I genuinely think freebsd would be more popular with a purely bsd eye-candy desktop environment

An attractive and focused desktop is a great thing, but if you want popularity you necessarily have to focus on other non-technical aspects, like help desk in your local language. That is not happening.
FreeBSD's Handbook is available in Japanese and German, if you like. KDE has been localized to those same languages. We can help you put it all together, just ask here at the Forums, and we can point you to the Handbook, and even help with connecting the dots. :) It does help to remember, FreeBSD and KDE are separate projects.
 
[sarcasm] Apple's default color scheme is this ugly stainless steel look with rounded corners on buttons! I want the default to be magenta, looking edgier, or else I'm not gonna buy Apple stuff any more! [/sarcasm]
That is not sarcasm. That is what Deutsche Telecom told to Steve. And they wanted the browser to hardwire the home button to their home page. Steve gave them the 132 answer.
 
Historically, Apple did not suffer from the combination of stability + a superb desktop environment.
Yes, but MacOS is also a commercial OS. Which in this case I think is a fairly significant consideration. It is used by people paying to have everything done for them.
 
What you're not considering is that perhaps people are not complete morons, but they would rather just not want to learn how to get X working so they can get on with their life continue doing their compiler, network, or any other thing that is not sysadmin work.
You barely have to learn how to get X working to just type a few simple pkg commands. If they are "too busy" for that with their high flying life, then how do they expect to be able to install development / networking tools you mentioned they might be interested in?

FreeBSD is the wrong tool to choose if you just want a quick rush job.

combination of stability + a superb desktop environment.
Ultimately that is an opinion. My opinion is that I actively choose to use the command line, just to avoid Apple's naff awkward interface. Opinions are effectively worthless but unfortunately computing environments aren't easily something that can be measured.

Guys that want UIs are almost overwhelmingly well catered for with Windows, macOS, Linux, NomadBSD, iPhone, Android, etc, etc. FreeBSD obviously supports them because it is a very capable OS but there is more to computing than all this.
 
Probably repeating someone's thoughts here, but: FreeBSD doesn't need/want to be "popular" as some click-once-to-install desktop OS. If it was, it wouldn't be popular any more for what it is: A general-purpose OS aimed towards (semi-)professional users who need the full flexibility of some modern *nix system.

That said, there's nothing wrong with distributions or forks building something like that on top of FreeBSD. Well, many of them died, so there is something wrong: keeping it alive and having enough contributors to do so, but that's more of an organizational problem.

There's also nothing wrong with developing some "DE" specifically for BSD systems, it might actually be kind of nice, but keep it a separate project. Well, for lumina, see above. You also need the manpower to get somewhere.

Trying to do that as part of the FreeBSD project would just draw manpower from much more important things: Maintaining and developing a well-designed (and general-purpose) base system and maintaining ports to build and run 3rd party software on FreeBSD. There's no justification to do so: Only few FreeBSD users would benefit from that work. Some don't even use FreeBSD on a desktop, and from those who do, most are perfectly happy with all the existing window managers and desktop environments available from ports.

Finally, some thought about "eye-candy": Although "good looks" can improve your user experience, and some things (like carefully applied translucency and drop-shadows, NOT over-doing it) can help with quick orientation on a crowded screen, people tend to really over-do it (someone remembers those weird themes for enlightenment wm?). The consequences are worse UX:
  • Many things are a matter of taste, over-doing it will lead to designs that most people will just hate
  • A lot of "bells and whistles" will draw your attention away from whatever screen content actually matters
So, just be careful with that.
 
I think a conventional desktop is best defined by the hardware support and the applications it runs over anything else. If you use FreeBSD for this, it's because you have contact with this system. There will always be easier, more extended and better supported alternatives. It's a choice.

Describe FreeBSD in three words only, seems like a good comprehension exercise:

 
"My Daily Driver" :)

I agree with alot of what Zirias wrote in #57. DE preference is like car preference, house color, heck even choice of spouse.
"eyecandy" in a DE seems to be defined now by how much visual overload we can put on the screen.
For years my preferred desktop wallpaper is defined as "solid, #333333" because I find it easiest on my eyes and not losing anything in clutter.
"themes" (for lack of a better term) tend towards dark backgrounds with contrasting foreground again, because it's easier on my eyes.
All the spinning twirling exploding visual effects when I close a window, get disabled. I want the window to close/iconize, I don't need to see it black hole into an explosion.

Forks/distributions focused on the desktop user have a lot of good stuff with them, watched a youtube of latest GhostBSD install and amazing how simple it was. Of course I don't like their default choice of MATE DE, but that's just me. What it tells me is some of the effort put into detecting hardware, loading the correct drivers, etc by default are very useful. That is the kind of stuff I would like to see rolled up into a port (new or existing) that would let the handbook simply say "For graphical environment/desktop use, doas pkg install blah && run blah.install and follow the prompts".
 
I was just sharing a simple thought and what a ride reading all of your comments. So here's my conclusion.
Surely freebsd can have the best of both worlds. From my perspective to think only in absolutes is just not right.
Maybe from your angle the BSD desktop doesn't sound smart, but it can be an option for all of those who just want an out of the box experience. No tweaks or headaches. Plain simple eye candy and responsive UI with all the standard features.

Anyway in the end it's just a vague idea but fun to try.
Here's a serious suggestion - you wouldn't need to create yet another DE or WM, just write a theme pack that customizes the main ones being used, then make it as a port. For example, I use primarily XFCE, and sometimes KDE. I almost always replace the little mouse or the "K" in the "start" buttons with a FreeBSD logo. Just a nice little touch that makes this instantly recognizable for not being just another OSS Desktop system. Someone with some more artistic flair and patience for making themes could easily create a theme pack that would give some consistency across the various DEs. Then you just have to advertise it for people to use, promote it, etc.
 
Here's a serious suggestion - you wouldn't need to create yet another DE or WM, just write a theme pack that customizes the main ones being used, then make it as a port. For example, I use primarily XFCE, and sometimes KDE. I almost always replace the little mouse or the "K" in the "start" buttons with a FreeBSD logo. Just a nice little touch that makes this instantly recognizable for not being just another OSS Desktop system. Someone with some more artistic flair and patience for making themes could easily create a theme pack that would give some consistency across the various DEs. Then you just have to advertise it for people to use, promote it, etc.
I'd rather not add yet another port to an already rather unwieldy collection. We have truckloads of language modules as ports (Perl, Ruby, PHP, etc). Those themes should be installable to $HOME, and available through the DE's channels, rather than crowding the Ports at FreeBSD. Just my opinion here. ?
 
When you buy a Lego model, do you expect it to come pre-built in the box? Or is the biggest reason why you bought that Lego kit the joy (and sometimes frustration) of building it yourself? Same with FreeBSD, you get all the building blocks and a big fat manual that tells you how to build a specific model. But nothing's stopping you from using those building blocks and build something creative that isn't in the manual. And that's why we like it.
 
When you buy a Lego model, do you expect it to come pre-built in the box? Or is the biggest reason why you bought that Lego kit the joy (and sometimes frustration) of building it yourself? Same with FreeBSD, you get all the building blocks and a big fat manual that tells you how to build a specific model. But nothing's stopping you from using those building blocks and build something creative that isn't in the manual. And that's why we like it.
Some Lego kits do come pre-built in the box if you shop around. But those are a pain to get out of the box, and the one I linked to is $680 USD. I like to compare FreeBSD to cooking - Just try combining regular flour, yeast, water/milk, eggs, salt and sugar for a good texture. Temperature makes a difference, so do relative amounts, the possibilities are endless with just those raw ingredients. FreeBSD is the farm for just a few high quality ingredients, not a cooking research school.
 
Sometimes around here it is said how Linux is just a kernel, other times it is bloated.
FreeBSD, on the other hand, is "perfectly" in the middle - perhaps not the word folks use, but you'd get that impression reading messages - both a complete system and not bloated...

There is also a lot of "in my day <ramble bumble something about it being better and the kids today don't know what they've got>", which makes me think that what FreeBSD users really want is for everything to stand still for just five minutes and stop pissing about with progressing.

I'm not saying progression/change for progressions/change sake is a good thing (nor that progression and change are the same thing). Far from it. But neither is standing still for standing stills sake.
 
There is also a lot of "in my day <ramble bumble something about it being better and the kids today don't know what they've got>", which makes me think that what FreeBSD users really want is for everything to stand still for just five minutes and stop pissing about with progressing.
Hmm, that sounds familiar. See #3 here:
I'm not saying progression/change for progressions/change sake is a good thing (nor that progression and change are the same thing). Far from it. But neither is standing still for standing stills sake.
Nice straw man you got there.
 
You barely have to learn how to get X working to just type a few simple pkg commands. If they are "too busy" for that with their high flying life, then how do they expect to be able to install development / networking tools you mentioned they might be interested in?

FreeBSD is the wrong tool to choose if you just want a quick rush job.
1. I get X working, but it is never a "few simple pkg commands." Having to memorize, look up, or disassemble to read the model on whatever video card you have in your system in order to pkg install the correct driver when the computer is capable of just telling you via PCI ids is frankly ridiculous.

2. You didn't read my comment. It's not that they're too busy, is that they don't want to, and the computer is completely capable of doing it for them.

FreeBSD is inconvenient and it does not have to be that way. To argue to keep inconvenience because you like it that way is to gatekeep. I'm just saying that a script or installer to do these things would go a long way. People can just choose not to use it if they like it the way it is. THAT is the FreeBSD way.
 
I get X working, but it is never a "few simple pkg commands."
Yes it is.

Having to memorize, look up, or disassemble to read the model on whatever video card you have
No you don't

FreeBSD is inconvenient and it does not have to be that way.
No it's not. I reinstalled everything a few months ago and it took me maybe 20 minutes including all the software packages I use daily. Far better than Linux where it installs too many things I don't want and have to uninstall them, find out that broke something, leaves unused packages behind, corrupts the OS and I have to reinstall all over again, possibly repeating the same thing.
 
in order to pkg install the correct driver when the computer is capable of just telling you via PCI ids is frankly ridiculous.
I believe the only awkward one where you need to do that here is Nvidia because it is a proprietary (and out of kernel) blob. But I am not entirely sure that is any different whether installing the driver for Windows or Linux.
 
Windows detects and downloads the correct driver at least to get a framebuffer working, if not modest acceleration. I don't know what linux does.
In the same way that vesa or scfb is always available from the xorg meta package. In all cases a user will still need to manually install the actual correct driver for power management (suspend / resume, brightness on laptops, etc), wider range of native resolutions and 2D/3D acceleration.
 
  • Like
Reactions: mer
The biggest "user issue" seems to be "what driver do I install". Do I use modeset, xorg driver, nvidia, which version, blah blah blah.

This is where things like GhostBSD and others have their value for "less inclined" users. Hardware detection. Probe what's available, look up in a database, present the user with a choice. The downside is newer hardware and getting the database updated to choose the correct driver. Much like updating pci ids for new hardware.
 
Back
Top