SparkyLinux harbors a flamboyant array of desktops SparkyLinux is a lightweight distro based on Debian, but it offers some choices that few if any others do. This year's first new version of SparkyLinux has two development streams: a rolling-release version 8, based on Debian "Trixie," release 2024.01 of which just landed. There's also a stable-release version: SparkyLinux 7 …
How about "consistency of interface"? In another thread, a poster lamented the failure of Linux to take off on the desktop, and one can point to the vast and confusing array of distributions and desktops as one likely cause. While Linux fanboys like to point to this level of choice as a virtue, it really only appeals to the sort of person who keeps pee in jars; for many other people, the question of deciding which of these innumerable desktops with their individual foibles, features, and issues is not even an interesting one, just an obstacle to productivity.
[Author here]
I think there are two ways to look at and approach this question.
[1]
Ubuntu's Unity is a good example. It looks like macOS but it fully supports Windows keystrokes, so you can drive it from the keyboard like Windows without needing to learn new controls. It follows CUA, but it also embraces Mac mouse controls and extends them. E.g. middle click a dock icon opens a new instance -- the Mac itself can't do that. Then the icon shows _two_ activity indicators -- the Mac itself can't do that.
And yet you can also use the Win98-WinXP keystrokes: Super+1 to 9 opens the _n_th icon on the taskbar. Not many Windows users know that.
It shows that the Unity devs knew the platform they were moving their design away from well. Better than the KDE, Cinnamon or LX[DE/Qt] folks did, for instance.
Good design means you can adapt someone else's design _and_ keep a good familiar UI.
[2]
The flipside of that coin: instead of 1 or 2 _really good_ versions of the Win9x UI, we have a dozen rubbish ones. That is not diversity. That is just lack of originality.
Want to copy someone else's design? Step 1: learn it absolutely inside out, know where the edges of its envelope are, then either expand that envelope or using that knowledge intelligently design something that retains the good bits and discards the bad.
FWIW it's clear MS no longer knows its own design well because since Win8 it's progressively degraded with every version.
"The flipside of that coin: instead of 1 or 2 _really good_ versions of the Win9x UI, we have a dozen rubbish ones. That is not diversity. That is just lack of originality."
I think we're in violent agreement here. Rather than focusing on a finished and delightful experience, many, if not most, UX developers in the Linux world seem way more interested in putting their imprimatur on a new distro or slightly tweaked user interface. The result is as dull and unsatisfactory as the many brands of corn flakes in the cereal aisle dressed up in different packaging.
Assuming that question means what I think it means (it can be hard to tell), remember that all this code did not write itself. It exists because somebody felt the need for it, and sat down to do the work. Or hired somebody else to do so.
You are free to criticize their work, of course. But a more productive use of your time would be to offer an alternative. If you haven’t got the skills to do it yourself, why not get together a group of people who agree with you, and hire somebody to implement your ideas for you?
[Author here]
You are making many assumptions that are wrong, and based on those, you are making recommendations which the same are bad ideas *that caused this situation in the first place.*
> remember that all this code did not write itself. It exists because somebody felt the need for it, and sat down to do the work. Or hired somebody else to do so.
This is wrong. You assume someone needed it. That is false. It is much more likely that the code *did* exist but that the person who wrote it, or paid for it, either:
* did not know
* could not use the existing code (e.g. it was not FOSS or it was under a license that didn't permit the use -- such as OpenZFS)
* didn't like it (e.g. GNOME exists because [a] KDE is C++ and [b] Qt wasn't GPL)
Of these, #1 -- ignorance -- is the likeliest.
> You are free to criticize their work, of course.
Actually, no. I am getting paid for it.
But this is not about freedom: that's another false assumption.
I think there is a lot of value in having editors and commentators. Someone who can lift their head up from the codeface and look around and learn the landscape. Just like some grazing animals have "spotters" who forego the chance to eat in order to watch for threats.
I think it's important to have people who try _all_ the desktops on _all_ the distros and compare them.
> But a more productive use of your time would be to offer an alternative.
No. This is not only wrong it's backwards.
Problem: there are too many half-done alternatives.
Your answer: "write another better alternative!"
This is such a bad answer it's literally a joke:
https://xkcd.com/927/
I am saying that, because I have my head up and I am looking around, I can see massive effort wasted on duplication.
Your response is to say "get your head down and do more duplication!"
That is beyond bad advice: it's absurd and it's tragic that you think that's an answer.
> If you haven’t got the skills to do it yourself, why not get together a group of people who agree with you, and hire somebody to implement your ideas for you?
No. Absolutely not. This is the reverse of the reasonable, pragmatic, sensible course of action.
>> remember that all this code did not write itself. It exists because somebody felt the need for it, and sat down to do the work. Or hired somebody else to do so.
>This is wrong. You assume someone needed it. That is false. It is much more likely that the code *did* exist but that the person who wrote it, or paid for it, either:
Of course the code did not write itself. Of course somebody needed it, even if it was only the person who wrote it in the first place. Do you really think code can spring into existence just by itself?
>> You are free to criticize their work, of course.
>Actually, no. I am getting paid for it.
So you get to make money off their work? Have you thought of contributing any of it back to them, to help them improve it? Like I said, code doesn’t write itself.
>Your response is to say "get your head down and do more duplication!"
>That is beyond bad advice: it's absurd and it's tragic that you think that's an answer.
That is too bad for you. You seem to think there are too many alternatives already: that has to be an economic argument. The large number of alternatives simply reflects the low barrier to entry. If you don’t like that, you are free to stick to the proprietary alternatives: now they are properly expensive to develop and maintain, with a properly high barrier to new entrants, and they need multi-million-dollar advertising campaigns to tell everyone how wonderful they are; is that the sort of thing you need to be able to take a product seriously? You know where to find it.
[Author here]
> Maybe you should put together your own UI design/distro, to show us how it’s done.
I have of course considered it, but as Alan Kay put it very well:
https://www.youtube.com/watch?v=rmsIZUuBoQs
You can spent a lifetime fiddling around with Linux and not learn anything.
If I were to do it, what I'd do would be something akin to:
* GoboLinux' filestsystem layout
* All binaries statically linked, no shared libraries at all (cf. Suckless sta.li or Oasis)
* Automatic CI/CD driven rebuilds whenever a system library is updated
* ROX-desktop and ROX-session as the GUI
* Only Appimage packages, or maybe Snaps. Flatpak is too hairy underneath, the swan of cross-distro packaging: looks elegant on top, but underneath, frantic flailing
If one were going to do it, be radical. Don't waste time with a trivial variant of existing stuff.
* All binaries statically linked, no shared libraries at all (cf. Suckless sta.li or Oasis)
* Only Appimage packages, or maybe Snaps. Flatpak is too hairy underneath, the swan of cross-distro packaging: looks elegant on top, but underneath, frantic flailing
Wouldn't the first of these preclude any need for the second?
Back in the days of the Cold War, it was entertaining to hear stories of defectors from the Soviet Union, encountering Western free-market capitalism for the first time.
They’d step into a supermarket, and be completely bewildered about why there needed to be half a dozen different brands of laundry powder, or a dozen different kinds of toothpaste. Soap is soap after all, and toothpaste is toothpaste; why do you need more than one?
Nowadays, we see the same attitude in people accustomed to living under the Microsoft/Apple regime, suddenly discovering the wide-open home-on-the-range that is Linux and Open Source. People want choice, and so we have the choice. Don’t like any of the existing choices? Or don’t even like having a choice? You know what you can do.
I hear it’s the reason why cars don’t sell as well as PCs—all those different models, with different numbers of seats and doors, some with extra luggage space at the back and some with not, some built high up and some low to the ground, different numbers of cylinders in the engine and some with no cylinders at all—they can’t even agree which end of the car the driving wheels should be on!
If they could get rid of all this diversity, and just settle on one design of car, maybe they would be more popular.
If they could get rid of all this diversity, and just settle on one design of car, maybe they would be more popular.
Seen many cars with the clutch, brake and accelerator pedals in a different order? With the driver's seat at the back instead of the front? There really is very little difference in fundamental designs.
> Seen many cars with the clutch, brake and accelerator pedals in a different order?
>With the driver's seat at the back instead of the front?
>There really is very little difference in fundamental designs.
You can think of those aspects as the “kernel” of the driving experience. A common kernel can drive very different hardware configurations—engine type/power, transmission, front/rear-wheel-drive etc. Then there are things like infotainment systems, driver aids, upholstery paint jobs, which one may think of as the “userland” and “package management” parts.
Do you see the parallels with Linux distros starting to emerge?
In Spain you will find that, even though potatoes loom large in Spanish cusine, your local supermarket (or just market - this is Spain) has only one type of potato on offer. It's a fine potato - tasty, works in any dish, great for tortillas, salads, roasted, baked, steamed, chipped, mashed, ... We've had Spanish visitors who are genuinely perplexed by the variety on offer here in the UK. Nowadays I tend to just buy whatever is sold as "white potatoes" (usually the cheapest). They're fine. My life is still good.
The choice you think you see in supermarkets is sometimes illusory, toothpaste being a case in point. I once counted about 17 different varieties of toothpaste from one brand. Taking the time to look at the ingredients, many of these were identical. They were simply the same thing marketed using different fear factors.
There's always a whiner.
Standing on a plain: it's too flat.
Standing on a mountain top: it's too mountainous.
Amidst hills and valley plains: there's too much variety.
Oh woe is us: some people live by only ever mooning over what they haven't got, and deriding what they have.
'I don’t think it’s whining when people are offering constructive suggestions that would promote the wider adoption of Linux'
Who says this is the primary goal of 'Linux'? Where is that goal written in stone?
Yeah, okay 'wider adoption', 'the [fabled] Year of the Linux Desktop', yadda, yadda, yadda, matters to some people, but I always understood that the primary and essential principle embodied by 'Linux' (and FLOSS in general) is 'freedom'.
In which case the 'number of desktops'—good, bad, and indifferent—is simply an expression of that freedom. We might not appreciate that plethora of choice, but then others may not appreciate our particular interest/obsession. Hopefully we all appreciate the freedom to pursue/indulge the aspects of freedom that Linux/FLOSS enable.
Who says this is the primary goal of 'Linux'? Where is that goal written in stone?
> we feel that the Linux desktop world badly needs more diversity of design.
I understand the idea. But ...
People go to a zoo to see all the animals and to marvel at the diversity, their different shapes, sizes and habits. But that doesn't mean they would want any of them in their home.
ISTM what people want from either a domestic pet, or a Linux desktop is something clean, well-behaved and easy to look after.
I realise there are always a few "look at me" types, who feel the need to display their individualism by choosing something exotic - just like all the other individualists do. However, the lessons of which O/S's are successes and which have marginalised themselves by offering a slew of alternatives - each one sufficiently different from the others to be a right PITA to learn, program and maintain - those lessons are clear for all to see,
[Author here]
> ISTM what people want from either a domestic pet, or a Linux desktop is something clean, well-behaved and easy to look after.
I don't think so, actually, no.
I think that's what a *lot* of people want, yes. But all of them? No. It is an overgeneralisation.
If someone wanted a dead easy OS that just does the one core task, then there is one. It's ChromeOS.
But it's not for everyone.
Some people want to keep snakes or tarantulas or a tropical reef tank, demanding things which are hard to keep, *because* it's a challenge. Same as some people climb mountains. Not for the view, not because it's fun, but because it's very very difficult and few do it.
While "diversity" is fun and gives the impression to being good, it acts as a drag on software development.
When each version of an O/S has a UI that needs it's own special coding to get stuff done then it rapidly becomes expensive for software makers to support all those variants.
Likewise, when every user is running their own special, customised, desktop, Linux version or hardware, then support for all that stuff becomes a nightmare too.
It seems to me that the operating systems that are successful: that have non-negligible market share, are the ones that offer few or no, options for variation. We see the same thing in mass production of consumable durables: few options, but cheaper and selling more - than for bespoke, custom, boutique products where every one of the is different from every other.
> ... a UI that needs it's own special coding to get stuff done then it rapidly becomes expensive for software makers to support all those variants.
But what UI needs "it's own special coding to get stuff done"? Surely most GUI coding uses one of a handful of standardised toolkits (GTK+, Qt, WxWidgets, Fltk,...) which will function perfectly well -- i.e., "get stuff done" -- with pretty much any UI*? Sure, your application may jar with the look and feel of your UI, but that's always been an issue, and not just for Linux.
*At least X11-based UIs... I'll concede that this may become more problematic with the introduction (take-over?) of Wayland; though even then, I'm guessing the major toolkit developers will be working on Wayland compatibility as I write. As an aside, I had the joy quite recently of having to develop a GUI scientific application in pure low-level Xlib (there was an imperative to get bits on the screen at high frequencies with minimal latency). I will never, ever do that again.
> ... rapidly becomes expensive for software makers to support all those variants.
Make that “... rapidly becomes expensive for proprietary software makers to support all those variants”. Open-source developers don’t have to care, because they can leave it to the distro maintainers to package up their work in some suitable form.
This is part of what makes it so easy to distro-hop: there is no “vendor lock-in” associated with different Linux distros.
The multitude of desktops is not really putting a drag on software development. I used Unity, Cinnamon, Plasma and Xfce in the last years (based on the use case) and was always able to install the software i needed, no matter if had been developed with a different desktop in mind. OK, I lack the strange fetish of needing all software to use the same toolkit. Software created using Gtk works perfectly on my KDE. It looks different and and stuff like file selector dialogs look differently, but is this really such a big thing? If have also seen a big variety on Windows and even Mac, so basically this ideal seamless look & feel is an ideal that does not exist. Those Electron or Flutter based UIs also do not really work towards that ideal.
People should just get their heads out of their behinds and accept that other people do other choices. Just develop their software as they like to, it will most likely run on all major distributions and most of the more obscure ones.
"It seems to me that the operating systems that are successful: that have non-negligible market share, are the ones that offer few or no, options for variation"
The larger the market share the easier it is to dictate to users that that's what you get, whether you like it or not.
It always ends up badly with abuse of a monopoly position.
The freedom of choice is a double edged sword. Raspbian is fantastic out-of-the-box; with the exception of the terrible package manager front end.
Mint is great until you realise how behind the curve some software patches are; which isn't a problem for most users. It did become an issue for me when I wanted some more bleeding edge features and software.
Manjaro is very nice (and the daily drive) though it does leave the user to do some of the bleeding on the bleeding edge. Pacman is an acquired taste compared to rpm and apt. The AUR functionality is both simultaneously amazing and frustrating at same time.
CentOS proved useful for certain esoteric hardware. It's pretty dire as a daily drive desktop however. (And IBM are hell bent on ruining it).
For all the foibles; I still spend a tiny fraction of the time administering any of the above than Windows demands. Different tools have different places; one size does not fit all. I for one welcome the options.
It is similar reasoning as to why the OH has an Apple laptop. Yes, the hardware has it's foibles and store lock in a pain, but from my POV; I never have to deal with the inevitable hissy fit when "the computers broken". This is a not inconsiderable chunk of my time back on my terms.
It's not compulsory, you know!
Over a {cough} few years, I've refined my setup so it behaves pretty much exactly as I want, and is as lightweight as possible. No doubt other's would hate it. Fine, they can (if they want to) set up theirs differently.
They can, of course, simply use what they are given... and moan about it.
Some of the options need to be good for choice to be worthwhile.
From http://downloads.bbc.co.uk/aboutthebbc/insidethebbc/howwework/reports/pdf/creative_book.pdf
"I remember Hugh and I wrote a sketch in which I played a waiter who recognised a diner in my restaurant as a Tory broadcasting minister. I clapped him on the shoulder and told him how much I admired his policies of choice, consumer choice, freedom of choice. I then was horrified to notice that he had only a silver knife and fork for cutlery at his table. ‘No, no, they’re fine,’ said the puzzled politician. But my character the waiter raced off and soon returned with an enormous bin liner which I emptied over his table. It contained thousands and thousands of those white plastic coffee-stirrers. ‘There you are,’ I screamed dementedly at him, virtually rubbing his face in the heap of white plastic,‘now you’ve got choice. Look at all that choice.They may all be shit, but look at the choice!’" - Stephen Fry
You missed the entire point of that joke. The diversity of choice offered by corporate production is an *illusion* they're all the same, maybe with different paint.
This has nothing at all in common with the Linux desktop landscape, and would only describe something that has been managed in a similar way, like Mac OS.
Sparky seems particularly well geared to many devices otherwise destined for landfill. A friend passed me a Win7 sub notebook a while ago, which couldn't update due to disk space issues and was functionally useless. Certainly could not update to later versions of Windows either.
Sparky, with only limited help from me, was able to pick up the somewhat proprietary software needed to get everything working on said hardware; and not only runs but runs very well indeed.
I'll put my grumpy old man hat on for a minute, and start an "in my day..." response.
Since when is a 3.5GB OS considered "light weight"?
When Windows hit 1 GB, we all thought it was typical bloated Microsoft crap, and now a "light weight" Linux install is 3.5GB!
Yes, I know there are some distros out there that are very small. I'm just remarking on the re-calibration of "light weight"...
You guys are just used to distributions like the more commercial ones. Community based distributions usually provide just about any software available for the platform. This is not a flamboyant array.
For example,
https://wiki.archlinux.org/title/Desktop_environment
Take a look at "officially supported" and "unofficially supported". I'd paste, but it's too much :-)
Arch's functionality is pretty awesome, however, the status of "community maintained packages" is a mixed bag. Manjaro makes some of the Arch framework a bit more accessible - this is a good thing for those of us that don't have 10 hours a week to worry about sysadmin. The big and frequently used stuff is fine. Oddballs with at most 5 users not so much!
The latter is the reason I keep a copy of CentOS around despite it's age and IBM.
For the record, I have done a Gentoo walkthrough; that's a useful exercise in it's own right but only the sadistic with a lot of time to spare are going to daily drive that.
[Author here]
> This is not a flamboyant array.
I presume that you did not study the list closely. Perhaps you find 30 an overwhelming number?
Did you look at the pictures in the article? Because then you might have noticed that one of the Sparky desktops for which I _provided a picture_ is not in Arch's list, let alone that three others mentioned in the text are not there.
The OpenBox-based Minimal GUI, a 1.5 GB download which installs into 3.5 GB of disk, and Minimal CLI, which is half that size.
Minimal is 1.5GB now? Mind you, even the latest Lubuntu is 2.7GB and it's not that long since it fitted on a CD.
As we've said before, we feel that the Linux desktop world badly needs more diversity of design.
"We"?
Before I had ever heard about computers, I didn't even want a computer.
I didn't even know I wanted a portrait display until I walked into a business computer store with my dad in the 80s, and saw something that I think was by Xerox.
I most passionately came to want a software company to stay out of my business when Microsoft forcibly introduced Windows 10.
Funny about choice (and expectations management) in implanted chips from someone on Mastodon https://social.makerforums.info/@NedMan/111848623311348167
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
