Does Desktop have a future on BSD?

SR_Ind

Active Member

Reaction score: 17
Messages: 138

(Snip a bunch of off-topic whining about amateurish system building.)

None of those things are in any way related to the topic of this thread, so the question remains, why are you here?
Are you really interested in Freebsd as a viable desktop operating system?
Or do you want to spread a bunch of FUD about how much it sucks?
I don't think you even understand what you are talking about.
Everything I said is related.

I already use FreeBSD as a secondary desktop. Thank you.

Quality of development tools drive platform adoption, not the other way around.
 

SR_Ind

Active Member

Reaction score: 17
Messages: 138

SR_Ind
I get some of what you're saying. I once pointed out that 14 hours of additional build time was excessive, to simply add a GCC compiler, and they fixed that. I use the same kind of logic to point out inefficiencies. It has gotten much better. However, also there's some funny stuff when trying to upgrade a program, and seeing the amount of programs that need to be deinstalled, and reinstalled. Trying to untangle this is difficult.

Only shell and C are more universal as a language than Python, so I am completely good with Python. Lua probably matches Python in terms of how universal they are. I was thinking of Sphinx being optional, it can be, but it is so essential, I would rather leave it in. If it makes more people happy, make Sphinx optional.

However, the arguments you're making aren't helpful to your view.
The official installer from CMake for Windows doesn't seem to have any python dependency.
That's the official installer.
IIRC it is there in Linux however (have to check that on a Debian based SBC).
So, from where does this whole kitchen sink approach has encroached the FOSS world, I have no logical explanation.
 

sidetone

Daemon

Reaction score: 909
Messages: 1,864

The official installer from CMake for Windows doesn't seem to have any python dependency.
That's the official installer.
IIRC it is there in Linux however (have to check that on a Debian based SBC).
So, from where does this whole kitchen sink approach has encroached the FOSS world, I have no logical explanation.
Python is needed for Sphinx documentation. That might be all it's needed for with cmake. A mathematical scripting language is expected to be highly efficient.
I always install ports with examples and docs turned off. Why should I have to look into Makefiles?
Makefiles are the best way to figure things out and improve things.
 

SR_Ind

Active Member

Reaction score: 17
Messages: 138

It is expected a FreeBSD user can use pkg install to trivially add this stuff. It will never be in the base install because it doesn't make sense for so many use-cases of FreeBSD.

As a C developer, you should also probably realize the issue with Qt and non-standard MOC for an OS primarily focussing on C. The very fact we have a C++ compiler in base is "luck".
I hear you.
Nowadays, I use minimal Qt for GUI development only. And that's my only use for C++.
I can't wrestle with ABI incompatibilities, whims of compilers. And and top of that the standards committee has embarked on very rapid refresh cycle. Not sustainable.

All said and done, other GUI toolkits in the *NIX world cannot compete with Windows and Mac OS. Not fair, but that's what it is.
 

bakul

Active Member

Reaction score: 57
Messages: 119

Forget about desktop on FreeBSD.

Is FreeBSD run by sane people in the first place?

I just reformatted a hard disk that was running the latest FreeBSD.

The trigger?

Why on earth should the port build for CMake pull in Python?
And I checked with a downloaded CMake source tar ball. It just builds fine outside of port system.

I'm an old C programmer and I usually select most optimized operating systems for our embedded systems. One rule is that simple self contained BSD build tools.

FreeBSD is turning into a mess. Port system is brokenup. BlueTooth doesn't work. WiFi is iffy even with well known drivers.

I don't like the bloated Linux distros with ever complex package dependencies.

Time to migrate to something more exotic.
While some of your criticism may be valid, this is the nature of open source, volunteer driven projects. You can complain all you want but that won't fix anything.

If you want to make a real difference, get involved. Write code, write documentation, test/review stuff, port software, rationally argue your point of view about some relevant topic, promote FreeBSD (or open source s/w you care about) in your workplace, get others interested in OSS projects, etc. etc. In the process you will learn many new things, improve your coding, writing and communication skills, make new friends, earn the respect of your peers etc. These things will even help when you look for your next job or consulting gig!
 

Trihexagonal

Son of Beastie

Reaction score: 2,425
Messages: 2,946

Desktop is simple, if and only if FreeBSD adopts a GUI/Widget/WM stack as the first class citizen. Just bundle Qt GUI and Widgets in the base system with a bare bone window manager with a set of configuration tools.
But the catch here is that no developer will spend their time to build these simple GUI tools, if the underlying OS has somewhat sub par hardware support.
The catch is the vast majority of people who use FreeBSD do not want a bundle of "Qt GUI and Widgets in the base system with a bare bone window manager".
Chicken and egg situation.
Exactly.

I build my desktop from scratch. I use my desktop for development work.
I don't require Python and any of the interpreted runtimes.

If I start changing the Makefiles that came with the ports, then it will bring the whole question of using the ports.
I'm glad to hear you build your desktops from scratch and use ports. That's my preference as well.

Then you know how to work through the problems you refer to with python 2.7, what you do and what you do not want on your system.

But please don't try to force what you want on your system on mine with a Qt GUI, Widgets and a bare bone WM in the Base System.
I know what programs I do and do not want on my desktops/laptops, too.
 

kpedersen

Son of Beastie

Reaction score: 2,091
Messages: 2,951

All said and done, other GUI toolkits in the *NIX world cannot compete with Windows and Mac OS. Not fair, but that's what it is.
Absolutely. However I wonder if we had great UI experiences, would this not detract from UNIX where the command prompt is the first class citizen?

Whilst it may be up for debate which is more efficient (CLI or GUI), what isn't really up for debate is whether we need more GUI platforms. We have tonnes of them. Why would we damage and reduce the number of CLI platforms to add yet another weak GUI platform to the world?
 

mark_j

Daemon

Reaction score: 711
Messages: 1,236

Whoa. Seems like a lot of ruffled feathers. Sincerest apologies if it meant personally to some of you.

Maybe this was wrong thread to complain about (sharing my frustration) about the state of radio interfaces in FreeBSD.

Actually any thread is the wrong thread to complain about hardware drivers. You need to go the mailing lists.

You may have a problem with bluetooth, so be it, but when you open your message with:

Forget about desktop on FreeBSD.

Is FreeBSD run by sane people in the first place?

I just reformatted a hard disk that was running the latest FreeBSD.

you're just attacking users here. It's uncalled for, and a stupid approach to attempting to engage in conversation.

FreeBSD is better suited for embedded space.
It will take a fraction of effort to get state of the art radio networking capabilities in FreeBSD.

You're kidding, right? Have you looked at the bluetooth specification?
You realise Linux got theirs from Qualcomm, where a swag of programmers (paid programmers) produced it then open sourced it?

Seriously, if it's so easy, knock yourself out. I can point you to the correct people to speak to if you wish and get you started on the project?

Unfortunately, that effort doesn't come cheap. Radio networking stack development is less common skill.
No idea what are the funding priorities of the FreeBSD foundation.

The "effort" is neither easy or cheap.

You are correct that there is a total disconnect between the foundation and its users. Absolute.
It's not a presence here or anywhere (such as mailing lists) and takes its cues solely from core, it seems.
I believe they attempt to gauge user requirements through the annual survey. Nothing much has ever come of it, in my opinion. It took ages for them to fund 802.11ac.
I surmise their funding approach is aimed at the large benefactors.

Desktop will remain a shifting goalpost, unless the approach changes.
Desktop is simple, if and only if FreeBSD adopts a GUI/Widget/WM stack as the first class citizen. Just bundle Qt GUI and Widgets in the base system with
That is never happening (well, OK, it did sort of with Xorg in base). FreeBSD is not a "distribution" cobbled together by someone, it's an integrated system where they give you that base OS and everything to function in it. If you want to add stuff to it, then you use ports and/or packages. If you don't want to put in that effort, then you can use a derivation like GhostBSD.

a bare bone window manager with a set of configuration tools.
But the catch here is that no developer will spend their time to build these simple GUI tools, if the underlying OS has somewhat sub par hardware support.
So, it has average to poor support for wifi and bluetooth and this is a reason not to use a GUI with FreeBSD? Fine, then install Windows 10 or systemd/linux and move on.
Life's short, too short to be annoyed about an OS.
 

kpedersen

Son of Beastie

Reaction score: 2,091
Messages: 2,951

You are correct that there is a total disconnect between the foundation and its users. Absolute.
It's not a presence here or anywhere (such as mailing lists) and takes its cues solely from core, it seems.
I believe they attempt to gauge user requirements through the annual survey. Nothing much has ever come of it, in my opinion. It took ages for them to fund 802.11ac.
I surmise their funding approach is aimed at the large benefactors.
You seem to not be thrilled by the foundation.

I do possibly agree with you in that they don't really seem to be acting for their "little" users. Though oddly enough this kind of has been a saving grace because users don't really know what they want and ask for dumb things like GUI installers or default desktop environments. At least with the foundation completely ignoring us and going for what (probably Netflix) wants, it ensures FreeBSD stays competitive in important areas.

It is a little weird though. Perhaps the discussion and reasoning is on a specific mailing list.
 

Menelkir

Well-Known Member

Reaction score: 340
Messages: 277

You seem to not be thrilled by the foundation.

I do possibly agree with you in that they don't really seem to be acting for their "little" users. Though oddly enough this kind of has been a saving grace because users don't really know what they want and ask for dumb things like GUI installers or default desktop environments. At least with the foundation completely ignoring us and going for what (probably Netflix) wants, it ensures FreeBSD stays competitive in a few areas.

It is a little weird though. Perhaps the discussion and reasoning is on a specific mailing list.
I think the foundation is more turned into the core of FreeBSD than anything else that resides in ports, since the core is the most important part for everyone.
 

sidetone

Daemon

Reaction score: 909
Messages: 1,864

You're kidding, right? Have you looked at the bluetooth specification?
You realise Linux got theirs from Qualcomm, where a swag of programmers (paid programmers) produced it then open sourced it?
I would be happy with having that in ports, no matter if it's GPL. Just to have the functionality. It doesn't have to be Bluetooth, just something where I could use 1 wifi usb stick with multiple devices.
 

vigole

Daemon

Reaction score: 1,462
Messages: 1,271

I do possibly agree with you in that they don't really seem to be acting for their "little" users.

I think the foundation is more turned into the core of FreeBSD than anything else that resides in ports, since the core is the most important part for everyone.

My take on it: in projects, source and its quality are essential, i.e. committers. Without committers, there's no source, hence no project.

The rest are auxiliary, they come and go. Their policies and statements have an effect, but it’s limited. They can improve or destroy a project. In worst case scenario, the latter can lead to fork, e.g. NetBSD->OpenBSD and FreeBSD->DragonFly BSD.

They're mainly a legal front, hopefully PR, and a mean to attract more capital. Capital is important. People have to eat! But a project can hold hostage by its investors. They can and will dictate policies, sometime through delicate peer pressure. But they don't have full control over a good working project. Good committers produce good code/project, and eventually, building up a happy user base. Happy users can save a project, even in the face of strong foreign forces. It's not a myth, e.g. OpenBSD and DARPA funding.

In some systems, the leader is the system, e.g. FSF. Take him/her out and the system is over. OS is not an organisation. It's code, and depends on merits of its committers. If someone is not fond of management, he/she still can supports the project itself, by supporting its committers.
 

sidetone

Daemon

Reaction score: 909
Messages: 1,864

The base is great. It's better than any other operating system. That everything from big contributors and companies goes into it doesn't negatively affect anything.

The ports tree is good. There's room for improvement, and there are too many project suites in it, some are owned by project entities. The ports tree is heavily leaned towards heavy desktop use, and it affects everything. The committers are great and do a good job. No one understands how it can be better. I think of making a port with different options for default packages, and I'm told, I don't understand the ports tree.

I used to see something that needed improvement. And then, I was told, it's fine, leave it be. Kind of like fear to adjust it, not seeing how it would be better. Then I made an improvement on it, and they're like, oh, that's better.
 

wolffnx

Aspiring Daemon

Reaction score: 231
Messages: 677

Screenshot from 2021-06-21 10-38-17.png


I have a taskbar, It counts? 🤣
 

mark_j

Daemon

Reaction score: 711
Messages: 1,236

You seem to not be thrilled by the foundation.

Well, yes and no. As you rightly point out, the end user often doesn't know what they need. Conversely, often they do. Does that demand a big say in the development of the OS? No.

As I see it with the foundation: Ignore the users, do what we want. Sure, its big benefactors are corporate and that necessarily means they cater for their needs. However, this surely leads to the trap where you just become the operating system arm of Netflix, for example.


I do possibly agree with you in that they don't really seem to be acting for their "little" users. Though oddly enough this kind of has been a saving grace because users don't really know what they want and ask for dumb things like GUI installers or default desktop environments. At least with the foundation completely ignoring us and going for what (probably Netflix) wants, it ensures FreeBSD stays competitive in important areas.

(Firstly, I am talking ONLY about hardware. People wanting a GUI and other non-base software installed by default are using the wrong OS.
I did not make that evident originally, sorry.)

I understand their rationale, I just don't understand the opaque process. If users are not important to them, why do these forums exist? Why did it take them so long to pick up the need for 802.11ac (which is still not working)?

Why, if a user has concern with, for example, Bluetooth, there's not a way to communicate that to "someone"? Does FreeBSD core/foundation even know Bluetooth support is awful? The USB stack is bad and behind the times? Would they like to fix these?

I am sure the foundation does a great job. I am sure they suck as an interface to the users.

To quote them:
The Foundation advocates for FreeBSD by promoting FreeBSD at technical conferences and events around the world.
...
Operating System Improvements: Providing staff to immediately respond to urgent problems and implement new features and functionality allowing for the innovation and stability you’ve come to rely on.
...
New User Experience: Improving the process and documentation for getting new people involved with FreeBSD, and supporting those people as they become integrated into the FreeBSD Community.

Hmmm...
 

ralphbsz

Son of Beastie

Reaction score: 2,353
Messages: 3,241

As I see it with the foundation: Ignore the users, do what we want. Sure, its big benefactors are corporate and that necessarily means they cater for their needs. However, this surely leads to the trap where you just become the operating system arm of Netflix, for example.
Reality check: The foundations goal for this year is $1.25M. Current donors are: Over $25K: Facebook and VeriSign. Over $10K up to $25K: Stormshield, Tarsnap, and VMWare. No other large donors to be seen.

Let's check 2020: Over $250K: The Koum Family Foundation (that's the foundation set up by the founder of WhatsApp). $100K-$250K: Arm, NetApp, NGINX. $50K-$100K: NetFlix. $25K-$50K: Juniper. $10K-$25K: Beckhoff, Fidelity Charitable (probably an anonymous donor using a donor-advised funds), Microsoft, Mozilla, Stormshield, Tarsnap, and VMWare.

I don't see any one donor dominating the foundation, in particular not the big three companies that use FreeBSD as the basis of their product (Juniper, NetApp, NetFlix).

Also: Consider the size of the foundation. It's typical annual revenue is $1.1M - $1.5M. The typical cost of an average Silicon Valley software engineer is $250K to $300K (they are somewhat cheaper in other parts of the US, and in developing countries). Even if the foundation spent all of its money on software engineers, they could afford 4-6 of them. In reality, a significant fraction of the income is going to be spent on foundation-internal costs (such as executive director and administrative functions). Anyone who thinks that the FreeBSD foundation can hire enough people to do all the wonderful things (like 802.11xxx for some value of xxx, or Bluetooth) fails the reality check.

In comparison: Annual revenue of the Linux foundation is about $80M to $100M. Of that, a significant fraction goes into things like Linus' Torvalds salary (over $500K), but it still leaves money for hundreds of software engineers. In addition, companies like Intel, IBM, Oracle, AMD, nVidia and in particular RedHat employ hundreds to thousands of Linux developers each.
 

Phishfry

Beastie's Twin

Reaction score: 2,668
Messages: 5,588

Thanks ralphbsz for the Tarsnap mention. I think that company and Colin Percival really exemplify FreeBSD.
Here is a guy who is a developer and contributes to FreeBSD in general.
He contributes some corporate profit to the foundation for the project that he volunteers for.
That is over the top. Who else does that. Talk about giving back...
 

vigole

Daemon

Reaction score: 1,462
Messages: 1,271

Reality check: The foundations goal for this year is $1.25M. Current donors are: Over $25K: Facebook and VeriSign. Over $10K up to $25K: Stormshield, Tarsnap, and VMWare. No other large donors to be seen.
What's the cause of low turnout?
 

mark_j

Daemon

Reaction score: 711
Messages: 1,236

Reality check: The foundations goal for this year is $1.25M. Current donors are: Over $25K: Facebook and VeriSign. Over $10K up to $25K: Stormshield, Tarsnap, and VMWare. No other large donors to be seen.
Sigh. It was an example to make a point.
I never stated all the hardware goals could be met, I am merely asking that what goals are to be met are stated. If there are none, let's hear it.
I certainly wasn't comparing FreeBSD to Linux so I don't know where that came from.
 
Top