Why the Distinction Between "Open Source" and "Source Available" is Important
When looking at scenarios when open source is misrepresented I’ve commonly come across cases where a project downplays the distinction between open source and source-available-style licenses, choosing to advertise the project and/or themselves as open source while licensed in a non-OSD adhering manner.
I’m not talking about ignorance of open source and licensing details, but instead about scenarios where a project acknowledges they’re not OSD-adhering, but make a choice to continue advertising themselves as open source.
In such scenarios, I’ve seen a variety of responses & reasonings including:
“[…] practically for end users there is little to no difference” ref
“[…] the source code is available and meets many of the elements of the Open Source Definition […] but only forbids the hosting companies from “selling”” ref
“[…] there are many benefits of open source that go far beyond the licensing specifics. The community of contributors, being able to inspect the code, transparency, reliability just to name a few.” ref
“[…] for nearly every use case the license applied gives them the practical implementation of what they want out of the idea of Sentry being open source.” ref
These statements are not really false, but they do tend to miss the point. There’s a lot of similarities to draw between house-cats and lions, but you’d probably only want one of those napping on your bed.
The differences and their importance
One of the great things about open source as per the OSD, is the strength of that definition. This is a definition that has been used since its origins back as the DFSG in 1997, and is what the reputation of open source has been built & popularised upon. It does a great job at defining reasonable bounds, in a way that fundamentally ensures users have open rights of use, modification and distribution. These are the rights which give open source its weight.
Comparatively, source available licenses like the ELv2 and the BUSL add limits to these freedoms to prevent certain use-cases. It no longer ensures open rights to the user, but prevents rights to instead focus on protecting the author. These have a fundamentally different dynamic and spirit, that will often not align with people fond of open source and familiar with rights it affords via the OSD. Advertising as open source could easily mislead someone into thinking the project is something which it is not. It’s attempting to use the good-will of open source without taking the same risks as those that have built its reputation.
While the limiting differences of source available are often dismissed as something along the lines of “no difference to pretty much all users”, this is often specific to the viewpoint and concerns of the author, not the audience, at a point in time. As an example, many source available licenses prevent competitive or SASS usage. While that may only hinder competition, and not the users initially, it could have a greater affect down the line. Say that the authors make a large negative change to their pricing model or development direction which screws over its user-base. With open source this can provide the opportunity for a fork if there’s enough user demand. Business can be formed around the use-case the project is already suited for, to help feasibly drive that fork, whereas anti-competition/SASS licenses can significantly hinder such efforts specifically for the author to retain an advantage.
Take OpenTufo as a recent example. Would such efforts been possible if the original Terraform had a license preventing production & competitive use? The ecosystem and businesses built around Terraform, thanks to it being open source, which allowed there to be momentum and resource to keep it going in way the keeps the originally provided rights to its user-base. While some may look to the risk to the original author, which there inherently is, the focus of freedoms is kept upon the user. This is a massively important part of open source.
Source available is not bad
To be clear, I don’t think source available licenses are bad in any way. I completely understand and respect the right to license your work how you wish to protect your efforts and time. It’s better to provide users many extra additional rights than none, which these licenses often do, I just think it’s misleading to advertise as open source.
Why attempt to confound?
Something that I’ve never really understood is why there’s often such as desire to use “open source” adversely to the OSD. I think maybe some don’t like to be asked to consider their wording, becoming dug-in against change while viewing folks like me as pedants. Otherwise I feel like most cases I see are in it for the advertising. I’ve often seen this in VC-funded groups that are desperately trying to growth-hack and market their way to relevance & sustainability. The allure of labelling themselves as open source may be just too much to give up.
But open source is no longer relevant/sustainable/functional as it is!
These kinds of statements often pop up, but are commonly from those for which open source does not suit. When someone says “open source is not sustainable”, they often actually mean “open source is not compatible with our desired business model and goals”. Open source doesn’t work for everyone, and it can be really tough going, but it’s not a business model and it doesn’t have to suit everyone. There are many projects, companies and people out there sustaining on and/or working compatibly with open source. I don’t see why there’s often a desire to then redefine open source, instead of using or creating an alternative. Tomatoes don’t align with my taste-buds but I don’t grab an apple and call it a tomato instead.
This post was an attempt to tangentially gather my thoughts on the subject after providing input to Sentry as they prepare to launch their new source available license. A shout-out of thanks to Chad Whitacre at Sentry for listening to my feedback.