Replies: 5 comments 5 replies
-
And then to solve the Linux IT Mom/Dad use case we'd have to have nice Cockpit configs between your images and someone would want to solve that and then voila, personal Chromebook killer with full blown "Linux" distros. And then maybe the Joe Ressington, the shaking fist at cloud guy would have no choice but to tell us how to design it, because it'd be literally be "run your own everything", and with ublue-os/forge you can do 100% self hosted. |
Beta Was this translation helpful? Give feedback.
-
And then for the implementation, what about a github action that took inputs of:
Then reuse the existing build process we have. Then everyone can just use it with minimal maintenance on their part, and then it'd also be usable outside of ublue-os for people who want to do their own thing. |
Beta Was this translation helpful? Give feedback.
-
Also purposely putting this as nice-to-have, medium-to-low as we're trying to chillops a bit more, so let's let this one bake for a while. |
Beta Was this translation helpful? Give feedback.
-
I like the idea a lot! I just want to make sure we have the foundation solid... Sounds like this will be some set of images building on top of.... Also, we have akmods in progress right now... I'd like to land that before moving forward here. It's part of the core, hardware-enablement, goal, but still has some things to work out... cough cough like signing keys. 😉 |
Beta Was this translation helpful? Give feedback.
-
I don't want to ruin yall's fun, but I personally would not consider using such an image (unless it added something crazy cool useful) due to the following reasons:
This isn't denouncing the whole thing, but I feel like even a |
Beta Was this translation helpful? Give feedback.
-
(This is labelled as a
concept
, which is an inprogress idea or proposal, it is not resourced or scoped so mostly just a collection of ideas. Over time it might move to aproposal
)This was initially my attempt to pitch that we should be making "developer edition" versions of our images for ourselves. You run the one with all your libvirt, vscode (because people want it this way anyway, but that's a discussion for later) and some other heavy stuff, and like, why not.
The idea being is that this fixes the "Linux IT Mom/Dad" use case. If you're like me then shout out! I need more bling but I also have a bunch of family and friends PC that I get stuck doing IT for, so it's basically a list of packages is the only difference, since we have every desktop image anyway. So your image ends in -dx, and we double the amount of images we make, yolo!
But then discussion became more about just solving the use case. I see most of the forks that people are making and while you think most of them would be outlandish most of them are people who just need a handful of changes.
Someone on this forum has asked if we should make an image just for machine learning and our default has always been like "nah we want to be vanilla Fedora mostly". But now we have that, so like, why not just let people clump around images for specific use cases, and if we run into resource limits we'll figure something out. But we should have some level of commitment or if you're an expert in the field.
And I brought an example:
@A-Lovett ended up being a beta tester. Like maybe you know how to Linux but you just don't know how to map that to this new thing, but if Universal Blue gave you the setup and you could team up with people, it sounds like a powered-up version of what SIGs do today!
So we made Cinnamon in not a lot of time, and it was "bursty" work, just some initial setup and then it's basically done, and then everyone shares in the maintenance of the thing upstream which so far has been pretty automatic now that it's set up:
Beta Was this translation helpful? Give feedback.
All reactions