The only reason I and a lot of other people use Linux on desktops is because everything we work on runs on Linux servers and we want the same OS on our machines that runs on the server (even if we also use virtualization & docker when needed).
Even if someone made a perfectly polished Linux desktop system I'd still use some Ubuntu Server with an XFCE DE thrown on top because I like developing on a system most similar to what my software runs on. Incidentally, if I need to develop Windows software, I develop it on Windows. Only exceptions to this are Android and iOS since I couldn't find any good enough dev env setup for these systems.
If you ask for a "polished Linux desktop not based on a server-Linux distro", you are putting yourself into a very narrow niche.... a niche too small be worth developing for :)
That may be true, but there are also a lot of people (myself included) that just prefer a Linux OS. I like having a choice of desktop environments. I like that a huge amount of software is freely available and just a "suda apt-get install" away. I like that there's no shitty bloat-ware on my systems.
Or maybe even an 'easy to use' theming system?
I know you can change window managers but thats technical and dense. It would be nice to have a way for designers and front end devs to mess with the UX.
I know the Ubuntu and Linux team have way too much going on but this a wish post lol.
200 bucks for a well UXed OS with support I would pay that tomm!!
Not really, you're a few clicks a way from installing KDE, Gnome or XFCE on any Linux and you can start hacking on it after some looks at the docs. Nowadays there should be plenty of HTML & CSS inside both Gnome and KDE themes... you just need the free time to waste on this!
Problem is not making the most gorgeous looking and butter-smooth animated desktop environment. Problem is you can't freaking expect drag'n'drop two work between any 2 applications because they are all so different and incompatible, or you can't have nice experiences with anything that need to integrate with the file manager because you don't just have one "windows explorer" or "finder", you have Nautilus, Dolphin, Thunar etc. And the zillion things that "half work": like, try copy pasting a folder open by ssh in Nautilus file manager... the damn idiotic thing will do a round trip through your machine and back instead of sensibly translating your operation into ssh commands...
A Linux Desktop Environment I'd pay for wouldn't have more diversity and hackability, it would have less, but all possible UI interactions and inter-program integration would be thoroughly tested and debugged. Also things like hot-plugging in/out your machine into displays and projector. All UI things that now work would work reliably! And customizatons would be a simple right-click/cmd+click + "options..." away on any UI element, not having to dig through pages and pages of settings in a damn "Control-panel-like" thinggy, or to install a buggy "Tweak tool".
As stable as Linux is for server applications, all Linux Desktop Environments I've ever used are unstable as fuck and I can get them to crash/freeze/delay-for-minutes probably once a day.
Ubuntu's unity had a right goal, but their head is up their asses - it's buggy, annoying to both power-users and new users, and pretty unhackable/obscure for non-professionals underneath. I don't event have a simple on/off toggle for "group windows in launcher task bar" which is 2 clicks away on Windows.
Their vision but "done right" and with enough respect for "regular power-users" ("power-users" who don't want to edit config files or install buggy tweak tools, that just want the advanced options baked in and thorughly tested!) is something to pay for...
Just a minimal bug desktop user experience with a reason amount of configurability for power users and if you want to go deeper there's always the command line.
There is huge problem with open-source design. The kind of design you are talking about needs to be centralized. Design is about bigger picture, about consistency and overall aim. There are so many design solutions to achieve the same thing... usualy what ends up happening is that the dev implementing it just does it in his way. When this piles up it harms the project. I understand the dev implementing it, he has all the right. But overall its realy hard. And its a social problem, not technical one.
The problem is that if you start giving your money to some group that could be working on the things that are most important to you, but they aren't working in that direction yet, then your support is understood to be a mandate for what they're already doing. Waiting until they're actually doing the things you want them to do before you start sending money their way doesn't work, because the point is to be the thing that incentivizes them to move in the direction you want.
This isn't so subtle that it's impossible to talk about and be understood, and it might work if your case is one that involves you being a patron for an individual or small group (e.g., through Patron). But for large, shambling, Red Hat-sized organizations, the message is one that's nuanced enough that you can pretty much count on it going uncaught.
But even for small groups, there needs to be a way to send the signal "please don't take this to mean that you can't stop doing the exact thing that you're doing now or else you'll disappoint; by all means, experiment on new stuff".
As someone who has worked as both a sysadmin and software developer for many years, RHEL and SuSE are not any better than the other Linux distros in terms of polish and usability.
RHEL/SuSE are "enterprise" Linux, meaning you pay the vendor for a phone number to call when things don't work according to the documentation.
As primarily server-based operating systems, they receive no extra UI polishing than their free counterparts (e.g. Fedora, Ubuntu). I would argue that they often receive even less because as server operating systems, the vendors expect you to run them mostly in a headless environment.
In fact, since they are so "enterprise" and have long support/release cycles, they frequently featured software which is years out of date even on the day of release. Because the software has had several months/years since release, all the obvious show stopping bugs have been quashed, so now it's "ready for enterprise."
I would say that back in the day (mid-2000s), I would probably have considered Mandrake Linux to be the most polished. They had a graphical software center years before Ubuntu did. Unfortunately they found, as many others did over the years, that unless you're #1 or #2 in the desktop Linux distribution list, eventually the money dries up and you're toast.
Not to knock "enterprise Linux" too much. It has it's time and place: when you need a stable operating system with vendor support that you plan to develop and release a product on and expect that product to work for several years without major attention.
But for polished desktop use, they're worse than the other free distributions like Ubuntu/Debian/Arch/Fedora, both because enterprise Linux distributions lack recent software, and because the documentation is frequently behind a pay wall and only produced by the vendor (instead of say the Arch Linux or Gentoo wiki).
I know, it takes 1000 customers at 200€/yr to hire the first developer... Too much, too little.
I develop with Mac OS and deploy my cloud services on Debian. I don't know much about Linux beyond apt-get upgrade, nginx, Ansible and Java apps. I know when I install Mac OS X that the progress bar is gray with no text on the screen, which means I don't need to dive into the technical details at any point from installation to writing a presentation, displaying it on an external screen through HDMI, tuning a mouse or plugging a printer. Has RHEL improved that much?
On step further is of course being able to install a good clone of Keynote (slideware), iMovie and an image editor for $40-80 each, but that's after the ecosystem starts gathering around the OS.
Ps: during the transition, you could support both. And of course the touch aspect needs to be cleaned up but it would still be a much better codebase to start from.
It's already a great OS, to be honest, but what about adding 20-100 employees and making it a blockbuster?
I'd love it if you gave it a go and told us how you felt the pain points where.
I'm not involved with the project I just want more users on Linux-based OS.
Interestingly, I had similar thoughts years back - https://www.quora.com/Has-any-company-tried-the-Apple-busine...
From 9Gag to HN, Windows 10 is a recurring meme of disrespect for the -cattle- consumer, in a world of Snowden and Facebook where we don't even own our computers. I say there's a demand (with money available) for a trustworthy OS in the PC world, and for a open-source in the Apple world.
Selling today's OSes might be impossible indeed. But a lot of people would switch to Linux even if it had fewer features, just because of Linux' values (which would be properly marketed around values of compatibility, open-source, reusability, privacy, cryptography, distributed services, ownership, offline work, etc). What about marketing the idea that Linux will bring pervasive effects in our democracy, like a global understanding that security systems can only be verified if they're open-source? What about the community value of VW and Tesla being required to open-source their security features, generating better security for them, for competitors and pull requests to upstream projects? We can make customers dream about a lot of values in addition to adhering to a great-UX Linux, and this value is not currently captured by neither Microsoft nor Apple. If only we had the money to kickstart a paid Linux...
People care about features and convenience. Getting stuff done and being entertained.
Seriously, talk to "normal people" and see for yourself.
So I'd bet people have been bitten enough by Android, Facebook, LinkedIn and Windows 10 that they're ripe to understand that their OS is worth 50$/yr.
- the forced upgrade to everyone who didn't agree with the upgrade,
- the telemetry
- the OEM "drivers" that display ads/change 404 pages/install a CA certificate/slow down an otherwise good the computer/install the OEM's wallpaper/break the trust between the customer and their computer.
For Android it's having coarse-grained permissions, the fact that "Ok Google" means the mic is on all the time, having Dropbox suggest to upload every time you take a picture. iOS is quite good but it's a closed garden, non-USB plugs, the impossibility of mounting the iPhone and just put mp3s on it, and requiring iTunes for music sync, which tries to push you to use their iTunes Store. It's all related to phones, but the point is, people do realize that they're providers don't care about them.
"people do realize that they're providers don't care about them"
Again, I can just suggest talking to average users. They don't even think about this sort of thing. They are annoyed if they can't get things done. If the app freezes, or drops the wifi connection etc. They don't care about permissions. They want their stuff to "just work".
You're reminding me that my parents (63 years old) and my sister might belong to a bubble made of 10% of the people. I'm highlighting that those 10% still make above 100 million people. There's definitely concern for privacy all across the board, but people won't switch to Linux because open-source software is ugly and dense today.
I'm not saying we'll sell a paid Linux to everyone. But good UX + respect for the user is still a huge market. The market will grow as the image of Linux improves and people will end up switching "because it just works better".
Someone needs to take some Linux and focus solely on the desktop environment. Make it more beautiful than OSX and Win10 and that may drive attention. That's an awful lot of work (not only the apps but getting the ecosystem to cooperate).
We've driven ourselves into the ground with our desire for free things. And we paid with our privacy, willingly or not. I think that reverting this trend will take a very long time now.
Second, people have paid enough with their privacy ("Ok Google, were you listening to me?") that they've started to understand that it's worth paying $50/yr for an OS.
Anyway I'm not a Linux expert, so I can't even tingle with a business plan. Someone needs to do that for me.
Apple bundles their OS very tightly to their hardware, so it's not as if there's really a choice between getting the OS or getting the hardware. Even the OS EULA forbids you from running OS X on any non Apple brand hardware.
To expect a new company to come out with something that is competitive with Mac or Windows is to ignore the tremendous ecosystems that took decades to build around them.
A few open source projects have had success with the crowdfunding model, when something non-trivial is in demand. If there is a possibility to get someone with a history of stunning design and UX on board, I could see a ton of money being donated for that persons time to build a rival desktop system.
[1] http://web.archive.org/web/20030207074123/http://www.lindows...
[2] https://web.archive.org/web/20040803022103/http://www.suse.d...
To get stunning UX design, upstream bugfixing and excellent marketing.
Let me explain myself. I love the levels of ergonomy and polish of Mac OS X. But it's closed-source software. If I use (and pay) Ubuntu, then great patches are sent upstream, which I can use in Debian on my servers and Arduinos. It becomes useful to everyone. With Mac OS X, we're not advancing the world. But when I used Ubuntu for work, I was impaired compared to my colleagues. Blame it on a lack of seniority, but a steep learning curve for my OS is the last thing I want at work. So no Ubuntu, no elementary.io, nothing that has rough edges.
What allows Apple to hire UX designers and do bugfixing is the revenue. Which in turns gets them a good marketing team, which persuades the world of adopting their software. Linux misses advocates towards B2B, B2C and B2Gov. If we want adoption, we need a stable income, to improve UX patterns, bugfixing and marketing.
The FSF says it's ok to sell open-source software, but you just can't prevent people from redistributing. So it's possible to design a system that requires a yearly fee to access the upgrade repositories. Of course hackers will find ways around, publish a torrent, or choose not to upgrade. But the majority of people want a system that "just works" and have money to put down for this service. Businesses, programmer shops, owners of Teslas and iMacs don't want to download their OS from an unsecure source: They want the top-of-the-art, official, upgraded releases.
My own threshold is €200 per year for my work computer. We pay that much of IntelliJ. The OS is the most important service in our stack, it deserves paid workers. My parents' threshold would probably be 50€/yr.
I think OSS volunteers will feel cheated at first sight, but the software should really push changes upstream and show the value in having a much bigger Linux community.
NB: For those curious, this comment has 18 points so far (10:43 GMT).