I encourage you to read about SPDX license identifiers and the REUSE specification:
=> https://spdx.dev/ | https://reuse.software/
In questions of software licensing (and how to talk about it), I usually follow the Free Software Foundation and the GNU project. For example, I prefer the term "free (libre) software" over "open-source software".
=> https://www.gnu.org/licenses/license-recommendations.html | https://www.gnu.org/licenses/license-list.html.en
Informed by the recommendations of the GNU project, I prefer the following licenses:
In rare cases when I want to reserve all rights, I don't use a license at all.
The GNU project recommends GPL-3.0-or-later for "big" programs and Apache-2.0 for "small" programs. (They recommend a benchmark of 300 source lines for the distinction, but I prefer to look at whether a program consists of a single file or multiple files instead.) The reason for their recommendation is that they consider it too cumbersome that everyone who redistributes even a small program should send a copy of the license text along with it (which the GPL requires, whereas Apache-2.0 doesn't).
I used to follow that recommendation and licensed my single-file scripts under Apache-2.0. However, I no longer agree with it. I believe that it is sufficiently easy nowadays to send a copy of the GPL along with even a small program. This leaves me no reason to prefer the inferior, non-copyleft Apache-2.0 over the superior, copyleft GPL-3.0-or-later.
The OpenBSD project recommends an old version of the ISC license which has "and" instead of "and/or". I like to refer to the old version as "the OpenBSD license" for this reason. I prefer the new version, and I do not consider the change a mistake (the MIT license also has "and/or" and no one complains about that).
=> https://www.openbsd.org/policy.html
I feel like the OpenBSD project does a lot of things in a simpler way than GNU/Linux, and I feel that it also shows in the license. The fact that the entire ISC/OpenBSD license fits comfortably inside a comment header (where other licenses recommend a paragraph of a similar length just to point to the license text in a separate file) is impressive. But ultimately, I prefer GPL-3.0-or-later over ISC because ISC is not copyleft. Also, with SPDX license identifiers, the question of how much space a license takes up in a source comment has become practically irrelevant.
Although Apache-2.0 and ISC both aren't copyleft, the GNU project recommends Apache-2.0 over ISC because Apache-2.0 apparently contains a useful patent termination provision, which ISC does not have.
Because it ensures that all copies of my works will be shared under free terms, and because it is satisfying to know that it frustrates those who want to take free software and share it under non-free terms. I agree with Richard Stallman on this matter, who has spent many years talking about it. Luke Smith illustrates the point in more brazen terms:
=> https://lukesmith.xyz/articles/why-i-use-the-gpl-and-not-cuck-licenses/
Similar to CC-BY-ND-4.0, the GNU verbatim copying license allows copying a document without changing it, but it is much simpler (it only contains one sentence).
=> https://www.gnu.org/licenses/licenses.html.en#VerbatimCopying
Like with the ISC license, the simplicity of the GNU verbatim copying license is impressive. Theo de Raadt (the OpenBSD patriarch) also argues against Creative Commons licenses because of their verbosity:
=> https://marc.info/?l=openbsd-misc&m=120618313520730
But the GNU project recommends CC-BY-ND-4.0 over the GNU verbatim copying license because CC-BY-ND-4.0 is clearer about works other than text. And with SPDX license identifiers, I think that using a Creative Commons license is just as simple as, or even simpler than, using the one-sentence GNU verbatim copying license.
When I first heard about the Creative Commons licenses, I preferred their "Non-Commercial" variants (which prohibit selling licensed works), because I didn't want people to profit financially off my work when I was giving it away for free. When I first read the definitions of free software as used by the Free Software Foundation and the Debian project, I was confused why they wouldn't prohibit selling free software.
But the purpose of free software licenses is not to prevent gains, but to preserve freedom. And a license that prohibits exchanging the software for money takes away the freedom to ask for compensation, e.g. for the costs and the effort of preparing the software for distribution, which I consider to be a fair thing to ask for.
On the other hand, if someone were to sell free software at a horrendously large price, only one person would have to make the purchase and could then provide the software for download for free again. This way, market pressure ensures that prices for free software tend to cover only fair compensation, nothing more.
If you don't license your work, it will be protected by your copyright and no one will be allowed to share or change it without your permission. You might like the idea of that if you fear that people might abuse your work for evil purposes. However, keep the following in mind:
However, good people with good intentions might care about your work, and you might hinder them with your decision. For example, suppose you have an internet site where you publish opinions, personal stories, code, photos, or other things, and suppose that by a sudden and unexpected accident, you can no longer provide any licenses or pay for your site. Now admirers of your work won't be allowed to keep your site running, even in unchanged form. Instead, people will have to wait something between 50 and 100 years after your death until your copyright expires. It is unlikely that your admirers will still care about your work after such a long time, because most likely, they will be dead as well. All the work that you wanted the world to see -- at least to some extent at some point in your life, because otherwise you wouldn't have published it -- will be locked away by your copyright as soon as you are no longer there to be prying it open.
I, at least, find this scenario motivating and convincing. When I decide to publish something, then at least some part of me wants the world (or at least some part of the world) to see it. Usually it feels like talking into the void anyway, which is fine, because I like to listen to what I might have to say; but I always remember in the back of my head that the probability that someone else might listen in as well some day is not zero. If it were zero, I probably wouldn't be publishing anything at all. And if I "tolerate", or even expect, others to look at my work at some point, then I don't want to prevent good people from making sure that that probability doesn't drop to zero if they already do me the honor of thinking it worth their effort and their time.
Copyright (C) 2023 Daniel Kalak
Licensed under CC-BY-ND-4.0
text/gemini; lang=enThis content has been proxied by September (ba2dc).