Our recent blog post 2 introducing the upcoming Ubuntu Core Desktop explains what Ubuntu Core Desktop is and outlined many of the features and, more importantly, our modular approach using the capabilities of snap. This will be the first post in a series detailing the architecture of Ubuntu Core Desktop, explaining the challenges we’ve had to overcome to make this a reality, and where we’re going.
We designed Core Desktop with composability in mind, but what does that really mean? It means the OS is made up of discrete components, or building blocks, that you can add or remove.
You must log in or register to comment.
The main thing I have against snap is that they force it on people (second is the single store). This seems like a perfect use case, especially for snaps strengths vs alternatives, like the kernel snap, to shine.
We’ll see how it plays out, but I would much prefer a more modular base to the monolithic images used by Fedora Silverblue. I don’t like that it puts so much of the control and decision making into the upstream.
I personally dislike snaps, but I’ve used the old Ubuntu Core before and liked it a great deal. I might take the new one for a spin just to see if it can change my mind on snaps.
Why make the filesystem so contrived just to bow to snaps, just seems all wrong to me.
But what do I know they dropped the kiss principle years ago.
With Ubuntu Core, the kernel is installed as a snap rather than being built into the base system.
snaps might be a bright future, but I imagine it’s going to be a rocky transition. I’ll wait it out on some other distribution…
What do snapd and the “gadget” run if they aren’t on top of the kernel?
Are they something akin to uboot running on bare metal?
