Terraform and the Open Source Trap

In a way, HashiCorp’s migration to the BUSL (originally from MariaDB) is hardly news. It’s a common playbook adopted by struggling companies based around widely adopted but difficult to monetize open-source software. If value-add services and features aren’t enough, the company’s remaining gamble is to extract more revenue from users. The license change is simply the administrivia that facilitates putting this squeeze on the ecosystem.

I say it’s a gamble because success depends on two factors, both of which are cast into doubt by the necessity of the license change in the first place:

1. The tool must deliver sufficient value to support the revenue. Simply put, customers have to be willing to pay for what they’re presently getting for free.
2. The company and its services must be an essential component of the ecosystem. If the broader community would persist in its absence, the company is likely not in a position to squeeze anyone.

As a fan of HashiCorp’s work, it pains me to say that I don’t think they’re in a great position here. Terraform isn’t MariaDB; this is not an online service that is critical for business functions (unless, of course, your business is hosting Terraform). I could easily see the OpenTF fork carrying the torch from here.

I feel like open source can be a trap. On the business side, you miss out on the signals that communicate product-market fit. For end-users, it’s easy to assume that an open-source license provides some kind of security. In reality, the governing structure behind a software product always matters, and open-source software requires the same risk analysis as any other.