Is there demand for a "FreeBSD Kommunity Edition"?

  • Yes, sure

    Votes: 19 18.6%
  • Likely

    Votes: 11 10.8%
  • Maybe

    Votes: 10 9.8%
  • Doubtfully

    Votes: 10 9.8%
  • No

    Votes: 43 42.2%
  • Don't know

    Votes: 9 8.8%

  • Total voters
    102
FreeBSD actively removed Xorg as one of the base sets. This means we are even further from a default desktop now than we were in FreeBSD 6.x(?).
Not only would we need to reinstate Xorg back in base but then we would need to greatly strip down KDE to make it suitable for base (and remove unsuitable licenses).

It is a lot of work for little reward compared to pkg install xorg kde5

Well, that’s interesting to know. At least it’s been tried, future attempts can be improved I guess. Unfortunately my package manager doesn’t configure ACPI integration, CPU-aware power management, and display enumeration. These are all issues beyond KDE I find. I ain’t going down that rabbit hole again. Maybe the hellosystem guys can shout loud enough and club the committers in the head a bit to get these issues fixed.
 
They dropped the ball and lost their way looking for it when they saw money in the form of Xsystems. They even offered WindowLabs WM as an option to KDE on installation after they saw me trying it out in the forums, but their eyes were glazed over by Love of the Root of all Evil by that time.

When the blind shepards led their flock too close to the edge of the cliff and put them in danger, I snuck up from behind and kicked them over the side. When they tried to make Weixiong a ghost it was PC-BSD that disappeared without a trace down to the wiki and the forums.

I'm still here beating that dead horse mercilessly and the riders can only watch.
I know the pain. I cry at night too.
 
I think TrueOS tried to initiate something like this with the Lumina desktop.
The KDE project appears extremely robust, almost like an entire OS in of itself, they even have their own semi-distro based upon Ubuntu. The Lumina project seems rather spartan by comparison. I don't know...

Microsoft has been a fairly consistent donor for the last several years, that may have something to do with it.

I liked PC-BSD, and used it during it's 9.X and early 10.X days.

And with respect to the completeness of KDE, that's one of its appeals for a standardized desktop system. I like the idea behind Lumina to make a BSD focused DE, but it's spartan and has a limited developer base. KDE has a large developer base already, and its Frameworks (KF5 and Qt5) abstract away most of the OS dependent bits from the application developers. Using it would leverage most of the work the Linux army is already doing and we just have to make sure the backend keeps working.

On a slightly unrelated note: I was also intrigued with what PC-BSD tried to do with the PBI package system. I did something like that with a Linux From Scratch system in the late 00s to manage dependencies by packaging required shared libraries with the package itself, so I had an attraction to the PBI system they used before deprecating it. I wouldn't recommend doing something like that again, though. If someone wanted to create a more desktop-oriented FreeBSD now, I would think tweaking default USE flags in the ports tree would be easier to maintain.
 
Can you clarify?

With which version of FreeBSD do you find trouble with KDE with displays?

These are issues I encountered before I gave up (temporarily) on a BSD/KDE daily driver:
  • Gamma management (restoring brightness)
  • Color management
  • Multi-monitor configuration
  • HiDPI, high resolution support
  • RandR brokenness, and tweaking
  • GPU parameter control (fan curves, voltages, frequency)

...and that doesn't even touch basic desktop I/O (ie. WiFi, Bluetooth, etc). This was on a fresh FreeBSD 12.2 install, on a T480.
 
I liked PC-BSD, and used it during it's 9.X and early 10.X days.
Then I bet you didn't see me post this:
PC-BSD 9.0 Isotope Firewall Manager breaks pf firewall

I was also intrigued with what PC-BSD tried to do with the PBI package system. I did something like that with a Linux From Scratch system in the late 00s to manage dependencies by packaging required shared libraries with the package itself, so I had an attraction to the PBI system they used before deprecating it.
Their .pbi was too much like a Windows .exe for me and I didn't want a Windows version of FreeBSD. That's why I decided to teach myself to use ports.

The matter at the top is why I moved up in the world to vanilla FreeBSD. It had become too much like Windows.
 
I think xorg in base would be a good first step. Honestly, that would save us of having GPL licensed dependencies in xorg like mentioned in a previous topic. I personally would love to see a fully permissive xorg in freebsd. That would give us real windowing and graphics support but without the GPL bs. I realize the GPL deps would need either relicensing from authors or replaced completely, the latter would be more likely, but that I do fantasize about is having a fully permissive OS. OpenBSD has several GPL elements in base such as gcc and that defeats the purpose of their similar goals. (Why though, when they also have clang?)
 
(Why though, when they also have clang?)
There are a few platforms that Clang doesn't support. I think however GCC (and GDB) is one of the very few GPL dependencies. I could be wrong though. I think Xorg (i.e their Xenocara "semi-fork") is quite free from GPL.

Admittedly Xorg in base again, would be convenient for 50% of my usage and it would define a usable workstation base a lot more. However it does open us up to "we should have a default desktop" which I don't think is a good idea for upstream FreeBSD. So I would still vote against it. Installing Xorg is a lot easier than ripping out Gnome-isms, KDE-isms and other crap freedesktop-isms.

Though I wonder if this will be re-evaluated with this new "pkg base" stuff that might be coming in.
 
That would give us real windowing and graphics support but without the GPL bs.
Honestly, you don't have to be a GPL fanboy to find that slightly amusing ;) Why would you care about the license of software you just use on your desktop box, as long as the license grants you that usage? Does it give you a "bad feeling"?

You care about licenses when
  • picking one for your own work
  • looking for projects to use in your own work
Therefore, it makes sense trying to keep GPL out of a base system. FreeBSD should be usable in any environment.

But please don't start with ideas like integrating Xorg or similar things in base! For a general-purpose OS, any GUI stuff is clearly optional (even Microsoft is slowly starting to understand that…) – the goal should be to keep base as small as possible to still provide everything you need for a full (Unix-like) system, and that's it.

If you're looking for kind of a "Desktop base", look somewhere else. I don't want to install Xorg on my servers, and I can use it on my desktops with no issues…
 
Honestly, you don't have to be a GPL fanboy to find that slightly amusing ;) Why would you care about the license of software you just use on your desktop box, as long as the license grants you that usage? Does it give you a "bad feeling"?

You care about licenses when
  • picking one for your own work
  • looking for projects to use in your own work
Therefore, it makes sense trying to keep GPL out of a base system. FreeBSD should be usable in any environment.

But please don't start with ideas like integrating Xorg or similar things in base! For a general-purpose OS, any GUI stuff is clearly optional (even Microsoft is slowly starting to understand that…) – the goal should be to keep base as small as possible to still provide everything you need for a full (Unix-like) system, and that's it.

If you're looking for kind of a "Desktop base", look somewhere else. I don't want to install Xorg on my servers, and I can use it on my desktops with no issues…
I think it would help a lot of people be able to make their own viable commercial products and much quicker with the help of such a license covering that much ground. Especially with things like the Raspberry Pi, it makes me want to dabble with making my own "products" down the line, as a software and game developer, this would really appeal to me. Even if it would be an optional thing - perhaps include it in base, but offer a means to remove it for those who don't want or need it.

Edit: back on topic, like i said, this would be a nice step to making a community edition with a full desktop to work with, and KDE in specific is nice but has a lot of things that need to be revamped before it would be ready for posing as a real desktop to the average user. In particular, there needs to be a means to remove the Hibernate option since FreeBSD doesn't have that, and also omit the option to have Suspend/Resume on systems that don't support it, auto-detected, etc and have graphics drivers auto-detect which to use at boot time, and a lot of other things to consider here. NomadBSD and GhostBSD do some of these things, but in Plasma desktop it may be a bit more challenging and forking the source code might even be necessary for some of it, such as the settings app having Hibernate/Suspend/Resume deeply integrated as "real" options, and only Suspend/Resume is, and a lot of the time even that isn't.
 
I think it would help a lot of people be able to make their own viable commercial products and much quicker with the help of such a license covering that much ground. Especially with things like the Raspberry Pi, it makes me want to dabble with making my own "products" down the line, as a software and game developer, this would really appeal to me. Even if it would be an optional thing - perhaps include it in base, but offer a means to remove it for those who don't want or need it.
Well, that makes sense from the developer's POV, you want to be careful chosing projects you build upon for sure. Still I'd oppose the idea to integrate Xorg in base (for the reason explained above). What you really want is an Xorg free of GPL'd dependencies. I guess this should somehow be possible with the LICENSES framework of the ports tree? But I never looked into that…
 
I dream of a system with no Xorg and an integrated freebsd install that ends with at least a minimal multitasking desktop. Of course, it would need flexiblity for customization like we have now.

Xorg has a client-server model invented when there were dumb workstations that could do graphics after download from a nearby server. I serviced these things but know little about X. I just suspect it has some cruft which contaminates layers of code upto the desktop.

Now life without Xorg would be difficult to achieve. It would take an R&D effort, see Skunkworks Project.
 
I dream of a system with no Xorg and an integrated freebsd install that ends with at least a minimal multitasking desktop.

I've been trying to convey this as an appropriate compromise without having maintain desktop infra. in base. But it fell on deaf ears. The difference between this (and the former), and a spork like GhostBSD is that it's managed by the same community.
 
… remove the Hibernate option …

Conditionally, not absolutely.


… wonder if this will be re-evaluated with this new "pkg base" stuff …


– for FreeBSD-based GhostBSD, the broader-than-base approach (broader than PkgBase) works nicely.

For FreeBSD, the separation works nicely. I don't foresee X.Org in base.

I dream of a system with no Xorg and an integrated freebsd install that ends with at least a minimal multitasking desktop. …

I'm reading, with interest, responses to your <https://forums.freebsd.org/threads/80287/> …
 
FreeBSD certainly does not need any "Editions" or "Distributions", maybe just more up-to-date info and tutorials to get from the base installation to a working KDE Plasma desktop.

Not sure what's the state and recommendation on sysutils/desktop-installer, but something like that should make more sense. I'm coming from the Arch Linux approach, because I never found any other satisfying Linux distribution that tries to setup a DE in a certain way with apps preinstalled that I would not have chosen.
 
What is FreeBSD? | FreeBSD Foundation – there's GhostBSD, and so on. Needs exist.
Such needs may exist and they may be reasonable, but I just don't see it that way, so this is my personal opinion only.

GhostBSD and the likes surely could have achieved the same with a port like sysutils/desktop-installer instead of trying to create an own distribution.

That's why I find the centralized FreeBSD approach much more appealing, because I'm sick of Linux-land with a gazillion of different distribution each doing its own thing in its own way.
 
I doubt that there'll ever be many distributions of FreeBSD.

… the same …

GhostBSD is ahead of FreeBSD in some areas.

<https://forums.FreeBSD.org/threads/80287/post-511611> in praise of macOS, in addition to what ralphbsz listed:
  • minor and major updates/upgrades to the system
– not simple with FreeBSD, simpler with GhostBSD, simple with macOS.

When PkgBase reaches maturity:
  • a KDE-based distribution could default to using PkgBase
– with FreeBSD continuing to default to freebsd-update(8) for as long as is appropriate.
 
Back
Top