible.

    It’s great another CMS has chosen to go open source, but I’m struggling to see how they’ve only just realised open source is a big player in the open web. It’s been this way for nearly a decade, arguably more than that in fact.

    I think it’s probably still a little too late though, albeit better than nothing. Those licencing terms are pretty passive aggressive, no wonder they decided to remove them.

    Report


  5. WP still my overall best CMS. Also the community behind is way greater than all the plugin makers for prestashop and drupal.

    Report


  6. It’s great another CMS has chosen to go open source, but I’m struggling to see how they’ve only just realised open source is a big player in the open web. It’s been this way for nearly a decade, arguably more than that in fact.

    Report


  7. I was a dev/designer with EE years ago (before I ever did a WP site). The community was small then. This is a too-little-too-late scenario, and I feel for those out there who wanted it well before now. It turned me to WordPress.

    Report


  8. “Although open source was a viable licensing model when we launched our first CMS back in 2002, it was not apparent then just how dominant open source would become on the web,” Ellis said. “It wasn’t until Eric Raymond wrote The Cathedral & The Bazaar that open source would even begin to enter the general public’s consciousness. Since then we’ve watched the open source market grow rapidly and continuously.

    ESR wrote CatB and published it online in 1997. The book was published in 1999. Have to call BS on the above quote. If they knew it was a viable model in 2002, they HAD to know about CatB before rejecting a FOSS model. If they didn’t, they were grossly uninformed about the FOSS community that was already very significant at the time.

    Report

Comments are closed.