@glitzersachen @stefano This last Q is for @stefano and others.

My hardware runs #FreeBSD. When I need Net or Open, it is in a VM.

But given their doctrines, I would suspect neither will have an issue with UEFI.

@stefano I can't wait for FreeBSD 15+ to include a live KDE option. People are going to go crazy.
@glitzersachen @nuintari

@ricardo @stefano @glitzersachen Personally, I wish #FreeBSD would leave this sort of thing to offshoots such as GhostBSD.

We don't need to emulate Linux. We need to continue being the best at what we do.

Shocker: I'm also not sold on pkgbase.

...

I suppose I should join the core team if I am going to use the first person plural....

@nuintari IMO, the foundation is just responding to criticism from FreeBSD users. Development and attracting more users couldn’t happen without improving laptop support, which includes better WiFi and optionally a GUI.
You can take my tiling windows and TUIs from my cold, dead hands
@stefano @glitzersachen

@ricardo @stefano @glitzersachen Preaching to the choir. I still run ion, which I last successfully compiled on 13.1-RELEASE. I'm running 14.3-RELEASE now.

I don't give a frack about laptop compatibility. I want an OS I can deploy that I know won't lose stability on the altar of the almighty CADT development model.

@nuintari YMMV, and your or my priorities don't have to align with the rest of the world, and that's a good thing
@stefano @glitzersachen

@nuintari @glitzersachen I can confirm that all the BSDs work great with UEFI (without any if) and booting from external disks. I have a complete, geli protected FreeBSD desktop installation on an external USB ssd drive. No problems.

@ricardo @stefano @glitzersachen Yes, but I've been in the world where an OS was, "what I wanted," and then it wasn't, due to populism.

I'm old, I'm tired of the dumbassery. #FreeBSD's simplicity is what I want. I can build on that.

I don't want to have to deal with another decade of failed UI/UX theory because popular thought dictates that _this_ time, we got usability right.

@nuintari Let me know the day you can't build your own kernel, a custom install image, or something that truly affects usability. Because the way I see it, if having a graphical installer and a GUI helps FreeBSD gain more traction, users, port maintainers, and funding, that would be a compromise I’d be willing to accept, even if it affected me, which it doesn’t.
@stefano @glitzersachen

@ricardo @stefano @glitzersachen We said all of that about Linux, and look what we got.

I don't want to see it happen again.

@nuintari @ricardo @stefano @glitzersachen

The first Linux distros I used were RedHat 4.x and 5.x (4.x was too buggy, 5 was usable). I could compile a kernel, but that broke any of the automated upgrade things. Compiling an entire distro was way too hard. A few years later Ximian provided their own set of desktop packages. Installing the, gave a better desktop, but completely hosed the RPM database and required a clean install for each upgrade.

FreeBSD at the time required building everything from source. Today, I build the base system the same way, except instead of the install steps I do make packages and then upgrade the packages (the old way still works, but going via a package set makes it easy to install exactly the same thing somewhere else). Poudriere makes it easier to build ports because they are never accidentally picking up dependencies from my ambient environment and I don’t end up with a load of build dependencies installed after the build. It also makes it easy to create VM images with packages preinstalled.

All of these things have become easier in FreeBSD.

I think FreeBSD and Linux have the opposite problem. The Linux world (mostly the non-Linux parts, ironically) tends to copy ideas from elsewhere without fully understanding them, because they’re popular. Often they’re popular because of a big marketing push, which the Linux version lacks. Sometimes they’re popular because of some structural properties that a superficial copy lacks.

In contrast, FreeBSD avoids copying things that it doesn’t understand, even when they are a good idea, rather than learning about them and picking up the rest. For example, the way that crun builds isolated environments with namespaces, cgroups, seccomp-bpf, and prayer on Linux is an abomination and is vastly inferior to jails. But the rest of the OCI model on top of this is much nicer to use than any FreeBSD jail manager.

Just because something is popular, doesn’t mean it’s a good idea.

Just because something is popular, doesn’t mean it’s a bad idea.