Endless Linux Contributions: A Legacy of Innovation and Collaboration

Endless OS 5 is live! I wrote a little bit about how important Endless OS 5 is here. Today I want to write about the ecosystem that made it possible, and our contributions to that ecosystem.

To give a little bit of context, you can think of a PC running Linux as two things. It is the engine under the hood and it is the interface that users interact with. GNOME is the most widely deployed and influential desktop interface, and it is the one that we chose to use as our base. It is widely used by many popular Linux distributions, including Ubuntu, Fedora, and Debian. We have had to make changes to GNOME to make it work for people in our target audience, first time computer users around the world. With that said, we have been close collaborators within the GNOME community. Most of our engineers are hired from the GNOME community. We have had many board members. The CEO of the Endless OS Foundation is the current GNOME board President. In some sense we are an extension of the GNOME community, along with many other very important collaborators like Red Hat, Canonical, SUSE and countless other individuals and businesses.

I am proud of the many contributions that we have made to the Linux desktop through the years. A recent article on Endless OS had the subtitle, “This may show a future direction for Linux desktops.” Our impact has been deep enough that a Twitter user recently tweeted asking us if GNOME OS is “basically just a modified version of Endless OS.” Our Director of OS Engineering at the Foundation said, “It’s not!” But the question itself is a statement of how deep our contributions have been. Pride is a funny thing. I’m not sure whether I’m proud of the fact that I feel proud, but I feel it. We’ve been pouring our soul into lifting the Linux ecosystem and I want to describe why.

The Innovation Waiting To Happen

First, why care? The open source desktop must remain strong. Why? There are two desktop operating systems that the whole world runs their work life through: Windows and Mac. Billions of people use them every day. Neither has innovated since Steve Jobs launched the iPhone.

The only way the platform that we run our entire work life through will advance is if one of those companies innovates (which they are not) or if there is an open source alternative that allows others, like ourselves, to innovate. We believe in a world in which that innovation flourishes.

In the short-term, we won’t notice the difference if Apple and Microsoft dominate the desktop. Over the long arc of history, we will. In software, innovation unlocks innovation. It is exponential. The magic of open source software is the power of collective innovation, where diverse groups of individuals and corporations collaborate to build upon a shared foundation, creating software that is more robust and adaptable than anything that any one organization could produce. Today, something like 98% of all running software is powered by open source software. Open source refers to software or technology that is released under a license that allows users to access, use, modify, and distribute the source code. This unlocks the power of the collective. The web itself is built on open standards such as HTML, CSS, and JavaScript. Most of the world’s servers run on open source software. Even Android is powered by open source. The lead of one of the world’s largest cloud providers told me that they could never have built their entire cloud infrastructure on their own. That’s true for so much software. The desktop is basically the biggest major platform that isn’t open source. There is an exponentially better desktop that will never happen if it is locked by two companies. The PC is the jumping off point from which almost every other digital technology is developed. Imagine if the collective innovation that we have seen on the web could be applied to the PC.

I will give you one example of a product at Endless that was simply not possible on Windows. The first is our vision of apps as the heart of the PC, with asynchronous content downloads. This idea of being able to harness storage to build an ecosystem of applications that can change your life regardless of whether you have an internet connection unlocks the ability for billions of people who live beyond the reach of the internet to have access to a useful PC. Building it has required us to alter software everywhere in this open source operating system.

As those of you who follow Endless know, our mission has been to empower the disconnected. In the process, I have come to adopt another mission: to make the open source desktop viable.

There are a number of important preconditions that Endless is trying to advance for that to be possible, including an easy interface, a strong app center, and the right technologies under the hood that solve some of the barriers that Linux has historically had. Today, all of the barriers that prevented Linux from gaining ground in the early 2000’s have been resolved. And Endless has contributed a string of important insights and technologies that allow the Linux desktop to lead the pack to be the best PC platform, especially for youth. For a decade, we have largely contributed silently in the background. Today, I want to share some of those contributions.

The Desktop Design

GNOME, the primary desktop stack in the Linux community, was a great interface for engineers. But typical users and certainly first time PC users struggled to use it. For years, we have been advocating for them to improve their design. Our studies showed it was essentially unusable for non-expert users, and especially people in emerging markets. They struggled to find their games, access their settings, install apps, and set up wifi. Many literally couldn't do anything. And so we had to advance the GNOME design. Our years of research showed that we needed to lean into the apps-first smartphone design paradigm, which has been the basis of Endless OS. In 2019, when we transitioned from a for-profit to a non-profit, we formally made a strategic decision to merge our technology stack as much as we could, including our interface.

We went about this in three ways:

1. We funded a study through a third party agency, run by the GNOME Foundation, to get an objective analysis of user research comparing various operating systems including GNOME’s default interface, Endless OS and updated designs. The evidence was clear: the interface was designed for engineers, not for the mass market.

1. We worked with Tobias Bernard from the GNOME design team to look at how we might modernize the desktop interface. He evaluated alternative designs for how we might be able to bridge the gap between what we were building and what GNOME had.

1. One of our engineers, Georges Stavracas, was the primary person who was prototyping new designs, even though at the time they weren’t destined for Endless OS). We lent him to GNOME to do the prototyping to increase their capacity to modernize.

Our actions catalyzed a redesign of GNOME in its biggest refresh in a decade. As one review put it, with this update, “GNOME 40 looks delightful right out of the box.”

It’s important to note that we couldn't convince the community to make what we believe were the most important changes, primarily having a default home screen with all of your apps, as you have on a smartphone, which we found to be an incredibly important part of any design. But it was still enough for us to be able to bring our own design a lot closer and advance GNOME.

GTK 4 App Toolkit

The tools that developers use to build upon a platform shape the quality of the platform itself. GTK is the toolkit that the GNOME developers use to build applications. We consistently encountered barriers in trying to design delightful experiences with this antiquated toolkit. This toolkit couldn’t use “graphics acceleration", which is the ability to use your GPU to power interface animations and changes, which is fundamental for building modern applications. Having it unlocks all sorts of things, like the ability to resize, rotate, and move things. Basically, to have a modern OS, you need a modern toolkit. To have a modern toolkit, you need graphics acceleration. This was the key to unlocking the next version of that toolkit, GTK 4.

The Linux community is made up of a collection of important individuals. I happened to get introduced to one of them, Emmanuele Bassi, in the very early days of Endless, and he got excited about our vision and joined Endless. What a blessing. It also happened that he was the very best person in the Linux world to build this key piece of technology because he had built all of the predecessor technology that this would be built on. When I asked people when we would have GTK 4, they would always answer, “whenever Emmanuele can start building it.” And it was my decision about whether we could prioritize it. I remember the conversation with Jonathan Blandford, our VP of engineering at the time, in 2015, when we decided that we had enough capacity for it. It wouldn't bear fruit for years. But we knew how important it would be for the GNOME desktop. So we funded him to work on it. Fast forward five years and it finally launched. The launch post called it a “colossal effort,” involving 10 employers, 256 developers, 16,852 code commits and a total of 1,053,492 lines of code added. Colossal indeed. I remember that conversation with Jonathan, walking around South Park in San Francisco, ultimately unlocking GTK 4. GNOME had a problem. It couldn’t build modern applications. That problem is now solved. The proof is right there in the dozens of new applications in GNOME Circle, built with GTK 4, with more being added all the time.

App Center

You cannot have a successful platform if you don't have a successful app ecosystem. You cannot have a successful app ecosystem if you don't have a successful app store. If Linux is going to thrive, its app ecosystem needs to thrive. Historically, in the case of both Endless and GNOME, a useful and usable app store hasn’t existed. Endless had built a half-hearted attempt to build a consumer friendly app center with a set of specific features that we needed on top of GNOME Software. And GNOME Software was buggy and looked like it was designed in the '90s, because it was.

To modernize the app store, we asked Tobias to build out new designs and re-envision what it could look like. As we are trying to build an app-based PC, it is also important to us that web apps be treated as first class citizens, so that the app paradigm is present across every activity you care to do. We backed the designs up with the contributions of engineers like Philip Withnall, Phaedrus Leeds, Joaquim Rocha, Georges Stavracas and others, who have written a ton of code over a span of years. Collectively, this has now been baked into the core of GNOME Software today. The result is a more stable, snappier, and more beautiful app center. It needs some sprucing up visually but, on the heels of years of work, GNOME Software is solid.

Flatpak and Flathub

One of Linux’s great challenges has always been application distribution and installation. Every application had to be customized to specific Linux distributions. Every time someone built an application, they needed to build their application for basically every one of those distributions. For a user to install those custom versions, they had to install apps from the command line. Obviously, this was a challenge to “normal” users and even more so for first time users.

There was a nascent technology at the time that bundled everything that was needed for an application to run in one common library that traveled across all of the distributions. This meant that people could write one application file and it could run on any enabled Linux distribution. We immediately embraced this technology, one of the first companies to do so and take it to production in Endless OS 3.0 in 2016, and contributed to its development in its earliest days. We also saw that the community was only as good as its distribution hub. This technology didn't have one. There was no single repository of Flatpaks. So we built Flathub. Nick Richards, Jonathan Blandford, Robert McQueen and Cosimo Cecchi launched it on a trip to Indonesia, looking across the rice paddies and seeing the people they wanted to reach planting rice on the other side.

Flathub is now on stacks of Linux distributions, and consumer products like the Steam Deck. It is the core technology that makes it easy to install apps across Linux distributions and is the community's repository for Flatpaks across GNOME, KDE and all of the other desktops. It would not exist without our team building and advocating for it.

Parental Controls

If you are following Endless, you will know that one of our core beliefs is that computers should be safe for kids. We have built an offline ecosystem of educational applications that work on any internet connection. Parents who are concerned about their kids being on the internet, with social media and all of the other scary things on the internet, can simply turn off the browser.  Our ecosystem of content is full of delightful apps that are safe for any child. It was important for us to enable this along with a number of additional parental controls features. We developed all of this working closely with upstream GNOME. If you use parental controls in GNOME, you can thank our team for your child’s safety.

OSTree

When you update your operating system on any other major platform, from Mac to Windows to your smartphone, the updates take place with a single press of a button. When it comes to Linux, you have to open the terminal and update it manually in pieces, fixing it when it breaks. If you’re an engineer, it’s no problem. First time users? Good luck.

When we started developing Endless, we adopted a technology that a small corner of the community was using that allows users to seamlessly update from one version to the other. This means that you have a really simple way to keep an operating system up to date without having to think about it. It’s so simple and yet so important. We were the first OS to adopt OSTree in production from the very start of Endless OS in 2014, and help to prove and mature it alongside the community. Since our early bet, OSTree now sits at the heart of Red Hat’s Core OS, the largest Open Source software company behind cloud computing.

The GNOME Foundation

I believe that the GNOME Foundation needs to be strong for the GNOME platform to be strong. It needs funding to be strong. Endless has been GNOME’s largest donor for many years now. We have worked with them to fund important initiatives. We have enabled them to hire a team. Funding the biggest project in the Linux desktop is important because it enables it to accelerate development and support its community, which ultimately benefits the entire Linux ecosystem.

Into The Great Wide Open

When we switched from being a company to being a nonprofit, it allowed us to change our strategy from having to own everything to being able to push software into the broader world. We were able to build software purely for our mission. We were able to dream about better technologies, build them, and put them into a platform that reaches millions of people.

In the process of pursuing a mission of digital equity, we have had to develop this intermediary goal of ensuring a free and open world-class desktop. That led us to another powerful mission: ensuring a future for the open source desktop. I remember a venture capitalist telling me to leave his home. He told me I would not be able to build Endless OS and I refused to listen. This mission was too important not to find a way. Here we are, years later. We have been one of the great contributors to the Linux desktop. So yes, I am proud of the hard work of our team. Amazing people will get you a long way. And, more than that, in each of these cases, we found community members who also cared about what we were building. In each of these cases, we ended up being minority contributors among a community of other people who brought their energy to manifest a common vision. The GNOME desktop, Flathub, GTK 4, the app center, Flatpak and Flathub, OSTree, parental controls, these are all the polished pieces of software that they are today thanks to the efforts of a community. Thank you, all of you.

That is precisely the power of open source. And why I love and fight for a flourishing Linux.