While planning the Boxcar Architecture recently I've been considering what features should ship with 1.0.
From my research I feel that the best feature Boxcar could offer is addons. The unRAID community truly needs it more than any new "pretty" UI.
I've decided to shift my efforts from replicating many features in the existing webGUI to getting the addon architecture stable and production ready. Once that's done, I'll return to adding features present in other solutions.
There are a few feature milestones for 1.0:
- Core stability in Boxcar including the installer, updater, and persistence through reboots. This is crucial before user testing can begin.
- Addon server needs a public API for package/dependency management, creating new and updated addon versions, and security for author's releases.
- Client tool for creating and releasing addons (the npm to Node, the gem for Rubygems)
I've already made major strides on the unRAID side of things: the installer and updater work fantastic and Boxcar runs smoothly and quickly. I'm going to back out some exisiting features and return to them later in development.
The Boxcar roadmap
- Boxcar installable on unRAID 5.0
- Boxcar easily updatable for subsequent releases
- Boxcar Basic array and power management
- Boxcar addon management
- Addon server
- Client CLI addon tool for authors
- Addon documentation for authors
- Marketing site for Boxcar (includes docs)
That's 4 new apps for 1.0! This will mark the first official release.
- Hooks for addons have views under settings or utilities
- Visualized system stats with D3.js
- Disk management
- Share management
- User management
- Design refresh (likely)
- Shift to data collected directly from system instead of emhttp's services
1.5 and beyond
- Richer addon functionality (likely)