HashiCorp CEO Predicts OSS-Free Silicon Valley | Hacker News Reader

HashiCorp CEO Predicts OSS-Free Silicon Valley

thestack.technology

68 pointsbcantrill103 comments

This is a dumb title for a good article. In that it makes is seem like he is talking about OSS software. The software that powers the entire Silicon Valley tech empire. Linux and OSS power Google, Facebook, Amazon, etc, right down to your 1-2 person startups.

He's not talking about that -- he's talking about open source software sales model used by companies like Redhat and HashiCorp. The threat here is that you're developing in the open, trying to up-sell enterprise features, professional services, and support. Then you have XYZ startup appear, they take your open-source code, repackages it and then sells enterprise features, professional services, and support. They basically leech off your OSS code and community, provide nothing back, and actively compete with you. It's brutal.

They are probably safe from XYZ startup. The real threat comes from XYZ enterprise. Like what Amazon did to Elastic.

OTOH, if they licensed their software under the AGPL, Amazon would not be able to offer it as a service without opening their secret sauce, whatever it is.

> Then you have XYZ startup appears, they takes your open-source code, repackages it and then sells enterprise features, professional services, and support.

That is a feature of the model, not a bug.

Exactly that happened to Git. Good. Happening at last in social media with Mastodon

The point is to have a lot of smaller companies rather than one or two huge companies playing "lock in"

Reality is it is sort of mixed. For the OSS company actually building the product, you end up playing whack-a-mole in that anyone can spin up a competitor over the weekend. Look at what happened with Docker. They completely changed the landscape of running code locally and in production and was run out of town. Everyone and their dog took the code and re-packaged it and competed with Docker. Then compare that with VMware, they completely changed the landscape of running code locally and in production but they were able to protect that value and thrive as a company. There is definitely a lesson in there.

Sure, maybe this is a feature but if you're about to launch an open source sales model company you probably need to think really hard about that idea.

The result is that the industry got better container run times because they had to compete with each other. Competition is a good thing - it gives us better products and services at lower prices.

And Docker still had the ecosystem, which they could have dug a deeper moat around. But they didn't, and they failed.

And not for nothing, VMWare has a significantly deeper technological moat than Docker. Containers are not hard to implement, virtualization is.

> Then compare that with VMware,

Their virtualization tools are niche, being squeezed by whatever the cloud provider offers for "free" on one side, and KVM (and platform-native hypervisors) on the other. You rarely hear them mentioned.

> Exactly that happened to Git. Good. Happening at last in social media with Mastodon.

No, it's a different thing. Git is not based on OSS code from BitKeeper and Mastodon isn't based on OSS code from FB/X/whatever.

git or mastodon are not 'silicon valley'. Yes, it is a feature for others to build upon Mastodon.

It's a bit more problematic for, say, hasicorp. They don't like that.

They also benefit from making their software open source because it drives people to adot it.

They also benefited from a large ecosystem of third party contributors, whom were not (at least, visibly?) considered as stakeholders when Hashicorp implemented this license change.

I'm pretty sure no-one contributing to a product's ecosystem likes it when the previous agreements related to IP underpinning the ecosystem is rugpulled out from under them.

> Exactly that happened to Git.

what a bizarre claim. Git was written by Linus because Larry McVoy kicked him off BitKeeper, and then the biggest Git ecosystem software spinoffs (GitHub) are entirely closed source.

HashiCorp and its CEO are too focused on their competitors while they really should be focusing on their customers

“HashiCorp’s technology was critical to the modern cloud”

I would challenge that. Their tech gained a stronghold because it had a liberal OSS license and thus a community formed around it.

Their technology itself isn’t critical, it was just popular to such a degree that it was almost standard in a lot of settings.

Now they’ve upset that community we see they’re not as critical as they think they are - OpenTofu exists and people are shunning Vault and the rest of their offerings.

If I was starting a software business today I simply wouldn’t open source on day one. I might lose some of the crowd that only uses OSS but I’d just have to focus on others for a while.

> If I was starting a software business today I simply wouldn’t open source on day one. I might lose some of the crowd that only uses OSS but I’d just have to focus on others for a while.

Or you could use GPL3 (for client-side) or AGPL (if it could be hosted). OSS crowd will be still happy and "leechers" are prevented.

I worked in way too many places that wouldn’t let me use AGPL and it was then harder to procure vs something totally closed source.

Don’t ask me how or why but then I went to work at a vendor and encountered similar experiences

It usually comes with an option: those that don't like AGPL can still get a non-AGPL licensed version for a fee. You would be still better off than with totally closed source.

The fact that a community-run fork is causing them to sweat so much shows that HashiCorp themselves may not be as essential to the Terraform ecosystem as they thought.

> Says Linux Foundation's OpenTofu adoption "tragic"

Yeah, for them, not for the community that they tried to fuck over.

Yup. The fact that they even mentioned this shows their true motivation behind the piece. It’s not that OSS is bad, it’s THAT OSS is bad (for us)

It's also essentially a public admission that OpenTofu REALLY worries them, instead of just something to sit out or combat by making a better product.

Outside of Vault, Terraform was their bread and butter. It’s understandable they would be scared of OpenTofu. Certain things that are novel have a tendency to leave the nest. This is one of those cases. They could go the Docker route, provide paid support and extra tooling around it, but decided to wage war against OSS instead.

Where's this same vitriol directed at Amazon, Google, et al.?

They've built huge cloud platforms on the backs of open source. These are highly profitable machines that siphon up so much free labor and give very little back. Instead we attack HashiCorp, whose entire product is going to fall into disuse over a fork of the very thing they built. Why is that?

I'm angry at AWS - it's literally just open source software run at scale. None of the infra and software behind AWS is open, though, so Amazon gets to lord over all of it and soak up all of the benefits.

I see it hard to attack smaller companies when the giant is in the corner taxing all of us and making fat margins.

Small companies struggle. Big companies have all the advantages, and we're giving them a pass.

Ugh. I don't get this.

Big Tech throws free stuff people all the time and they usually being the one winning product / lib everybody is using. People just started to expect the same thing from all companies no matter the size and profitability. It's madness.

HashiCorp is seeking a monopoly on hosting Terraform, which is bad for me as a user if I think another vendor might do a better job (or even just to keep prices in line with costs). It would have been better if they had been paid for writing Terraform and now hosting services could compete on their own merits.

> HashiCorp is seeking a monopoly on hosting Terraform

HashiCorp is trying to see an upside to the thing they've put a lot of effort into building. They're seeing lots of other companies -- including Amazon -- use it to enhance their own bottom lines off of HashiCorp's hard work.

These companies pose an existential threat to HashiCorp, and the HashiCorp stakeholders are getting nothing in return.

Free repackaging lowers the fitness of HashiCorp as an organization and cuts short HashiCorp's growth potential, resulting in lower revenue, less hiring, and more competition -- all of which ultimately inhibits HashiCorp's growth into a well-rounded company with a rich set of offerings. They're effectively being knee-capped and fenced in by low-margin competition.

Meanwhile you're eating for free.

While this is certainly true, there is also another part of the puzzle.

There are plenty contributors in the ecosystem building providers, submitting PRs under the assumption that the ecosystem will benefit and not solely Hashicorp.

As it stands most providers are not maintained by Hashicorp, e.g. AWS, Azure, Google, Hetzner, GitLab, ...

While the license change does not directly affect the providers it limits the ecosystems use of those providers.

HCP chose to open source Terraform. They didn't have to. They chose to market the OSS aspect of their products and use it as a selling point.

It may well be true that HCP is having trouble hitting their revenue goals because they open sourced Terraform but that doesn't mean the competition is the bad actor, for using an OSS product entirely in the spirit of OSS.

As it stands, my opinion is that your framing is incorrect. There is Terraform, which was OSS, and there is Terraform Cloud/Terraform Enterprise, which is not OSS. A lot of the development energy goes into TFC/TFE, and there is where the attractive, for pay, features are. Terraform has not been receiving a lot of development time. Only a few people from HCP, part time, working on it. I believe that the reality is that HCP is unable to keep up with their TFC/TFE product. They aren't innovating, and they are attempting to use the legal system to remove competition rather than compete on product. And now that OpenTofu exists and is part of LF, you don't even need HCP to succeed for your Terraform code to continue to live. Maybe one feels that is unfair, but it was their decision to OSS the product, and their decisions to change that. And capitalism isn't fair. Shrug emoji.

Right, I wonder how many people are insulting Hashicorp while deploying to AWS.

If Meta announced that every company hosting a React website needed to pay them a licensing fee starting next month or shut it down, they would (rightfully) get vitriol for it. That is what this is equivalent to, not big companies using open source projects. HashiCorp is free to use whatever open source they want, including in proprietary products (that's the entire point of open source). It's the rug pull and blindsiding of the community that is the problem.

I see it more as both sides taking calculated risk. Developers contributing to a corporate-led OSS project with particular licensing should know that they can be rug-pulled at any time. They benefit during the project and afterwards from a good fork that otherwise wouldn't exist (a healthy fork taking shape or not is additional risk). OpenTofu is a great win for the community.

For the company the risk is that the stronger community they build, the higher the chances of a serious backlash when they rug pull. Weigh the reputation damage, and upon emergence of a good fork having a serious competitor, to the benefits of community contributions.

Don't want to be rug pulled? Choose community-led projects with good governance and copyleft licensing to contribute to.

... if you want the whole pie.

The open source business model works if the service you provide is to sell consulting, implementation, certification...

If you want the whole SaaS pie then why are you going open source in the first place? To have the community develop the ecosystem for free for a decade making an SaaS business viable before you pull the rug?

Pretty much, this article should read "man shocked to discover he cannot have his SaaS cake and eat it"

The middle ground gets lost here.

What if you like the idea of collaboration, letting list people use your work for free, and making some money on proserv with the people who need it?

Well, too fucking bad, because Megacorps (Amazon) or others can create an alternative to your product at a loss just to curbstomp you out of existence.

I don't blame a company for trying to earn a living for them and their family while trying to find a middle ground that allows for open source and free community usage.

> making some money on proserv with the people who need it?

If you invest in developing the tool, you'll still be the best option for supporting the tool.

> Megacorps (Amazon) or others can create an alternative to your product at a loss

This is why things like AGPL exist.

They have a business model that they cannot sustain - instead of changing the license (and pulling the rug under their community's feet), they should look into changing the business model.

The only thing that will die is the open source rug pull model where companies dominate their space by having an open ecosystem and a large community of volunteers and then unilaterally switch the license terms from under everyone so that the founding team and VCs can make billions. Efforts like OpenTofu have thankfully shown us that the community can organize and take back control of their projects, and I really hope we see more such forks.

Going forward as a contributor your best bet is to direct your efforts towards a project sponsored by a reputable open source foundation, not a VC-funded startup.

I mean, HashiCorp is one of the companies leading the charge on that, are they not? They chose to stop using open source licenses for new versions of their software. That’s certainly their prerogative — they obviously decided that being an open source company didn’t fit their business model — but don’t then turn around and blame the Linux Foundation for picking up the ball that you dropped when you couldn’t make the numbers work.

I guess the lesson is to never open source your code? Or maybe to not trust "open source" and demand GPL?

Or maybe make the paid portion of your product (TFC/TFE) good enough that people want to use it?

The lesson is not to sign a copyright assignment unless you’re comfortable with the maintainers potentially doing this with your work later.

GPL would prevent others from modifying their work and turning it into a competing product. AGPL would have prevented others from modifying their work and turning it into a competing service.

BSD and other "liberal" licenses are great if you want to make adoption easy, but makes you vulnerable to Amazonification.

Hashicorp wouldn't have a successful product if they hadn't open sourced their code.

Now that they are entrenched, they want to change rules (and whine about it) because it makes them more money.

The title is editorialized

> there would be “no more open source companies in Silicon Valley” unless the community rethinks how it protects innovation

This is a sentiment I've heard from VC friends of mine for the past year now after their experience seeing the inability of Confluent, Sysdig, and even Hashicorp fail to successfully monetize their core offerings by undervaluing them.

They shouldn’t have gone all OSS if monetisation was their goal. It’s their own fault for thinking they could build a base off open source and then monetise it later.

They raised and went public in a different era.

Public Investors have significantly higher expectations for rates of return, and plenty of OSS-core companies raised significantly larger late stage rounds. That said, try IPOing a tech company with $15-70M a year in revenue in the 2020s.

Those numbers were the norms in the late 90s/early 2000s, which allowed companies like RedHat or Sun Microsystems to IPO, while still being able to ramp up their own vision and product lines.

Until almost a year ago you could raise IPO-sized valuations in the private market, while dealing with less regulatory scrutiny and overhead.

> in a different era.

It may seem like the early 2000s were a different era, but it's better to think of it as an earlier stage in this era. These companies built on OSS with varying degrees of giving back, so it seemed like they were all about the sharing economy. A decade later they'd built out their business (RIP Sun, though) and had solid positions in the market, not easily displaced. A decade later and the clawbacks begin. RedHat and HashCorp are just two examples of companies closing the doors, changing the terms of the deal, and harvesting the revenue they forwent in the early years. AKA the rug-pull

> A decade later they'd built out their business (RIP Sun, though) and had solid positions in the market, not easily displaced

They were given the ability to build out their business.

Companies like Hashicorp and other OSS core companies have revenue numbers roughly comparable to what would give you an IPO even 10 years ago. Mind you, Hashicorp is public, and their YoY revenue is around $520M a year.

The issue was they listed to minimize their IPO listing "Pop" (probably due to pressure from late stage investors IME), and this doomed them as a $70-80 per share value is difficult for any company to hold. Companies like Salesforce ($3.95) and RedHat ($14) listed at much lower share prices.

Lots of companies IPOed/SPACed during that time period as a number of the funds matured and needed to return profits to LPs. I know Social Capital had a similar thing happen due to Chamath's acrimonious divorce leading LPs to call their funds earlier. That's probably the most prominent example, but there were plenty of other funds that had a drawdown during a similar period.

None of which is especially relevant to the rug pull business, except in the sense that investors have seen the rug pull work and they are seeing that wild ride of the SV VCs is coming to an end, either through regulation or collapse, so they are doing it sooner.

That "rug pull" you and other HN commentators talk about has happened after there is a need for cost cutting (eg. IBM wanting RoI after an expensive M&A deal, late stage Confluent investors wanting stronger returns, pissed off activist investors in MongoDB and Hashicorp annoyed that the share value has dropped despite core revenue fundamentals being consistent).

A lot of these OSS-core companies are extremely late stage or already public, so they are at the stage where investors wish to get an RoI out of them.

> seeing that wild ride of the SV VCs is coming to an end

Most of these OSS-core companies switching to restrictive licensing are towards the later stage/age of a company - around 10-15 years old. Their late stage investors (if their a startup) and largest shareholders (if public) tend to be Growth Funds and Mutual Funds - both financial instruments that have different goals and return criterias compared to Venture Capital.

> a need for cost cutting

But then you go on to mention investor returns

> the stage where investors wish to get an RoI out of them.

Exactly describing the rug pull or, as Cory Doctorow deemed it, the final stage of enshittification. Clawing back all the value provided to customers to hand to the investors. See for example Google's two $70 billion buybacks at the same time they were joining the rest of Big Tech in Big Layoffs.

I think there is a lot of confusion (or misrepresentation) around Terraform and its license change:

1. There is Terraform project (that used to be open source and now BUSL), and there is Terraform Cloud SaaS (that was never open source)

2. Terraform project - includes CLI tool and HCL language. Terraform Cloud - is a hosted service that adds "enterprise" capabilities around Terraform.

3. If Terraform Cloud would also be an open source and people host it / repackage it, etc., and Hashi would change its license, that would be a valid comparison to Elastic and other license changes and would be reasonable. Changing a license on a CLI tool / language interpreter - is unprecedented. IANAL, but it's also unprecedented to make your own definitions of what embedded or hosted means - and if you read Hashi FAQ about license - it's just a bunch of nonsense (specifically in regards to what embedded and hosted means in the context of Terraform)

4. The only reason there are Terraform Cloud competitors (Like Scalr, Spacelift, Env0, Terrateam, and others) is because of how deficient Terraform Cloud is and Hashi's inability to execute properly. If Hashi would listen to their customers and customer success team and do what customers are asking - there is no way any of TFC's competitors would gain any ground. They are in full control of Terraform roadmap and prioritization - this gives a huge unfair advantage to them compared to any other competitor.

You "kill" competition and win the market by innovating and creating a superior product (talking about Terraform Cloud) and not by monopolizing market via license change and legal nonsense.

Disclaimer: I work for Scalr.

"CEO predicts that his unpopular gambit will pay off"

In other news, water is wet.

It has not gone well for Unity

Hopefully not for Terraform

I think what a lot of people are missing is to do the backwards induction from where companies predict how things will go back to the present where they develop open source software and release it.

Previously, the belief was that you'd open source the software and then sell support since you knew the code the best (even though it's open source, the knowledge in the brains of the developers working on the code was not open, and you could extract value by leveraging that knowledge). That's the RedHat model and... well it's pretty dead. No startups today are predicting that will be the profitability scenario for their company since the paragon of the model, RedHat, had to sell themselves to IBM.

The next profitability model was open core, where you open source the main thing and sell integrated services and closed source enterprise features, etc. This model was more promising but it is squashed by companies like Amazon, or even other startups who can wait to see which open source products win, and then compete on those services etc. The main difficulty here is that open core isn't "commoditizing your complement" it's commoditizing your main thing and then being left with nice-to-haves as your business.

So a startup today has to look at this future and decide "open or closed source?" and if they look down the open source future and see no profitability, they just... never take that path. People are looking at companies switching their licenses today and saying it's greed, etc, but the difference is that those open source products exist at all due to a company building it. In the future, without an answer to open source profitability, there just won't be any companies producing open source software.

Open source software might still come from:

* bigcos with extra cash around commoditizing their complements: open sourcing non-core pieces for the purposes of competitive advantage. (e.g. kubernetes and bazel)

* hobbyists building things they personally enjoy working on. This usually precludes solving gnarly inelegant problem spaces like terraform (keep up with all cloud APIs in near real time... at night after your real job!). It also tends to mean there's no particular plan for maintenance or support. I love hobbyist open source, it's wonderful, but there are definitely things it consistently doesn't do well.

>That's the RedHat model and... well it's pretty dead. No startups today are predicting that will be the profitability scenario for their company since the paragon of the model, RedHat, had to sell themselves to IBM.

RedHat had $3 billion in gross revenue ($434 million net) in 2018 before the IBM acquisition and that was UP 15% from the previous year.

They didn't have to sell themselves. They chose to.

env0 CEO here. From the article:

He added that it offered to carry on working with the four main companies affected by the switch, saying “You just have to bear some of the r&d costs. And they were like, ‘No, no, we're gonna do something else’. Which is fine.”

I want to clarify that nobody ever approached us (env0) from Hashicorp.

Same for Spacelift (co-founder here).

This has been brewing for a very long time, and he's by no means the only person or even the best person making these points.

Open source dates back to the 1980s but really took flight in the 1990s with Linux, Linux distributions, and the mainstreaming of the Internet. Back then it was mostly a movement for software freedom and openness in contrast to closed source software like Windows and macOS. While it obviously didn't displace closed source software, it definitely played a huge role in preventing the Internet from being "embraced, extended, and extinguished" by Microsoft among others.

Without the open source movement we might be using the Microsoft Enterprise Internet with IIS as the only viable web server and NT as the only viable server OS. Imagine that hellscape. We also wouldn't have modern cloud, single board computers with reasonably open software, and loads of other things.

Yet the world has changed radically since the 1990s. Today the major form of closed software is cloud SaaS. SaaS usually runs on top of open source software but in terms of openness and freedom it's a profoundly more closed model than old-school closed source. SaaS is the ultimate in closed. You get to understand and control nothing, not even your own data in most cases.

Against this model open source software is offering zero resistance. In fact it's helping facilitate this ultimate lock-down of compute by providing free labor to SaaS companies. That's its primary role now.

I've seen this situation developing for going on fifteen years, but it seems like it's difficult to get the open source community to even consider the issue. The mentality is completely stuck in the 1990s.

> You get to understand and control nothing, not even your own data in most cases.

Well said.

80's or 90's open source? There was no effective internet. There were only BBSes and Compuserve. FOSS took off in the 00's.

> 80's or 90's open source? There was no effective internet. There were only BBSes and Compuserve. FOSS took off in the 00's.

https://en.wikipedia.org/wiki/Berkeley_Software_Distribution

Initial release: 1978

Just because there was no effective internet doesn't mean computers didn't exist and people weren't working on projects together.

> Against this model open source software is offering zero resistance.

Pushover licenses like MIT offer zero resistance to it, but that's just a reason to use better licenses like AGPLv3 instead, not to abandon open source entirely.

AGPL is very clearly a nonfree license, despite what the FSF says.

Please elaborate

https://news.ycombinator.com/item?id=35352641

https://news.ycombinator.com/item?id=30495647

https://news.ycombinator.com/item?id=30044019

The OSI and DFSG are both fine with the AGPL too, so it's hard to take that claim seriously, especially as a bare assertion with zero evidence.

AGPL: If you want to host it you have to pay.

SSPL: If you want to host it you have to pay.

BUSL: If you want to host it you have to pay.

Yet one is OSI-approved and the others aren't.

More like:

AGPL: If you want to host it, go ahead. If you want to host a modified version of it, provide source.

SSPL: If you want to host it, release your entire stack under the SSPL first. Good luck with that, lol.

BUSL: If you want to host it, you better not be making any money from it.

... and this is the problem.

Any license that doesn't help feed the non-free SaaS ecosystem is a non-free license.

I agree. SaaS isn’t a loophole, it’s one of the legitimate things you can do with software freedom, in letter and in spirit.

Software is not services, and attempting to conflate them is, at the very least, logically flawed.

The structural economy we're creating is one where privacy, autonomy, and software freedom are only for nerds. Everyone else uses services and surrenders all privacy and usually control of their own data.

That's because making software usable is extremely difficult and expensive. If software must be free but services are paid, all the funding goes to make services usable and not software. This results in an ecosystem where only highly technical people can own their own compute.

Is that what we want?

You have to think in a whole-systems way not in terms of single isolated issues. We're making free bricks to build a prison and reasoning that this is fine because bricks should be free and people should be free to build prisons with them.

It might come down to the question of what open source is about. If it's about creating a software ecosystem for nerds to have freedom, I'd argue that it's been successful in that endeavor. If there is any goal of freedom for the rest of humanity, it has been a failure.

Are we a guild that cares only for ourselves and our profession?

> Microsoft Enterprise Internet with IIS as the only viable web server and NT as the only viable server OS. Imagine that hellscap

Easy to imagine. That is the mobile space. Two suppliers, but a hellscape nonetheless

The Open Innovation model described in the book by Henry Chesbrough was super charged by Open Source. I think moving out of the open source model is a mistake that would make things move slow for the ones not playing it. Nowadays, when I think in software that I have to build with and have no access to source, all I am reminded is the bureaucracy of dealing with licenses where a much preferable approach is to have payments for support disconnected from this. If this is really some way things will go, I wonder if someone will make standardized drm for software libraries. Honestly, I hope I don’t have to develop with these things.

I sure wonder what they expected to happen when they changed the license. They were aiming to destroy the businesses of competing services that had grown up around their open source project. Were they imagining that the people running those companies would just sadly close up shop and the customers would eagerly jump to HashiCorp's bosom?

The reality was that those companies exist because there's a gap in their offerings and plenty of demand for alternatives. Once that dynamic exists it's going to try to protect itself too.

My point is this was all so far predictable. I would love to know what they thought would happen instead.

OSS and free is used by big tech as predatory pricing to kill or make unviable most of the competition.

Many things out there could be medium or big sized companies if Big Tech did not kill their business model by offering a good enough solution for free in order to drive the customer to other products/ services.

The problem is that HashiCorp also benefits a lot from the OSS community. But somehow I think they want to have their cake and eat it too. Which I don't think it is going to happen. It is not ok to take in contributions from everyone but only allow the usage for non-commercial purposes.

What is most likely to happen is to happen is that the terraform fork will become a new provider for cloud provisioning and the Hashi maintained one will become another provider, and the features and capabilities will diverge over time.

"Benefits a lot" is an understatement. Terraform would be nothing today without the countless hours spent by the community writing 5000+ providers for free (Hashicorp themselves have written 35 first party ones).

Note: the title is missing "unless the open source model evolves".

That is the interesting part indeed. Perhaps there is a place for another type of license in between gpl/mit and full proprietary.

I wonder what that would look like. Perhaps it’s time. But the grey area is just so vast.

I'd imagine it looks a lot like BUSL: source is available, and fine for use in non-competing ways, but using it to compete against the commercialisation efforts that pay for primary development is not allowed. Basically every successful open source company other than Red Hat at this point has the model in one way or another.

That sounds reasonable indeed. But competing against the commercialization efforts is just so vague. It’s also a huge turn off for enterprises, they are already afraid of the GPL, I think enterprises will not touch BUSL code.

It’s nice for self hosters perhaps. But I for one am always thinking about how I can commercialize my self hosting skills (and have been somewhat successful here and there).

I don't know I'd agree that it's vague - the triggering conditions in BUSL are instance specific so each instance requires evaluation. The HashiCorp provisions are perfectly clear.

What I do wish is that all companies who have this kind of license would also have a clearly priced, clearly defined "pay us X and you can use the OSS bits under this commercial license instead" model, for exactly the reason you describe though.

Enterprises are afraid of the GPL for libraries or OSS stuff they’re running but not paying for.

If they’re paying for a service because they need it (not so they can resell or repackage it in some way), I don’t think they’re going to be so picky about the license. They pay for plenty of 100% closed and proprietary software after all. The benefit of OSS/SA licenses in that case is debugging, being able to contribute patches, and continuity if the vendor goes bust. For those purposes, having access to the source is what matters, not the particular license.

I submitted this, and opted to elide that part of the title, both due to wordiness and also because it doesn't change the headline: the "evolution of model" that McJannet is advocating is not in fact open source. He is, therefore, predicting an "OSS-free Silicon Valley" -- one way or the other. (I also think he's entirely wrong.)

The title is misleading.

> He said the Linux Foundation’s adoption of Open Tofu raised serious questions. “What does it say for the future of open source, if foundations will just take it and give it a home. That is tragic for open source innovation. I will tell you, if that were to happen, there'll be no more open source companies in Silicon Valley.”

When I think of OSS companies, I think of financially-precarious companies like Mozilla (nonprofit so it’s kind of maneuverable), and companies like RedHat and Canonical that rely solely on donations/enterprise support contracts.

So I would suppose that his take warrants a “duh.”

> RedHat

This is now IBM.

Also donations =/= support contracts from a revenue generation standpoint (mostly looking at Canonical in this case).

Companies like Hashicorp are better compared with other OSS core companies like Confluent, Kong, Sysdig, and Hortonworks (talk about a throwback).

It's a subsidiary of IBM, and if I had said "IBM" instead of "RedHat" nobody would understand what I was referring to.

Fair enough!

Twitter, Slack, Discord, Youtube, Gmail, Google search, JetBrains software and all the other good daily drivers aren't open source. Games we play usually aren't opensource either. PS5 and Xbox aren't opensource, as well as Nintendo. As long as those tools provides you with some "free" version (doesn't matter if you're product) then it's perfectly fine. And it seems that with all the non opensource versions you get a better product overall. And I'd rather play the better game, use the better search and write in better editor, than go for opensource version, which source I would probably never read.

> JetBrains software

I think this falls into the "we know what you meant," but that's not systemically true of JetBrains; quite a bit of their stuff is Apache 2: https://github.com/orgs/JetBrains/repositories

They have done some underhanded shit with some of the plugins but via acqui-hires, not rug-pull-ing their repos from an open license, so I wouldn't lump them into the same category as closed source software, nor the vile behavior of Hashicorp, Sentry, or Sourcegraph

I think the key point here is actually from the first paragraph.

> HashiCorp’s CEO predicted there would be “no more open source companies in Silicon Valley” unless the community rethinks how it protects innovation.

Open source is literally a license to give the power to the people, the community, such that we allow anyone to contribute and give back. It is not a tool to "protect (your) innovation". It is a very self oriented move built on the greed inherent in capitalism. Basically the opposite of open source and tangentially copyleft.

I think a lot of people forget, especially the newer generations, that open source is not some "good boy points" or graceful gesture you do to others. It's not about you doing a "kind gesture" to the community, open source is a political move to use copyright law against itself to give the ownership right to anyone. Many seem to look past that, and try to seek out a middle ground between "my innovation" and getting the usage/praise from having the code be labelled as open-source. You can even see the Hashicorp CEO misuse the open-source label pretty regularly.

There is nothing inherently wrong with HashiCorp's license. What IS wrong is building a product and community and accepting contributions based on an OSS license and THEN changing that license when it no longer suits them. The reality is that whether it was HashiCorp, Elastic, Mongo or any of the others that have gone down this path, if they had just started out with their BSL-type license they likely never would have become successful and none of this would matter. Instead, they achieve success with an OSS license, then demonize their community and other companies that comply with that license but take away some of their revenue opportunities.

I have been involved in OSS companies. These risks are all well known when the product is starting its life, and yeah it sucks when someone comes along and does a better job monetizing the product you created. That said, it is just so disingenuous for these companies to whine about it later. Do a better job. Build a better product and service. Find ways to add value you can monetize. Do not blame your community because your VC-funded company is not hitting the numbers they want.

> There is nothing inherently wrong with HashiCorp's license.

Proprietary software licenses are inherently wrong, though.

This is contested.

Reminds me of this post about how open source means surrendering your monopoly over commercial exploitation.

https://drewdevault.com/2021/01/20/FOSS-is-to-surrender-your...

Ex- HashiCorp customer here. HashiCorp has some great products. but as a company they were bad over the years, and even worse over time. Bad culture, some stupid employees, bad terms, increasing pricing and not being transparent. I don't think their CEO understands the market. or understand what customers need. I think the company is dying , or will be acquired next year.

As for us - we just moved to their OSS Terraform + Vault and it's even better than how it was when we had their Enterprise. Have I said that company has no future?

It would be so lovely to see a return of direct to consumer OSS. OSS being so industrially purposed has kind of obscured some of the noble cause of the project, has made the virtue of sharing & cooperation less evident. That's easy to let ride when it's all just devs consuming packages. But once software starts being user facing I feel like the closed proprietary path is much more noticeable of a boundary.

It seems like there are a crop of new major OSS projects that are thriving, particularly direct to user projects like: Godot, Krita, multiple projects in the fediverse, activity pub, mastodon space.

I believe most of Dave McJannet's opinions are targeted to $HCP traders to have them believe the company is going to skyrocket its value.

Technical people knows that HashiCorp tools are good but there are alternatives and any tool can be replaced with moderated effort.

And I can predict he'll end up like that recent Unity CEO: fired by a small board meeting, because Hashicorp basically killed their own product through OSS rejection.

    IaaC

Do they mean "Infrastructure as as Code"?

I think "infrastructure as a code". I've seen this multiple times already, eg. [here](https://tenesys.io/en/technology/infrastructure-as-a-code-ia...).

Reply on news.ycombinator.com