10

Free software, sometimes including source, has been a common feature of computing for as long as I can remember, and probably longer. But the terminology and "legalese" seemed to change at some point.

In the 1980s and early '90s, free-to-use-or-modify software was often released as "Public Domain". It basically means the author gave up all copyright. The modern term "Open Source" was not used back then, to my knowledge, nor do Open Source authors truly give up copyright.

When and why did the common practice of releasing software into the "Public Domain" end?

NOTE: I'm not claiming that Public Domain has "disappeared". Just that a tiny fraction of modern free software uses that approach, and the term has almost completely fallen out of favor.

  • For one thing, the GPL requires that derivative works also be open source. – snips-n-snails Apr 25 at 22:17
  • 2
    It might be useful to point out the difference between "Public Domain" and "Open Source", which has in my opinion been invented by the FSF. Much of that would probably answer the "why". – tofro Apr 25 at 22:18
  • 1
    The OSI invented "open source." What FSF invented was "free software" which has a lot of overlap but isn't entirely the same thing. Permissive licensing dates all the way back to the CTSS project at MIT in the 1960s though. I'll write all this up in a proper answer if nobody beats me to the punch. – Matthew Barber Apr 25 at 22:56
  • 2
    Open Source might be a better venue for this. – Stephen Kitt Apr 26 at 7:40
  • 2
    As someone who used to write for magazines with names like Shareware Shopper and PD Software, we used PD because it was short and made a good logo, not because it was legally accurate (the UK doesn't even have the concept of public domain because of moral right). The readers didn't care about the differences, and we weren't paid enough to educate them. Mind you, trying to sell glossy mags in newsagents to a demographic with little or no interest in spending money may not have been the best business plan ... – scruss Apr 26 at 12:13
2

The TL;DR answer is that they've co-existed for as long as home/micro computers have been around.

Open Source software is a comparatively modern term - the Open Source Initiative (OSI) was founded in 1998 - but the principles of delivering software along with the source code under a permissive licence date back much earlier. The CTSS operating system, developed at MIT in the early 1960s, the GNU Project from 1983, and the Free Software Foundation (FSF) from 1985 would be some obvious examples. Still, if you want a date when the big switch started to happen, 1998 is it.

Public Domain, on the other hand, was a well established term in copyright law prior to the development of the first electronic home computers. And yes, it means you give up all rights. There's still much PD software made, but the general opinion is that it's better to retain copyright and go with a permissive license for collaborative projects if you want them to keep going beyond the original creator's direct involvement.

13

But the terminology and "legalese" changed at some point.

Terminology, as used by the general public, yes, legalese no.

In the 1980s or early '90s, free-to-use-and-modify software was referred to as "Public Domain". It basically means the author gave up all copyright. The modern term "Open Source" was not used back then, to my knowledge.

No, yes, err, this is a mixup of different concepts. Public Domain and Open Source are neither the same nor interchangeable. In fact, they are in some variants the exact opposite. And your question feels as if you also mix in the idea of Copyleft.

So let's first look at the items before answering the historic part.

  • Public Domain is a basically US legal term (though similar concepts are in many legal systems), describing that some piece of work is free from any rights and can be used without any restriction. (Software-) Items in the Public Domain do not have to be available in source. They can as well be just binary.

  • In addition, the Free Software mentioned is again none (or any) of the above. "Free" is not an in-itself defined term here, but more of a concept - one easily misunderstood, as it's about freedom to use, not free as in free beer.

  • Open Source at the core just describes the fact that the source code of an application is available. It can be available to everyone or just for a restricted user group. It doesn't have to be free at all. In fact, GPLed Source isn't free, its protected by freedom restrictions. Then again, it's also a name coined to remove the ambiguity of 'Free'.

  • Copyleft is the idea to use the copyright system to keep (software-) items open and return every change made back to the public. Copyleft is the exact opposite of public domain. GPL is eventually the best known variation. It only works because the item licensed via GPL is not in the public domain (*1).

Terminology is of course changing, especially when some area is new and in rapid development - do I hear software? In the beginning people tend to use well known terms, and "Public Domain" is one such. Programmers with no intention to sell put their works out into the Public Domain, binary only as well as some including source code - if the language used even made this difference at all.

Giving a work into Public Domain means giving up all rights. So others can take this, modify it (or not at all), and resell the result.

Not everyone wanted to give up all rights, so the idea of various kinds of Free Software were conceived.

Some (aka RMS) wanted to take that even a step further to secure freedom of use by using the copyright system to protect this freedom. Thus Copyleft was born - and GPL is the prime license to protect Free Software.

While the basic idea did gain a strong following, the term 'Free' was way too easily misunderstood - and also repelling to any commercial/professional usage. So the Open Source term was used by people promoting professional use. A term RMS fought, as it allows others to use it without guaranteeing the freedoms he envisioned. (*2)

These fights over the right naming were quite active during the late 90s into the mid naughties, when most settled for FOSS (Free Open Source Software) as compromise. Somewhen in the 10s this got extended to FLOSS (Free/Libre OSS) to make it even more clear that it's about freedom, not free beer.

The good side of this fight was that OSS is now a term not only known by some insiders, but known and used by a majority of developers and distributors and anyone reporting about such. As a result OS(S) has reached the general public, far beyond the usual circles.

Well, and amidst all of this is the common man with his need to use a term describing what he wants, a term that can easily be used and understood. Not an easy task, as reading the fine print is still an issue.

Using OSS might as well be just an intermediate step (*3).


*1 - The usage of copyright to protect Copyleft results in a great twist: When the copyright expires, it will move into the public domain. I really would love to see what's happening then. But thanks to Mickey Mouse this day may never happen - or at least so far in the future that I won't live to see it :(

*2 - Notably, MS with various initiatives to 'open their sources' without giving real freedom to modify and redistribute.

*3 - Hopefully one where the crap about IP gets removed as a whole -
You may say I'm a dreamer, but I'm not the only one, I hope someday you will join us, and the world will be as one.

  • 1
    "GPLed Source isn't free, its protected by freedom restrictions" It's not necessarily free as in beer but it is most definitely "Free Software". – JeremyP Apr 26 at 10:08
  • 1
    @JeremyP Help me, I fail to see the point you try to make by separating this sentence from its context. – Raffzahn Apr 26 at 10:33
  • 1
    @JeremyP I still don't see your point. Have you read the text at whole?Eventually even while considering to set asumptions aside? Or maybe more to your point: GPL does restrict freedom - for example the freedom to use the source licenced without redistributing it again, including the changes. That's the whole point here and how it works: GPL restricts the freedom in general to support the existence of some freedom. The sentence ou are cyting is in no relation to free beer. – Raffzahn Apr 26 at 11:28
  • 2
    @Raffzahn You have one sentence that claims GPL source code is not free. This one sentence is false regardless of whatever you wrote elsewhere (especially since you have now clarified that you do not mean free as in beer). You can do whatever you like with GPL software without any obligation to share your changes to anybody including the copyright owner. The only conditions that apply do so when you choose to redistribute the software. In those cases, you must extend the same rights to the people who receive the software as you received from the copyright owner. – JeremyP Apr 26 at 11:55
  • 3
    @JeremyP But that's exactly the point. GPL limits freedom by not allowing to redistribute it without source code and granting the same rights it came with. This sentence is not about good or bad, or what my stance is. It's the simple fact that GPL restricts freedom - quite as you mentioned it - to reach its goal. – Raffzahn Apr 26 at 22:25
7

I don't think there actually been any big move from public domain software to open source software, as these terms are actually defined. Instead it's just become much less common to use the term public domain inappropriately.

The terms "public domain" and "open source" mean two different things with little overlap between the two. Public domain means that no one owns the copyright on a work, and can be applied to any kind of work, whether a computer program, book or a sculpture. Open source means that the source code for a computer program is freely available under permissive terms. Open source don't imply public domain, and in practice almost never is. Public domain software is often open source, but not necessarily so(*).

Software that is both in public domain and open source has been rare for a long time now. Back in 1991 when I released mytinfo as a source code distribution and put it in the public domain, I was doing something well outside the norm. Most what we would now call open source software released back then had some sort of copyright notice. While the GPL (Gnu Public License) was still uncommon outside of FSF projects, simple notices asserting copyright while permitting copying under certain conditions were very common.

I think what has changed that is people are no longer using the term public domain incorrectly to refer to any software that they can obtain legally without paying for it. Back in the mid-to-late 80's and 90's you'd often see both closed source shareware, like PKZIP, and open source software, like GCC, being called public domain by people who only cared that it was free (as in beer). The amount of actual public domain software in common use back then was pretty small. The most popular public domain software I can think of from that era were the original versions of rz and sz, which implemented the ZMODEM protocol. They were only in the public domain because they were originally created for the US government.

These days public domain software has only become even more rare, while in comparison the term open source is better known and much better understood. There really isn't as much opportunity to mistakenly call open source software "public domain" if it actually isn't.


(*) for example, my mymc utility is in the public domain, but I haven't released the C++ part of the source code. The Python part is freely available and implements almost all its functionality, but it was released just so people could use it on Linux.

  • 4
    I think this hits the nail on the head — it’s more about incorrect use of “public domain” than anything else (especially with “PD” collections for people with Amiga and ST backgrounds). Note that public domain can only be applied in certain jurisdictions. Nowadays PD is making a sort-of comeback with PD-like licenses (relying on copyright of course) such as CC-0 or WTFPL. In the 80s and 90s, quite a lot of freeware was released as public domain, but without source code. – Stephen Kitt Apr 26 at 8:20
  • 1
  • @JdeBP - nowhere in Paul Vixie's cron distribution does it say it's public domain. The licence language seems a little naïve these days, but was fairly typical of its time – scruss Apr 26 at 12:04
  • 1
    You've not understood the point, or read the name of the software it seems. – JdeBP Apr 26 at 14:24
  • 1
    This answer seems plausible, but I'm skeptical that there was no real change in the preferred legal status of software works by the authors. There was so much available software tagged as PD, it's hard to believe the authors didn't intend it as such. Perhaps the loophole was the source being closed, gave them the ability to control their free product while opening it up to wide distribution and use - same end effect (for the author) as an OSS license. – Brian H Apr 26 at 15:41
5

Rather than argue about the premise of the question or the distinctions between "public domain" and "open source", I'm going to answer the question. Several factors seem to have occurred:

  1. In late 1980s to the 1990s, Richard M. Stallman and the GNU project introduced the idea of a software license that allowed people to use software at no cost, but not completely putting it in the public domain. Prior to this, most software was either proprietary or in the public domain, with nothing in between. A key aspect of Stallman's license was that the source code be publicly available. Other developers took notice and used either GNU's licenses or their own open-source licenses.

  2. Also during the late 1980s to 1990s, the growth of the Internet made it much easier to distribute not only program binaries, but also source code. Anonymous FTP, NNTP newsgroups, and later the Web made these possible.

  3. Source code often wasn't included with programs on cassette tape, floppies, or CDs. This was largely because of space, as well as low demand from end users.

  4. A significant amount of public domain software in the late 1970s to 1980s was written in BASIC, which has its own issues. Programs were usually stored in tokenized form, which was not necessarily readable between different flavors of BASIC. With some exceptions, a user could simply LIST the program, so possessing a program largely implied that you had the source code. People (incorrectly) equated public domain with available source code. By the 1990s, BASIC was out of vogue.

  5. By the late 1990s, hardware and software companies started to see the advantages of using other people's open-source software as a component of their own products. Why spend time and money developing your own solution, when one is already available for free? It can just be part of your product, and you can keep the rest of your product proprietary. The impact to the company is minor: append the open-source license to your own license, and run an FTP server with the source of the non-proprietary components.

So I'd say the change occurred in the 1990s, due to the reasons above.

  • Some very good points, especially #2 and #3. With PD, the author knew it would be spread (disk collections, BBSs, casual sharing) without cost to them. Commercial distribution of software at the time was an expensive gamble. Plus, many PD programs had only marginal value - easily worth the cost of the media, but not necessarily worth enough to be commercially viable. – Brian H Apr 27 at 4:30
  • Re point 1, there was something in between before the GPL: the BSD license, used since the late 70s AFAIK. Also, the GPL doesn’t imply “at no cost” (you can sell GPL-licensed software). – Stephen Kitt Apr 27 at 10:40
  • And I’m flat-out wrong re the above, BSD dates back to 1988 apparently and the X11 license is slightly older (see this surprisingly timely article). Still earlier than GPLv1 though. – Stephen Kitt Apr 29 at 16:38
  • @StephenKitt I think you're off by a couple of years. The first BSD license was issued in 1980 with the release of 4BSD Unix in that year. 1988 would correspond to the release of 4.3 BSD. – doneal24 May 6 at 17:11
  • @Doug I’m not sure about the license being used in 1980; 4BSD had a rather different license. 4.3BSD had the now-familiar license, in 1986, and the MIT license was in use by 1985, so the GNU project wasn’t the first which was my main point. (And yes, I know others have also mentioned 1980 for the BSD license, so I could well be wrong; but I haven’t found any direct evidence either way.) – Stephen Kitt May 6 at 18:54
4

A lot of software written in the 1940s, 1950s, and even 1960s was never copyrighted, because people didn't realize the copyright protection extended to software. Even when copyright began to be recognized, there was some debate about whether algorithms should be governed by patents rather than copyrights, while code should be copyrighted instead. There's a subtle point here. If the code is copyrighted, but the copyright doesn't extend to the algorithm, then perhaps an engineer could read the code, learn the algorithm, and write some new code that implements the algorithm, without using any of the intellectual property of the copyright holder.

Another thing that started happening is that authors would release a piece of code to the public domain, only to find that commercial programmers would cut and paste their code into products for sale. That's not illegal. Many PD offerings came with a proviso not to charge anybody for PD work, but that can't be enforced. So authors began reserving their rights to control the software even when they didn't plan on making a dime off of it.

The other answers have got a good outline of what the other alternatives are, and what purpose they serve. The main purpose boils down to preventing somebody else from appropriating property that was released to the public domain. There is still plenty of PD stuff out there. I've posted work on Stackexchange that has become, in effect, PD stuff. They are little tinkertoys, and I don't even care if a professional pastes it into their work.

  • 1
    "read the code, learn the algorithm, and write some new code that implements the algorithm" - This method, with the use of 'clean room design', is how Compaq managed to create the first IBM-compatible clone PC. – Kaz Apr 26 at 12:22
2

I see a lot of technical answers here, but not a lot about why. Why copyright your work if you don't intend to exploit it for money?

One possible answer: It's no fun to say, "I am the author of X" if X has been turned into crap and widely disseminated by people you don't know. Copyright lets you tell other people to cease and desist from calling their crappy derivative works by the same name as your beautiful original.

  • 3
    "Why copyright your work if you don't intend to exploit it for money?" Because in any country that is not a signee of the Berne convention, you cannot not copyright a work? – a CVn Apr 26 at 14:54
  • @CVn It sounds like you think the timing of the Berne convention affected the history of this transition. If that is so, then how can this be just a change in people using the right terms and not a change in preferred legal status of software works? I ask because most answers above indicate there was no "real change". People just got brainwashed into using the right "language". – Brian H Apr 26 at 15:31
0

Another factor that was important was the advent of gcc (GNU Compiler Collection, formerly GNU C Compiler). Up to the late 1980s, compilers for home and small business computers were either fairly expensive, or fairly limited. Turbo Pascal was a notable exception, but it was limited to CP/M and MS-DOS.

Richard M. Stallman seems to have been well aware that getting people to write free software would be a lot easier with free development tools, and the appearance of GCC in 1987, and its being ported to several architectures and operating systems within a few years, helped a great deal with getting the Open Source movement running.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.