People don't make choices by exhaustive analysis they consider a small quantity of options and by popularity or proximity select one and run with it. Successful ecosystems make it reasonably to pick a highly visible option and come out with something good enough. The proliferation of Linux distros is largely small difference that don't meaningfully decrease fitness because only a tiny minority who themselves had little problem finding a distro think its an issue.

By contrast issues like Wayland wherein your plumbing and hardware suddenly matter and you have to become at least somewhat familiar with the plumbing in order to select a usable set of options DO matter because they increase the chance of crapping out when you stroll down the aisle metaphorically and pick a box and end up somewhat satisfied.

It honestly makes little sense to compare Mac or Windows to Linux. In the eyes of 95% of consumers an OS either an immutable characteristic of a computer or a feature thereof and they aren't interested in it as a separate product. They might well buy a computer with that OS if the overall package makes sense but they sure as hell wont install it on their windows or mac machine. Those who are technically literate enough to be satisfied are unlikely to be confused if people add 3 more Ubuntu derivatives and 2 more arch ones.

But writing a tiling wm is fun. Writing software collaboratively is the point in itself, not just the means of getting some kind of artifact in the end.

Being an open source developer means having agency over your goals and means to get to them, which you often don’t have when doing commercial software development.

You sure end up with roughly edges, but the. IKEA-effect kicks in and you like it anyway.

Right and this holds the platform back. I'm all for hobby development- I've been an supporter of open source software for over 2 decades. But nothing much has changed since that time in terms of UX and user-friendliness. (Yes I am aware a lot of things have changed, but as much as things have changed - much of it is still the same.)

It doesn't hold the platform back because if the hobbyist didn't scratch his own itch he was never going to spend 1000x as much development effort creating a Photoshop competitor or even yet join the effort to triage gnome bugs. We would almost certainly just be less their contribution for no benefit. This whole faulty analysis relies on treating open source developers as a fungible resource like employees of a firm that ought to be tasked with something different. It's not so.

Agreed. If I couldn't write my own packages for my OS and scratch my own itch, I'd never bother contributing to the big stuff.

The appeal behind libre software is the power to go your own way. If the only way to access it was to interact with groups and get your things added through a political process, I'd have found some other thing. Too much of our world is wrapped up in processes and busywork and other empty interaction, it's a waste of time.

Give me the code, the license, a text editor, and a nice cup of coffee, and I'm good. Hold the interpersonal drama. :P

This is a problem that is meant to be solved by distros. People who don't want many choices should be using a distro that makes a bunch of choices for them, instead of pushing to remove choice from the ecosysytem altogether.

Everybody using the same software creates a monoculture which is more prone to attack than a diverse ecosystem. There's a reason Windows was targeted so much by viruses: not only was the target BIG, but it was also consistent. If you broke into one normie's Windows machine, you could probably break into the rest. You're going to have a harder time of that if you target Linux.

Also, it's libre software. Unless you pay someone to do something for you, most of it is volunteer activity. And who are we to tell someone what they should do with their volunteer time?

It's also an indictment of the dominant social structures if there are all these groups and nobody wants to join them. They could start by getting rid of Codes of Conduct that give them permission to punish you for out-of-project behavior. Try that.

> People who don't want many choices should be using a distro that makes a bunch of choices for them, instead of pushing to remove choice from the ecosystem altogether.

You don't need to remove choice to push better UX and software. It's not a zero-sum game here; we can foster better conversations with people in the Linux community and start talking about what would make the experience better for the average user, not just technically proficient power-users.

> Everybody using the same software creates a monoculture which is more prone to attack than a diverse ecosystem.

That's wrong - you're using a sort of strawman argument. What makes Windows a monoculture? There's a much larger, thriving developer community on that platform with many more developers making free and/or open-source software.

> You're going to have a harder time of that if you target Linux.

"Achkshully..." Linux has had some serious bugs, but the damage was relatively low to end users because of it's low adoption rates. It has very little to do with "Linux is more secure because we're diverse" and everything to do with "1% of end users use Linux, so the target is less likely to be worth my time." And furthermore, that 1% - the majority will be more technically proficient and less likely to be tricked into running arbitrary software.

> one normie's Windows machine

This. Is. The. Problem. Framing an end-user as a mentally deficient 'normie'. Why? What's the point and what's to be gained from this?

> And who are we to tell someone what they should do with their volunteer time?

No one is suggesting we tell hobbyists / volunteers what to do with their time. What people are suggesting is that we spend a little time talking about tackling common problems together, rather than re-inventing the wheel every single time.

It's how the Bazaar model of development works. You're free to join a Cathedral and follow their policies if you want.

I would say there are even more launchers and docks then tiling WMs.

Also I don't know what would need to be better about them. Something like ulaucher or Albert is much more powerful than what is available in other platforms and I don't know how one would improve their UI.

Maybe the work needs to go into developing better UX, not so much UI or what the various tools can do. Look at PySimpleGUI (of which I was reminded of from a recent post here) for a good example of a nice project for the masses - we need better software for users.

We need better applications that have better UX.

> How many different tiling window managers do we need?

As many as people want to create.