Preferred DE of the FreeBSD users

Which is your current DE or WM? If not listed please specify!


  • Total voters
    192
This is a screenshot of plasma on my old thinkpad, with no other apps running it's used about 1.7 GB of RAM. Admittedly much more than windowmaker, fvwm etc. But if you consider that a typical modern laptop has 16GB ram and upwards, it doesn't seem too bad. Yes it was bloated on something like an X60 with 2GB RAM.

Of course if you're going to run this on lower-powered hardware like a pi then perhaps its not the best choice. And obviously you wouldn't use it on a server, I'm thinking about desktop here.
 

Attachments

  • Screenshot_20230402_193753.png
    Screenshot_20230402_193753.png
    81.6 KB · Views: 64
Considering the entirety of Windows XP (kernel + services + UI) ran in ~128MB of RAM and had considerably more functionality than KDE4+ does; I feel that 1.7GB is bloated regardless of the machine you run it on.

How much RAM does macOS need these days? I recall a few revisions ago, it ran quite nicely on a ThinkPad X61 with ~2GB of ram. With the added functionality and integration (and commercial support) that macOS's desktop offers; I don't think KDE in its current direction is going to be an asset at attracting new users. Gnome 2 was starting to do a very nice job until it was killed; alongside what seems like the success of the open-source desktop along with it.

I get it. The old saying goes; "unused RAM is wasted ram" but... yuck! ;)
 
I guess nearly 2 gig is quite a lot! :) I suspect my sense of what is big and small has been warped by multi-terabyte SSD's that fit in the palm of my hand, 64GB RAM in two SODIMM's, 16-core cpus, etc.. Reality seems to have shifted left. Or shifted right. I can't work out which.
 
As for gnome, yes, '2' wasn't too bad, but once '3' came out and I couldn't even resize a window by the border without fudging around with "gnome-tweaks"... hideous. Please, not gnome 3.
 
Personally I think something so bloated will scare people away!
Absolutely!
Above all, the pros.

Any default DE will be more easy for newbs to usher in.
They are attracted by the many colorful blinking feature stuff they are used to.
But at the same time it'll scare off many established users,
who know this is mostly just sparkling, but useless, and bloated crap - the pros 😎

I agree with you:
the more bloated the DE the bigger this effect.
(Especially KDE 😁)

...........

The pull at the top of this thread proves it (as far as it can be extrapolated to be representive):
If they have a choice,
most users do not even use a DE, but a WM only.
Second most use XFCE,
which I personally liked most until I decided for me not to use any DE anymore, which of course is my personal taste, only.

However,
the pull is not representing the full truth,
since "not using any GUI" is not an option one can chose.
I wouldn't be surprised if there were as many votes as for WM only, if not more.

Of course the OP specifically asked for desktop users.
But face it:
GUI-less systems are one of the most important main pillars FreeBSD stands on.
Sometimes you still read FreeBSD may a primarily, even only server system.
The idea of a modular, multiplatform, adaptable, multipurpose system is not grasped by everybody.

So, with any decision for a default DE/WM
you'll not only lose the established users,
the non-GUI applications, and their users, only,
but the whole idea of a highly adaptable multipurpose OS would be killed.

The crucial point of any WM/DE installation is always (any BSD, any Linux)
to get X (or wayland) running,
which always depends on knowing exactly which graphics adapter you have,
and chose the right kernel-module (driver) for it.
Which is rouhgly assumed the reason for half of installation issues.
One may also say:
Not reading the documentation (HB and supported hardware list.)

The other half is the fact that not all graphics adapter's drivers are available for open source OS immediatly.

Deciding for a WM/DE (doesn't matter which one) to become default
will not change that.

The installation cannot be done more easily,
as it is realized under FreeBSD.

Under FreeBSD the whole installation of a DE/WM is done
by one, or two lines of
pkg install ...
I don't find that so hard.

Who finds even that unacceptable may revisit his idea
why not stick to Windows, Apple, or a Linux turn-key distri.

And yes, of course, one has to find, and decide for the DE/WM one wants to use, too.
This can be solved by a default WM/DE:
to avoid seek, and decide.
But again, if one doesn't want that either, there are several other choices where this heavy burden is already made.

To me the WM/DE is not my main concern.
The work with the machine is.
I decided for one I like,
configured it to my needs,
so I have something I can and like to work with,
so the issue WM/DE is settled for me,
I move on, and work with my system.

After all within the world commonly known as unix,
most of the work is, and best shall be done as text within a shell anyway.

The usage of programs you use with a GUI (browser, gimp, blender, libreoffice, CAD, EDA,...whatever)
is mostly assigned by the design of the program, and not so much by the DE/WM.



There are so many ways of getting food:
restaurants, fast-food chains, supermarkets with convenient food, kitchens of all sizes, and kinds of equipment.
One likes cheeseburger, another one likes pizza, or prefer sushi.

If someone is trying to convert a sushi-bar into a burger grill, because he's not satisfied with what the existing fast-food stores offer,
but at the same time too lazy to cook by himself,
and believes eventually one day there will be a restaurant serve exactly the food he expects, tailored to his wishes to the point,
then in my eyes the whole idea needs to be revisited from its very origin.

In my eyes this is a matter of attidude,
not an issue of to few possibilities, or wrong offered options.

So, bottom line:
Since this thread originally aimed to find a WM/DE which may be elected to become FreeBSD's default one, the whole discussion about which DE is better or worse is mostly about personal taste,
and so missing the point.
 
Just so long as I can still run wmaker on it. The problem will be when X11 is deprecated and only wayland remains. Perhaps X will survive longest in openbsd (xenocara). Yeah, peace... lots of peace.
 
I really dig wayfire. I think it's great to see wobbly windows and old compiz effects working on modern systems again. :D After plasma dropped the old desktop cube code it's good to see wayfire has preserved the cube. I didn't even know about wayfire until FreeBSD.
 
I'm not sure this day will ever happen.
Me neither.
X may a dinosaur, but is sophisticated,
may complicated for WM developers, but is powerful and flexible.

A new windowing system could also appear and eventually replace both.
In theory, yes,
practically not really.

Wayland itself proves both:
Neither X will die soon, nor there will be some replacement killing both.

As far as I knew about wayland and may judge,
wayland is more stable, faster, ... better than X (maybe, I don't really know.)
BUT:
According to last time I've checked,
with wayland you may pick from 2 or 3 (4?) DE, only.
If you're not happy with one of those,
you either completely develop, and program your own one ....,
or pick one of the vast choices which run on X, and so using X.
 
Me neither.
X is a dinosaur, but sophisticated,
complicated for WM developers, but flexible.


In theory, yes.
But wayland proves: not really.

Point is:
as long as the most used and most popular DE/WM are not converted to it,
which is an effort by its own additionally to write such a system,
X will survive.
Isn't Wayland that thing that substitutes network transparency for wobbeling window effects?
 
Isn't Wayland that thing that substitutes network transparency for wobbeling window effects?
Haha. Thats a good way of putting it.

What I think will be a good compromise is a decent low level library like Xlib -> Wlib to provide most of the abstraction.

And then a decent wayland compositor that allows Wlib / wcb based Window Managers to interface with it.

The kids get wobbly windows and the grown-ups get usability.

Unfortunately at this point Red Hat is focusing on maximum market penetration rather than something actually useful.
 
I think the developers of Enlightenment have misunderstood that Linux is just a kernel and thus can't be a primary platform ;)
Ahh... I'm under the impression that kpedersen is messing around...
From the homepage: ”Enlightenment is a Window Manager, Compositor and Minimal Desktop for Linux (the primary platform), BSD and any other compatible UNIX system.” (emphasis by me).
this quote, I'd read to mean that Enlightenment devs primarily target Linux when writing code for Enlightenment... and that's why they call Linux a 'primary platform'. So it's nice to know that Enlightenment is also available for FreeBSD.

You can call x11/kde5 a platform, or www/firefox can be a platform, too. Heck, even GTK can be called a 'platform'. 'Platform' is a very loosely defined term:
 
Isn't Wayland that thing that substitutes network transparency for wobbeling window effects?
I don't know.
I never got any deep into wayland.
I just read the wikipedia page about wayland,
considered for a moment, if it could be a choice to use it ('cause it all sounds better than X),
found out, neither WM I like to use runs on wayland.
End of story.
 
The kids get wobbly windows and the grown-ups get usability.
My point exactly.

I just use more words
to be not misunderstood as being rude,
to substantiate the said with plausible points,
so eventually convincing some kid (of any age) to become a grown-up 😁

Good example was this 3D-Cube one may use with KDE,
which had the windows as surfaces of the cube.
I never tried it, because I already knew,
this will not, because can not enhance the usability in any way,
but being just pure optical fiddle-faddle to show off, only.

If you switch from one window (e.g. browser) to another (e.g. shell),
it does not enhance usability, not to say it's of no practical use whatsoever,
if the window wobbles, or flushes away, dissolves, be drained, or turning to another cube's face,
or just simply switches.
The latter is way faster, uses less resources one may use or miss for more important things,
and cannot become boring,
because it already is boring in the first place 😂

You decoy kids with not being boring by optical jugglery.
Grown-ups learned:
Any optical fiddle-faddle will become boring sooner or later.
Changing to some other optical fiddle-faddle, then,
will not change anything about that in the long term.
But to find
the real interesting, exciting, thrilling stuff lies deeper, will.
So the look only matters as far as it actually improves something useful.

Example I learned it:
The file lists ls produces within a shell are boring to look at.
It's pure text only.
But you get any information, way more information, and way - way - faster than with any filemanager.
And the information (size, name, date, permissions...) is all you really want.

If one knows a file is within a directory,
it's way quicker to find it within searching yourself within a sorted list, even if the list contains hundreds of entries,
than finding the same within a filemanager, especially when set to this absurd tiling view, even with twenty entries,
or to trouble a search function first.
The use of a shell requires learning to use ls, and keep your system in order.
FMs are for the disguise there was no need to keep order,
to say support laziness, and produce chaos, for the cost of being slower.
(how many times you helped some with computer probs, and found their machines are the total mess?)

But of course, who never had a taste of this, only knew the concept of MS Windows, may not understand, what I'm talking about,
but only fear one may take away his crutch 🤓

MS made poeple addicted to crutches:
'Everybody uses them. The are so comfy, and look cool.
Anything else is so boring and grinding.'🦽

unix pros learned:
If you really want to walk, even run,
first thing you have to do is get rid of the damned thing. 🏃‍♀️
:cool:
 
Haha. Thats a good way of putting it.

What I think will be a good compromise is a decent low level library like Xlib -> Wlib to provide most of the abstraction.

And then a decent wayland compositor that allows Wlib / wcb based Window Managers to interface with it.

The kids get wobbly windows and the grown-ups get usability.

Unfortunately at this point Red Hat is focusing on maximum market penetration rather than something actually useful.
I completely agree that the wobbly windows are childish and do nothing to increase usability. For me they are nostalgic. I still remember the first Novell demo video circulated on the internet. It was something new and interesting. A bit like floppy disks in regard to usefulness but I hate to see things go. 😄
 
It's possible I guess. Wayland seems to have been coming real soon now for the the last 10 years at least.
I haven't experienced any issues with Wayland. In fact the only reason I use it is because on another unix like os I was using at the time I was getting an average boost of 11fps on games haha. I'm not sure why that is but it was the case.
 
Then there is this.


Oh dear.
That was in 2020. Since then there has been one more X.org release. In 2022 Gnome were debating dropping X11 support in gtk5. Who can tell the future, "for if a stone is cast, there is no fortelling of where it may land" (Flann Obrien).

Personally I hope X stays in development for many years to come. Don't throw out your old hardware in too much of a hurry. Especially old stinkpads.
 
Then there is this.


Oh dear.
That was in 2020. Since then there has been one more X.org release. In 2022 Gnome were debating dropping X11 support in gtk5. Who can tell the future, "for if a stone is cast, there is no fortelling of where it may land" (Flann Obrien).

Personally I hope X stays in development for many years to come. Don't throw out your old hardware in too much of a hurry.
I kind of want to go back to xfrre86. Such a legit sounding name. I have a massive amount of older machines. I had never considered the question of if they will work with Wayland. I hope Xorg sticks around too.
 
Worst case scenario: xenocara become the default graphical server on BSD's systems.

Edit.: Explanation, xenocara would become the default on BSD's because xorg died and wayland became the default on the Linux world.
I can see that happening.

If anything to ease the maintenance burden on the developers / port maintainers. I am of the opinion that Xorg splitting up into many modules years ago was a mistake.

I can also see it happening on Linux too. If Xorg really does see a lack of development, it is only natural that distros migrate to Xenocara. Or at the very least if development stagnates on both, then there is really no need to keep it split up into tiny fragments anymore.
 
Back
Top