Wayland 0.85 soon, 1.0 by end of year

The Wayland developers are heading towards a release of a first stable version 1.0 of the Wayland Display Server, saying it should be out by the end of this year. The plans for Wayland's first major release were presented by project creator Kristian Høgsberg at FOSDEM in Brussels over the weekend.

Version 0.85, a developer snapshot, has been tagged for release in the project's git repository and will be followed through the rest of this year by a number of 0.90 release candidates with a freshly frozen API, eventually culminating in a 1.0 release. Høgsberg says that he hopes to see the 1.0 release out in the second half of the year, though there is a possibility, currently unlikely, of it slipping into early 2013.

Wayland is a possible replacement for the ubiquitous X Window System used throughout the Linux and Unix world. The project was started in 2008 by Høgsberg as a part time project while he was at Red Hat; Høgsberg currently works at Intel. Wayland promises to be more appropriate for modern software architectures, compared to X's model of server systems as clients to networked client systems running graphics servers, and to the modern demands of user interfaces, such as compositing.

Applications with Wayland draw their graphical output to their own buffers and the Wayland display server and window manager pulls those buffers together onscreen. The Wayland project also includes Weston, a reference implementation of the Wayland compositor, which can run as an X client or directly using Linux KMS. Weston is described as minimal and fast and suitable for embedded and mobile users. The Wayland protocol currently sacrifices X's network transparency, but an X server can run under Wayland to allow X applications to be displayed on systems.

The 1.0 release itself will probably not presage a sudden switch over to Wayland though, despite the long term commitments of a number of Linux distributions. Instead, it should provide the stable basis for filling out the Wayland ecosystem in terms of toolkit, driver and other support to enable a switch over in the longer term.

(djwm)