What is Open Source and Why is it Good for Business

We are called Django Web Studio because we make web applications on top of the Django web framework. Django is an example of open source software. Open source software and its philosophy -- free to use -- enables the internet. And it enables us, and thousands upon thousand of software developers and software entrepreneurs, to build and make money from internet software applications.

But what is that, open source? What is free to use? And what does "free" mean anyway?

This article aims to be a guided tour into open source, in particular open source software, and wants to show how open source has generated billions upon billions in revenue, despite being "free to use".

Before we start, let's get a few terms straight about things that are NOT "free to use".

First, patents. A patent is a right conferred upon a private person or a company by a state. A patent is an invention that is described in enough detail to satisfy the patent office that it is is a unique "useful art", that would further humanity. According to the European Patent Office, a patent must be described in terms sufficiently clear and complete for it to be carried out by a person skilled in the art or craft. It is basically an idea for an implementation, but not the implementation itself.

A patent is the right to forbid others from implementing the idea without paying the original inventor.

As a case in point, let's look at the the patent infringement law suit of Ford vs Selden.

Ford was Henry Ford, who needs no introduction. Selden was George Baldwin Selden, born into a prominent American family, who became a lawyer at his fathers bidding, but was also something of an engineering enthousiast. In the course of his work as attorney at law, he handled patents, among others for George Eastman, of Kodak.

In 1894 Selden filed a patent himself. For what was then called a "road engine" and what we would now call an automobile. He patented the idea of the motor car.

That was a very smart move. For a number of years, he collected royalties from any company building motor cars, without ever having built one himself.

It was only after a hard legal battle initiated by Henry Ford did Seldon finally lose the protection offered by his patent. On a technicality. Ford won because the judges agreed that the motor powering the "road engine" was sufficiently different from the motor powering Fords cars. And the rest is history.

Now let's turn to copyright. What makes copyright law different from patent law is that copyright exists to protect the rights of the author of an original work that is fixed in a tangible medium like a photograph, a book, or a computer file. So it is not the "invention" or idea that is protected, it is the work as it may be disseminated in physical or digital form. It is the implementation that is protected.

So when is something called an invention and when a work to be copyrighted? For that, there is a concept called the idea-expression dichotomy, which states that copyright forbids reproducing a work, but does not forbid anyone from doing what is described in a copyrighted work. An example might be that it is forbidden to reproduce a copyrighted work about yoga, but it is not forbidden to practise yoga as described in the book.

There is an important case called Baker vs Selden that led to this being codified into law. (This Selden seems to be no relation of the above-mentioned.)

Selden had written a book on a way of improving bookkeeping, copyrighted it, but had failed to sell it. Note that the improvement was the idea that his book espoused.

Baker had later written about a very similar idea, but succeeded in selling his book. In fact, it had international following.

In Baker vs Selden, Selden's lawyers argued that by copying the idea, Baker had infringed upon a copyrighted work. But judges decided that it was permissible for to write and publish works that described that idea, even though an earlier work describing that same idea existed.

It is the published work that is copyrighted, not the idea. That is contrary to the patent, which basically copyrights the idea.

A computer program is generally not patentable. This is the stance in most countries across the world, including the European Union.

But it may be copyrighted. In many countries, the author of a computer program is automatically assigned the exclusive rights to publish the program the moment it is produced. Again, note that only the dissemination of an existing program is subject to copyright. Not the idea that underlies it.

But what what is a program as opposed to its constitute parts? You can't copyright a computer programming language any more than you can copyright a spoken language. However a computer program makes use of ever more pre-packaged components or libraries. Each might have it's own copyright license, and the author of the program might have to pay the original owner of the package.

Components are copyrighted by default. If you want to use it, you must pay the author, or you must pay the author's employer.

In the Microsoft world this is common practise. Sometimes a component can cost several thousands of dollars, so much that only enterprise would even think of buying it. Others than enterprises with huge budgets then, would need to code the library themselves, reproducing the copyrighted library by basically copying the idea.

Adhering to copyright law would make for countless hours wasted on copying the idea, or paying for the component. Either way it would add a lot of cost to the creation of the software.

Sharing is at the core of all invention. Computer programs are no exception. In the early days of digital computers, programs were developed in the course of research. That research was shared. Indeed, it is mandatory to publish scientific research it such a way that it may be peer-reviewed, so the outcomes may be reproduced.

Consider email. Though not without a fight, email was consolidated by the 1980's. Email is based on open standards, making it possible for authors to create email programs to create, send and receive emails. Protocols such as SMTP, for interoperability of email programs, and FTP, for interoperability of file systems, and HTTP, for the viewing of HTML documents, make for the internet as we know it today. All of these are open standards.

But email programs might be copyrighted. Apple Mail and Outlook make use of the open protocols, but are copyrighted programs. Companies like Apple and Microsoft use copyrights to give themselves exclusive rights to publishing the programs, therefore excluding all others, and keep the source code of the programs under locks. The name we give to these types of programs is proprietary software.

A program is called open source when a company chooses to open up the source code to the general public. An example is Mozilla Thunderbird, which is a desktop email client that runs on Windows, MacOS X and other systems. The source code of this program is available for anyone to use, and to copy, under a license that is (far) more permissive than the licenses of proprietary software. Another, perhaps more well-known example is Mozilla Firefox.

In fact, the internet runs for a very, very large part on open source software. Counting only servers that run web applications, 95% of all internet servers run the open source operating system Linux.

But perhaps more importantly, many components that software programs use are open source. And many of them are considered so ubiquitous that they seem more like part of the general environment than a specific program. An example is openSSH, which is takes care of security on the open source servers mentioned above, but also enables me to access servers on the internet using my operating system of choice, MacOS X.

So how do companies survive of they give away their source code for free? Well, let's first define "free".

Free as in speech, not free as in beer, that is heard often in open source circles. Open source is free as in speech. Partly because of the background of programming in scientific research, but also to a large extent because it was found early on that creating a program involves much of what we call "boilerplate", which can be understood as functions that are the same for any program. After programmers found themselves creating the same component for every program, it was the logical next step to share such components among programmers.

But sharing is publishing, and the very act of publishing makes the creator a copyright holder. So, to make intentions clear, and prevent copyright lawsuits, alternative copyright systems have been devised in order to define the scope of the software and the rights of the authors. GNU, MIT, BSD are some of these licenses.

So how DO companies make money when their software is free to use, free to copy, or both?

Mozilla Firefox has a modest market share, but still needs millions to pay developers and other staff. In fact, the 2022 revenue of the Mozilla Corporation was just shy of 400 million US dollars. So, while Firefox is free as in beer, as in no costs involved to use the browser, and free as in speech, as in the source code is open to the public, the Mozilla Corporation can still generate revenue through licensing deals, such as using Google as the default search engine.

OpenProject is another initiative I would like to mention. This is an open source alternative to proprietary ticketing systems such as Jira. OpenProject offers a broad range of project management solutions from software development management to inspection software for municipalities. They make money in a similar way to closed source, proprietary SaaS solutions, as in that they may charge a subscription fee per seat, but they also make money from consulting, training and support, in addition to customisations for specific clients. And also by hosting, managing and supporting OpenProject for those that do not wish to host the application themselves.

There are many of open source solutions out there, many household names by now. Think of WordPress, Magento, PrestaShop, Drupal, Joomla. But there is an open source alternative for many others: SuiteCRM instead of SalesForce, NextCloud instead of Google Drive, Mautic instead of Marketo, Odoo instead of QuickBooks. All make their money from a combination of hosting, managing and supporting subscription-based SaaS platforms, training, and implementing custom solutions.

All of these open source solutions are built using open source components. Made available on the internet, itself built using open source components.

Allowing others to use components within a licensing system that promotes rather than restricts sharing has enabled us -- us as a company, but also "us" as the software development community -- to create solutions that would otherwise be impossible to fund, leaving it only available to big players. We would end up with an even greater reliance on big tech than we already have.

So, to answer the question: open source is good for business because it promotes rather than restricts competition; because it enables small players not only the big; enables us to create automation solutions for small companies with small budgets; enables the software community to build with blocks devised by the brightest minds instead of having to figure out the tiniest problems from scratch.

It's proof that sharing knowledge is the best "useful art" we have.