Hacker News Reader

Comment by Andrew_nenakhov | Hacker News Reader

You can use whatever no-evil license for your open-source project, but if you want your project to be FOSS [1], you can't. If you restrict your project from being used for evil, it'll violate the basic freedoms that proper Free licenses grant to their users.

Also, check this video by Douglas Crockford [2]. ("I give permission for IBM, its customers, partners, and minions, to use JSLint for evil.")

[1]: https://en.wikipedia.org/wiki/Free_and_open-source_software

[2]: https://youtu.be/-C-JoyNuQJs?t=39m45s

View on HN

The "Good, not Evil"[1] licence:

Crockford started releasing his JSMin software under a custom license, which he created by adding the requirement "The Software shall be used for Good, not Evil" to the open source MIT License.

[1] https://en.wikipedia.org/wiki/Douglas_Crockford#%22Good,_not...

Who cares if it's FOSS or not? Neither FSF or OSI existed when Stallman enshrined his morals into the GPL, that didn't stop him.

A no-evil clause makes your license not Open Source either. There's no practical difference between "Open Source" and "FOSS". You probably meant something like "source available".

I think what you are actually talking about is permissioned vs permissionless licenses. We typically think of open source as the later, but licenses like GPLv3, the licenses like Elastic & Mongo (commercial restrictions), and ethics are all licenses that only grant you permission in certain situations.

OSI couldn't trademark "Open Source" and the common usage of the term (source available) has steered away from the OSI definition.

[citation needed]

Common usage has matched the OSI definition for decades. Even Wikipedia acknowledges it [0]. There are always some people who are confused about the terms, sure, but those usually come simply from lack of exposure and experience around FLOSS communities. Trademarks have nothing to do with it; it's not OSI that owns the term, but the community that uses it and widely agrees with OSI's definition.

[0] https://en.wikipedia.org/wiki/Open-source_software#Definitio...

In my experience, some people are genuinely confused, but also a lot of people who misuse it and argue that OSI shouldn't be the authority because they really want to call their license "open source" even though it isn't.

i don't think that's quite fair. for example take stallman's comment, linked from your wikipedia reference above,

  the obvious meaning for the expression “open source software” is “You can look at the source code.” This is a much weaker criterion than free software; it includes free software, but also some proprietary programs, including Xv, and Qt under its original license (before the QPL).

  That obvious meaning for “open source” is not the meaning that its advocates intend. The result is that most people misunderstand what those advocates are advocating. Here is how writer Neal Stephenson defined “open source”:

  Linux is “open source” software meaning, simply, that anyone can get copies of its source code files."

i wouldn't call Stephenson's confusion here a rare occurrence; "Source available" hasn't really caught on, but instead open source remains ambiguous

I agree that the term isn't self-explanatory, which is why people get easily confused. I also happen to agree with Stallman there and I strongly prefer to use the term "Free Software" instead, as it highlights what's actually important. Nevertheless, the "obvious meaning" isn't how it's actually being used in communities that produce and use such software, so its actual meaning is still worth pointing out.

> There's no practical difference between "Open Source" and "FOSS".

These terms are often used as if they are interchangeable, but they are not. Source code can be open, and yet, its license might not grant the users necessary freedoms for the software to be classified as FOSS.

I do agree that "Open Source" is a flashier term and much more popular, yet, when we specifically talk about project licenses, I prefer to be specific.

The term "Open Source" has a specific definition which covers practically the same set of software that "Free Software" does: https://opensource.org/osd

If the license doesn't grant the users freedoms, it won't match the definition. A commonly used term for software that has its source code public but isn't FLOSS is "source available".

I've had the (mis?)understanding that these two were not in fact synonymous as you claim.

Let's just collapse on some specifics, then? Please correct any inaccuracies in the below:

Licenses promoted by FSF qualify as "Free/Libre software" - GPLv2, GPLv3, AGPL. These also qualify as "Open source" per OSI.

Licenses which qualify as "Open source" but not "Free": Apache, MIT, BSD-like.

All the licenses that you mentioned qualify as both Open Source and Free Software. A license doesn't have to be viral (copyleft) in order to be free.

See: https://www.gnu.org/licenses/license-list.html and https://opensource.org/licenses

Okay, great. So I've learned something new. But I feel like this distinction didn't come from nowhere. Did the FSF not feel the same way at some point in the past?

Not that I'm aware of. The licenses you mentioned as "free" are copyleft licenses which require modifications to retain the same license terms. The rest are usually called permissive licenses, as they don't have such requirement. Nevertheless, both kinds of licenses are and AFAIK have always been considered FLOSS and I'm not aware of any controversy around that.

The distinction came basically because of political reasons. The term "Free Software" is about user freedom; the term "Open Source" has been coined so you don't have to refer to scary freedoms when presenting practical aspects of FLOSS to businesses. That's more or less the whole reason why those are separate terms, and also why FSF doesn't like the latter one.

I won't argue about FOSS, as a specific term of art. But "open source" is a generic thing, without a single rightful authority that controls what it means.

As a member of the open source community, I think "open source" is what we make of it. Open source didn't stem from political or technical passivity, and it doesn't need to remain passive or static. What that means is up to individuals who actually do it.

Both Free Software and Open Source Software (commonly abbreviated together as FOSS) does have a established meaning though, authority or not. Starting to include things like "do no evil" in something already established, adds confusion. "Do you mean old FOSS or new FOSS?"

What you can do instead, is to come up with a new term for what you want to describe. So instead of just saying your software has a open source license, say it had a open-but-only-good-users license (whatever that means), so the rest of us don't get confused.

For bonus points, that term already exists; the original article we're commenting on says that they're calling it "ethical source".

I don't think an appreciable amount of confusion is added: there's already a wide range of "approved" open source licenses that impose varying restrictions or requirements on end users, packagers, and potential contributors. AGPL, for example, is materially different in its requirements from Apache-2, and yet both are "open source."

Put another way: it's not clear why the current complexity is okay, but other kinds of complexity (that accurately reflect the current ethical landscape, instead of the ethical landscape of software in the 1990s and 2000s) are not.

> that accurately reflect the current ethical landscape, instead of the ethical landscape of software in the 1990s and 2000s

This seems like it provides exactly the reason why we shouldn't have such a thing. What is the ethical landscape of the 2040s going to look like compared to today?

By that token, we shouldn’t have open source at all, even in the OSI’s approved form. After all, it’s a direct product of a historical ethical landscape, one that shaped our current landscape.

Have you not heard of the OSI? I think they may beg to differ on that.

Of course I've heard of the OSI. I don't understand why I'm particularly bound to their definition of open source, given that both the phrase and the concept predate the OSI, likely by decades.

Why do you care? You have a particular vision of how your software should be able to be used. Why do you care if it's "open source" or not? At least with companies, I understand the reason for marketing purposes to be perceived as open source. I'm not sure I understand why an individual maintainer of a project is so desperate to be seen as being open source if it doesn't conform to their vision.

Why does anybody care about anything? It’s certainly not life or death for me; I guess I feel like it’s an unnecessary restriction of what “open source” could (and maybe even ought to) mean.

I’m not desperate to be perceived any way in the OSS community; I do plenty of work on projects that satisfy the OSI’s definition, and that will never change. But I also don’t see why the work I do, which I do for the public and not for corporations, which is open in the most literal sense of the word, can’t be rightfully called open source.

Reply on news.ycombinator.com