Funding, Focus, and What's Next? #17977
Replies: 60 comments 198 replies
-
|
Sharing my two cents here.. I am also a user of https://www.odoo.com/ which is an open source product too. What odoo team is offering are :
Running an open source project is beyond my experience, but I'm willing to pay to get services if Directus plays a great role in the projects of my company. I think Directus will. |
Beta Was this translation helpful? Give feedback.
-
|
@rijkvanzanten @benhaynes , additional info to @xu4wang Odoo.sh is exactly what you're looking for if you're interested. We're doing ERP customizations to Odoo so I found it's really useful, business customer focus. Basically, it allow you to do customization on your own, make pipelines,...etc. Whenever you commit, it will run build new code and serve to customer. So, no need to worries about CICD, auto scale, DB,...infrastructure in general. Pricing scheme:
|
Beta Was this translation helpful? Give feedback.
-
|
We've played a very similar game up until now actually 🙂 That being said, that does pull the focus away from building the core product into focussing on parts that only a (small) percentage of the total user base will ever see 🤔 A great example is the kanban layout. We spent quite a good amount of time on designing and building one, but people don't really get to use it #10380 which is then additionally painful when I see comments like 😅 |
Beta Was this translation helpful? Give feedback.
-
|
Odoo community version is crippled beyond usefulness. https://www.odoo.com/page/editions Also Odoo is not a good example as it is software tailored exclusively to companies. It's an ERP after all. |
Beta Was this translation helpful? Give feedback.
-
True. Odoo is with a lot of vertical/industry related solutions which can be used by a company without too many customization. This will make odoo relatively easier in finding customers. Directus is more developer focused. |
Beta Was this translation helpful? Give feedback.
-
|
Just throwing in some of my thoughts... When I first started using Directus I was so impressed with the features and how easy it was to build my backend. A key thing for me (and I'm sure a lot of others) was the fact that it was open source and I'm able to extend upon it and have access to my raw data/db whenever I needed. Compared to other Backend as a Service companies where you have less control and you aren't able to export your data easily (8base, Hygraph etc) If I'm being honest, I've (shamefully) never donated toward an open source project. Not because I don't appreciate the software or that I don't want to contribute in any way, I just find it a bit alien because it's something I've never done before. If I were to contribute to something, it would likely be a library with a very small team (1-2 people). Not to take anything away from you guys, but if I see a company has had funding then I'm even less likely to contribute because on the surface it looks like you already have the money you need to grow. I think Directus Cloud is a great idea. People don't have to worry about the hassle of managing a server, backups etc. And they can feel like they're somewhat supporting the growth of the business by subscribing. I'm currently paying $30+ a month for basic app hosting on Digital Ocean where I would be much happier paying it to you. However, your lowest cloud hosting price is currently triple the price and we'd be unable to add our own custom extensions. I may be alone in my opinions, but you guys have done such a great job building the software that I wanted to share my thoughts to see if it could help. |
Beta Was this translation helpful? Give feedback.
-
Thanks! ❤️
I think this perfectly perfectly perfectly highlight one of the massive challenges of making open source sustainable at scale! It's a chicken-egg problem isn't it? 😄 You need funding to be able to maintain the project, but by getting funding, people stop funding because you "already have funding" 🙈
I still think so too! That latter bit doesn't really happen though.. People (especially businesses) don't pay for things out of the goodness of their hearts.. 😬
Yes! The answer is both complicated and very simple: the infrastructure we can offer at that price point isn't very reliable. This might feel like a "well yeah, cause you're on the cheap plan" type of deal (we thought the same!) but what ended up happening is that everybody defaulted to the cheaper version (because it's cheap!) and then run into the limitations of the infra that you get on that plan, and then get annoyed with the service. Now that problem can partially be solved by educating the users up front more (by having Directus itself gracefully fail earlier, or enforcing harder limits to prevent the servers from being overloaded), but that doesn't necessarily fully solve that negative feedback loop we were creating on the cheap plan. As of right now it made more sense to only offer the "production ready" tier that includes all the scaling and securities we need to be able to properly support it at scale.
Yeah same here! But.. see the answer above 😓
Fully agree! Funnily enough this isn't as much of an upsell as it is a technical limit that we're solving 🙂 The main (tech) reason that only enterprise has extensions atm is that the enterprise tier gets fully dedicated infrastructure, where standard is sharing some resources. It's the sharing of resources that makes allowing blanket extensions a tad security sensitive, so we wanna make sure we really nail the security aspect of that before GA launching it to the whole of cloud 🙂
Thanks! And yes it helps a lot 🙂 |
Beta Was this translation helpful? Give feedback.
-
Transparency might mitigate that problem. https://buffer.com/transparent-metrics |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
hope you will solve this problem. great product. |
Beta Was this translation helpful? Give feedback.
-
|
The best way to help the ecosystem based on market data is to switch to a Business Source License (BSL 1.1) that allows developers to still get free production usage, while requiring a paid license from large companies and governments. This incentivizes contribution, while still providing free access to hobby projects and small businesses, while generating revenue to maintain the open source project. |
Beta Was this translation helpful? Give feedback.
-
|
I think this thread highlights one of the trickiest questions to answer for this discussion: How can you objectively measure the "value" of a Directus instance? |
Beta Was this translation helpful? Give feedback.
-
|
@rijkvanzanten , I guess it's time for you ask "hey, ChatGPT! Can you, please, give a formula to find a free tier limit of a product considering numbers of users and the revenue of the company using it?" |
Beta Was this translation helpful? Give feedback.
-
|
Revenue seems like a good match for BSL.
How would you measure that for a self-hosted instance? |
Beta Was this translation helpful? Give feedback.
-
Great great question. Tbh I think this whole thing would realistically have to start as an "honor" system. Once you self-host it, there's 10000 ways to workaround any sort of metrics gathering, as you own the systems 🤷🏻 For larger companies, the risk of legal action should be enough of an incentive to be honest |
Beta Was this translation helpful? Give feedback.
-
Plus most enterprises are behind firewalls so no need to cheat. |
Beta Was this translation helpful? Give feedback.
-
|
Directus is becoming popular among other things because of its openness and solutions, making it a paid system will make it lose popularity. Directus cloud would be great if it didn't cost a fortune, maybe in the U.S. or other richer countries $99 is not a lot, but for countries that are developing it is 1/4 of some people's paycheck, so at the start they give up using this solution. If on a private server for $ 30 it is possible to set up a large number of instances, why not give it for $ 2-3-5? Of course, with limits or pay per use. Under small projects it can give a lot and count on the effect of scale. An idea, maybe instead of pushing this type of solution it is better to design and finish extensions makretplace and charge 10-20-30% from the sale of licenses for add-ons. This will develop the project and also allow everyone to make money. I myself would be happy to buy extensions for board view like on directus cloud so the topic can be developed. |
Beta Was this translation helpful? Give feedback.
-
|
That's literally the Apple's business model... |
Beta Was this translation helpful? Give feedback.
-
.. Except they also have hardware sales, other services, and a monopoly on billions of devices 😄 |
Beta Was this translation helpful? Give feedback.
-
|
It is still a significant portion of their revenue and even more of profits. https://www.theverge.com/2021/1/8/22220873/apple-2020-app-store-revenue-60-billion-dollars |
Beta Was this translation helpful? Give feedback.
-
For sure! Welp, if the Directus Marketplace could just make *checks notes* 60 billion in revenue from the start, I think we know our answer 😄 |
Beta Was this translation helpful? Give feedback.
-
|
Wouldn't it be awesome? 😎 |
Beta Was this translation helpful? Give feedback.
-
|
Dual license trickery. I've paid for amcharts, fontawesome, and shopify themes after giving them a spin and realizing the pro features are worth it if they work as well as the free features. |
Beta Was this translation helpful? Give feedback.
-
|
My vote is for free personal use, and a license for commercial projects with e-mail support for a year – then a smaller charge for subsequent years to continue getting updates/support. Same licensing for extensions. Also echo some comments about cloud – I'd much rather use this than self host. But need a way to extend it and sync with local environment (all settings). Love Directus, I want to see you thrive! |
Beta Was this translation helpful? Give feedback.
-
|
Thanks! ❤️
Yesssssss! Also make sure to take a peak at #17495 and #13041 (comment) and leave your thoughts if you haven't already 😁 |
Beta Was this translation helpful? Give feedback.
-
|
Brainstorm a crazy idea I had. What if Directus put sale the features. Something like that: You announce, if 200 companies pay 100 dollars for feature x, it will start to be developed in the next cycle. I know that there is an option to pay for features, but what I'm proposing here is something similar to the collective purchase of a feature. I don't know the viability of this in your business model. But I wanted to leave this idea here, so that who knows, maybe it can be evolved into something that can contribute financially. |
Beta Was this translation helpful? Give feedback.
-
|
Good idea! I'll just have to stop myself from building something cool until enough people chime in haha 😁 |
Beta Was this translation helpful? Give feedback.
-
|
I've always agreed on collective funding of open source solutions. As of right now I would put between $500 and $1000 towards "future drafts", for sure. You could offer anonymous funding of features, or even "sponsored" public funding for companies who want it. I'm very confident that there are features many companies would sponsor if given the opportunity. |
Beta Was this translation helpful? Give feedback.
-
|
I'm grateful for every day Directus exists and I sincerely hope you find a satisfying solution. I feel however you can be more aggressive in asking users for funding. E.g. You could show a big friendly banner telling users (of a self-hosted admin application) that they are using free and open-source software. I'm thinking of the yearly Wikipedia funding rounds. I find them especially effective and I think they have proven to be very profitable for the Wikimedia Foundation. There is still such a small environment around Directus. I believe that's because there's a lot of agencies using Directus. It's their life-blood to sell their stuff. Which, for my part, belongs on the yet non-existent marketplace. If you would offer Directus Hosting with the same flexibility as open source and yet the same benefits as paid – we'd be your customer. Is a Docker style licensing something for Directus? Yearly revenue as a requirement for obtaining a commercial license? |
Beta Was this translation helpful? Give feedback.
-
|
"Congratulations! Your website reached 1 million users this month! It's time to give back to the community and buy a Directus license!" Of course, you can't simply turn of features until the license buying. |
Beta Was this translation helpful? Give feedback.
-
This is actually a very good point and something I'm hoping we can achieve through that BSL1.1 license. We could add a clause that allows any party under some sort of yearly financial threshold to use it completely for free (as per usual), but require the parties that can afford it to contribute. |
Beta Was this translation helpful? Give feedback.
-
|
I think there is a huge difference on how to make money between big companies and small companies / individual persons. For large companies the easiest would be to require them to pay (by license). For individuals it get's a bit more complicated but there's a lot of opportunities (and based on my experience most people are happy to pay if it saves time): Looking at cloud exclusive stuff like the kanban layout I'd be more than happy to pay for the extension. Same would be for more specific relational interfaces with predefined filters, and integrations (sync Data with xy, payment providers, ...). Another product could be starter-templates and complete (internal) apps. 1.) Starter-Templates 2.) Full Apps 3.) Third-party starters for directus
Another thing would be "verified" extensions. Those would be extensions that aren't build by core but verified to work and be secure. These could be sold in the upcoming marketplace and take a comision from every sell. For inspiration you'd just have to have a look on all the WordPress Plugins and Notion templates. |
Beta Was this translation helpful? Give feedback.
-
|
Honestly, to me rn, it's more important to make sure that the larger organizations contribute to the project. The individual devs (like you and me!) are contributing a lot in time and effort (even if not financially), but there's a vast majority of companies that use it for free, get a lot of value out of it, and never contribute back in any way (financial or otherwise).. |
Beta Was this translation helpful? Give feedback.
-
|
For those you can to the same. For example migrations are most likely still a problem in 98% of bigger project as there's no way to sync roles, permissions, settings, translations, etc. When building an advanced migration / sync tool I'm pretty sure there's a big need on larger projects where companies would pay for (or in this case have to pay for). ... these are just some Ideas that came into my mind (next to some custom license that required bigger companies to contribute. But this is already mentioned and you all probably have way more knowledge about those opportunities than I do 😂) |
Beta Was this translation helpful? Give feedback.
-
|
Rijk, I think this comment dovetails my my suggestion below about a Tailwind style license. You could extract fair value from the large companies who are free-riding, in a way they'd understand (pay to reduce risk) and in so doing wouldn't limit Directus from taking any future actions to tighten licenses furthe, such as subscription models |
Beta Was this translation helpful? Give feedback.
-
|
Maybe an option that has already been named. Next to that, maybe the marketplace can generate some revenue? |
Beta Was this translation helpful? Give feedback.
-
I think this kinda circles back to my original post above:
I'd really like to find a middleground where our core team can focus on.. building core 😄 Love that second idea tho! |
Beta Was this translation helpful? Give feedback.
-
|
I really like directus and want to give some advices on how to make the project financeable and continously grow.
Hope it helped 💪 |
Beta Was this translation helpful? Give feedback.
-
|
The marketplace aspect worries me because, over time, they are often filled with poorly maintained or unmaintained add-ons due to low profitability / lost of interests. I use Directus precisely because I can easily extend its functionality and maintain the extensions myself. On the other hand, it's up to the user to use it or not |
Beta Was this translation helpful? Give feedback.
-
|
If implemented correctly, a marketplace has a huge potential. Obviously there must be some level of security and quality control, especially for paid extensions. |
Beta Was this translation helpful? Give feedback.
-
|
I knew it was coming... was too good to be true :) My personal startup fully relies on Directus for its headless CMS part. I also work on a couple of projects at my full-time job where we use Directus (I introduced it to the company). While the company that I work for may be able to afford a paid plan (although it's a non profit org, so the funds are limited) my startup will certainly not be able to do it. I realize how much work you guys do and any good work must be compensated accordingly. For that reason I personally made small contributions in a form of PRs and a small donation, to show my appreciation. But it is very unlikely I will be able to pay for Enterprise plan for my startup to have custom extensions, which I currently have. Besides there are many use cases where you have to host your data on premises, not in the cloud. Data sovereignty rules in some of the countries where my nonprofit company operates is a good example. The ability to use the product for free for small startups, companies with a small revenue, or nonprofits sounds great. It will be interesting to see the criteria for qualifying for this. Not allowing to use the new versions for a few years on the other hand is a terrible idea on my opinion. In the era of technology when products are evolving on a literally hourly-basis, no one will want to use a few YEARS old software in production. A few years ago Directus was PHP based, who would want to use that version of Directus today except the legacy platforms that had that version installed few years ago and never migrated? Just think about it... I think a pricing model would be better if it
With all that said, my impression is that the problem with funding is not caused by your current monetization model as by marketing. There are several fully open source CMSs that make a huge profit from hosting and support. Drupal with its Acquia.com and Wordpress (wordpress.com) are great examples. In fact Aquia was the fastest growing company on East Coast a few years ago. If the pricing model was the cause of your funding problems, there would be more installations of self-hosted Directus instances than the paid subscriptions to your competitors, such as Strapi, Sanity, Contentful, and others. I don't think there is an easy way to count all active production instances of self-hosted Directus, but something tells me that number is way smaller than the number of paid subscriptions to those competitor services. And to be fair, Directus is superior to all those competitors when it comes to features and UX. So why is Directus struggling with funding? When I was looking for a headless CMS for my projects a couple of years ago, I saw all those competitors in Google search results and in various articles with reviews and comparisons, but it took me about a week of research to see the name "Directus" in of the articles. And I believe it was the only place where I saw it back then. The situation may have improved since then, but even today when someone searches for a headless CMS they would get all sorts of products in search results, but for some reason Directus is very often not one of them. So there is clearly a discoverability issue. Therefore I think the number one problem that Directus needs to solve is the discoverability and effective marketing. Being an owner of a startup myself I know very well how important this is. Once you improve your discoverability your funding will also improve IMHO. |
Beta Was this translation helpful? Give feedback.
-
Big thing to figure out! I'd think it could be around some concept of "total finances" to measure a companies size (kinda like how Unity prices their gaming-engine license), but we'll have to figure out what's fair.
Totally! I don't really see it as "not allowing" though, it would mean that code would be under BSL1.1 for however long we choose to go with, after which it becomes GPL-3 again. It's mostly just a tool to make sure that people that hit that financial threshold actually have a reason to pay for it! If we'd make that window too short, you'd end up with a "Why pay for it if I can just use a version that's a couple months old?" 🤔 |
Beta Was this translation helpful? Give feedback.
-
|
Yeah, that makes sense. But if that period is more than 1 year then the code will become obsolete. |
Beta Was this translation helpful? Give feedback.
-
|
I think @adanielyan has a good point on marketing. I'm only able to find Directus via Jamstack Survey 2022 -> Headless CMS -> 3rd options (https://jamstack.org/headless-cms/) while I am researching CMS for my company to use yesterday. It was a random hit where I feel other options are too constraint with just around content. Even Google doesn't have any result for Directus when I was searching (maybe it's in 2nd or 3rd page but you get my point). |
Beta Was this translation helpful? Give feedback.
-
|
Hi All,I think it's a good idea to keep Directus open source, many people put their time into this project to make it better. Still not everything works properly and it will be difficult to sell such a product. The software base you have achieved at the moment should remain under this license. Let's focus on the next features, price them and collect funds to implement them. Marketplace can be a source of income, or extended projects, for example: someone needs a system to manage a car fleet, rent apartments, or maintain tools in the company, then you offer the service of implementation and data transfer, for which a fee will be charged. |
Beta Was this translation helpful? Give feedback.
-
|
@koksikus this is also what I thought. ERPs are doing often the same, they offer the base software for free in open source and then (as they are the creators) leverage projects and offer implementation services or consulting or custom development. Might be the easiest version to scale. And with every custom project also new functionalities might be financed as the customer wants it and pays for it. @rijkvanzanten we (https://morethandigital.info/en/) are an NGO and we would be also looking for a solution to follow up on the current platform we have with a better platform to enable better knowledge sharing. Maybe a project for you that would also support your development? |
Beta Was this translation helpful? Give feedback.
-
|
The downside of that approach is what I outlined in my original post as well: it means we have to put our time and effort in things that aren't the Directus platform in order to fund the maintenance of the platform. We've been doing this for the last 6-8 months or so, and it doesn't really work. The Directus project itself needs way more love and attention to grow and thrive, which we're not able to give it if we have to spend our time on other things to generate revenue. Besides, by making parts paid, or making revenue from implementation / support, we're incentivized to make the core platform itself worse, to ensure that the revenue stream is protected. That's not really a situation I'd like to be in! |
Beta Was this translation helpful? Give feedback.
-
|
I do not agree that the soft in such a case will be especially difficult to implement or that the source will become worse. It will be up to you if you want to lose customers. The same can be said now that the wait for changes is very long, and this is because there would be more people willing to pay etc. You can look at it differently - will a large company want to implement the soft without a guarantee and stop production for a day or week in case of problems ... My company is not big but it is not thinking about introducing this soft because it has other licenses and no one knows it well enough to adapt it to the requirements. Perhaps a good option would be to have this software and, as part of the project and cooperation, implement it to a company that would otherwise never implement it. |
Beta Was this translation helpful? Give feedback.
-
|
Just thinking out loud, and this may have been mentioned already, but what about a Jetbrains style subscription approach for roll-your-own commercial instances, with a perpetual fallback license after a certain subscription length, leaving you with access to the last major version of Directus once your subscription ends? People tend to be more accepting of the model because you know once you're subscription ends, you're still left with a product. |
Beta Was this translation helpful? Give feedback.
-
|
Ohh kinda like Sketch where you only get "updates" as long as you're subscribed, but you can keep whatever the last version was you paid for after that forever? |
Beta Was this translation helpful? Give feedback.
-
|
Something like that. Jetbrains has a minimum subscription length (I believe 12 months) before you're entitled to the perpetual fallback license, which is a good way to discourage subscribing for a month just to get a license. |
Beta Was this translation helpful? Give feedback.
-
|
Ahhh I suppose in actual use that'd be very similar to having a license that "converts" to a proper open source license after a window, and then making sure that the minimum "duration" of a commercial license aligns with that window 🤔 |
Beta Was this translation helpful? Give feedback.
-
|
Now I put an ounce of thought into it, they are very similar. But instead of the codebase as a whole becoming open source after a period, the licensees license rolls over to open source. So you would have to have bought in at some point to qualify, as it were. EDIT: Maybe not rolls over to a traditional open source license, since they'd be able to freely release it at that point. But ya know... something more permissive/perpetual. |
Beta Was this translation helpful? Give feedback.
-
|
Yeah! The main difference being that you need to have bought in at least once in order to access the "forever" state of the project, rather than "just" using the outdated version 🤔 |
Beta Was this translation helpful? Give feedback.
-
|
Just getting into Directus & it is a total breath of fresh air amongst a sea of mediocre paid products & far less developed open source options. My use case would be a paid SaaS offering, which is currently utilising a very restrictive paid platform. I would be totally fine with paying for Directus, most likely by way of Directus Cloud. I think it is important to retain free commercial dev & hobbyist usage, but beyond that you either have to go down the path of Directus Cloud or pay for a license/subscription to host yourself (if using for commercial use). As much as open source is amazing, you guys have put a lot of time & money into this project, so it makes sense to be able to recover costs if someone is using it for profit. Whilst there are always new products entering the market, I currently do not see any that compete on the full feature set that Directus offers in an open source capacity or even paid. |
Beta Was this translation helpful? Give feedback.
-
|
I really like that you‘re putting your thoughts out in the open. 👍 Have you considered the Fair-Code option, developed by Jan Oberhauser of N8N? |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for your feedback! 👍 How would say this model compares to the BSL 1.1 license (advantages / disadvantages)? Sounds like the intentions are very similar. |
Beta Was this translation helpful? Give feedback.
-
|
Disclaimer: I‘m not at all a licensing specialist. 😉 It seems to me that, indeed, both are pursuing the same goal. The key difference appears to be the Change License used in BSL 1.1, whereas in the Fair Code model, the license remains the same, forever. Fair Code seems a bit more permissive (non-commercial use is also permitted in production), but such a right might also be permitted as an Additional Use Grant in BSL 1.1. Overall, I feel that BSL 1.1 is more flexible and probably a better fit for your goals. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you very much for sharing your insights and thoughts! This is much appreciated! |
Beta Was this translation helpful? Give feedback.
-
|
Hello. Directus is one of the coolest open source projects I've seen. I want to share my opinion about this whole situation. It seems to me that Directus should use a business model similar to that used by Vercel. they also has an open-source project that we all know called next.js. So, in my opinion, the success of Vercel lies in two things, what kind of user experience they provide, and how they push users to use their platform. Let's start with the first one, When the user sees "Create a collection" (maybe implement onboarding?) in front of him when opening the application, he understands that he now needs to learn something, do something, the whole wow effect disappears. A demo project that could previously be created in the free version of Directus cloud produced some kind of wow effect, but this could not be put into practice, so this effect quickly subsided. That is, my solution to the first point, obviously, as many have already said $129 per month, this is a lot compared to what competitors offer. I can use railway.app, get a Vercel-like experience, and host it almost for free. Directus cloud should obviously become usage based, similar to railway. The user experience should look like this: authorization with Github -> deploy template -> use in production. Three clicks and Directus solved the user's problem, and only then does the user learn about the enormous power of Directus (auto-generated API, extensions, and many more). About the second point, Vercel motivates the user to use them in a very elegant way, many features in next.js are specifically tied to platform. Yes, you can use the middleware just in the self host option, but it will not be an edge function. Yes, you can just store static assets on the simple server, but Vercel will place it on the CDN for you. Edge runtime, caching, cron jobs are just a few examples of why it's just easier to pay money to them than to host next.js yourself, although this is also very simple at first glance. If Directus Cloud makes life easier, doesn’t make users think about scaling, offers some bonuses like Vercel offers with next.js and has the ability to deploy a working and useful project in three clicks, then this will be a real success. And with such a service, it is much easier to conduct marketing, when you can easily demonstrate such a user experience, then, accordingly, it is easier to advertise it. Directus itself has long been ahead of competitors in terms of features, but the service we have now is far behind what we have on the market. Yes, I know, I mentioned Vercel here so many times, but in my opinion, these guys have done the best to build a business around an open-source project and I wish Directus the same success |
Beta Was this translation helpful? Give feedback.
-
|
The disconnect between development and production is the missing bit of the puzzle for me with Directus. Having a way to |
Beta Was this translation helpful? Give feedback.
-
While I understand where the comparison comes from, I don't really believe it's an apples-to-apples comparison. Vercel is a cloud-hosting Platform as a Service company first. Their focus and efforts have been to get as many people to host whatever they want on Vercel. They (correctly) noticed that websites / jamstack was the most popular thing that people deployed, so they took over maintenance/development of Next.js to own that pipeline end-to-end (similar to how Netlify took over Gatsby). Directus Cloud is different though: it's a SaaS version of the Directus platform. Next.js is a funnel to get people to use the paid product Vercel, whereas Directus would be that funnel and the paid product 🙂 It also circles back to the original post above though! By making Cloud hosting the primary business, we have to spend most of our time and effort on making cloud really good, instead of focussing that same effort on the project itself. This is effectively what we've been having to do for the last 6-8 months or so, which we immediately saw the downsides of (issues raised quick, couldn't keep up with PRs, had to reduce the amount of questions we can answer etc etc etc). Not a great spot to be in |
Beta Was this translation helpful? Give feedback.
-
|
I think your 3rd option sounds like a good one, Rijk ... makes simple economic sense to find a way to earn from those who are truly earning from implementing the software, ie enterprise users. How that is achieved is not simple I know. But as it dawns on people what an elegant and dynamic solution Directus is , for devs and for users, I feel sure huge success will come. |
Beta Was this translation helpful? Give feedback.
-
|
Option 3 sounds good. Option 2 is also fine, particularly if Directus continues its pattern of maintaining new features for Directus Cloud. It would have been fine for a feature like Flows to have been held in Directus Cloud for 2-3 years. Maybe that wouldn't have been possible, or maybe Flows were exclusive to Cloud for some time, but just as an example. An Extensions marketplace is desired. Totally understand the worry about its costs outstripping revenues. Since Directus is in part about democratization of data management, it would be particularly awesome to charge on API requests to a Directus Enterprise Cloud instance. The developer opens paid APIs on the Marketplace, Directus Cloud charges per request, and Directus receives a fraction of the revenues. This would create a positive feedback loop between development of Marketplace Extensions, and use of those Extensions with paid APIs. Would love this functionality for the purposes of publishing federated data models, where customers can use a Directus schema/model maintained by my company, self-host their data on Directus Cloud, and pay their fair share for transactions within the federation. Directus Cloud is then, ultimately, providing usage-based billing and data privacy to the customers of federations on your marketplace, while also collecting regular hosting fees. Super complex, but Directus is in one of the best positions to do this sort of thing. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks again for all the invaluable insights all ❤️ The BSL option has been released under v10.0.0, so I'll close this thread 🙂🐰✨ |
Beta Was this translation helpful? Give feedback.
-
|
Neither did I! I thought we made an error - TIL. |
Beta Was this translation helpful? Give feedback.
-
|
Price section for self-hosted version - think is good idea. To get in touch properly. Also, it would be good to know preliminary pricing details. Perhaps someone will want to compare this solution with others. |
Beta Was this translation helpful? Give feedback.
-
|
Price assumptions are not known i.e. |
Beta Was this translation helpful? Give feedback.
-
Self-hosted pricing is interesting. For 99% of users, it would say "Free". Then for the companies that pass the $5M threshold, it would say "Contact Us" since that essentially uses our Enterprise pricing (minus any hosting costs on our Cloud). We don't price based on users, etc... so there's a whole matrix we use to calculate project size. And we offer different terms (quarterly, yearly, multi-year) which also affect price. So that column might get really complex! |
Beta Was this translation helpful? Give feedback.
-
|
In the past, when I was looking at this type of projects, I always looked at the price section to see if I could afford it. |
Beta Was this translation helpful? Give feedback.
-
|
Congratulations on the step forward. I really appreciate the terms you’re offering. They seem fair to me ($5mil, 3 years to license change) and hopefully the community will agree! |
Beta Was this translation helpful? Give feedback.
-
|
Does this mean that Directus 9.x still uses the GPLv3 and this new license only applies with version 10.x forward? This means that interested parties can fork this old version and develop it further? Finally, I would like to thank you for the awesome product and congratulate on the good choice of license. Hopefully, this will inject some needed cash and make Directus development more sustainable. |
Beta Was this translation helpful? Give feedback.
-
That's correct, we're not retroactively relicensing older versions and we'll keep them up on all the usual distribution channels at well (it'd be unfair to try to remove those imo!)
Sure! You technically could, but in practice it's a ton of work to properly maintain the platform, and you wouldn't be able to sync in any changes from upstream which would mean you end up doing a lot of "double" work
Thanks! I'm hopeful 🙂 |
Beta Was this translation helpful? Give feedback.
-
|
@rijkvanzanten Do you plan to continue to publish security updates for Directus 9.x? |
Beta Was this translation helpful? Give feedback.
-
|
Due to our small team size and limited resources, the core team is only maintaining "Directus"... which is to say the leading release. In my opinion, there's no reason to be on anything except the latest version (other than to avoid the license). Not only is this important for features and fixes, but more importantly, to stay up to date with all security patches. |
Beta Was this translation helpful? Give feedback.
-
|
@benhaynes Thanks for your reply. To add some context, for consultancies this leads to quite some complex and potentially drawn out discussions with their clients (and ultimately you). By not releasing security releases for v9 this puts quite some time pressure on the situation where switching to another CMS or forking v9 might actually make more sense due to the immediate security considerations. The fact that this switch happens like this (without grace period so to speak) might also make some people or clients wary about wanting to license v10 if it's an indication of how future big changes are being handled. This also leads me to another question, do you have a support and maintenance roadmap or strategy with published EOL dates for future versions of Directus starting with 10.0? I haven't found anything on this. What about LTS versions? |
Beta Was this translation helpful? Give feedback.
-
|
Regarding the time pressure, there are no issues below the $5M threshold... and above that, our team is prepared to discuss very flexible terms that ensure there is no downtime or lag in staying on the latest release. We're certainly not holding anything hostage in v10... which is why we only pushed a license change so far. I'd urge you or anyone else curious about the commercial licenses above $5M total finances to talk to our Sales team. There's plenty of room to get users situated on v10/BSL in a timely way. Again, we're not trying to rush things, but as a small startup, we need to move quickly with these changes to actually see an outcome in a reasonable period of time. We've communicated our thoughts/intentions in real time... the whole process was just fast. As for the LTS/EOL side, as I said, we don't operate in that fashion due to limited resources. Much like a web browser, you simply stay up to date and patches aren't typically released for previous versions. I know Chrome technically has LTS support... but that's Google, we don't have their resources. When we're a bigger company, we can look into some of what you've mentioned... but as of now, we don't operate all that differently from the 2-person OSS org we were two years ago. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks again, to your latter point, I totally understand the issue of limited resources. In the past, as far as I'm informed, major version bumps often led to breaking changes with Directus. As such, since staying on "latest" is the only way to receive security updates there was some additional risk associated with choosing Directus for projects. Until now, for some this risk might have been worth it by Directus being free and the GPL license. From my perspective and experience, providing EOL timelines with regards to security patches is definitely something to consider when moving into the enterprise space because other solutions provide that. |
Beta Was this translation helpful? Give feedback.
-
|
Hello @hokkaido , |
Beta Was this translation helpful? Give feedback.
-
|
First and foremost, I want to express my gratitude for all the hard work you and the team have put into Directus so far. It is undeniable that the open-source licensing has played a significant role in reaching such a broad audience, and the collaborative spirit of the community has truly elevated the project. It is entirely understandable that, in order to maintain and grow a project of this scale, it must be financially sustainable. The need to adopt a different licensing strategy is both reasonable and, perhaps, inevitable. Your transparency in sharing the challenges you face and the options you've considered is greatly appreciated. The re-licensing approach after three years is an interesting solution. However, I would like to point out that such a lengthy period may potentially disengage open-source advocates who value unrestricted access to the software. This aspect might be worth reconsidering to strike the right balance between financial sustainability and maintaining the open-source spirit. Lastly, I believe that continuing to provide support and development for patches in version 9.X, addressing bugs and security issues, would be highly beneficial for the entire community. This would not only maintain the trust and support of existing users but also demonstrate your commitment to the project and its open-source roots. Once again, thank you for your dedication and for keeping the community informed about these critical decisions. Wishing you the best of luck in finding the right path forward for Directus. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you — I appreciate your kind words and thoughtful response. I'm curious, with the "Additional Use Grant" that allows anyone under $5M USD in total finances to use the platform completely free and open-source... what are your thoughts on how the typical Directus user will be impacted by this change? Or, are you more concerned about the affect on larger (greater than $5M) companies? |
Beta Was this translation helpful? Give feedback.
-
|
The solution you've found to maintain free usage for smaller businesses and individual users is truly ideal, especially if they represent the majority of your platform's user base. However, I would like to point out that there are institutions like the one I work for, a public university, which exceed the $5M USD threshold in annual budget but constantly need to cut costs due to a majority of the budget being allocated to unavoidable expenses. Open-source licensing can serve as an effective gateway to attract institutions like public universities and other organizations that may be more inclined to support and adopt platforms with a strong open-source foundation. By fostering an open-source community and maintaining a certain level of support for the GPL version, you can encourage these institutions to consider Directus as a viable solution and get them in the path to upgrade to the BSL version. I am not sure if Directus has received support from such entities, but their involvement in projects like Moodle or Sakai is significant. |
Beta Was this translation helpful? Give feedback.
-
What you're describing is effectively my proposed option 2! Have an open source "core", and limit certain parts/features under a different license. It's a shame GitHub auto-hides a large chunk of comments, but I do really recommend giving them a glance through. This topic has come up a couple times before in this thread, and there's some great back-and-forth discussions on the pros and cons of this approach 🙂 |
Beta Was this translation helpful? Give feedback.
-
|
Hello @sinmsinm , |
Beta Was this translation helpful? Give feedback.
-
|
I am in a very similar situation as @sinmsinm. I work for a large international development nonprofit organization that has hundreds of small projects in developing countries around the world, funded by different foundations and the US Government. Although our total annual revenue is way more than $5M, the funding for individual projects are usually small and most of the time do not cover software licensing fees, especially if the project was planned and funded before these fees were even known (as in case with directus v10). It means that a small project with a limited budget that has a large data collection activity in Uganda cannot afford paying $4,500 a year for storing their data on a platform that was supposed to be open source. The fact that the annual federal budget of the USA is 6 trillion dollars doesn't mean that every US citizen can afford a Lamborghini. |
Beta Was this translation helpful? Give feedback.
-
|
As someone who has done many different software and platform trade studies for the companies I work(ed) for (requirements, use cases, etc.), here are my two cents (as of now, price may change due to inflation). First, let me say that your software is awesome! As someone who puts a heavy weight on user experience and learning curves, you guys did a terrific job. The money problem is an understandable one and is definitely one that needs to be solved early to minimize impacts (after all, people gotta eat). While you might have sponsors and individual buyers, realistically speaking, your main source of income will come from companies that want to use your tool for data management and publishing. Taking that into account, here are some points that I've learned dealing with FOSS while at the same time trying to make the legal, contracts, procurement, and security folks happy:
A lot of FOSS project make the mistake of starting to adjust their licenses to "protect" their work from large companies. That is a bad idea. The last thing I want as an engineer is to go through a trade study, select a platform, and then have legal tell me no because the license of the software is restricted (a thing I learned the hard way with licenses such as the AGPL). For example, as part of the trade study that we run, we also do controlled (usually somewhat short) pilot runs to ensure that the teams like the software and it meets their needs without causing any unforeseen issues. Directus would be automatically removed from the list of potential software because of your license. My company's revenue is higher than the $5M limit and a pilot run, even though done for analysis, runs in a production environment. Even if I can convince legal to allow us to use the 12-month version, security would never allow it (not to mention the possibility of losing out on testing new features). If you're okay with one of the "standard" FOSS licenses (GPL, Apache, MIT, etc.), just use that without any asterisks.
Your Enterprise Cloud solution has exactly the stuff that enterprises are looking for and most individual users don't really care about (SSO, extended and long term support, etc.). The only issue (at least from what I see on the website) is that none of those features are present in the self-hosted version. Some companies (such as the one I work for) cannot use cloud solutions due to different reasons (including the type of data that is being process and where it's stored). If you haven't done so already, creating an addon/module to add those features to the self-hosted version for enterprise customers would be a good idea.
Compliance is big these days. Very big. Going off the previous point, since your platform is designed to deal with data, it would be a good idea for you to start looking into the different data accreditations for the different industries out there (FedRAM, HIPPA, etc.). Since there are a LOT of accreditations out there, you should first start with looking into integrating with platforms that are already accredited (e.g. AWS GovCloud). Compliance is a deep rabbit hole that people spend entire careers in so it's not something I'd recommend a startup to do (unless you have the resources), which brings up the point about using accredited platforms. These are just a few of the items I can think about now while drinking my morning coffee in the afternoon. If you have questions or would like to discuss further, let me know :) |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, @IvanDrag0 — I really appreciate your comprehensive thoughts on all of this. I'll give a few quick responses:
So, I guess my question to you is what are your thoughts on generating revenue in a meaningful way, and in a shorter timeframe? |
Beta Was this translation helpful? Give feedback.
-
My question would be, who are your general users/audience and who are your revenue generating users/audience? If you've added enterprise features such SSO to the FOSS version, then the question you have to ask is "Why would a company pay us?". Your enterprise version should contain the things that an enterprise needs/wants (RBAC, branding, auditing, SSO, rate limiting, etc.) and the average user can do without. Hiring developers is very expensive, so if SLA support is part of your enterprise offering, are your prices/processes competitive with just hiring a few devs? If you want to protect your software from companies that just use it locally without paying you for support, why the $5M cap? $5M is a decent size company. You'd need to either heavily drop the cap so that only individuals and small mom and pop shops would fall under that limit. Also, I would spend some time revamping your self-hosted FAQ section since I think that some of your answers/definitions are not helping your issue. Mainly your definition of "Production Use". For example:
Why is the second sentence there? Whether I modify your product or not has nothing to do with "production use". You should use a more "accepted" definition of production use (one that's similar to the ones found here).
Based on your definition above, a Fortune 100 company can use it internally (as long as they don't modify the software) and they won't violate the BSL clause (since they won't be using it in production).
By specifying "human" you're entering a gray area when you're starting to touch automation. In your case, an end-user should be any entity that interacts with your software (including software, platforms, etc.). Hope those points make some sense. |
Beta Was this translation helpful? Give feedback.
-
|
I wouldn’t have considered Directus if they paywalled security features like SSO, RBAC, auditing or rate limiting. Open source or not, making the internet less secure is not the way. |
Beta Was this translation helpful? Give feedback.
-
|
For use the cloud solution has always been a no go due to it not supporting extensions. I am very aware that Directus needs to earn food somehow, but I think touching the core might have been a hit and miss. Looking at other popular FOSS, wordpress, laravel etc. The core is fully FOSS, whilst the plugins, extensions and additional products are charged for. I think Directus core has a lot in place to support purchased extensions, (interfaces, modules, panels, layouts etc.) And could just lock it down and get the current set of features working 100%. Then spend time on creating quality extensions, which I think would be easier for an internal team to do since they already know a lot about how to create extensions. Then one could sell extensions at a premium to exactly the users that need. But in the end, making a proper living with FOSS is hard work, and in the current time and abundance of FOSS offerings it is hard so I'll say this that I respect Directus for even having this discussion in the open and taking the time to answer all these questions and "what if this" and "what about that". Good luck. |
Beta Was this translation helpful? Give feedback.
-
I think you've just answered why the license change has happened. Because they now legally have to...
Not adding SSO (and I'd add RBAC here as well) to the FOSS version is both counter-culture to the spirit of FOSS and makes software less secure. I am biased on this topic, but my hackles get raised when I see suggestions of SSO being an enterprise only feature. Small mum and pop stores, and even families hosting software for personal use can benefit from security that SSO brings. Check out https://sso.tax/ for more discourse on this. |
Beta Was this translation helpful? Give feedback.
-
|
any updates on this ? |
Beta Was this translation helpful? Give feedback.
-
|
Updates in what sense? The Directus project has been bumped to version 10 under a BSL with very permissive allowances for free use. Several previously closed-source features have already made their way to the public codebase and are available for use to entities that don't exceed $5mil USD total finances per year. |
Beta Was this translation helpful? Give feedback.
-
|
thanks for the information ! |
Beta Was this translation helpful? Give feedback.
-
|
I’m not sure it’s easy to build amazing software.
… On Aug 17, 2023, at 1:29 PM, Nathan Doyle ***@***.***> wrote:
I have a question that will sound rhetorical, but it's not my intention and although it sounds like a simple question (and possibly a stupid, uninformed one), it's actually much deeper than that.
Why can't you go all in?
The further you move away from OSS models and its fundamental paradigm and values, the less people are going to want to contribute, in-turn the more the Directus team is going to have to take on, in-turn the more money you have to make from outside entities, in-turn the further they'll want you to move away from OSS (because you have to pay them back and they have to make their profits and fair enough.) This thread existing being case and point I think, and eventually you'll die because you can't sustain the product (okay, maybe you'll survive as is, but hear me out), because you don't have the funds to do so, because the community you could have built to take on the monumental task of making the World's Best CMS finds a new home. A home where they feel like their contributions are better served and more appreciated. A place where they feel more included and a part of something greater than themselves. A place where they can say, hey, "we're all in, and it's hard, but we're all in together".
WordPress and Drupal didn't grow the way they did because their software was the best - well, maybe at one time they were, but that only lasted for a short time. Either way, they still have some of the best communities out there, because they were all in and stayed the course no matter how difficult it was. And so the people were all in, in return and then everyone profited. Even to this day, their communities are mega-all-in (although with so many better options out there the last 10 years it's is becoming harder and harder to ignore said better options; and all the tech dept etc, but that's another topic.) Either way, people still want to be apart of the World's Best Community and they want the best CMS too, of course, but community is stronger then the best (WP case in point.) And there isn't anyone stepping up to the plate in that regard; most of you all start, but then someone flashes something shiny and you go the direction you're going now.
In my mind you guys are just another Strapi at this point. No offence (but that should sting some, regarless.) And there is nothing wrong with that, if that's what you want to be. Just be honest with yourselves about it. And be honest with us (not to imply you aren't.) But if you want to build the World's Best CMS it's going to require the World's Best Contributors and for that you'll need the World's Best Community and you can't have the World's Best Community if you yourselves aren't really all in.
You know, I could just use Payload. It's great software. But you know what, I'm not sure they are all in either. Maybe they see OSS and its contributors as cogs in a wheel and people can sense that and eventually those people aren't going to go all in either. Especially if the Mothership isn't all in themselves. With their cloud offering now, they're just another Strapi too; there are other reasons I say this aside from cloud servies, but I digress.
It's easy to build amazing software, it's easy to grow a community just enough to show a VC you're a sustainable investment. But in comparison, to build the World's Best CMS, now that's not easy at all and I don't blame you guys for not wanting to do it. But at least be honest with yourselves, because you can't have your cake and eat it too. It just doesn't work that way.
If you want to eat your cake, great, go eat your cake, no one is going to blame you for that. But if you want to save your cake for the future party, the heavenly reward, so to speak, if you really want people to go all in too and wait for the party promised and not eat their cake now, save it for a better day, then you have to wait. And sure, it sucks, it's risky, the whole thing might not work out, and like, how much more all in do you have to be, right? Is it even possible to more all in than you guys? I don't know, tbh.
But either way, at some point you're going to have to decide if you're OSS, heart mind and soul or if you're just another good piece of software and a service, along a sea of software and service.
I don't envy you guys. I sure don't have the answers. It's easy for me to spout all this stuff off like some OSS fundamentalist or something. You could even say you were all in a couple years ago and that didn't work and now we're here and I'd be like, I totally get understand that. But what can I say, what I wrote is what came to mind and for better or worse, I hope I'm wrong and I hope what I said is just hogwash. But I just don't know that it is and although I didn't even want to hit the send button here, I'm going to do it anyway. Because maybe there is something here that resonates and may you'll pull it off and truely become the World's Best CMS.
Whatever you decide and whatever happens, I wish you guys the best...
—
Reply to this email directly, view it on GitHub <#17977 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AOM56SAJDUOWRVVEFIXE6ULXVZWK5ANCNFSM6AAAAAAWKY7PXY>.
You are receiving this because you were mentioned.
|
Beta Was this translation helpful? Give feedback.
-
|
@turbolaserhawk I decided to remove the whole post, in case it was a bit too callus, which I think it may have been, at the expense of the points I was trying to make and I apologize to the team if that was the case. |
Beta Was this translation helpful? Give feedback.
-
|
Noooooooooooo dangit!! I just spent 30 minutes writing a response 😂 |
Beta Was this translation helpful? Give feedback.
-
|
I'm just gonna paste it here anyways haha 😅 You made some interesting takes that I feel obliged to give some context on. If you don't want your response on the internet anymore, just lmk and I'll delete my comment as well 👍🏻 Heya! Great questions and takes! I think your final paragraph actually contains the short answer to a long response: You could even say you were all in a couple years ago and that didn't work and now we're here and I'd be like, I totally get & understand that. I'm also sensing a hidden assumption that more contributions equals less overall work for the maintainers, which unfortunately isn't the case at scale. But let me try to answer individual points inline to give some more context. 🙂
In my experience, this hasn't been the case! The amount of contributions or community engagement really hasn't slowed down at all. I think that's because we've spent a lot of time and care in making sure that this next step is as open source as it possibly can be, while having a way to ensure the longevity of the project. I think our novel usage of BSL allows for for the majority of the OSS community to keep using the platform as full free (as in beer) OSS, while simultaneously making sure that larger organizations that extract a lot of value from the project contribute back meaningfully.
I think this is a great example of the hidden assumption of more contributions = less cost. In practice, there's a lot of resources required to properly moderate and steer open source contributions. The work involved in reviewing and help seeing a PR through to completion is not something to underestimate, especially in a larger connected system like Directus. Similarly, the work involved in then maintaining / documenting those additions in the long term adds up quickly. PRs for small actionable items with a known scope are fantastic and can help speed up the development process of the project, but PRs for new features that aren't discussed in depth will actually quickly take more time to review and nourish than it would've taken to just build it by the maintainers themselves..
Fully agree! It sucks if a community project stops to exist and you have to move on. If the requirement for your home is "free (as in beer) or I'm out!" then I'm personally afraid you might find yourself in an endless loop of having to find new homes.. Once a project hits a certain size, the work involved to moderate and steer the project and it's community will cost actual money, at which point that money will have to come from somewhere. Experience has taught me that donations from that same community can't pay for the cost involved in maintaining said community (we currently make $452.00 a month from donations). Some projects sell out completely, some projects start having to do some dual-license trickery, some projects heavily focus on a paid-proprietary SaaS with more features to differentiate, etc, but at the end of the day, something has to be done to make sure that cost is covered. Otherwise, the project will end up as abandonware (like so many great projects are nowadays..). I strongly feel like our current approach to explicitly not differentiate between a "community" and "enterprise" version, or making the SaaS superior on purpose, helps keep the project aligned with it's open source heart, while making sure it doesn't disappear.
Agree to disagree! When they came out, the competition for the type of platform they were building was incredibly small, meaning that communities formed rather quickly.
This is very anecdotal of course, but I'm personally curious how much of that is because of the status quo versus actual all-in ness. (Hot take:) the same could be said for the current React ecosystem. It's still massively popular and large, even though many better options have since come up. I'd argue that a lot of that is from the lifecycle dev-brings-tool->company-adopts->dev-leaves->company-only-hires-dev-with-that-experience and education-focusses-on-what-companies-ask-for->new-dev-brings-tool 🔁 I'm also very curious to see what the ratio is between WordPress (et all) users versus contributors. While the ecosystem is enormously large, I'd bet that the actual amount of people actively contributing back to the open source project is only a tiny fraction. I also think WordPress is a great one to bring up for another reason: while the core WordPress project has remained open source under the WordPress foundation, I believe it was pretty early on when the paid hosted version was launched alongside it as (presumably) a way to help fund the development of the open core. At present time, if I'm not mistaken, the version of WP running wordpress.com is a superior platform to WordPress (self-hosted).
Nothing wrong with opinions! I personally disagree as I the open-core-but-paid-features doesn't vibe with my own ideology of building open source software. If you meant it as cms-like open source platform that had to find some sort of funding to stay around and grow, then sure!
If I'm understanding correctly, you seem to imply that this license change would somehow mean that I'm not all in, which.. I'm not sure where that's coming from! This license change idea came from discussions about funding with the world's best contributors, and the conclusion to go this direction was a direct result of discussions like the one you can read above!
I personally won't shittalk other projects, cause I respect the struggle that's growing an OSS project, but I don't disagree with you! The "standard" track that you see most often when a project like ours grows to a point where funding becomes an issue is either dual-license to lock certain features / impose limits and require everybody to pay no matter what, or offer a hosted version that's way superior to the free core. To me, the former hurts open source more than it helps, as it's a little freemium bait-and-switchy IMO, whereas the latter effectively means you're sooner or later forced to enshittify the free offerings to ensure the revenue generating part of the business keeps growing. That's also why I'm personally so happy with the direction we've currently taken. It ensures that the core platform—the one available to all—is the core focus, as improving the core project helps the sustainability rather than hurt it.
Ehhhh agree to disagree there! Though if it comes easy to you, I currently am looking for people to help out so hmu!
Am I correctly reading between the lines here that in order to World's Best, it has to be free (as in beer) and free (as in speech)? The cake I'm trying to have here is to have this project that we started from the ground up sustain for as long as possible and be available to as many people as possible while doing so.
I'm a little lost as to what the cake analogy is referring to now 😬
This I can answer! We were completely free (as in beer) and open source, and what happened is: Ben and I were all-in to as much as humanly possible, which meant about a 20/80 split between agency-website-work (to make sure we can pay rent) and building Directus. More and more people started helping out, which meant more and more time was required steering and moderating those contributions. Then loads more (large) companies started using the platform, and therefore requesting more resources in bug fix requests etc, and we started to notice a trend. In general, and yes this is anecdotal, we noticed that it's most often the individual hobbyist that is heavily contributing, whereas it's the large organizations who—generally speaking—aren't contributing back. Neither in code nor donations. Could we have given more? No. Humanly impossible. I highly recommend reading https://github.com/zloirock/core-js/blob/master/docs/2023-02-14-so-whats-next.md to get a good look of the possible human cost of open source projects that don't have a sustainable way to be funded.
I don't think there has to be a difference between the two. You can have a great piece of software and services, that are paid, and still be OSS. Keep in mind that the free in free and open source originated from the idea that the source code should be freely available and modifiable. Even though we require the small subset of users who've historically not contributed back to the project in any meaningful way to contribute financially, I still strongly believe we're an OSS project at heart. (And yes, that's all-in 😉 )
I wouldn't call you names, but ehh yeah! And I was and even am one of those "fundamentalists" myself! If the goal is to get rich quick, don't do open source. If your goal is to make a meaningful change, open source is fantastic. I hope my answers above have been able to shed some light of the "under-the-hood" stuff that happens at scale which cause some of those fundamental ideas to just not work once you hit a certain scale. (I have https://xkcd.com/2347/ bookmarked for a reason!)
Yes! I could and I am saying exactly that 🙂 It was GPL-3 "pure" "fundamentalist" open source, but as the Dutch say Only the sun rises for free. Maintaining the project at this scale takes a full time engineering team of at least a dozen, which is very expensive. Companies generally speaking don't let their staff contribute to open source projects for fixed number of days (I wish this was the norm..), and it's still unusual for large enterprises to donate / finance projects directly. If the choice is between adding a license threshold for large enterprises vs abandoned, I know what I pick (and picked!).
Resonates strongly! I do think it's very important to shed some context to the made points though! While I agree from that OSS fundamentalist mindset that the world would be a better place if everything was freely available to use for no matter what, that's unfortunately proven impossible at this scale. |
Beta Was this translation helpful? Give feedback.
-
@Natetronn maybe a hair callus, but no hard feelings! I've struggled with the same feelings while growing this project as well. I did feel it was very important to give you some counter-views context around those views though coming now that I've experienced what it's like taking a hobby oss project from 0 to a big multi-year international effort 🙂 |
Beta Was this translation helpful? Give feedback.
-
|
@rijkvanzanten oh no! hahaha thanks for the response. I'll read it now. I don't actually have an issue with my post being up, per se, because for better or worse it was how I felt at the time I wrote it. But then I got to thinking, there was probably a much better way to express how I felt about things, you know, without being a jerk lol. |
Beta Was this translation helpful? Give feedback.
-
|
very informative content |
Beta Was this translation helpful? Give feedback.
-
|
Hi, I read that Directus project has been bumped to version 10 under a BSL, this could solve the long term monetization problem, but I would like to contribute with some ideas to monetize Directus. You could combine different methods/platforms to suit the needs. Basic ideas to implement in short term:
Explore other methods/platforms:
Source: |
Beta Was this translation helpful? Give feedback.
and others
-
Open Source is fantastic. Creating and contributing to software that enriches the industry is fantastic. Elevating the tech industry as a whole, without any egos getting in the way (mostly, lol), is fantastic. The ability to collaborate with developers across the planet is fantastic. However—as I’m sure many of you are aware—open source has a bit of a funding problem. Who would’ve guessed that giving away software for free would be a bad way to earn money? 😆
Building and maintaining a project the size of Directus is (obviously) not something that can be done in off-hours or weekends, or something that can be done by one person (I’ve tried 🙋). As a matter of fact, maintaining this repo currently takes at least 10 full-time developers. In order to be able to afford those crucial core maintainers, we operate Directus as both a fully open source repo as well as a paid SaaS version, a strategy that’s increasingly popular within the industry. We’ve even raised a bit of financing to enable us to grow into that larger model more reliably than bootstrapping would allow. However, as you might’ve noticed, over the past few months things have seemingly gone slower rather than faster on this repo: issues have crept up to a record 200, PRs aren’t getting the love and attention they deserve, releases are few and far between, and new features (like the Kanban layout) have been made as exclusives for Cloud. What gives?! Well, the answer is both painful and simple: Open Source has a bit of a funding problem.
The revenue we’re getting from GitHub Sponsors (our primary OSS funding source) (thank you to all who donate! <3) is about 1% of what it costs to actually maintain the repo at its current scale. This in turn means we’re forced to focus on things that generate actual revenue outside of the core codebase to make sure the project can be sustained, which quite frankly sucks.
So, what’s next? Great question! If I had the perfect answer, the OSS funding problem wouldn’t exist in the first place. However, we have done a lot of thinking and research on this topic, which led me to a few different approaches, one of which (3) to me feels like a fair and sustainable path forwards:
Option 1: Make the whole thing private
“Wait you’re giving it away for free?!” is often the first thing friends and family ask me when I explain what it is that I do. When discussing the Open Source Funding Problem™ the follow-up question quickly becomes “So why don’t you just stop giving it away for free?”. I don’t think I’ll have to explain to this audience why that is a terrible terrible idea 😅 (Don't worry: this is not a thing we're actually considering)
Option 2: Add some dual-license trickery
I’ve seen this used in other “open” source projects to great success: make the whole codebase available, but limit a piece of critical functionality behind a paywall, and make the paywall a different license (to make it illegal to hack your way around it). While this approach is certainly a great way to make money, it feels... weird. I’m personally not a big fan of locking away the things that make Directus useful behind a paywall, blanket limiting the platform for all projects and users.
Option 3: Switch license to encourage/require more contributions
While researching our options, and seeing what other projects have done in the past, I came across the BSL 1.1 license. A BSL 1.1 licensed project is effectively an open source project with one usage limit: You can’t use the software in production for the first n years after a release is made (after which it becomes the regular OSS license again). That’s very restrictive (and kinda worse than the dual-license approach if you ask me), but there’s one important part to the license that makes this a pretty interesting direction: we can add exceptions to that rule in the form of “additional use grants”. Those additional use grants allow us to still provide free production usage for individual developers, growing businesses/ngos, hobby projects, etc., but require a paid license from large enterprise companies/governments using Directus in production. There have been other big open source projects that’ve made this switch in the past (including MariaDB, Sentry, and Couchbase) and it feels like it would be a great fit for Directus as well 🤔
Any thoughts, feelings, concerns, opinions, ideas? We’d love to hear your “Option 4” down below! 👇🏻
Beta Was this translation helpful? Give feedback.
All reactions