FreeBSD system and its share of worldwide use - July 2024

I'm now listening to the 2022 fireside chat

Around 00:58 on the timeline, whilst talking about ARM, he described his passion for Raspberry Pi, wanting to run FreeBSD with a full desktop. Emphatically a desktop solution "because that's where people are going to start". Raspbian was described as truly awesome. "Don't diss the $20 computer …".
 
The only large cloud company that uses any FreeBSD is Netflix, and rumor has it (the Netflix HQ is in the same town I live in, and we know people) that they have been replaced by Linux.
By that do you mean openConnect? That's a standalone system sent out to all the cable and internet providers and talked about widely so I'm surprised quite a bit if that's what they're talking about.
 
It's not very often one would care about USB mass storage on servers but earlier this week we had a unique situation where having UASP would have been extremely helpful. I'd say that is a much higher priority than wireless or a default desktop (which hopefully there will never be).
 
You missed the even more fundamental question: Do we want to raise those numbers? Does it benefit us if many other people use the same OS as we do?
There's no doubt about it. If it had 5% market share, they wouldn't have dropped ARM and i386. Man pages and the handbook would be more recent. More forum questions would be solved sooner...
 
More forum questions would be solved sooner...
I don't think so.
Compare to the Linux fora I know, signal/noise is much better here. You might get an answer sooner there, but it would likely be of lower quality, or even outright wrong. Also, finding a question you have already answered also needs the time/date as a key there since the landscape keeps shifting.
 
It is very difficult to advance FBSD as a desktop OS if most of the devs and/or maintainers are using Windows or MacOS.
I had run FreeBSD with some GUIs in i86, later short in aarch64 version in VMs.
After a certain FreeBSD update, GUIs no longer work. I've been through pretty much all GNOME based desktops. Old instructions no longer work.
Nobody could help me in various forums and on Reddit.
On i86 I used Virtual Box, on M1 Mac Mini now UTM (QEMU), VMware Fusion.
It somehow seems to fail because of choosing the right monitor - or do modern GUIs now require Wayland instead of X?

PS: I know this is not the right thread for this!
 
I don't think so.
Compare to the Linux fora I know, signal/noise is much better here. You might get an answer sooner there, but it would likely be of lower quality, or even outright wrong. Also, finding a question you have already answered also needs the time/date as a key there since the landscape keeps shifting.
Sounds like no graphics at all...
How about X.org without GUI and no xorg.conf? Then try to start xterm from a ctrl-alt-f? system terminal or remote ssh.
 
If you add?
  • iOS
  • Android
Shouldn't affect those stats. They aren't desktop operating systems.

There's no doubt about it. If it had 5% market share, they wouldn't have dropped ARM and i386. Man pages and the handbook would be more recent. More forum questions would be solved sooner...
Linux, Windows and macOS dropped ARM and i386 years ago. Once might suggest that the larger the market share, the quicker you run out of manpower to maintain niche platforms.

man-pages are far superior to Linux and macOS.

Check out the manpage for macOS powerd(1). It is laughable.
 

Apple​

macOS

… users are completely not involved in developing. They can't even file bug reports in the usual sense of the word (where you get feedback about the bug), …

You're misinformed …

Feedback

Trivia: years before public betas became a thing, I was AppleSeed member 405 (that's the 405 in my avatar here); and part of a much smaller group (membership unknown) that tested builds of Feedback Assistant before the Assistant was made available to other members of AppleSeed.

From Apple's site:

"… check for updates or resolutions on feedback you’ve filed. We may also email you asking for additional information or files to investigate the issue. … we can’t reply to every submission, we review each one and monitor the amount of feedback submitted so we can better understand the scope of the problem. …"

A random pick, something that I reported ten years ago: "Resolution:Investigation complete - Works as currently designed".
 
Shouldn't affect those stats. They aren't desktop operating systems.


Linux, Windows and macOS dropped ARM and i386 years ago. Once might suggest that the larger the market share, the quicker you run out of manpower to maintain niche platforms.

man-pages are far superior to Linux and macOS.

Check out the manpage for macOS powerd(1). It is laughable.
Just guessing: they don't like CLI operations and have their own graphical power settings program. Ofcourse requiring Xcode if you want to use it inside something else.
 
Sounds like no graphics at all...
How about X.org without GUI and no xorg.conf? Then try to start xterm from a ctrl-alt-f? system terminal or remote ssh.
No. Sounds like "leave default as simple as possible". You can install a desktop trough ports, you can submit a port that does some preconfig or has a graphical setup. And you can build your own iso starting your tool at startup and run bsdconfig from your tool - but do not burden the system with keeping your port running.

There is already a port for desktop installation, I have heard good things about it.
 
No. Sounds like "leave default as simple as possible". You can install a desktop trough ports, you can submit a port that does some preconfig or has a graphical setup. And you can build your own iso starting your tool at startup and run bsdconfig from your tool - but do not burden the system with keeping your port running.

There is already a port for desktop installation, I have heard good things about it.
Having my own FreeBSD 'distro' that's commercially used in cash register systems. This is my exclusion approach. Eliminate possible problems by temporarily taking out the system parts that could cause it. The whole system is only a stack of different components of which only few are actually required to keep the system running.
My idea is that if you have a problem with a full desktop meta-port not working, first check if X is working without 1. It should.
 
… none of its derivatives have shone as an attractive desktop environment system to gain users in the world …

I found PC-BSD shiny and attractive; GhostBSD shines in various ways; and so on.

The goal is to maintain continuity of the project.

It's not …

There really isn't a specific "goal" per se. …

… there is a goal (it's no longer what's in the FreeBSD Handbook).

From polls a few months ago, I gained an impression that the vast majority of people:
  • simply don't care about the goal that's expressed by the Project
  • will treat any suggested goal as a Project goal, if the suggestion has a feel-good factor.
This willingness to believe anything that sounds reasonable, without fact-checking, is hardly surprising for a project that's more than thirty years old. A lot of talk :)
 
If it had 5% market share, they wouldn't have dropped ARM and i386. Man pages and the handbook would be more recent. More forum questions would be solved sooner...
I wouldn't bet on this.

You see those 7.something% for Linux are shared by all Linux distris.
If FreeBSD gets to 5% it would be one OS/'distri', only, having those 5%.

That means it would be very interesting to become some form of business.
And that means:
Goodbye foundation, hello management, CEOs etc.
Goodbye to almost any free spirit, creative, experimenting,... just focus on where the (most) money is.
If this would be servers, goodbye desktop, and embedded.
If this would be desktop, goodbye ... - you get the picture.

Will a business improve forums?
I think not.
The live source, and the power of this forum comes from its vital community of experts, and enthusiasts, pros, and hobbyists (an amateur means 'loves it'), greybeards, and newbs, young, and old.
Many know each other through this forum for many years, some even in real life e.g. from Cons.
People connecting people because they care, want to share their interest, their knowledge, their passion.
Spend (lots of) time to help, invite people, open the gates to FreeBSD, and discuss topics with others.

Imagine this will get into the hands of some business.

See what Wikipedia became.
The foundation hoards billions of dollars, starting every year again in december their opera of the whining beggar, while the authors of the articles, the producers of what Wikipedia actually is, work for free.
It's a core flaw in human thinking:
More money does not makes you feel more secure.
In the contrary it makes you more insecure, because you may lose it. All you receive by lots of money is greed for get even more money.

A forum like that then will become some support outpost, only, mostly done by 'trained support drones', people do it as long as an underpaid job until they get something better. Who know exactly three answers:
"Read the FAQ!"
"Did you updated?"
"Please contact support." (the real one costs, of course)

Or some enthusiasts may stay, continue support for free.
While 'the company' advertices with their good support, and the CEO and CFO congratulate each other for having support without paying a penny for it.

Otherwise the forums is nothing more as for statistically evaluate count customers fancy ideas. Tell the managment the most named three ones as top customer wishes.
And the management will tell the developers to leave the rest, focus on customer demands - in tight schedules, of course:
'we will see to eleminate flaws, and fix bugs later, when we will have some spare time.' (yeah, yeah, naturally. There is no such thing as 'spare time' in business. That's called 'too much staff')
Not listen to them, what they think is needed most, or does best.

...so get completely out of developer's hands.

What seems to be a neat idea has a crucial flaw:
Most customers are idiots.
They do not know what's best for them.
They don't think.
They dream. They wish. They hope. They believe. They want. They demand.
But they don't think.
For selling that is best.
Until you also need to sell your company; in time as long as you can find some moron you mock on paying money for a company drained below zero.

But to get the best:
Don't listen to the majority of customers.
Don't listen to salesguys.
Listen to experts.
And listen to yourself.
Stay loyal to your founding ideals!



Besides the fact the HB ain't that bad
a business will not improve that either.

Do show me a HB on an operating system
that is at least as good as the official FreeBSD one.

However I agree the HB could use some improvement.
A HB always could need some improvemt - except it's for some steam engine development stopped on seventy years ago.

My idea on how to improve the HB would be
to make it as Wikipedia does it - open it up for more authors.
(A lot of potential ones can be found within this forums [not me, of course - too few expertise, too long articles :cool:])

The core idea of Wikipedia is that anybody can write, and edit articles.

The question to be solved would be:
Who gets writing access?
Besides developers maybe automatically anyway e.g. established members of this forum was my idea.

And one need some moderators - of course also in Wikipedia every change needs to be verified first.

I know there already is a FreeBSD Wiki
But besides I have the feeling it could use a bit more help, to get more articles, and some updated,
I like the idea for an open the Handbook with more authors to be considered for a moment.
Because I believe - I don't know, I believe - doing the HB right is like all documentation in software:
Underestimated,
as in the needed effort,
such as in its meaning, importance, and value for the project.

When I recall right Cath O'Deray already is or was participated in that.
Maybe he can tell us if I'm wrong,
or this could be an idea to be considered more closely.
 
I wouldn't bet on this.
I'm very 'opinionated' about this: the only reason FreeBSD (and linux as well) have a inferior market share compared to the whole set of available operating systems is the lack of non-proprietary, yet full support of hardware-accellerated graphics. Not having the privilege and/or information to operate your graphics hardware on metal level is a really big deal. The most important 1, regarding OSS computing. Graphics is your visual output...
 
FreeBSD Wiki
But besides I have the feeling it could use a bit more help, to get more articles, and some updated,

1723312808198.pngClick About this Wiki.

I like the idea for an open the Handbook with more authors to be considered …

Certainly, there are discussions about how to grow and improve participation. None of these discussions use the email list for the Documentation Project, and (unless I'm missing something) I don't see the same discussions here in The FreeBSD Forums.

The most recent FreeBSD Journal includes Submitting GitHub Pull Requests to FreeBSD. Note:

"… This article focuses on commits to the base system, not the documentation or ports trees. These teams are still revising the details for these repositories. …"
 
That's what I tried to sketch.
I just wanted to point out, that FreeBSD doesn't neccassarily become better just by gaining 5% of the market.
There is a saying in Germany : "Shit tastes great. Trillions of flies can't be wrong".

It's horses for courses, tools for the job. And I am proud not to use the tools everyone else has.
 
It's horses for courses
There are exactly two ways of doing business right:

1. Selling crap. Focus on masses, real large numbers, and lowest selling prices; for that keep quality, and all costs as low as possible, which means barely close to refunds, and customers turning their backs on you eat too much of your revenue.
If you master this, you may become rich.
That's why this way is so alluring.
Downside: It's not a walk on a tightrope, it's a battle on a raze blade.
Here is most, and hardest competition.
And because you have to continue reducing costs, the rope becomes thinner - the blade even sharper.
If you don't master it - and most don't, you will hit rock bottom before you even realize you're falling.
And you'll be lucky if you may ever get up again.

2. Doing top quality. Become the best.
Downside: You will not become rich. Money is, where the masses are. And the masses buy crap.
But instead of monetary wealth you may enjoy your work, gain satisfaction by satisfied customers.
You have a solid foundation of convinced customers you can rely on, therefor no much cost for advertising.
Way not that hard competition (if even), so fewer stress, more time for considerations, which are needed for top quality. Being part of something special, something exclusive.
And having good chances your business/project/company will last long,
while those crap merchants always need to fear to be blown away within a few days, no matter how big they are.

And there are only those two ways.
Trying to combine both ways, 'find some compromise between' is guaranteed to fail.
Because you get all downsides of both ways, but never reach a goal, since you don't go for one, but fluctuate between them.

Or, to put it into other words, and back to more context of this thread:
If you just gain more popularity to sell more, your product will suffer. 'cause you're trying to form some top quality business into masses-crap market, crossing the uncertain in-between way to start the tightrope walk.

Otherwise you may trust on the self-advertisement which comes by the popularity of a high quality product automatically.
This will also attract new customers. Fewer. You grow slower.
That gives you time to do things more considered, far-sighted, and gain some more solid values you can rely on.

Of course nobody admits to sell and buy crap.
Almost everybody prints 'high quality' on their packages.
Except those who actually produce it.
 
Since Ventura, macOS users are required to activate online via Apple's DRM servers in order to reinstall the OS

If activation locked was enabled prior to reinstallation, yes. I was able to reinstall Ventura on my OpenCore patched 2015 rMBP (which now runs Sonoma 😁) without any needing to ping Apples Servers. Your MacFu is lacking.
 
I think the trade off for using FreeBSD, is we test, improve and teach others what we've learned when we can, in return for using this free operating system. We don't have to, but that's if we want it to improve, by as little as we're each capable of contributing to that. This is in reference to quality, which the base system does have.

There may be a lack in drivers and abilities, but it has halfway caught up within the last decade. There's also a need for better standardization and a need for improvements of infrastructure of ports. Or the ports tree could be forked, and used on top of FreeBSD, if there were an organization capable and had a need for that.

It's also ok to expand the userbase, as long as FreeBSD principles are guaranteed. Then, for those who want FreeBSD to be Linux or something else like that, there's also the option of forking or providing a distribution which uses FreeBSD as its base.


On another note, 32bit hardware architecture needs to be left behind, except for on 32bit SBC's (ARM, Raspberry). This doesn't include using 32bit layer or emulation for using older software on modern computers, which will always be needed. There's not even an appropriate operating system for 32bit SBC's. There needs to be a hybrid NetBSD Minix operating system with parts from Public Domain OS (PDOS) for a direct DOS and 32bit Windows interpreter layer to be used from the Unix-like command line, and parts from Haiku and various BSD operating systems for a command line and/or menu driven OS for 32bit SBC's. It's pointless to put together old computers with legacy physical hardware components to use 32bit; that use needs to shift to the domain of 32bit SBC's.
 
Back
Top