Open-core Ain't Worth It.
Open-core companies should stop trying to have it both ways and either be closed source or all in on open-source.
On one hand, open-core has more open-source than closed source, but on the other hand it’s not lines of code count that matters, it is value delivered to users. Two different open-core cases to explore delivered value include Elasticsearch and GitLab.
Elasticsearch has benefited immensely from contributors and others building an ecosystem around the tech. Even AWS helped increase its popularity. This was done with the community promise that they would stay open source, “We did not change the license of any of the Apache 2.0 code of Elasticsearch, Kibana, Beats, and Logstash — and we never will.”
GitLab is open-core. I have sincerely enjoyed GitLab and advocated for its adoption at multiple organizations. Not that long ago, I purchased GitLab Enterprise despite the rigid licensing and billing only to stumble on bugs in Enterprise features. These bugs had (have?) issues that were years old and marked for the community to fix even though they were not open-source features. I finally had to give up on GitLab, cancel the contract, and go back to GitHub where it is clearly closed-source, but things actually work (and they have super flexible user licensing).
On one hand, open-core has more open-source than closed source, but on the other hand it’s not lines of code count the matters, it is value delivered to users.
Many have benefited from Elasticseach while it was open-source and it’s probably legit to give some credit to GitLab for pressuring GitHub’s product development. Is open-core a net gain to society, even if missing some open-source benefits? I don’t think so. What has the success of open-core companies kept from the market? If Elasticsearch was closed from the beginning, the community would have poured its energy into something such as Solr. If GitLab was closed source, there would be many more working on Gitea. The open-source companies behind many of those GitLab superpowers, such as Sentry, have their own managed solutions to monetize by integrating with GitHub and Gitea rather than be hosted and bundled “AWS style” by GitLab.
Lots of great open-core companies are doing great things, but they all seem to operate as proprietary companies focused on royalty extraction and restricting users, rather than maximizing delivered value. There is no free lunch and open-core seems to have hidden costs. Should we look at open-core as “opportunity cost” rather than “a little is better than nothing?”
CrateDB is returning to its roots, away from open-core, and I hope to see many more follow their lead. Who do you want to see ditch open-core?