Hacker News Reader

Comment by traveler01 | Hacker News Reader

This is one great example on why Linux do desktop is failing hard in my opinion. People spend more time arguing about BS than anything else. Guy wasn't happy with his pay, left to another company and now he's asking even more, from a failing company.

If he just wants to kill what's left he should just say it.

View on HN

Nah, the thing is there is no single way to create a Desktop for Linux, What I want from a Linux Desktop is opposite what GNOME users want, what Red Hat or Canonical want is also different from what volunteers want. Windows and Mac have 1 direction, the big ego dude on top will decide where things go and all the developers will have no choice and implement it.

Other big issue is money, volunteers that work a few hours in some weekends can't have same output as someone paid to work 8 hours a day.

If I had a ton of money I would hire a big team of developers from eastern europe, fork KDE and give them a clear direction(fix all major bugs, fix design issues, make it just work on certain hardware)

Edit: there will never be "One Linux Desktop" we will always have at least 2 or 3 very popular ones, one for power users, one for GNOME/Apple types and one for minimalists

> Edit: there will never be "One Linux Desktop" we will always have at least 2 or 3 very popular ones, one for power users, one for GNOME/Apple types and one for minimalists

And all those "very popular ones" put together with the less popular ones will have 1% of market share and will never have high quality common commercial applications or high quality first party drivers or games that run on all distributions, etc.

It's a sad story, and I say this as someone who used to be on desktop Linux for years and years and have been around FOSS for 15+ years.

It is what it is, we can't force people to work on GNOME and honestly Linux is not such a great of a kernel, the world needs something modern maybe Fuschia.

That is why when VMWare became good enough to run Linux, I just packaged my Linux zealot stuff and went back to Windows.

Yes, it is sad, but I wouldn't go as far as OP to claim that Linux desktop "is failing". That would be the case only if the goal of desktop Linux was to "get a considerably large market share". When I am contributing to FOSS, what I care about is that I improve it for whoever is using it now - maybe it is just 5 people in the whole world, that is fine. Sure I would _like_ if it was 5 thousand, and I will try my best to ensure that, but that is an extra, not the end goal.

This story is so strange to me. I get that if you read the papers and whatnot, you could think this.

But the charge of "low-quality" is just gobsmackingly daft in the face of the real life experience of people who use this stuff everyday. It may not have some of the bells and whistles, but it works. Every day.

The only "low-quality" I see is the struggles of the poor everyday folks in the Windows world who consistently have to deal with waking up in the morning once in a while and finding that Windows (or one of the big app makers like Adobe) has unilaterally decided to change or break your workflow and there's practically nothing you can do about it.

Random example. Find a big desktop publisher or any kind of big 2D graphics companies, I mean companies that create posters or stuff for websites, etc.

Disregard the OS and count how many graphic designers use non-proprietary applications for their main workflows.

I disagree. I've been using desktop Linux for 15+ years. I spent a lot of time a solo dev and yeah I was blown away at all the b.s. I had to deal with as far as making it run well. Then I got a job working for a larger company full of Macs and Windows. Man the problems those people were having? No freakin' way. I'll stick with Linux and its quirks. It's way less hassle than Windows or OSX.

Near as I can tell the high quality features that come from commercial applications center around: * Telemetry and ads embedded in the apps and the operating system * Upgrade costs * Abandonware * Lack of older hardware support * Lack of configurability * Not owning the software you run

> * Telemetry and ads embedded in the apps and the operating system

I imagine you're talking about Windows. It's annoying but for an experienced Windows user it's a 10 minute set up to get rid of them when you install the OS and at most another 5-10 minutes when you upgrade the OS.

> * Upgrade costs

Windows 10 upgrades are free :-)

> * Abandonware

Not sure what this is supposed to mean. Apple tends to ditch stuff but Microsoft's support for legacy stuff is legendary. That's why many things in the design of Windows are obiectively crap, because they really want to support old stuff.

> * Lack of older hardware support

I heard that Windows 11 bumped up the spec requirements so you're probably right. Windows 10, though, runs well on hardware from 7+ years ago.

> * Lack of configurability

Meh, after a while you just want something stable. And I say this as someone who was patching Conky and configuring FVWM for hours on end.

> * Not owning the software you run

This is primarily philosophical. In practice it doesn't matter much.

Especially since Linux distros are frequently self-serving too, see the many projects started and abandoned by Ubuntu.

A lot of your comment in practice is anti-proprietary FUD straight from Eric S. Raymond's book from 1998.

Everything has drawbacks in practice and not even Open Source devs are saints. See Gnome devs, Poettering, the glibc maintainer from a while ago, etc.

First, if you read what I said carefully I'm not just talking about operating systems but mainly "high quality commercial applications."

But to your point about Windows, why does it need to take an "experienced" Windows user to turn off ads at the OS level? In the U.S. at least all of our privacy laws center around a "reasonable expectation of privacy."[1] A private company adding telemetry and ads to an operating system erodes that expectation. If people accept that Microsoft can do it then it becomes easier to justify it state actors (for instance) do it. The same goes for what happens in software apps.

Upgrade costs: I'm talking about software here, not operating systems.

Abandonware: Cool for Microsoft.

Lack of older hardware support: Cool good for Windows. I still game on an even older laptop running Linux.

Configurability vs stability. You can have both, as I think you know.

Owning your own software: You've never run Adobe's Creative Cloud then. The SaaS model here isn't just a philosophical difference, it's a trap for anybody who can't afford spending > $120/year[2]

Yeah Linux distros (and their software) can definitely be self-serving, and that's awesome because anybody who doesn't like the direction of an open source project can fork it and go their own way. Yeah it leads to fragmentation, but with 7 billion people in the world there's plenty of room for that.

FUD: FUD implies disinformation. What did I say that isn't true?

And yeah I agree that everything has its own drawbacks. I'm just saying that there's plenty of high quality software in the non-commercial world, from operating systems all the way down to text editors.

[1] https://en.wikipedia.org/wiki/Expectation_of_privacy_(United... [2] https://www.adobe.com/creativecloud/plans.html

I can't help but think that if Linus had been more opinionated about what happens outside the kernel things would have gone a lot better for Linux Desktop. The kernel works because of his benevolent dictatorship and insistence on doing things certain ways, a certainty and direction that is lacking in the wider ecosystem.

> there will never be "One Linux Desktop" we will always have at least 2 or 3 very popular ones, one for power users, one for GNOME/Apple types and one for minimalists

Conceptually there is no reason these can not all be the same DE with different configurations. KDE likes configuration, but GNOME has more market share and seems to hate the idea that people might actually want to do things with it.

Personally, I think Linux Desktop missed a huge opportunity to have embraced GNUStep in the 90s as its one true path.

Probably the only one company that may direct Linux Desktop in any direction at all is Valve. They launched Steam Deck running on a Linux distro, they made SteamOS, etc. They clearly are taking steps in Linux, but obviously the community should get behind it, which isn't happening in a good scale.

> which isn't happening in a good scale.

I'm still waiting for my Steam Deck, so there must be hope

I admire your optimism, but almost every Steam Deck review complained about the UI. All the work under the hood looks really good, with Proton and pressure-vessel and so on, but they just don't seem to have the almighty yet tasteful UI dictator that a desktop distro needs. Unfortunately, that is also the part that the community cannot help with.

Still, can't wait to get mine and play around with the desktop. Ironically, I was planning to use mine for contributing to elementary, after having been locked out for a while (ARM Mac).

I completely agree but there have been always two forces against it: 1. People will do in their free time whatever they want. 2. They rationalize it with "competition is good".

And so here I am, the only Linux desktop user in a team of 20 developers with Macs. My faith is strong for I'm the guardian of freedom.

Started in the 90s with Slackware and have tried at least a dozen distros and right now my algorithm for installing a distro: choose a popular one, install, if drivers work keep it otherwise try another. Couldn't care less about Gnome vs KDE, pulseaudio vs whatever, systemv vs systemd, rpm vs deb, directory layout, bla bla bla...

So for example the Desktop is created from different independent parts (maybe now is different in GNOME case where it might be a big blob). So you have the Shell, the Window Manager, the toolkit for the applications and maybe some core libraries.

Now say you want to change the fonts to only use Comic Sans, you need to cnfigure the Shell, the toolkit and Window manager(for the font used in the window toolbar), a place to select Qt fonts and a place to configure GTK3,3 fonts so you have 3-4 different places. In Windows most apps would use whatever fonts they want, can't remember how is on Apple. On KDE the different components are threated as independent, your distro configures them and you have a Settings application that will dynamically plug the different stuff to give you configuration, this is why you get different places that allows you to change fonts. GNOME will not let you configure Qt fonts, and they will attempt to hide the options for you since they don't want to support anything then their hard coded stuff and the theme their designer loves.

You can't put more Settings in GNOME and obtain a KDE< KDE is different because on how it is created, small team of volunteers working on their small application and component without a big ego dude deciding if we should add a button or menu. It suffers of being designed by developers and actual users and not a pro designer that never uses the shit he designs for.

Always weird to me when people suggest that the Linux Desktop is a failure. Keeping a solid percent of all computers when your competition is effectively a coalition of the richest companies in the world should be seen as a rousing success. Anything thing short of complete annihilation is a great accomplishment.

>Guy wasn't happy with his pay, left to another company and now he's asking even more, from a failing company. If he just wants to kill what's left he should just say it.

That wasn't my understanding at all. Looks like, he left for another job to secure his income while also reducing Elementary's costs (due to not having to pay his salary). Other founder automatically assumes they should get the whole company now. End of story. Sounds to me like he just doesn't want to give up his share in the company, at the end of the day, which is 100% his right to do, since he _owns those shares_.

Reply on news.ycombinator.com