Progress so far
Several items in the Workstation PRD are aspirational and/or multi-release deliverables. They depend on availability and integration of technologies still under development or integration in Fedora, such as xdg-app, Wayland, and kdbus. However, the Workstation group has made progress on several other items, including:
- Third party offerings
- Standardization and unification
- Application guidelines/design
The following development work in F21/F22 addresses these portions of the PRD:
- Software tool. The Software tool has been improved with support for disabled repos, allowing us to make use of COPR content to improve developer experience with applications like the PyCharm IDE. We also work with members of the community and upstreams on curating appdata (application metadata) for a smooth experience in the software market, which has received universally positive reviews.
- Wayland. Wayland has been moving forward fairly quickly in Fedora for some time. We are now using it by default for the login screen, and optionally in user sessions. Bugs that emerge are being dealt with steadily. Developers are working with upstream to avoid serious behavioral regression, which has led to drag and drop and improved input capabilities.
- The DevAssistant tool is included in the Workstation edition to encourage developers to set up their environments in line with Fedora recommendations.
- Qt/Adwaita theme bridge. To improve experience for users of non-GTK apps on Workstation, there is now theming work that supports Qt apps to give them a more native look in the default GNOME environment. The uptake is not universal; it depends either on upstream making use of the work, or on packagers in Fedora. Some apps are already using this work, though.
- kdbus is included in Fedora Rawhide for easy consumption and testing by the community. Note this is not a Workstation deliverable, but its inclusion is relevant since messaging over this bus will be helpful for application stacks using xdg-apps. Thanks to the kdbus team and the kernel team for working out a consensus here.
See also the list of completed Workstation tasks for more details, bug/RFE references, etc.
Planned work ahead
Some of the targets of the PRD are still ahead, including:
- Better upgrade and rollback control
- Container based applications
On the near horizon supporting these targets:
- We are using the FESCo's third party repository policy change to ship metadata for specific COPR repositories in F23, such as the PyCharm repo. We have a review open for packaging the repo definitions.
- The Workstation group is working with the Anaconda team to find ways to optimize the installer in a modular way for this edition. For instance, certain spokes of the installation process can be further streamlined given certain assumptions about the target use cases for Workstation. Also, support for additional payloads will be relevant here.
- There is already a concept and design for integrating a full Fedora upgrade into the Software app. Some initial work is already underway (or close to underway, within weeks). With the need to move away from fedup, a supportable and well maintained upgrade method is needed for users. This is most important on the Workstation whose consumers are the least likely to understand the subtleties of upgraded vs. freshly installed systems.
- Other dependencies include on release engineering work to enable new deliverables such as rpm-ostree, xdg-apps, and/or containers, and collaboration with the Workstation WG to define content and delivery. This will probably be the F24+ roadmap.
See also the roadmap of tasks for Fedora 23+ on the Workstation task list for more details, outlook for viability, and bug/RFE references.