The PERFECT Laptop for FreeBSD

An alternative to being picky about the laptop supporting everything you need, is just sticking with the default OS that comes installed (Wind0ze, or MacOs), setup a FreeBSD virtual machine and use SSH in tandem with X-forwarding to run graphical applications.

Here is an X-server for windows: Xming

On MacOS there is XQuartz.

This works only with graphical applications that are sane, though …
 
I think, the only way to summarize this thread is, if person wants to use an O/S without a little massage or tweaking one should use the O/S the manufacturer installs on the laptop. They design their laptops for use with the O/S they've designed their laptop for. This includes writing drivers to work around bugs and warts in their ACPI implementation. This is the only guarantee. Any other O/S, regardless whether it's a BSD, Linux, or some other O/S may be incomplete.

BTW, while migrating my FreeBSD on my old Acer to my new HP 840 I also tried to migrate the Windows 10 from the old Acer to the new HP. FreeBSD, even with the need to tweak here or there, works fine. Migrating Windows, not so good. The partitions copied over. And it booted. But, it blue screened 15 minutes after it booted and failed to boot after that. The lesson here is Windows installed on vendor A's computer is not guaranteed to run on vendor B's computer. But Windows on vendor A's computer didn't require tweaking. And to get Windows working on vendor B's computer one needs to install Windows afresh. No tweaking required. For FreeBSD and Linux the same image is generic enough that with a little bit of tweaking it will run just fine.

Every UNIX system I have ever worked on and for that matter IBM mainframes before that too (I've been doing this stuff for over 50 years) required some kind of customization.

Of course you can always hire someone to tweak the O/S for you. I know a lot of people who even do that for Windows.
 
lcsavb said:
For example, after finding a laptop where all the drivers work correctly, what about reaching out to the company and offering: we'll make this model officially supported by FreeBSD?
Click to expand...

Whoa there buddy! Who’s going to write drivers for all those little chips? hope and dreams?

Any OEM isn’t going to waste resources on a platform with little to non-existent market share.
 
  • Like
Reactions: cy@
It would be commercial business like having particular hardware sponsored. 1 problem is that most if not all Microsoft OEM computers are actively working against a uniform OSS solution. I know HP workstations that use some hidden state and refuse to boot a non-Windows system and freeze, but only 1 in 30 times. Still unrreliable for any professional application.
 
cy@ said:
if person wants to use an O/S without a little massage or tweaking one should use the O/S the manufacturer installs on the laptop
Click to expand...
Well, a pre-built kernel, an installer and automated updates can also be considered a massage, don't you agree?
One can learn a lot by doing those things by himself.
 
Holger said:
is just sticking with the default OS that comes installed
Click to expand...
I like to disagree.
To me there can be laptops to buy without anything installed.
I personally would like to buy just the hardware, and install the OS myself - if there were any!
If I configure my desktop PC myself it also comes blank.
The top main point I've chosen my laptop I bought was
'no OS installed.'
You don't get any. Only with some Linux installed. Okay. I don't bother. I just wiped the nvd clean, and installed FreeBSD.

Almost all have Windows installed.
To me that's not to buy.
I simply do not buy a license I neither want, nor use, but just formatting the disk (and having this blot of a Windows sticker on my machine *shiver*.)
 
%pupil -sceptical said:
In theory: Of course.
In reality depends on the price, the configuration, and other points, like to estimate pretty accurately the minimum number of users will buy.
Click to expand...
When you consider that people buy laptops that don't work 100% with FreeBSD without some significant fiddling, and some people here run Macs with FreeBSD, I'm pretty sure that my hypothesis is pretty solid.
 
lcsavb said:
a guide that explains how to get all the drivers working
Click to expand...
It's not only people just don't know how to set up the 'drivers' (While an additional guide on that wouldn't hurt.)
Often enough there simply aren't any 'drivers' for FreeBSD.

How you're going to solve that by a guide?
Explain how to modify the source to work under FreeBSD? (While another guide on that wouldn't hurt either.)
Programmers lack the source code, or at least the addresses and the specs of a certain hardware, but not the knowledge how to do it.

If there is open source available, often there will be a Linux driver soon.
If there is a Linux driver chances are good FreeBSD will be supported also.
Sometimes it's vice versa.
But anyway one needs the source code.

Otherwise if there is no 'driver' and no source available, then what?
...just looking for hardware that is supported... We've been that road just within in this thread here at least already twice:
The list kpedersen and I already mentioned.

I also don't understand why many hardware manufacturers do not publish their driver's source code if they could sell more this way.
Maybe it's some kind of a conspiracy with Microsoft, maybe they are ashamed of their cheap code, maybe they fear some hackers may misuse it, maybe... - I do not know.
But many simply don't provide drivers for FreeBSD/Linux, not even open their source, nor publish at least the detailed hardware specs.

drhowarddrfine said:
my hypothesis is pretty solid.
Click to expand...
I didn't contradict.
It's not a hypothesis, since there are people will buy a FreeBSD laptop, including me.
That's not my point.

All I was saying is that people will buy FreeBSD laptops is not enough for a business to start.

My point is one needs to know pretty solid how many people will buy one for sure. And it has to be more than the minimum quantum a manufacturer considers to bring out a model. Plus a manufacturer needs to know what config the majority wishes. And the price they are willing to pay.

Provide those numbers, and if they suit, the first FreeBSD laptop will be available within half a year or less.

The point is:
How to figure out this numbers?
Get the config, and price willing to pay is easy. A thread containing a poll here might do it.
Get a reliable number of buyers is the crucial point.
How to get this?
This is the crucial core question on every question related to FreeBSD specific hardware.

Think about how to get this number. And get it.
Then, and only then there will be a FreeBSD laptop (and other hardware, like FreeBSD's own ARM boards.)

Otherwise I see only two ways:

1. Starting something like an altruistic non-profit project sponsored by some kind of a patron,
simply bring out an own one, taking a very high risk on losing money, because of
a) the number of buyers was overestimated too optimistic,
or b) there are way more than enough buyers, but commercial competitors join the train, and kick you out of business.

or 2. The way there already is:
Check the market which available laptops (hardware) may work how far with FreeBSD,
collect those info in a database and provide it to the public,
hence the list mentioned already twice here.
And simply live with the fact that almost all laptops cannot be 100% supported by FreeBSD.

Anything else is just throwing the cap over the mill. Day dreaming.
 
The likely reason for FreeBSD's limited popularity is the lack of marketing during the 1990s and later. In 1995, when I started using the FreeBSD, it was a fully functional and stable system, while Linux was far from it (complete crap). Both were freely available to users. However, Linux seemed to have invested in marketing, which FreeBSD lacked, leading to a larger user base for Linux despite its inferiority at the time. There might be other factors, but the absence of promotion was likely significant. For some reason, young and fresh sysadmins started to deploy Linux, rather than FBSD (why?). Larger user base, of course, motivates hardware manufacturers to write drivers. Therefore, IMHO, I believe that FreeBSD needs a larger user base, even if there is just a very small faction of these users who are actually contributing to the code.
 
Argentum said:
The likely reason for FreeBSD's
Click to expand...
I experienced the same.

In the mid 90's everybody was talking Linux. Nobody (I knew) ever heard of FreeBSD.
Why? I don't know.

I dimly remember to got me some 4pointanything(?) FreeBSD via ftp once, tried it; but gave up (too) qickly, because I thought it was just another messy Linux distro.

I wouldn't say 'invested' in marketing. This came later, when Red Hat, SUSE, etc. started using Linux on professional tasks.
Which also could had been done with FreeBSD, if not better.
In the 90's I experienced Linux simply as a gruesome mess, not really usable as a daily driver, but to tinker and develop on it, only, if not caring if it works reliable.
(It did not really changed, since it still is a mess, even it became quite usable...partially ? )

Maybe they simply got the bigger community faster.
Which to me could be explained by FreeBSD is more strictly organized, what code is being implemented, while within Linux (world, not the Kernel) anybody can do freely anything as long as its fun, not caring if anything is finished. Hence the mess.

Also maybe the name was deciding.
Linux has two syllabies. FreeBSD are four. From a marketing point of view everything more than two syllablies is too long to become a successful name. If they named it e.g. UX it may had better chances to became more popular.
I don't know. But the name has its origin, its reason, and was chosen right, not for marketing.
Now changing it just for marketing purposes made things worse, since it'd cause more confusion as it will gain more popularity. (e.g. twitter now is named X - whotf cares?! [besides I wonder why X.org didn't object. The other way around Elon had sued X.org for sure. But that's another topic.])

However,
I agree FreeBSD could use a wider user base. (Who objects?)
And I also see to have more hardware supported was a way to gain it.

But this leads to the 'hen and egg' problem.

First thing you always need before you can go anywhere is to know where you are.
 
%pupil -sceptical said:
In the mid 90's everybody was talking Linux. Nobody (I knew) ever heard of FreeBSD.
Click to expand...
Before that, I had an access to the BSD (not Free) machine (Motorola 68010 based) at my workplace. We had also a real System V UNIX machine in operation. So, basically I was happily surprised when I did hear about the FreeBSD. It was essentially the same system, only running on an affordable PC hardware. But, yes, for some reasons, everybody started to talk about Linux at that time.
%pupil -sceptical said:
Why? I don't know.
Click to expand...
I have wondered almost 30 years. FreeBSD is free and has always been better (so far). Perhaps now, when FreeBSD cannot keep pace with drivers, especially to the ever growing zoo of cheap chips, this is changing.

%pupil -sceptical said:
I dimly remember to got me some 4pointanything(?) FreeBSD via ftp once, tried it; but gave up (too) qickly, because I thought it was just another messy Linux distro.
Click to expand...
At that time I was setting up a web server and actually tried Linux first. Dropped that in a week and replaced with FreeBSD. Linux did not even have a boot loader at that time. There was a DOS partition for booting and a script to start Linux...

%pupil -sceptical said:
I wouldn't say 'invested' in marketing. This came later, when Red Hat, SUSE, etc. started using Linux on professional tasks.
Which also could had been done with FreeBSD, if not better.
Click to expand...
Agree. By investing I did mean to put in some effort or time. Linux fans probably made a loud noise. FreeBSD was mostly mature already and nobody rang any bells.

%pupil -sceptical said:
In the 90's I experienced Linux simply as a gruesome mess, not really usable as a daily driver, but to tinker and develop on it, only, if not caring if it works reliable.
(It did not really changed, since it still is a mess, even it became quite usable...partially ? )
Click to expand...
Agree. And the mystery remains, why does it have such a large user-base?
%pupil -sceptical said:
Maybe they simply got the bigger community faster.
Which to me could be explained by FreeBSD is more strictly organized, what code is being implemented, while within Linux (world, not the Kernel) anybody can do freely anything as long as its fun, not caring if anything is finished. Hence the mess.
Click to expand...
Somehow they gained a momentum very fast.
%pupil -sceptical said:
Also maybe the name was deciding.
Linux has two syllabies. FreeBSD are four. From a marketing point of view everything more than two syllablies is too long to become a successful name. If they named it e.g. UX it may had better chances to became more popular.
I don't know. But the name has its origin, its reason, and was chosen right, not for marketing.
Now changing it just for marketing purposes made things worse, since it'd cause more confusion as it will gain more popularity. (e.g. twitter now is named X - whotf cares?! [besides I wonder why X.org didn't object. The other way around Elon had sued X.org for sure. But that's another topic.])
Click to expand...
That has always been my another hypothesis. UX refers to UNIX and and many people knew this name. Few knew BSD. Another thing is that people may be suspicious when something is free. And even if it is free, they may not like when this is so explicit.

%pupil -sceptical said:
However,
I agree FreeBSD could use a wider user base. (Who objects?)
And I also see to have more hardware supported was a way to gain it.
Click to expand...
Back in old times, FreeBSD had support for some top notch hardware. For example I had a DPT Raid controller in my system. That means the development of drivers was the initiative of the HW manufacturer.
 
While marketing is key, FreeBSD wasn't available due to the lawsuit going on back then. As Linux Torvalds stated: "If BSD was available, I never would have started working on Linux." (paraphrased from memory)
 
lcsavb said:
Well, a pre-built kernel, an installer and automated updates can also be considered a massage, don't you agree?
Click to expand...
Uhhh, no.

lcsavb said:
One can learn a lot by doing those things by himself.
Click to expand...
This we can agree on.

If anyone wants to learn more FreeBSD. Install -STABLE over -RELEASE. Do some buildworld/installworld. Build and install some ports. You never know. You might like it. ;)
 
What's "the" perfect laptop for any OS?

People have different requirements in terms of cost, storage, processing power, graphics (GPU, number of screens supported), audio, security, upgradability, etc., etc. depending on their use cases (light users, power users, video editors, audiophiles, developers, gamers etc.)
 
%pupil -sceptical said:
It's not only people just don't know how to set up the 'drivers' (While an additional guide on that wouldn't hurt.)
Often enough there simply aren't any 'drivers' for FreeBSD.
Click to expand...
Well, that's why I suggest picking newer hardware with mostly already functioning drivers. Someone reported in the Laptop compatibility guide that there is a Framework laptop in which everything works. That could be a good starting point.
It may be an option if the project /community chooses a laptop every five years and perfects it. Anyway, I am no one, and I am a newcomer. I have no clue how things work around here.


cy@ said:
Uhhh, no.
;)
Click to expand...
Gentoo guys would agree with my statement. In 2020, I wrote a shell script to update my kernel automatically, and they went crazy.
Where is the line in the sand? What is a "massage"?

My point in opening this topic is that I would love a FreeBSD laptop in which everything works, that I could tweak with things that matter to me, and that I could not fiddle to try to fix the suspend function.

For example, Debian didn't include proprietary code for a long time, so a lot of stuff didn't work out of the box. It is very straightforward to install drivers in Debian; adding a few repositories is all that is needed. Small things, even if they are easier to deal with—such as the Debian case—can hamper adoption. Bookworms changed that, attracted many new people, and received praise.

The question that matters is: Would that be one of the goals of the FreeBSD project? Everyone here knows FreeBSD is more than excellent, yet some of the discussion around here leads me to believe that it certainly is not one of the goals. And that is okay.

Again, I am not talking about writing thousands of drivers or supporting thousands of different computers.
 
  • Like
Reactions: mro
lcsavb said:
Framework
Click to expand...

2022: "Two Foundation staff members, Ed Maste and Mark Johnston, as well as a few developers and community members now each have access to Framework laptops, …. The goal of this work is to ensure that the experience running FreeBSD on the laptops matches the stability that FreeBSD users expect. …"


2024: "…the Foundation recently purchased a couple of the latest generation Lenovo X1 Carbon laptops and sponsored work to make sure that peripherals are supported out-of-the-box. …
 
%pupil -sceptical said:
… https://wiki.freebsd.org/Laptops … (before anyone thinks this thing was dead/outdated)
Click to expand...

The row for my previously used device is outdated.

There's no row for the device that I began (re)using around a year ago.

My recommendation was: If all FreeBSD on laptop users were giving their experiences into that list, way much more would be achieved than to wish for some fancy project may be started by others. …
Click to expand...

I do understand the wish, however if everyone shared their experiences the page could easily become difficult.

My experiences with a particular device might differ wildly from someone else's experiences with the same device; and so on.
 
Its true. You can have identical model laptops from the same year and they can often have different components / chips inside.

Usually it is whatever bargain bucket component is currently going cheap gets used in that batch.
 
Cath O'Deray said:
however if everyone shared their experiences the page could easily become difficult.
Click to expand...
kpedersen said:
You can have identical model laptops from the same year and they can often have different components / chips inside.
Click to expand...
Yes. You're right.
And the list is already too complex as it is.

One may think to reorganize this kind of data volume somehow else, with MySQL, or whatever...

However it's more as a start.
Better to simply collect all experiences with hardware FreeBSD is tested on in one single pool,
than to start a second, third... pool,
or to think of producing some kind of an own FreeBSD laptop, ARM,... without having the slightest idea of how many sales there were.
 
lcsavb said:
Well, it seems it's already done:

https://xyinn.org/blog/freebsd/framework_laptop
Click to expand...
Nope.

quote:'
Bluetooth
Status: Fails

Fingerprint Reader
Status: Untested

Micro SD Slot
Status: Fails'

By what I read here in the forums Bluetooth is the second most complaints about not working.
Fingerprint reader is untested - I don't need nor use any,
but with the micros SD slot not, and ethernet working 'slow'
it's definetly not 100%, ?

Simply another compromise you may find on many other laptops, too.

Face the reality:
With an free-to-use open-source OS there is no (current) laptop having absolutely 100% supported every single feature.
It's always a compromise of what you must have, and what you can dispense.
And to me the WLAN adapter is my least concern because I simply can place a dongle in the USB port.

But - again - having a database of hardware attended consequently lowers the risk to buy something on chance.

But I had said all this at least three times in this thread alone.
Repetition does not bring new knowledge.
So I'm off here.
 
You must log in or register to reply here.
Back
Top