I was talking to someone recently about the idea of an “open-source-first” company as a designation - that is, companies whose primary product is open source code, and who have built their marketing and capitalization strategies around utilizing that code from day 1. This seems to me to be a different brand of open source than, say, a large company open sourcing some sections of their code afterwards; university labs producing open source code as research artifacts; or non-capitalized open source.
I don’t know if there is a list of this sort of company, or where the bounds are. Would RedHat count, for instance? Or something like Socket? The PSF? I think having a list that includes these under different categories might also be helpful.
Before creating this, I want to know: Does anyone know of a list or a community of open-source-first companies and organizations?
If anyone is interesting in talking about it, as well, I want to propose a thirty minute call or so to spitball ideas. I’m currently thinking Thursday the 26th at noon.
Aye. But COSSI has a pretty clear objective, and tends to only look at large companies. I wonder if there are smaller companies out there which wouldn’t be included.
I’m not aware of a list but I often use a simple variation of this criterion when discovering a new company. Is the majority of the organization income derived from Free Software? Would this income vanish if the software was proprietary?
Is the majority of the organization income derived from Free Software? Would this income vanish if the software was proprietary?
I like this, but it also could apply to all a huge amount of companies that simply use FLOSS without making it.
Noon ET was my thinking. But I’ve just realized another, probably better time we can cover this is this Friday at the Sustain Weekly Roundup. Sustain Together Discussions - Google Docs
OSI could sponsor this but before putting resources in this effort I’d like to understand what the ultimate goal is.
My gut feeling is that this is a rats’ nest of hard definitions that don’t have a shared agreement, like @LawrenceHecht said. With so much SaaS offerings out there built on non-copyleft code, what exactly counts as Free Software income?
@dachary how do you classify a company like Discourse, for example? How about one like cal.com or plausible.io?
I can imagine tons of companies that would be less profitable. I really mean the income vanishes not the income is reduced by 30% because of proprietary licenses. In other words, companies whose income is actually dependent on the software being released under a Free Software license, not companies for which it’s nice to have but ultimately optional.
Discourse is one of the few companies for which I answer yes to both questions:
Is the majority of the organization income derived from Free Software? Yes
Would this income vanish if the software was proprietary? Yes
Of course that’s not a guarantee that it will remain so forever, as demonstrated when Discourse Teams was announced as a proprietary software extension in 2020. It was released as a Free Software plugin two years later and I can only speculate that the plan to compete with slack and in this way did not work out as intended. A few months later Discourse announced raising millions and promise to remaining uncompromisingly open source, therefore one can bet it won’t change overnight.
I did not know about the other two (thanks btw, they both look interesting ) and it would take me a while to answer these questions. For cal.com I would start by trying to figure out what is Free Software and what is not: I suspect integrations with external tools are predominantly proprietary plugins and I’ll start with this, hoping to be wrong. For plausible.io, which looks much more committed to Free Software, I would simply start to self host it and use it. Actually… I’m going to add that to my schedule for next week.
Of course this is my opinionated and personal way to look at things. Bottom line is: it needs time and effort, on a regular basis and if someone wanted to make such a list, it would be a significant undertaking. And since the vast majority of companies that are friendly to the OSI won’t pass the test, I suspect the OSI won’t get funding: what would be the incentive?
Do I understand correctly that your criteria ends up being a binary choice: either a company is a “Free Software company” or it’s not?
that’s why I asked what is the use and the ultimate purpose for such a list I’m not concerned about corporations sponsoring OSI at the moment. More interesting for me is understanding what criteria and definitions are available and assess what level of agreement there is for them in the community.
But first: Why would one want such a classification? Who is this for? What use case does it solve? It’s not clear to me.
Binary choices don’t seem right for this, to me.
Why would one want such a classification? Who is this for? What use case does it solve? It’s not clear to me.
I’m still finding a good amount of FUD in the ecosystem around how to set up a project as a business, or how to make a profit out of open source in an authentic way. Specifically, what I often see is people putting their project through the process of having GitHub Sponsors or Open Source Collective, and then not reaping a huge amount of reward from donations. By showcasing projects which have made the leap towards being profitable businesses, it’ll be easier to showcase to those project maintainers how they could potentially profit off of their work.
From this perspective, I’m less interested in hard-and-fast classifications and dichotomies, and more interested in stories of people who managed to have OSS as the core of their business, particularly for smaller projects and not for major organizations. Right now, the main way I see most maintainers profiting off their work is by including it on their CV and getting a job out of the experience, or by using their clout to get personal contracts. I’d be interested in hearing other stories.
Yes, it is a binary yes or no. The interpretation of this binary answer is left to the reader
For me it determines my behavior. I will frequently and publicly recommend and cite Discourse as an example that a purely Free Software business model is possible. And I will spend time learning the Discourse internals. I will create software that depends on it. I will also spend significant time setting it up for companies and non-profits I work with, thus creating a dependency on Discourse.
A contrario, I will use GitLab for myself but not recommend it publicly because the answer is no. I’m worried that organizations deploying GitLab will become dependent on it and be deceived into using proprietary extensions without realizing it.
I can’t speak for other people but hopefully you see why it matters to me
When @realaravinth and myself started Hostea in February this year, we looked for such stories to get inspiration, but in vain. The organization matured a little but is still struggling to figure that out. If you find stories, it would be great. Interestingly, I initially thought following the footsteps of Discourse or Weblate would be a good idea. And then I realized that not only were both projects created by exceptionally talented people (that helps a lot) but it also took them years to mature and become sustainable.
That guy is basically ripping off his users… hardly an example to follow and his comment about Fedora being “dual license” makes me question his knowledge of the topic. His code and his behavior have very little to do with best practice of any open source project.
Dual licensing copyleft software means that the income originates from selling proprietary software licenses. IMHO this is not sustainable because there is a conflict between the proprietary version and the Free Software version.
The organization will, eventually, have to choose between the two and the outcome will likely be to prefer proprietary over Free Software. MySQL had that dual licensing and was sold to its proprietary competitor Oracle, for instance. GitLab gradually becomes crippleware and features as simple as git push/pull mirroring are kept out of the Free Software version and only in the proprietary version.
I think his aim was to monetize the project one way or the other while also using a GPL license and like you said, some orgs would prefer the proprietary license.
I’m also wondering how he was able to get every contributor to sign off on dual licensing.
) and it would take me a while to answer these questions. For 