Spreading FreeBSD

[...]
Canonical did a huge pile of work in this area, and it really paid off for them. Install on any box and it comes up with a sane install where just about everything works, without the user having to do any work. Anyone could put the usb stick in, install it, and it would all come up working.
[...]
Ubuntu Linux is not a good example of a properly working OS. Sure, installation usually runs fine, but when you actually work with it (software updates, different non-native packages, etc.) it quickly becomes a mess where trying to fix a single bug confronts you with the whole machinery of “automation” that is underneath Ubuntu's hood. And yes, there are bugs; more than in Windows. No thanks.
 
I know this is a matter of perspective, but in my exp this is one of the most open community I saw. Even the RTFM culture here it is pretty polite.


there is one:
bsdconfig(8)


I suggest that you take a look at This awesome guide on how to simplify the boot process.

I used his tutorial for setting up a desktop years ago when I was just starting. I've never seen this one before. I appreciate you sharing it.


Yeah but I think something better could be done.

Ubuntu Linux is not a good example of a properly working OS. Sure, installation usually runs fine, but when you actually work with it (software updates, different non-native packages, etc.) it quickly becomes a mess where trying to fix a single bug confronts you with the whole machinery of “automation” that is underneath Ubuntu's hood. And yes, there are bugs; more than in Windows. No thanks.
Not in it's entirety but they did some great work to draw attention. I think with a more coherent system like FreeBSD and the "love" they put into Ubuntu it would work a lot better. I find Linux to be like spaghetti code. FreeBSD is much cleaner and more consistent.
 
Ubuntu Linux is not a good example of a properly working OS. Sure, installation usually runs fine, but when you actually work with it (software updates, different non-native packages, etc.) it quickly becomes a mess where trying to fix a single bug confronts you with the whole machinery of “automation” that is underneath Ubuntu's hood. And yes, there are bugs; more than in Windows. No thanks.
Well, I would hope it a bsd version would turn out better than ubuntu. All I'm saying is that what they did in that area bought them mass adoption. Everyone started using it because it was so easy to install. Sure, ubuntu ended up a mess, but it didn't have to be that way.
 
[regarding Ubuntu Linux]
Not in it's entirety but they did some great work to draw attention. I think with a more coherent system like FreeBSD and the "love" they put into Ubuntu it would work a lot better. I find Linux to be like spaghetti code. FreeBSD is much cleaner and more consistent.
Something like MacOS, then? This would be impossible without commercial financial support and leadership. And I, for one, don't like the consumerist nature of MacOS at all.

Maybe have a look at OpenBSD? They include a graphical desktop by default and even provide their custom window manger (CWM).
 
I do hate that we are such a small minority among operating systems yet we bicker and fight about the stupidest things.
Not that people here are not nice. It is hard to come off as humble and not brash.

I think its a FreeBSD forum user trait. The ability to debate a topic. We are great debaters.
 
Well, I would hope it a bsd version would turn out better than ubuntu. All I'm saying is that what they did in that area bought them mass adoption. Everyone started using it because it was so easy to install. Sure, it ended up a mess, but it didn't have to be that way.
I would argue: Yes, it had to be that way, in the long run. When you target consumerist, non-technical users, what do you expect? And btw: Everyone can become a technical user that can make a contribution, if he or she has the will. If that is not the case, what do you expect from those users except being some sort of cash-cows (like with M$ and Apple)?
 
Well, I would hope it a bsd version would turn out better than ubuntu. All I'm saying is that what they did in that area bought them mass adoption. Everyone started using it because it was so easy to install. Sure, it ended up a mess, but it didn't have to be that way.
Almost everyone I know who currently uses Linux started with one of their distros. I know way back in the day I got a free Kubuntu DVD in the mail. Not saying we should mail out DVDs but they definitely had a draw with it "just works" and making it easier to get into.

Something like MacOS, then? This would be impossible without commercial financial support and leadership. And I, for one, don't like the consumerist nature of MacOS at all.

Maybe have a look at OpenBSD? They include a graphical desktop by default and even provide their custom window manger (CWM).
I don't like MacOS either. Especially with what it's turning into as of lately. I've tried all of the BSDs. Keep coming back. FreeBSD is where I started with the BSDs and it feels like home, if that makes any sense.

I would argue: Yes, it had to be that way, in the long run. When you target consumerist, non-technical users, what do you expect? And btw: Everyone can become a technical user, if he or she has the will. If that is not the case, what do you expect from those users except being some sort of cash-cows (like M$ and Apple)?
I honestly think some people would just prefer to use something community driven without having to modify configuration files or all of the other things that technical users want. It wouldn't hurt to have the option available. The systems out there that do this, NomadBSD, GhostBSD, etc are nice but they still have the small things that have been mentioned here. I've gotten people using FreeBSD and they like it but to them those little things made all the difference.
 
I would argue: Yes, it had to be that way, in the long run. When you target consumerist, non-technical users, what do you expect? And btw: Everyone can become a technical user that can make a contribution, if he or she has the will. If that is not the case, what do you expect from those users except being some sort of cash-cows (like with M$ and Apple)?
Hmm, I guess you have a point.
 
[...]
I honestly think some people would just prefer to use something community driven without having to modify configuration files or all of the other things that technical users want. [...]
When I studied at the university, during the mid 2000s , there was this old lady working as a secretary. She was a completely non-technical person who started her career in the typewriter-era. When the faculty started using IT-systems (a Unix mainframe) she was trained for using a terminal.

In the 1990s these terminals were replaced by graphical desktops (the “easy” way). And what did this non-technical lady do? She opened a terminal because those graphical things didn't make sense to her, but the Unix philosophy with text and terminals, etc., did.

That was when it became clear to me that it is all about education and training, not about trying to make things “simple”.
 
When I studied at the university, during the mid 2000s , there was this old lady working as a secretary. The was a completely non-technical person that started her career in the typewriter-era. When the faculty started using IT-systems (a Unix mainframe) she was trained for using a terminal.

In the 1990s these terminals were replaced by graphical desktops (the “easy” way). And what did this non-technical lady do? She opened a terminal because those graphical things didn't make sense to her, but the Unix philosophy with text and terminals, etc., did.

That was when it became clear to me that it is all about education and training, not about trying to make things “simple”.
I'm not disagreeing but I also think that most people don't care to go through the education need to begin using the terminal when they don't have to. It's education, yes, but also comfort and the current mindset of "I want it now." that societies have fallen into.
 
I'm not disagreeing but I also think that most people don't care to go through the education need to begin using the terminal when they don't have to. It's education, yes, but also comfort and the current mindset of "I want it now." that societies have fallen into.
I agree that the “I want it now” is some sort of zeitgeist. And I personally enjoy that FreeBSD is decidedly not that way, because it has empowered me in many ways, most importantly along the lines of “Yeah, I can do it! It's not that hard, not on FreeBSD!”

But, I agree with you that there are people who have other goals or are under pressure (”I need to get this server up till next week, yikes!”).
 
What ways would you all say would be good to spread the reach of FreeBSD to more users? More users could bring with it more developers and hardware manufacturers. This could mean better hardware support and more software aimed for the OS. Here are my thoughts:

1. Friendlier atmosphere. I've noticed some of the developers and administrators tend to have a negative view and response to users. (You know who you are. ;)).
2. Something like a settings screen. Even a terminal based one to start. Instead of having to go through configuration files or commands and reading man pages. Settings in one location. I know when I first started this took a lot of trial and error.
3. Faster boot times. I know waiting on dhclient has slowed my boot times down and attempting to switch it to run in the background doesn't always seem to work.

Before I hear something along the lines of "If only someone was intrepid enough to work on these things.". Yes, I am aware they need someone to work on them. My thoughts are, if there's a list, I could work on what I can and maybe others who want to can work on what they can.
Well, you can start with ports that have no maintainer... but if you want a list, there is one: https://wiki.freebsd.org/IdeasPage

This is probably not the most organized approach to try and do 'Development for FreeBSD'. Just trying to pull ideas out of random places can result in a mess with no clear structure of what comes first, second, and so on.

The way I got even familiar with the FreeBSD development process is by first compiling ports on my systems. Then I set the options I want, started playing with that, and now I'm doing a personal project to try and address the dependency hell by compiling progressively newer KDE packages against an existing base. It's quite a rabbit hole, and it takes a LOT of discipline to make myself get a handle on different components on that process, and yet not lose sight of forest for the trees. But it's my personal project...
 
...


I honestly think some people would just prefer to use something community driven without having to modify configuration files or all of the other things that technical users want. It wouldn't hurt to have the option available. The systems out there that do this, NomadBSD, GhostBSD, etc are nice but they still have the small things that have been mentioned here. I've gotten people using FreeBSD and they like it but to them those little things made all the difference.

I also think improving FreeBSD experience on desktops/mobile systems is a worthy goal (I am using FreeBSD every day on my work & personal notebook for many years now).

Food for thought:

Maybe it would be an idea to merge back the great work NomadBSD is doing into the FreeBSD installer as an „opinionated desktop install“?

Practically everything that is setting NomadBSD apart from stock FreeBSD afaik is implemented by installing a set of packages, some configuration and a custom kernel configuration.

There could be more preconfigured setups like this one?
 
Maybe it would be an idea to merge back the great work NomadBSD is doing into the FreeBSD installer as an „opinionated desktop install“?
I don't really see the point. These things are just perfect for 3rd-party distributions :-/

I mean, FreeBSD itself offers anything you need to configure anything you like yourself, one of the points is that it's not "opinionated". And I'd argue those who use "vanilla" FreeBSD want it just that way, at least that's the case for me (using it on both servers and desktops).

What I do agree with is some things could be technically improved for desktops/laptops. IMHO especially the wifi support could really need some love...
 
I don't really see the point. These things are just perfect for 3rd-party distributions :-/

I mean, FreeBSD itself offers anything you need to configure anything you like yourself, one of the points is that it's not "opinionated". And I'd argue those who use "vanilla" FreeBSD want it just that way, at least that's the case for me (using it on both servers and desktops).

What I do agree with is some things could be technically improved for desktops/laptops. IMHO especially the wifi support could really need some love...
I think there are quite a number of people that don't know about NomadBSD but who have heard of FreeBSD.

For many of them installing and configuring FreeBSD may be too complicated.

In my opinion, it would be useful to make installing and using FreeBSD on your desktop/notebook as easy as possible: Some of the users FreeBSD might gain that way will stick around, become more and more experienced, spread the word and they even may become contributors sooner or later.

P.S. (added in edit): Yes, I also want the installer not to be opinionated. But the installer could support both: Individual install/configuration as it is now for those who want it that way and additionally predefined setups for others.
 
In my opinion, it would be useful to make installing and using FreeBSD on your desktop/notebook as easy as possible: Some of the users FreeBSD might gain that way will stick around, become more and more experienced, spread the word and they even may become contributors sooner or later.
I don't agree, and don't think that's a realistic hope for the following reasons:
  • There are different definitions of "easy". FreeBSD installation is easy, as in: there's almost no complexity in the process. It involves several steps, a) because ports are kept strictly separate from the base system and b) because the design principles (POLA) don't allow any "hidden magic" to happen. Therefore, doing it the first time, you will certainly need the handbook. Still, this is "easy", it's as straight-forward and transparent as it can possibly get.
  • For a very different definition of "easy" that involves clickable GUIs and "magic" like automated configurations, you'd probably expect everything up to a fully configured desktop system handled by one single installer tool. That's already problematic, as you'd have to tie together base and ports. Doesn't mean it would be impossible, but it would introduce a lot of unnecessary complexity, another thing that's pretty much against the FreeBSD ideas.
  • Even then, users would probably just run into a wall once they need to "maintain" their system, as they never had the transparency during install, so they have no idea how it really works. So, the next idea would be to even extend the magic, providing "management tools", further spoiling what IMHO makes FreeBSD great.
  • Once you've done all that, maybe you'll have a few more desktop users. At the same time, you have alienated experienced administrators. And talking about developers/contributors, you'll have more or less the same people, under the optimistic assumption that these people are willing to dig through all the "magic" layers to gain enough understanding of the system to actually work on it.
 
... users would probably just run into a wall once they need to "maintain" their system, as they never had the transparency during install, so they have no idea how it really works.
Excellent point! This is what to expect when attracting users who are not willing/being able to learn before doing.
 
Make it more attractive to use as a desktop. Servers are fine, but I don't want one on my desk.

Above all, improve the auto-configuration to target hardware. When I install it on a box, I want it to "just work"(tm). I don't want to have to manually port it to my hardware.

For example, I don't want to to have to fudge around reading forums and manpages just to get the speakers and headphones in a laptop to switch over when I plug the phones in and out. Slackware had that working that out of the box back in 1995, freebsd still can't do it now. Even today, I end up trawling through peoples sites (cf vermaden!), lots of tweaks in loader.conf, sysctl.conf, rc.conf, device.hints, etc.

Canonical did a huge pile of work in this area, and it really paid off for them. Install on any box and it comes up with a sane install where just about everything works, without the user having to do any work. Anyone could put the usb stick in, install it, and it would all come up working.
Xorg have also made huge strides in auto-configuring X. No-one bothers about having to work out modelines any more. Imagine what a drag it would be if you had to go back to the old 90's way of configuring X.

IMHO that characteristic is an essential prerequite to make FreeBSD compelling. It's a lot of work, I know.
absolutely agree, I use Slackware too, and FreeBSD having everything working as not everyone uses it for a server only. headphones, bluetooth, usb hot plug, I don't even know if my USB ports work because I don't use them on freebsd, had to burn a stick on my laptop and just deiced to jump out of Freebsd because I don't even know if they work on 13.1 had to do all kinds of editing to get them to work on 12.x so I rebooted into Slackware instead because I already know what to do on Linux

the installer should install a GUI with an option os a wm/dt Slackware has plenty to pick from on install and afterwords already there just waiting to be used.

wifi, bluetooth, usb, sound - headphone/speakers they have not fixed this yet? what have they been working on?
 
what have they been working on?
On more important, fundamental aspects so projects like GhostBSD have a solid base to build a desktop experience on or so you have a solid base to build your own desktop experience based on your requirements.

If you're looking for an out-of-the-box FreeBSD desktop experience (i.e. you're not willing to do it yourself), choose something like GhostBSD. That is exactly what it does.
 
Ubuntu Linux is not a good example of a properly working OS. Sure, installation usually runs fine, but when you actually work with it (software updates, different non-native packages, etc.) it quickly becomes a mess where trying to fix a single bug confronts you with the whole machinery of “automation” that is underneath Ubuntu's hood. And yes, there are bugs; more than in Windows. No thanks.
Ubunutu sucks ? I installed there studio version and they're using snap and apt to install apts and I'm like wtf you have to update them indeviually I keep getting messages on that they're a want-a-be Windows version of Linux - the only reason I have it is because the table I have only supports Ubuntu Linux as if that is a good thing. there install is not open source so I cannot even hate it to make it work on something else. That is the thing about them is they have that foot in the door where outside software venders only support Ubuntu Linux and not Linux itself. which sucks because that is the last linux distro I want installed on my laptop.
 
Let's not lose sight of the fact that you can run almost all Linux software on FreeBSD through the linuxulator or a VM.
This is a very personal opinion but I have that thing going on where (at least for private/personal stuff): If I can't get an open-source software to work natively under FreeBSD I don't want to use it anyway as it is either poorly designed/written or it's relying on Linuxism aspects which tend to change more frequently than the British PM.
 
absolutely agree, I use Slackware too, and FreeBSD having everything working as not everyone uses it for a server only. headphones, bluetooth, usb hot plug, I don't even know if my USB ports work because I don't use them on freebsd, had to burn a stick on my laptop and just deiced to jump out of Freebsd because I don't even know if they work on 13.1 had to do all kinds of editing to get them to work on 12.x so I rebooted into Slackware instead because I already know what to do on Linux

the installer should install a GUI with an option os a wm/dt Slackware has plenty to pick from on install and afterwords already there just waiting to be used.

wifi, bluetooth, usb, sound - headphone/speakers they have not fixed this yet? what have they been working on?
If I may recommend something to your concerns on audio, wifi and Bluetooth. My suggestion would be to program something that works for you and license it so that others can contribute to your project. If someone else finds it useful then they can use and modify it to make it work how they would like. Waiting on others to do the work doesn't always satisfy what your needs are. That is my suggestion regard to those issues. I'm sure it won't be a very fast solution and you may have people asking what you're working on if it's not to their liking. But you can always recommend they do the same. ?
 
If I may recommend something to your concerns on audio, wifi and Bluetooth. My suggestion would be to program something that works for you and license it so that others can contribute to your project. If someone else finds it useful then they can use and modify it to make it work how they would like. Waiting on others to do the work doesn't always satisfy what your needs are. That is my suggestion regard to those issues. I'm sure it won't be a very fast solution and you may have people asking what you're working on if it's not to their liking. But you can always recommend they do the same. ?
the same can be said about you and I am a novelty user not a developer that knows the in and out of freeBSD
 
Back
Top