I was setting up a new instance in my home lab, to replace an existing one, and was unable to configure OIDC. This was definitely disappointing as I’ve been using OIDC in the past with the previous install. I’ve worked around it by installing an older version.
Hi everyone,
I’ve been following this thread closely. While I completely understand the initial community reaction whenever a previously open feature gets fenced off, I feel the current debate is missing the elephant in the room. Calling this “enshittification” does a disservice to what Grist Labs has built and the reality of keeping independent open-source software alive.
Let’s look at the facts: Grist is a tiny team of roughly 17 people. Yet, they have engineered an extraordinarily mature, stable, and well-architected piece of software. In an era where venture-backed SaaS tools rush to ship half-baked, vendor-locked bloatware, Grist has delivered solid, gradual, Proxmox-like engineering quality. Better yet: they chose SQLite as their core data format. That alone means they have given us the ultimate guarantee of data sovereignty. Our data is never trapped in a proprietary black box, and it will still be readable in 2050.
Now to the controversial topic, SSO. Ask yourselves, if you were running an independent 17-person company trying to survive without selling out to predatory VC funds, what else could you possibly gate to convince institutional procurement departments to pay?
In the public sector and enterprise world (I work at a German public research institute), procurement rules are cynical. If a tool works 100% for free, administrators are legally and culturally discouraged from spending a single cent on support (I talk about this almost daily with our IT department - we hate it!). Institutes simply will not pay out of goodwill. Gating SSO is literally the only lever left in modern OSS that forces enterprise legal and procurement departments to actually write a purchase order. And let’s be fair:
- Grist Labs didn’t pull a rug-pull on existing deployments; existing instances keep working.
- Capable sysadmins and homelabbers who run small internal setups still have reliable, native alternatives like reverse-proxy forwarded headers (GRIST_FORWARD_AUTH_HEADER) behind Authelia, OAuth2-Proxy, or Nginx.
- Small teams and non-profits under $1M can still apply for free keys.
The real problem here isn’t Grist Labs trying to stay solvent. The real problem is that European public administration and enterprises expect a tiny boutique team across the Atlantic to fund our digital sovereignty for free.
Grist is quietly becoming critical infrastructure. In France, the DINUM and ANCT recognized this early and stepped up to co-develop features. In Germany, initiatives like ZenDiS (openDesk) are looking for sovereign building blocks to break free from Microsoft. Two major European states should easily be able to co-sponsor or set up structured public-sector support agreements for a 17-person team to secure this software as a public digital commons.
If we want Grist to remain independent - rather than being forced to sell to an aggressive private equity firm that will actually enshittify it - we have to accept that sustainability costs money.
Thank you, @anais-grist, @paul-grist, and the entire Grist team, for keeping the core open, keeping our data in open SQLite, and navigating this difficult topic. You have allies in the public and research sector who see the bigger picture.
*10-person team 
Thank you for this deeper thinking on the pressures that force such decisions.
I’d like to note one thing. You said
Grist is a tiny team of roughly 17* people. Yet, they have engineered an extraordinarily mature, stable, and well-architected piece of software. In an era where venture-backed SaaS tools rush to ship half-baked, vendor-locked bloatware, Grist has delivered solid, gradual, Proxmox-like engineering quality.
This has only been possible because of @dmitry-grist , co-founder and Head of Product. He initially bootstrapped the project through a long development phase. He had vision, commitment, patience, and care to make choices like using SQLite as the core data format.
Fun fact, from our about page (Bot Verification)
Grist Labs was founded shortly after mass surveillance disclosures exposed the vulnerability of even the most respected cloud services. The best remedy we knew was open-source code, transparent security practices, and enabling users to use their own infrastructure.
Grist was originally envisioned as installable software with end-to-end encrypted sharing – software that users can trust. Users would have full control of data, and would not need to trust the cloud. A SaaS service, however, is simpler for users, easier to start with, and easier to share with others. That’s what we released.
@Alexander_Dunkel You raised some valid points. I should have been more careful by using the word “enshittification”.
Grist is not “enshittified”. But the user base is, based on many previous experiences, working with a constant “fear of” of a product becoming this way.
You are right at identifying the core values and properties of Grist as a perfect fit for a more open digital infrastructure. Having Grist as as part of ZenDis would be a dream. We need sustainable ways of funding the development and general existence of Grist. Including it in a large state-wide initiative would ensure that.
But there is another side to it. @TomNit used the fitting words “grassroots movement”. We are in a phase where a part of the user base is working towards promoting and establishing Grist as a feasible alternative to closed-source software such as Microsoft products or others.
Nobody wants to lose the time, nerves and trust that has been invested so far.
Seeing something that even resembles “enshittification” activates a defensive behaviour. If that reaction is unjustified, it would be a good thing.
Personally, I am only an amateur, but also an enthusiast and idealist. I like Grist. I want this to work out.
I’ll keep it brief, let me just say this: Nobody is denying that GristLabs needs to make actual money. This should not be a discussion that’s effectively just the community clamouring for more free stuff. The point I was trying to make is that, the public sector situation being what it is at least in my country, Grist might have a shot at succeeding here the way it is in France - and by this I mean German public administration branches would become paying customers -, but only if the open-core promise remains solid and trustworthy until we’ve reached that point. Of course, GristLabs is in no position to change the system regarding digitization in our public administration system, and nor would anyone expect them to address the frankly ridiculous hurdles it imposes. I and many other small cogs in this machine aren’t in any position to do that either, unfortunately, and so the “grassroots” style efforts I was referring to earlier are pretty much the only option for getting a supremely useful product like Grist into service with us. I just think the Grist team need to be aware of this and tread correspondingly carefully around issues like this, if they’re interested in us as a market. It is entirely up to them, of course. I’m merely begging them to consider this point of view because I’d really like Grist to succeed over here. It’s a top-notch product with an admirable team of professionals behind it. We could really, really use it.
Best,
Tom
This sentence is the core of a debate that I think went completely off the rails in the past decade (the reason why I argue against using “open source sustainability” at all).
A way to fund Grist already exists, and that is to pay Grist Labs for the work done by our product team, our engineers, our marketers, our support, our lawyers, our doc writers, etc. Grist is not like Linux, Kubernetes or GNOME. Those are large collaborative consortiums that produce semi-finished products that can be refined and sold with different market strategies (RHEL, Rancher, Ubuntu etc). Grist is more like OpenProject, Element, and Passbolt: individual companies make the main investments to develop a product and bring it to market.
On the demand side, one must consider that corporations, public administrations, and individual people don’t pay for something that can be infinitely replicated at marginal cost (infinite supply of anything drives price to zero: it’s economics 101). Grist Labs, like other developers of applications in its class, needs to find the thing that cannot be infinitely replicated and that customers want to pay for. There is no generic “sustainability” problem: it’s a business model problem. I get it: as a user, I prefer to have everything for free (and now that I can afford it, I donate to projects I use daily). But open source can’t rely on donations only (I wrote about this topic, too).
In our experience, the path to a sale that you recommend (start free, show the value, ask the users to pay) works very rarely in practice. Too rarely to guarantee we pay our employees at the end of the month. How can the person in charge of IT budget justify a new expense to the head of procurement/VP Finance for something that has already proven to be working for €0/year? All buyers know that they have to pay for software, and we’re adjusting our tactics as we learn what customers want to pay for in exchange for the good value they get from Grist.
We also want this to work out! We know your advocacy helps. We’re aware of the value created by the members of the Grist community. We’re in sales conversations with many public administrations in Germany and other countries, thanks to users turned advocates: you found a good product that is easy to run and use, and you tell others about it. Thank you! We’re giving full Grist for free to most of you, on top of the powerful core open source application.
Of course that’s perfectly understandable. What I don’t quite understand is why the way to deal with this has to be gating off features that already exist. That’s basically the only issue here.
I get it and I respect the fact that you’re bound to have more experience with this than a lone public admin worker like me! 
But that said, I still find it hard to believe. Of course, organisations won’t start paying for something that already works, as you say. But then again, in an open-core model like Grist’s, shouldn’t the premise be that the commercial offering just offers more – as opposed to being the only really workable setup? Once the usefulness of a low-code DIY platform like this has been demonstrated (and has duly impressed people!), isn’t it a viable sales pitch to say “Look how much further you can go with this, now that we’ve opened your mind?” Before this gating move just now, that seemed to me the exact road GristLabs were taking. New and exciting features like automation were announced and placed behind the paywall. I fully understand and support that. But why gate off things that were there before?