Let’s not be pedantic about “Open Source”

Written on 3 Jul 2020

Another day, another “it’s not true open source, you’re not allowed to call it that!” thread on Hacker News.^[1] This seems to pop up every few weeks or so and today I’m sufficiently peeved by it to write something about it.

Here’s the thing: I don’t think you, the Open Source Initiative (OSI), or anyone else gets to single-handedly define a One True Definition of Open Source. Even if the OSI would have unambiguously coined the term Open Source (which is complicated, more on that later) then that still doesn’t mean they get to be the arbiter of all usage of it; it’s just not how language works. If trademarks can be come generic then so can neologisms such as Open Source.

Whether you like it or not for many people – especially those not deeply invested in the entire movement – Open Source means “there is access to the source code in some way”, with varying levels of things you are and aren’t allowed to do with it. They generally don’t mean “it fits the definition according to this checklist some organisation drafted”.

I have nothing against the Open Source Definition (OSD), but judging all usage of Open Source by it is unworkable. Correcting millions of people that “using it wrong” is not a good strategy for accomplishing anything meaningful – certainly not for promoting the use of Open Source.

Some say we need a new term for Open Source software that doesn’t meet all the OSD criteria. Supplanting a wide-used term with some neologism is a very hard task, and I’m doubtful that it will be successful. Wikipedia has Source-available software, but it seems like a clunky term and the difference is subtle and lost on many. I’ve rarely seen it used.

Is it messy, confusing, and annoying? Sure. But language employed by millions of people across different cultures and views always is. In 2020 the Free Software Foundation is still trying to get people to say “crackers” instead of “hackers” and have an entire list of common things you should be saying different. The entire thing is a massive turn-off for anyone concerned with things that actually matter.^[2]

---

So did the OSI invent the term “Open Source”? That’s a tricky question because what does “inventing a term” mean, especially when the term is composed of a common adjective and noun?

If we look at Google ngram we can see that open source was already gaining traction in the mid-90s.^[3] If you look at the case-sensitive version then you can see that “Open Source” specifically got a big boost in 2003. Below you can see some usage of “open source” I found in a quick Usenet search; I’m sure there are more examples to be found on mailing lists, old websites, etc. but searching that is not so easy.

I have no doubt that Christine Peterson is honest in her retelling of How I coined the term ‘Open Source’; she probably came up with it independently, or perhaps she had seen it before and subconsciously “remembered” the term.

“Open” was (and remains) a common adjective added to all sorts of stuff: Open Linux, OpenDOS, OpenGL, OpenBSD, OpenStep, X/Open (later: Open Software Foundation, The Open Group), Common Open Software Environment, OpenVMS, etc. etc. etc.

In short, the origins of “Open source” as a term are a lot more convoluted than is often made out to be. The OSI was certainly important popularizing the term, but this doesn’t mean they single-handedly invented it, much less have control over how anyone uses it.

Pre-1998 usage

The earliest reference I found – quite by accident – is in a 1985 broadcast of Computer Chronicles (13:50) on UNIX:

As for the future of UNIX, he [Bill Joy] says its Open Source Code, versatility, and ability to work on a variety of machines means it will be popular with scientists and engineers for some time.

Many more usages of open source can be found in Usenet search with Google Groups. I tried downloading the Internet Archives for better searching/numbers, but that doesn’t seem to work at the moment, so this will have to do.

1990:

BSD’s open source policy meant that user developed software

1990:

My primary concern is that those policies must comply with the U.S. Constitution and thereby allow the free dissemination of open-source/published material – including software (ESPECIALLY FREE SOFTWARE) which is developed directly from published algorithms.

1991:

For U**X, apparently some vendors distinguish in price or availability between human-readable-source, and encoded-but-compilable- source which isn’t very readable but is only intended to make the program portable to systems other than the vendor’s standard version. If you specified U**X source in #2, indicate interest and price for either or both versions. Note that because of version control, we may be slower to releasing an open-source version.

1991:

One can still choose to do the port from the open source in the traditional way.

[..]

Now it is perfectly reasonable for customers (and prospective customers) to lobby HP to release source code with its product.

[..]

Open source for a user-oriented package like ‘WordPerfect’ might create more problems than it would solve.

1993:

The GPL and the open source code have made Linux the success that it is.

1993:

Anyone else into “Source Code for NT”? The tools and stuff I’m writing for NT will be released with source. If there are “proprietary” tricks that MS wants to hide, the only way to subvert their hoarding is to post source that illuminates (and I don’t mean disclosing stuff obtained by a non-disclosure agreement). Open Source is best for everyone in the long run.

1993 – this one is interesting, as it very explicitly uses “open source” in a way that’s not compatible with the OSD:

The closest comparable product sold by Softek today costs $4000 including open source.

1994:

This agreement would prevent open distribution of source code, so the XFree86 team has chosen to not support Diamond products rather than give up open source code distribution.

1994:

I think that you are much better off with open source code than with .TPU files or obfuscated source. When you find bugs or need to make changes because conditions have changed, you can do the fixes. You won’t be relying on someone who may have gone out of business or lost interest in the code.

1994:

The main problem is that the mac doesn’t really have any large scale open source projects.

1995:

I need open-source code or open-software for KL(Karhunen and Loeve) transform of images. Does anyone know of site information to get this?

1995:

Can anyone direct me to a modeler that has an open source code. Anything will do.

1996:

Caldera announces open source code model for dos

[..]

Caldera believes an open source code model benefits the industry in many ways.

1997:

Canopy provides open source code for former Willows crossBplatform technology: Windows API/ABI