We Are Not Supporting OpenTF

I replaced “Terraform” with “XFree86” and “OpenTF” with “X.org” as I read it, and the arguments sounded so familiar.<p>“If they fork XF86, users won’t know which product to install. They’ll drift apart and ruin the whole market!” Or maybe everyone will (once again) adopt the truly open version that the majority of contributors switch to.<p>Either way, I’m not too worried about it. Forks of open source projects happen all the time for a thousand reasons. That certainly beats closed stagnation!

&gt; That being said, after HashiCorp’s CTO released “binding” answers to some of these FAQs, it seems clear that the only folks who this affects are those that sell software to other companies.<p>I’m not so sure about this. The Licensing FAQ doesn’t have the full legal weight of an update or addendum to the actual license. There’s no guarantee that they won’t renege these “binding” answers next year, and unlike the license agreement my company isn’t agreeing to them as terms of use.

OpenTF is a good thing. Hashicorp basically stopped reviewing community terraform contributions a while back:<p><a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=28425849">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=28425849</a><p>They have no skills in managing true open source projects. We’re better off in an OpenTF World. We don’t need their support.

Who is this dragon company and what weight does their opinion hold anyway? Most of these reasons to not fork are pretty half-baked.<p>&gt; A schism like this would be devastating for the tool’s adoption<p>Except when it isn’t, like in the case of LibreOffice. Terraform is already by far the most popular in its category. It’s pretty obvious that OpenTF will instantly be the superior choice by the merits of the license alone. HashiCorp will have the support of one company while OpenTF will be backed by multiple companies.<p>Cloud providers like AWS will have an automatic legal incentive to contribute to their providers as a part of the OpenTF project instead of HashiCorp. HashiCorp will be alone on an island.<p>Other parts of this article basically just say “don’t worry, HashiCorp won’t sue you.” Sorry, that’s not as good as free and open source.<p>Yes, HashiCorp is moving too slowly. That’s why they have to use the Oracle army of lawyers method, because most of the value in Terraform Cloud can be replicated in a weekend with a Gitlab template. Print the terraform plan out to the MR comments, use Gitlab’s variable management for secrets, manage your state in the backend of your choice, and that’s pretty much it.<p>Also, the terraform private registry gives you zero benefit over using repository URLs with references to tags&#x2F;commits&#x2F;branches. Yet another thing you don’t need terraform cloud for.

My biggest concern is the legally untested clause in their BUSL on ‘competing with hashicorp’. I do not know whether I can run Terraform from a production control panel to start new infra for customers. I do not know when this competing might change.<p>Their faq comments are useless: they had a faq item before saying they would be FOSS forever which they removed with this move.

The arguments presented by dragondrop are a classic example of FUD. The author is afraid of what Hashicorp can do if we don&#x27;t just ignore the license change and pretend nothing happened. Based on that fear, they present various scenarios that seem scary to them. But when you look closer, these fears are ridiculous.<p>For example, &quot;people will be confused whether they should use TF or OpenTF&quot;. Guess what: at the beginning everybody will be using TF except a small minority who will consciously chose OpenTF, mostly for development. And they will make sure that, at least at the beginning, the choice will be irrelevant because the compatibility between these two is a top priority. It&#x27;s like saying in 2004 &quot;You shouldn&#x27;t release CentOS because people will not be sure if they want to run CentOS or Red Hat&quot;. It simply makes no sense. All other arguments are like that.

I don’t agree at all with the arguments here. Terraform is so widely adopted that it has become a basic, almost built-in component of every cloud ecosystem. Hashicorp would have been better off by treating it as a Trojan horse into enterprises and focusing on its other assets. I expect there will be a newcomer that will take advantage of this mishap and we’ll be talking about their solution next year (unless OpenTF manages to win people over).

&gt; You may make production use of the Licensed Work, provided such use does not include offering the Licensed Work to third parties on a hosted or embedded basis which is competitive with HashiCorp&#x27;s products.<p>&gt; A “competitive offering” is a product that is sold to third parties, including through paid support arrangements, that significantly overlaps the capabilities of a HashiCorp commercial product.<p>We all get what they&#x27;re trying to do. The bad thing is that this isn&#x27;t only overly broad, but a moving target. I looked at HashiCorp&#x27;s product page, and they&#x27;re building up a wide portfolio of IaaS tools. Those are such fundamental tools that even if you&#x27;re not a cloud provider at all, you&#x27;re likely to have an overlap in &quot;capabilities&quot; at some point. IMO, IANAL and all that, that&#x27;s a significant business risk for anyone using Terraform, for any purpose.

&gt; HashiCorp could feel forced to strengthen its position and introduce incompatible differences in new releases of Terraform and major-cloud providers<p>I&#x27;m sure it&#x27;s hard to turn FOSS into money. I sympathize. But poisoning the well too discourage competitors will drive everybody away.<p>But could they really? Who actually developed the AWS provider?<p>My biggest problem with terraform over the last few years is that it always feels like they haven&#x27;t been investing enough development in it. This has been spelled out directly by Terraform employees in their github issues, and it&#x27;s visible by inference in the way that they disengage from certain issues.

&gt; Notably, the three largest companies that most-directly compete with HashiCorp’s managed Terraform offerings, Spacelift, Env0, and Scalr, all have made multi-million dollar pledges to support a fork of Terraform (13 FTEs for five years between these three companies).<p>Wow all these three companies have products offering automation using Terraform. Obviously they would be against the license change.

It’s absolutely crazy to me that other companies free-ride off of companies like Hashicorp and Mongo’s work and the free-riders are the ones many - particularly on this site - support.

TLDR; if we don’t keep TF open source in 5 years everything will become Terraform Cloud - horrible, painful to work with Enterprise-grade BS.<p>I don’t think forking is bad for terraform ecosystems. It might be bad for HashiCorp, but for community in general it can become that source of competition that will make terraform even better.<p>Let’s be honest, most of the modules are made by community, there are thousands of modules out there.<p>HashiCorp recently spent all their love on enterprise related products (like TF Cloud) and almost no love on terraform itself. Our team tried TF Cloud (because it’s HashiCorp, it should be a great product, huh?) and the experience was horrible.<p>If we let HashiCorp do that licensing trick they as a company won’t have any incentives to improve terraform.

&gt; The only companies that need to worry about the license change are those with production products that are “offering the Licensed Work to third parties on a hosted or embedded basis which is competitive with HashiCorp’s products.”<p>That&#x27;s might be the current situation but it does not mean that will last. If the license changes again or if they e.g. decide stop the development, it will be a lot harder to make a fork in order to e.g. keep up to date with security patches.<p>&gt; these problems with the Terraform core repository are not new<p>&gt; The only material change over the past ~14 days, in our opinion, is that certain business’s have had their margin profile change.<p>Often people deal with things until something breaks camel&#x27;s back. In this case it was the license change for many companies&#x2F;users and hopefully it will mean that OpenTF will end up fixing those earlier problems as well.

I think many start-ups use open source mainly as an argument for adoption in the ecosystem, and they have to consider the consequences. Many projects avoid choosing AGPL or GPLv3 licenses to facilitate easier enterprise adoption (&#x27;Hey, we can use X and create SaaS without hassle&#x27;) without generating revenue, while being funded by venture capital and without getting contributions back. Then, when they are adopted, they complain. While the contributions from corporations like HashiCorp are impressive, the overall situation is complex. There&#x27;s a reason why Linux is still around and growing.<p>From my point of view: Stick with a real, copyleft license that has less adoption by other enterprises and focuses on organic growth instead of VC-driven hypergrowth. Alternatively, be prepared for the consequences that Amazon or other hyperscalers will attract a large number of potential customers using your product without giving anything back. In that case, establish another source of income right away. One has to be realistic: Competing with your open-source product&#x27;s SaaS against a SaaS by Amazon or other hyperscalers will not work out if that&#x27;s your only way to make money.<p>Edit: typos, grammar

Oh My... 99% of users are not affected only companies which have profit motive.<p>Do not forget it is those companies which provide the choice to the users and Hashicorp strives to restrict their choice.<p>Hashicorp should not be seen in the vacuum. What is point of it all ? The point is Hashicorp is looking to monetize their ecosystem better, meaning making things more expensive for customers.<p>You probably seen RedHat&#x27;s play - kill CentOS to improve RHEL monetization, not enough ? Take the next step to try to kill off some alternatives

Hudson and Jenkins come to mind here