Quantum OS aims to build a new operating system based on Linux, with a user interface built on Qt and designed according to Google’s Material Design guidelines.
We plan to develop the desktop shell and applications primarily using Qt 5 and QML, which will allow us to build highly polished and dynamic user interfaces and will work well for implementing Material Design. If possible, we will build the desktop shell in as much QML as possible built on top of the QtCompositor API, which provides a Qt framework for building a Wayland compositor.
As for the base system, they’re still not sure if they’re going for Ubuntu or Arch.
We plan to initially leverage an existing operating system, most likely Arch or Ubuntu. Arch is a strong possibility because of the simple packaging manager, lightweight base system, and the rolling release concept. Our goal is to base our work on the latest upstream versions available, with no patches or modifications, so our work will run on any base Linux distro that supports Wayland.
Qt compositor, wayland … won’t be ubuntu then i assume?
Would be pretty funny if it was.
Think of the irony.
There would be a wayland ubuntu distro before a mir one.
Why not just build on top of KDE which has similar goals and users some of the same underlying technology?
It would allow them to focus on the design part and forget about reinveing the wheel when KDE already has a very strong technical foundation.
I would welcome very usability defaults and nicer icons on KDE, although this is pretty easy to fix by any user, but really a whole new desktop?
I wouldn’t be surprised if they are going to leverage the work that the KDE community has put into its libraries.
This is an initial annoucement, focusing on the primary goals of the project, not discussing implementation details in great length.
Using KDE Frameworks were appropriate is basically a no-brainer for any large Qt based undertaking, especially those working on desktop shells, e.g LXQt or the Maui project.
The problem with all these efforts to create yet another Linux based OS – is that they aren’t that different from one another.
At worst they are simply another established base (debian/ubuntu) with someone’s themes, desktop background and some preconfiguration. At best they are – like this – using some new custom apps. In the old days, you installed such custom apps / desktop environments onto your debian / ubuntu. You didn’t call it a new OS.
What is needed is a complete turn around. Instead of only worrying about systemd, package managers, wayland … start from the other end – the user.
What does the user need? What are the most common user stories – or what are the clusters if there are clusters of user types (developers, creatives, email-webbers, gamers, etc).
Then once you understand the users – what do you need to implement to make their lives easy? Most of these users don’t give a hoot if their system is based on rpm or dpkg – that is not a latent user need.
Given this user-oriented approach is paying dividends in other sectors – and in some sectors it is a necessary competitive edge – I’m really surprised no well funded organisation like Redhat or Suse or Canonical is doing this.
The failures of the win8 debacle were an opportunity for others to quickly fill the gap – by meeting user needs (with whatever technology) … by providing pleasant user experiences that are respectful of the human user.
Start button to shut down?
I really think the open source world is missing non-techie human-focused user experience people.
There is no lack of technical expertise and enthusiasm on the open source world. I just think the missing piece is crafting all that expertise and goodwill into something that works with normal human users.
I’d take this one step further and say that a user focused distro doesn’t even need continuity to do well, it just needs to work.
Users can deal with various interfaces and they do every day, WMP, iTunes, Office, Gmail, and Outlook all have completely different interfaces but my dad who is 64 can still figure them out more or less.
The missing ingredient seems to be something that works 99+ percent of the time. I have five calculator apps on my system by default (apropos calculator). Users need a wifi system that works, not another polished themed calculator. They need to be able to double-click to run JAR files rather than being prompted to change the executable bit.
Leaving files non-executable by default is a big piece of why Linux is so much less vulnerable to malware. That’s an antifeature you’re suggesting.
I completely agree that user safety makes a system much better and we shouldn’t sacrifice it at any cost. I do think “untrusted” applications could easily be run in a container, sandbox or as a very limited user.
I actually used the example because it came to mind as death by 1000 cuts that desktop Linux users face.
Don’t get me wrong, I love Linux and haven’t used any other OS since I switched over seven years ago. The fact that I can even suggest JARs be run in a container is what makes the OS so amazing. There is just a very clunky glue layer between visually striking desktop environments and beautiful core system that keeps them both from shining.
The probably isn’t that the system protects the user in that way. The problem is that, the last time I used Linux, you would either get, at best, a cryptic message about executable bits, or you would just get nothing.
On Mac OSX, the system will tell when you have a new executable. The first time you attempt to run it, you will get a message telling you this is a new application downloaded from the internet, and will ask you if you want to proceed with it. Any user can understand that.
Executable bit are jargon 90% of users wouldn’t understand.
I think you missed the point. I think we should explore what’s important to the people; which clearly is getting a mention of systemd in almost every recent osnews article.
Hi! I’m the developer of Quartz OS (now Quantum OS).
You’ve actually described exactly what our goal is – not to focus on the the technical aspect of an operating system, but to focus on the end user experience. That’s why we’re building a new OS, not just a set of themes around an existing distro. My goal is to build an operating system and user experience that is truly built for the typical computing user. The typical user shouldn’t worry about terminals, text configuration files, or package managers. It should “just work,” and that’s what I’d like to build.
It won’t be easy, especially for one person to do alone, but I hope others will join me in truly trying to build something that people actually WANT to use.
Sorry to say this. But this is the goal of many GNU/Linux distributions aswell.
Yeah, I hate to buy that guy, but:
http://xkcd.com/927
I dunno… maybe someone can succeed with YALD where everyone else has failed, but I don’t have high hopes.
I can only wish you luck.
I can’t tell you what to do but my suggestion would be to publish which users you’re researching – and what their user stores, prioritised, are.
http://www.mountaingoatsoftware.com/agile/user-stories
This of course should be done before developing any new software… otherwise the risk is that the cool new toys lead the way instead of the user needs.
Edited 2014-11-22 21:47 UTC
At least if TFS is correct, then it sounds like you’re building a new Plasma Shell too. I’d highly recommend working with the Qt and KDE/Plasma folks to perhaps build the Material Design functionality into Qt/KDE instead of writing a whole new shell yourself.
KDE/Plasma is already extremely built on QML; and if you really wanted to have something a little different from the normal Plasma Desktop interface you could do a spin, like the Plasma-Netbook or Plasma-Tablet spins, instead while at the same time leveraging all the capabilities and compatibilities of the existing KDE/Plasma system.
This would also just let your system work on any existing OS that KDE Supports (Mac, Linux, UNIX, and yes, even Windows), as well as the relevant graphics systems (e.g X11, Wayland) on each.
$0.02; HTH
The only link in the article is currently broken.
IMHO he should start with Manjaro.
Yeah right! Microsoft learned the hard way that a flat brain dead interface is useless and mostly unwanted. If you want simplistic, go to Apple
It’s their choice.
Although Material design is good for phones, and to a degree tablets, I don’t think it scales well.
Particularly, I don’t think it works at desktop resolutions.
So good luck to them, but I can’t see myself ever using it.
Flat design was the last thing I missed on any distro I used so far.
I couldn’t agree more. That is a very ugly UI and I’ll never ever use it. I would rather use Windows Vista or Windows ME than use an OS that sports that UI.
Why not just take Android and start fixing desktop related problems?
I am not on the flat side of things, actually, I prefer the interface OS X had before Ivy took over completely. It was not perfect, though (nothing is). I know Material is not a “flat” design but its visual cues are not that helpful, at least to me.
A point that should always be remembered, and the reason MS has such a strong prevalence on desktops, is that people use “your” platform because of the applications. As the architect of this “new” platform your objective is two fold: provide an interface to your users that have enough visual cues, flexibility and attractiveness; give developers the tools and guides to make their live easier. Really, good luck with both as no one seems to agree with the first factor and, looking to the landscape of tools on FOSS the second is even harder.
I think you really should consider working with KDE people.
Anyway, enjoy the journey. Learning is the most important outcome of any project.
I’m unaffiliated with the project, and just learned of it moments ago. Why not use BE::Shell for now?
http://sourceforge.net/p/be-shell/wiki/Home/