longhorn Proper releases, perhaps once per year, rather than rolling releases.
The model that is going to be leveraged is much more akin to snapshots than the big releases from my understanding. This will be useful for OEMs and those wanting to base on top of a "stable Serpent OS" their own custom images, spins, user experiences, while allowing a faster iterative cycle.
It is worth keeping in mind that while the "platform" can be productized (such as with commercial offerings, OEM support, etc to supplement OpenCollective), the focus is on building a constantly evolving platform that is designed for all verticals, not just desktop. The "product" would be whatever curated experience is build by others on top of SerpentOS, whether that is a spin, server images, OEM offerings, containerized systems, etc. This is no different than what I do with Budgie Desktop, which is treat it as a platform that others can leverage to build a curated desktop experience that they / their users want.
The iterative, asynchronous cycle for Serpent OS would be a mismatch with a "proper" release. It forces all these features into a large drop of changes, puts pressure on the team to put out deliverables when they may not be ready, and also doesn't match up with how hardware enablement in the kernel and how the majority of the open source community delivers software. You see this with Solus at the moment. It's been over a year since their last release. I don't know why it has been so long since the last release, but they could really do with a newer kernel and release for even just hardware enablement on newer CPUs.
In the Linux landscape, releases simply need to happen faster. You don't want to put people, OEMs, etc. in an awkward position of having to choose something else over their preference because they want to leverage their new hardware, new features or efficiencies, etc. from newer kernels.
There isn't a DE that I currently enjoy using. I want something usable. KDE Plasma is buggy to the point where I almost can't use it. GNOME is slighly less buggy, but still awful. Applications that aren't specifically made for GNOME (including Firefox) are broken inside GNOME.
If only there was Budgie 😛
A good Wine package. Appropriate CJK fonts, audio and video codecs, exe icons (usually requires installing icoutils), all of it should ideally work out-of-the-box after installing Wine. CJK fonts and IMEs should be either preinstalled or available without hassle and without mucking with the terminal. GTK, Qt, Steam and Wine applications should work with IMEs out-of-the-box.
While some of this would be package enablement, most of what you are saying with "out of the box" defaults would come down to any "product" built on top of Serpent OS and the defaults decided for it (e.g. a Spin).
A rock-solid, LTS operating system
Then the question is which LTS because you will inevitably have hardware that might work on an "old LTS" but not a "new" LTS. Or is the expectation that there will be multiple maintainers that are willing to maintain specific LTS kernels? @ReillyBrogan mentioned "A strict policy on when to update linux-lts. Something like "Always must be the second to most recent LTS release" (so 5.10 now) and accepting that if this means that end-user hardware falls out of support then so be it." which I think would be a good compromise.
Staudey When it comes to DEs, Budgie is my favorite and current daily driver. This would of course preclude Wayland until Budgie 11 gets released. But I can live with something else too.
I think we would need an X11 fallback for some hardware anyways, but being Wayland-first would be a good direction IMO.
ReillyBrogan An AUR-like system for building and installing custom packages locally. I imagine this would be for things that don't really fit in the core repos, like various themes and apps. This definitely shouldn't be one of the early deliverables though, more something that would be nice once everything else is running well.
You will finally get a SUR! The S will just mean Serpent OS 😁 (no I don't think it'd be called SUR they'd just be normal source builds 😛)
ReillyBrogan Using linux-zen as the default kernel choice. I've experimented with this a bit and though as always it may be hard to measurably quantify the "feel" of the system is more responsive when doing things that bottleneck the system.
Interesting idea. Would be nice to get @sunnyflunk and @Ikey's thoughts on it.
There has been some really good points on what capabilities we should be mindful of for things like security and CVE validation for Summit.
