Hello everyone, I’m part of the engineering and product group focused on graphics displays and GPUs for Red Hat Enterprise Linux (RHEL), and I’d like to share with you a decision we recently made. Product and technology. I would like to provide you with the context and explain the efforts we made to reach this decision.
The transition from the more than 30-year-old X Window System to the new Wayland-based stack has been occurring for about 15 years, and Red Hat has been there from the beginning. Over time, it became clear that the X11 protocol and the Xorg server had fundamental problems that needed to be addressed, and Wayland was the solution. Today, Wayland is considered the de facto solution for window and display infrastructure.
During this transition, Red Hat has supported the X.org and Wayland stacks. This shares the time we and the upstream community have to support new features and fix bugs.
However, the community developed new features and filled gaps in Wayland while development of the Xorg server and X11 infrastructure was halted. While it’s great that Wayland has been significantly improved, this does mean that there is an increased maintenance burden in both stacks, with a lot of new work being maintained in Wayland and a lot of older work being maintained in X.org. It has become difficult to maintain this situation.
As Wayland progressed and became more powerful, we worked upstream and internally with numerous hardware and software vendors, customers, the visual effects (VFX) industry, and upstream projects to understand and develop the projects required to to close the functionality gap and even extend Wayland stack. I am very proud of the work we have done:
- Lead part of the effort to support High Dynamic Range (HDR) and color management.
- Leading Xwayland as the cornerstone of backward compatibility with X11 clients
- Develop the infrastructure to support modern remote desktop solutions
- Review and development of explicit synchronization support in the Wayland protocol and relevant projects.
- Libei is designed to provide a solution for emulation and input capture
- Co-sponsoring the Wakefield Initiative to make OpenJDK work with (X)Wayland
- And dozens of other initiatives come and go.
We would like to acknowledge the significant efforts of all of these organizations and individuals, especially the rest of the upstream community, without which this project would never have reached such maturity. These efforts gave us the confidence to make Wayland the default for most use cases in RHEL 8, followed by the deprecation of the Xorg server in RHEL 9, with plans to remove it in a future release. Earlier this year (2023), as part of planning for RHEL 10, we completed a study to understand the status of Wayland not only from an infrastructure perspective, but also from an environmental perspective. The result of this assessment is that while there are still gaps and applications that require some degree of customization, we believe that the Wayland infrastructure and ecosystem are in good shape and that we are confident that we are on track to address the identified obstacles to the release of RHEL 10, scheduled for the first half of 2025.
We have therefore decided to remove the Xorg server and other X servers (except Xwayland) from RHEL 10 and subsequent versions. Xwayland should be able to handle most X11 clients that are not ported to Wayland out of the box. If necessary, our customers can remain on RHEL 9 throughout the lifecycle while clarifying the details required to transition to a Wayland ecosystem. It is important to note that “Xorg Server” and “X11” are not synonymous. X11 is a protocol that is still supported by Xwayland, while Xorg Server is one of the implementations of the X11 protocol.
While we recognize the energy behind some distributions and Fedora spins looking toward a similar future, this decision is limited to RHEL 10 – we recognize that other Linux distributions have different requirements and decision-making structures, and beyond that we are no known plans to do so We are not involved in similar efforts in Fedora, nor are we involved in similar efforts beyond sharing our knowledge.
We’ve worked to get feedback, but know we can’t speak to everyone directly. If you have any ideas or questions on this topic, we invite you to join the discussion we have set up on the customer portal.
This decision will allow us to focus our efforts on a modern stack and ecosystem, starting with RHEL 10. This means we can address issues such as HDR, increased security, mixed low- and high-density display configurations or very high-density displays, better GPU/display hot-connect, better gestures and scrolling, etc. We are confident that Wayland will provide a solid platform and look forward to working with the community and all of our partners and customers to shape the future of Linux.