Can OpenTofu Become the HTTP of Infrastructure as Code?

When OpenTofu was initially forked from Terraform, Kelsey Hightower, one of my favorite voices in the open source and Kubernetes community, chimed in to say:

I believe OpenTF, a fork of HashiCorp’s Terraform project, will end up growing Terraform adoption in the long run.

Take HTTP for example, which has many implementations, the adoption is higher than ever. TF has just become the HTTP of configuration management.

— Kelsey Hightower (@kelseyhightower) Aug. 25, 2023

Several months later, I joined a panel discussion to explore the impact of OpenTofu’s general availability. Throughout our conversation, we came back to Kelsey’s comment.

As we did, I felt compelled to write this post to share my thoughts on why OpenTofu could be the evolution the Terraform protocol needs to become the “HTTP of the cloud.”

Proof of Maturity

On Jan. 10, 2024, exactly five months to the day of the HashiCorp license change, a major milestone was achieved with the general availability of OpenTofu v1.6.

How we got to this point is incredible, but I’m not here to retell that story. It is sufficient to say that reaching this milestone demanded an immense amount of effort — not just in building the project but also in constructing its support environment, including the new public registry.

The 5-month marathon in which this was done stands as one of the greatest achievements in the history of Terraform. More importantly, this accomplishment belongs entirely to the Terraform community.

The GA did two important things:

1. It showed that Terraform could have an independent community-driven future.
2. It demonstrated Terraform’s technical maturity and stability — both of which are now baked into the DNA of OpenTofu.

Building upon these, OpenTofu now opens doors to a realm of new possibilities, offering a choice between the HashiCorp vendor-backed non-open source software (OSS) Terraform flavor and its equally proficient, community-backed and impartial open source counterpart.

Having this second option, I believe, will contribute to the widespread adoption of Terraform — not just as a specific solution, but as a foundational technology and a concept.

In turn, this paves the way to rethink what the “Terraform ecosystem” might mean — a group of users, but also a cluster of technologies where multiple binaries utilize the same core tech to implement various concepts.

Speaking from the perspective of this ecosystem, I’d like now to zoom in on two of OpenTofu’s characteristics that I believe will uniquely influence its future.

True Open Source

The first thing I want to touch upon is OpenTofu’s true open source nature.

I’m emphasizing “true” here for a reason. Even before the whole license shift deal, Terraform was already part of the commercial open source (COSS) domain, with HashiCorp having full control over the roadmap.

Naturally, that put Terraform in a position to be affected by the company’s business needs.

For instance, a long-standing feature request has been Terraform state encryption. However, due to the vendor’s discretion, it was never prioritized.

Today, the same feature is a part of OpenTofu’s official roadmap, after it was selected based on merit following a community proposal.

👁 Image

This example points to a larger truth. For Terraform technology to achieve universal HTTP-like adoption, it had to outgrow its commercial origins. In other words: Before it could belong to everyone, it needed to be owned by no one.

As a foundation-backed project, OpenTofu meets these neutrality criteria. As it does, it creates new possibilities of long-lasting, universal widespread adoption.

The Credibility Factor

The other thing I want to discuss is credibility — the other side of the coin.

This also relates to ownership. Over the years, we’ve witnessed early commercial open source projects. This includes HashiCorp, as well as others like Redis and Elastic, changing their licenses and moving away from their open source roots.

Each time this happened, it eroded a layer of trust in all non-foundation-owned open source software.

However, as a foundation-backed project, I’ve seen OpenTofu earn incredibly high levels of trust right from the start.

Observing the aftermath of moving it under the Linux Foundation, I see Alpine, Brew, Gitlab and others rapidly incorporating OpenTofu support (and even deprecating Terraform support).

It’s evident to me that none of these heavyweights did so on a whim. All of these organizations have plenty of ways to utilize their engineering resources, and they wouldn’t prioritize OpenTofu unless they trusted the team behind it and the reputation of the Linux Foundation.

The speed at which OpenTofu integrations are being introduced says a lot. Looking forward, I’m already seeing these votes of trust expand and compound, eventually snowballing to deliver a level of trust that a company-owned Terraform would never reach.

From the perspective of the ecosystem, the “credibility potential” of OpenTofu is a game-changing factor, and one that could play a critical role in ushering in a universal standard.

A Unique Possibility

Before the license change, Terraform had gained widespread adoption in the DevOps ecosystem.

Other tools such as Pulumi and Crossplane, and also platform-native frameworks like Amazon Web Services‘ CloudFormation and Microsoft Azure‘s ARM, saw success. Still, for the most part, those who needed a unified way to deploy to and manage multicloud did so with a HashiStack.

However, despite its yearslong undisputed dominance, for all of the reasons mentioned above, Terraform was never positioned to evolve into a true universal HTTP-like standard.

OpenTofu’s unique combination of open source neutrality, foundation-backed credibility, mature technology and community support has the potential to change this dynamic.

And no, I will not try to predict here whether a universal cloud native configuration standard is set to emerge.

What I will say with certainty is that, for the first time in Terraform’s history, there is room for this possibility to exist.

For me, this is enough to be excited about OpenTofu’s future, and grateful for the opportunity to be a part of its journey.