Bonum Certa Men Certa

Guest Article: Are We Still Removing Problems, or Removing Options?

By figosdev

Windows options
Windows options



Summary: Monoculture in GNU/Linux as explained by figosdev

To avoid a misunderstanding, here are two things this article is NOT referring to:



1. I'm not referring to situations where there is too much extra trouble to create an option. In other words, if you switch to (or from) elogind to another method of removing a systemd dependency, I'm NOT saying that you need to make both methods an option. That would lead to a potentially infinite amount of maintenance when applied across everything.

(Don't forget though, that a similar argument is made against providing alternatives to systemd. And maintaining alternatives to that is important.)

2. I'm also NOT saying that it is important to include "the option of having handcuffs." This is not about supporting non-free software.

"When I do promote FSF-endorsed distros, I focus mostly on Hyperbola and Guix -- Hyperbola is particularly important in that I feel Guix is more highly technical, and Hyperbola is necessary for the mainstream."So with that out of the way, I think a lot of the fully-free FSF distros fall short when it comes to user freedom. I'm referring of course, to newer problems like init freedom. When I do promote FSF-endorsed distros, I focus mostly on Hyperbola and Guix -- Hyperbola is particularly important in that I feel Guix is more highly technical, and Hyperbola is necessary for the mainstream.

So I support Hyperbola about as much as any FSF-approved distro. I'm still very happy that Hyperbola was accepted to the list (it's a huge step forward for the FSF.)

My critique is based on something someone has told me, which I will treat as a mostly-hypothetical problem that I hope Hyperbola devs will either mitigate or avoid.

Essentially, it's how we go about liberating the user. There are (at least) two ways we can make the user more free: We can become nannies, or we can assist the user in their freedom. I consider these two different approaches, and the difference is the point.

"There are (at least) two ways we can make the user more free: We can become nannies, or we can assist the user in their freedom."If we choose to be nannies, we make all the decisions, we enforce those decisions, we might even (and this is the worst of it I think) choose to enforce those decisions in a way that there is little recourse for the user if "we" (developers) make a mistake in our judgment.

Here are some things I would like to have removed:

* Gnome: As much as possible. I guess we are stuck with GTK at least -- I LOVE leafpad. I use IceWM as well. Year after year, it's the best. Qt is good, but will there ever be a Qt IceWM or Leafpad?

* Pulseawful: Not enough is done in other distros to make this monstrosity optional.

* Systemd: Obviously. I refer to it as a weapon against free software.

"Some of these technologies are bad for the user and don't even belong in the free software ecosystem."And I know the blacklist Hyperbola uses -- there are a lot more things.

Some of these technologies are bad for the user and don't even belong in the free software ecosystem. If they are under a free license, they are free software at least. If they are designed deliberately to limit what else we can do, perhaps the term "Open Source Proprietary Software" (OSPS) applies. I am also ready to promote the term "Punix" for, a reduced-modularity, reduced-user-respecting corporate overthrow kind of design.

Some of these technologies make it really difficult to remove them once they are entrenched. Many of us already agree that this is a major problem.

Getting back to the critique, the longer the blacklist gets, the higher the odds that we will make "a mistake" in what we should remove. And as much as other people are making mistakes about what to include, mistakes about what to remove are just as important.

"Because there are more users than developers; we want the user to have as much say as possible."It's very important that we not become nannies -- it goes against the entire spirit of free software in my opinion. I've spent a while now complaining that the distro concept itself is ripe for user abuse and limits user autonomy.

We should always have user autonomy as our goal -- a goal that is higher than autonomy for developers only. Because there are more users than developers; we want the user to have as much say as possible. The freedom isn't just for us, we are trying to bestow it on everybody.

That means the decisions aren't just for us.

So I think it's great that we provide services to remove harmful software. It's all about how we do it, and what the user is left with in terms of options. Yes, our first priority should be to minimise the impact of (OSPS) software designed to take freedom away. But as we remove more and more we should always try to empower the user in making their own decisions, and not to simply be protective.

I'm not saying distros are evil -- they can help a lot. There was a time when a distro was the only efficient way to provide freedom to the user. In an age of automation, we will find the distro is not the most liberating thing possible. We should transcend the distro, and consider options that are distro-agnostic when possible.

That's a long road ahead, and in the meantime we should never take too much power for ourselves.

I'm in favour of the blacklist, I think it's a great idea.

I'm not saying that everything on the blacklist should be hosted in the repos either. Perhaps some of it should be. I don't know if it is -- this is partly hypothetical, so the current status is partly irrelevant to what I'm saying.

"Defaults are defaults -- the very nature of a default is that it is a choice made for everybody. That's alright."What I am saying is that freedom is more important than choice, but choice matters as well. We cannot support every possible example of choice, but we can keep some choices open. That should remain the default, except when we honestly can't do everything.

We don't have to make every decision for the user. Defaults are defaults -- the very nature of a default is that it is a choice made for everybody. That's alright.

The decisions made that go beyond defaults, we need to be careful not create another distro that tells the user what to do by doing everything for them, and denying them recourse against our own decisions.

We really have to trust the user more than that, and not devalue their judgment.

This doesn't necessarily mean making systemd an option; this is about other things than systemd. Systemd itself is an endless growth of problems -- plural.

But we definitely can't treat every option we consider a problem, the way we treat systemd. We can't enforce every choice of ours as though "we know best." We have to leave distros more open than that, or we are ultimately making ourselves a slightly improved version of Lennart himself.

"What I do know, is that apart from Hyperbola, other distros are making mistakes like these."There are no accusations here. This is a comment on concerns brought to me personally, that I have not yet verified. I hope those concerns are unfounded.

What I do know, is that apart from Hyperbola, other distros are making mistakes like these.

So my advice to the Hyperbola team, and my advice to the entire culture of free software is: Let's do better than that. It's not about perfection, it's not about making "every single possible choice" an option.

It's just about our attitude towards users, and how highly we think of our own decision making and value theirs.

Let's abandon all the hubris that we can -- not just for Hyperbola, but as a higher goal for all free software development: to create freedom wherever we are able as a higher goal, and make our job go beyond removing things we consider a threat.

Finally I will, as an example, put to you a matter related to linux-libre.

"Many of us don't really care about non-free kernel modules, as we don't think it's good to load them in the first place."Linux-libre has a bug related to its loading of non-free kernel modules.

Many of us don't really care about non-free kernel modules, as we don't think it's good to load them in the first place.

But we don't work to forbid people from loading them -- not even in the FSF distros; meaning, we don't create whitelists or use DRM-like measures to stop people from loading non-free code on their computers.

Linux-libre has had a bug for years, that prevents it from loading non-free modules.

I used the word "bug" because that's the word the linux-libre developers use. It's actually not a desired effect that linux-libre prevents loading non-free modules. It is an unwanted side-effect, according to the developers.

They have sought a fix for years, and consider it unfixable. But it is a shortcoming.

We don't want to be in the business of creating shortcomings like that when they're avoidable. Linux-libre has that shortcoming because of its goal of not "enticing the user to install non-free software." It is related to an error that shows up during boot. In fixing that problem, linux-libre creates an unwanted side effect.

"Having this goal requires an entirely different attitude to one of a nanny. It requires the attitude of a devoted civil servant -- devoted not to institutions, but to everyday people."The point here, is that linux-libre does not have a goal of creating such side effects. They are not considered desirable and it may not be fixable for linux-libre, but when it is possible we should avoid such effects.

Having this goal requires an entirely different attitude to one of a nanny. It requires the attitude of a devoted civil servant -- devoted not to institutions, but to everyday people.

To the Hyperbola team and anybody who takes up a similar list of goals, keep up the good work. I realise you are volunteers -- this consists of a think piece stating opinions, there are no demands being made here.

Licence: Creative Commons CC0 1.0 (public domain)

Recent Techrights' Posts

KillerStartups.com is an LLM Spam Site That Sometimes Covers 'Linux' (Spams the Term)
It only serves to distract from real articles
 
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Thursday, November 21, 2024
IRC logs for Thursday, November 21, 2024
Gemini Links 21/11/2024: Alphabetising 400 Books and Giving the Internet up
Links for the day
Links 21/11/2024: TikTok Fighting Bans, Bluesky Failing Users
Links for the day
Links 21/11/2024: SpaceX Repeatedly Failing (Taxpayers Fund Failure), Russian Disinformation Spreading
Links for the day
Richard Stallman Earned Two More Honorary Doctorates Last Month
Two more doctorate degrees
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Wednesday, November 20, 2024
IRC logs for Wednesday, November 20, 2024
Gemini Links 20/11/2024: Game Recommendations, Schizo Language
Links for the day
Growing Older and Signs of the Site's Maturity
The EPO material remains our top priority
Did Microsoft 'Buy' Red Hat Without Paying for It? Does It Tell Canonical What to Do Now?
This is what Linus Torvalds once dubbed a "dick-sucking" competition or contest (alluding to Red Hat's promotion of UEFI 'secure boot')
Links 20/11/2024: Politics, Toolkits, and Gemini Journals
Links for the day
Links 20/11/2024: 'The Open Source Definition' and Further Escalations in Ukraine/Russia Battles
Links for the day
[Meme] Many Old Gemini Capsules Go Offline, But So Do Entire Web Sites
Problems cannot be addressed and resolved if merely talking about these problems isn't allowed
Links 20/11/2024: Standing Desks, Broken Cables, and Journalists Attacked Some More
Links for the day
Links 20/11/2024: Debt Issues and Fentanylware (TikTok) Ban
Links for the day
Jérémy Bobbio (Lunar), Magna Carta and Debian Freedoms: RIP
Reprinted with permission from Daniel Pocock
Jérémy Bobbio (Lunar) & Debian: from Frans Pop to Euthanasia
Reprinted with permission from Daniel Pocock
This Article About "AI-Powered" is Itself LLM-Generated Junk
Trying to meet quotas by making fake 'articles' that are - in effect - based on plagiarism?
Recognizing invalid legal judgments: rogue Debianists sought to deceive one of Europe's most neglected regions, Midlands-North-West
Reprinted with permission from Daniel Pocock
Google-funded group distributed invalid Swiss judgment to deceive Midlands-North-West
Reprinted with permission from Daniel Pocock
Gemini Links 20/11/2024: BeagleBone Black and Suicide Rates in Switzerland
Links for the day
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Tuesday, November 19, 2024
IRC logs for Tuesday, November 19, 2024