It’s been roughly two months since HashiCorp switched Terraform from the Mozilla Public License (MPL) to the Business Source License (BSL), and the community response has crystallised into something real: OpenTofu. What started as an open letter and a manifesto has become a Linux Foundation project with significant backing, active development, and its first alpha releases. The question is no longer “will there be a fork?” but “how will the fork and the original coexist?”
The Backstory, Briefly#
In August, HashiCorp announced that all future releases of Terraform (and other HashiCorp products) would be under the BSL 1.1 license. The BSL allows most use cases but restricts using the software to create competing products or services. For many users, nothing changed practically. But for companies building managed Terraform services, CI/CD platforms with Terraform integration, or consulting businesses, the implications were significant.
The response was swift. Within days, a coalition of companies and individual contributors launched the OpenTF manifesto, calling for Terraform to remain truly open source. When HashiCorp didn’t reverse course, the fork was announced. In September, the Linux Foundation formally adopted the project, renaming it OpenTofu to avoid trademark issues.
Where OpenTofu Stands Today#
OpenTofu has released its first alpha versions, tracking closely with Terraform 1.6. The project’s initial goal is straightforward: maintain compatibility with existing Terraform configurations while establishing an independent development path under a genuinely open-source license (MPL 2.0).
The OpenTofu repository is active, with contributions from engineers at Spacelift, Env0, Gruntwork, and other infrastructure companies. The steering committee includes representatives from several organisations, which helps prevent any single company from controlling the project’s direction.
From a practical standpoint, the current alpha is largely a drop-in replacement. You can point it at existing Terraform state files and configurations. The registry compatibility work is progressing, with OpenTofu setting up its own provider and module registry while maintaining compatibility with the existing ecosystem.
The Technical Divergence Question#
The most interesting question right now is how quickly OpenTofu will diverge from Terraform. In the short term, maintaining compatibility is essential — nobody wants to rewrite their infrastructure code. But over time, independent governance means independent technical decisions.
Some areas where divergence might occur:
State management: Terraform’s state handling has always been a pain point. Remote state, state locking, state migration — these are areas where significant improvements are possible. An open-source project with multiple commercial backers might be more willing to make breaking changes in state management if the community consensus supports it.
Provider ecosystem: Both Terraform and OpenTofu use the same provider protocol, so existing providers work with both. But new providers could theoretically be developed exclusively for one or the other. The OpenTofu registry is being built to remain compatible, but the long-term evolution is uncertain.
Language features: HashiCorp has been conservative with HCL evolution. An independent OpenTofu could potentially move faster on language features — better loops, first-class functions, improved module system — if the community demands it.
Personally, I think we’re looking at a period of compatibility lasting at least a year, possibly longer. The OpenTofu team is smart enough to know that forcing users to choose too early would be counterproductive.
What This Means for Your Infrastructure#
If you’re using Terraform in production today, don’t panic. Nothing requires immediate action. HashiCorp isn’t going to stop maintaining Terraform — if anything, the competition will motivate them to ship faster.
Here’s my practical advice:
Keep your configurations standard. Stick to well-documented HCL patterns and avoid relying on undocumented behaviour. This gives you maximum flexibility to switch between Terraform and OpenTofu if needed.
Watch the registry situation. Providers and modules are the ecosystem’s lifeblood. If your critical providers are well-maintained by their original vendors (AWS, Azure, Google Cloud, etc.), they’ll likely support both tools. Niche providers might take longer to explicitly support OpenTofu.
Test OpenTofu in non-production environments. It’s alpha quality right now, but testing it against your configurations helps you understand the compatibility story and provides valuable feedback to the project.
Don’t rewrite anything yet. If your Terraform setup works, keep it working. The time for potentially switching is when OpenTofu reaches a stable 1.0 release and your team has had time to evaluate it properly.
The Broader Open Source Lesson#
This saga is a case study in open-source sustainability tensions. HashiCorp built an incredible product and gave it away under an open license. Other companies built profitable businesses on top of that product. HashiCorp decided the value exchange was unfair and changed the license. The community decided the license change was unacceptable and forked.
Everyone involved has a legitimate point. HashiCorp invested hundreds of millions in R&D and watched competitors monetise their work. The community invested time, plugins, modules, and expertise into an ecosystem that was promised to be open source. There’s no clean villain here.
What I find encouraging is that the Linux Foundation stepped in quickly and gave OpenTofu institutional backing. Open-source projects need governance, funding, and legal support to survive long-term. A passionate community without institutional support tends to fragment. OpenTofu has both, which gives it a real chance.
My Take#
I’ve managed infrastructure with Terraform since the early 0.x days, and it transformed how I think about infrastructure provisioning. Whatever happens with the fork, Terraform’s impact on the industry is undeniable.
My bet is that both Terraform and OpenTofu will coexist for the foreseeable future, similar to how MySQL and MariaDB, or Elasticsearch and OpenSearch have settled into parallel existence. The market is big enough for both, and competition drives innovation.
For now, I’m keeping Terraform in production and running OpenTofu in a test environment. The alpha is promising, and the pace of development suggests a stable release isn’t too far off. When it arrives, I’ll evaluate the switch more seriously.
The infrastructure-as-code ecosystem is healthier with genuine competition, and that’s ultimately what we’ve gained from this situation, messy as the journey has been.