What is Free Software?

17 July 2024

The “free software movement” is a single-issue political movement beginning in 1986, with exactly one goal: “software freedom” for all software, and for all users. They define software freedom in terms of “the four freedoms” (no, not those four freedoms), which sound like rights.

  1. Use The right to use a program, for any purpose.
  2. Change The right to modify that program, to do things it doesn’t do already.
  3. Share The right to distribute programs to people who need them.
  4. Share The right to distribute your modifications to other people.

These are not unqualified rights. They operate within the laws and ethics of the surrounding society: if something is illegal, or wrong, it is not a matter of software freedom that you should be able to do it anyway. I think the reason people call them freedoms is that they refer to the freedom not to be coerced by the software owner. So freedom 0 is really the freedom from coercion by the owner of the software to use the software to do something. It says nothing about coercion from the state, except that that coercion requires justification.

In the modern world, none of these freedoms are guaranteed. Some limitations on software freedom are the result of state action. Software is covered by copyright in most of the world, which prohibits sharing and modifying software except with the permission of the owner. Software patents define certain algorithms or technical processes which, while publicly known, may not be implemented by anyone - except by permission of the patent holder. And owners often require the user to agree to an “End User License Agreement” (EULA) in order to use the software at all. The terms in an EULA can cover anything at all, and the only way to definitively find out a term is “unenforceable” is for a judge to say so in court, which is an expensive and intimidating hurdle.

Other restrictions happen passively, and it would require state intervention to prevent them. Most software comes to users not in the form it was written, but “compiled” into instructions that the computer can interpret directly. If a vendor distributes only this “binary” copy, nobody can force them to distribute the “source” version that they used to create it. Changing a binary in a way that results in a useful change to the program is very hard, often to the point of being impossible.

It used to be the case that even if the software could not be reverse-engineered, if someone wanted to do some equivalent task on the same data, then they could implement the software themselves, from scratch. It was later realised that the program could encrypt that data, using secret keys known only to the software or hardware owner, and thereby “lock” the information so that anyone who received a document in that format had no choice but to buy a copy of the software. The so-called “Digital Rights Management” (DRM) provides quite an extreme degree of power over the user: the owner could make files that can be read one day and not another, or be read on one computer but not another, or be read five times and then “self-destruct”. Attempting to circumvent such a restriction (to determine the keys, or to read the hidden data) is illegal, and carries a multi-year jail sentence.

Another related trend in the software industry is “Software as a Service” (SaaS), which performs a bait-and-switch: rather than being given any software at all, the user is merely permitted to access a web server, which is where the software actually runs. Since the server has total control over the software running, the user is entirely at their whim: the software could stop working at any time, for any reason (accidental or deliberate), with no “higher power” that can force the vendor to restore access. Often in this case, the data is also stored exclusively on the remote machine. After using such a system for multiple years, the “network effect” prevents many users from ever even considering changing vendor.

An advocate for Software Freedom in my sense would consider all of the aforementioned mechanisms not just annoying or inconvenient, but morally wrong in and of themselves. For instance, many people pay money in order to use software, and if the price is not too high, then they don’t see anything wrong with that. They might complain if the price goes up, but they don’t view themselves as a victim of some grand injustice. And in fact, this is not directly unjust. But in order to control the price of a copy, the owner must have control over who can share the software. The owner is free to choose not to sell the software to certain people at any price - after all, you can’t force a shop to sell you goods. If they don’t want to deal with you, then tough luck. Similarly with modifications: if you really want to add a feature to, say, Microsoft Word, then they have the right to refuse to allow it to be added. Not just refuse to add it, but refuse to let you do the work to add it yourself.

Here, the advocate says, you have gone too far. Software is too enmeshed with modern life to allow individuals such great control over how it is used.

Anyway, that’s what the advocate believes. But why should you feel the same way? I plan to write in more detail about this in the future.