Selling FreeBSD-preloaded Computers

[Samuel Venable]

I'd like to help out in selling FreeBSD pre-loaded computers. What I would do is create an automated installer script allowing the user to have the option of selecting a Desktop Environment, and have everything auto-configured to "just work" out of the box, and the script would be optionally run by the user the first power on of the machine. Since they would technically be prompted and would activate installation by their choice, this will avoid any legal concerns related to GPL'd and patented software, they would be downloading and installing it themselves.

In this topic I'd like to see who's interested in this, and whether there is any demand for it. This is not the original topic, I figured it would be better to change the discussion rather than creating a follow up topic.

I would also need someone to handle the hardware side of things. Not sure if this will get anywhere. I intend to do this for free btw. I don't think it would be enough work on my end to justify any sort of payment. The script will also be available on github when it is ready.

I'm sorry if I come across a little aimless in my goals, there is some truth to that lol there are a lot of ways I'd like to help out in this community and not all of my ideas will be good ones. I'm not in a position to donate to the FreeBSD foundation right now but I look forward to being able to down the road.

Edit:

Spoiler: Original Topic Contents

I've been toying with the idea of partnering with a local company which makes custom gaming computers, seeing if they'd be interested in making FreeBSD-preloaded computers with specific software and configurations I would specify like a custom distribution. The intent is not for me to make money, as i don't think they'd be willing to do it and pay me for it, but in particular, to promote several game development software that not many people know about, it would be a product very similar to the Pinebook Pro as it would be very experimental and generally only appeal to the FOSS and other game/software development communities.

My questions are:

1) Do I need permission to use FreeBSD as the base system? Who do i contact about this if so, and how?

2) This is sort of a two in one question. I'd like to have GPL software and possibly even some software that is patented, such as ffmpeg just so it can have OBS studio and a basic video player built-in. Do i need to package the source code along with the OS itself, or is linking to the source where it is available sufficient in the case of a custom distro? (downloaded from the pkg package manager, not installed from some other weird or obscure method). Is all of this perfectly legal without having to pay for all the patent licenses?

3) is there anything else not listed here i should be concerned about? I created a Virtual Machine of the OS i'd like to be preinstalled. Is that a license or patent violation on its own? Should I take it (the download link for it) down? These are some things I didn't think about until after releasing my VM, so if you don't feel like answering the other questions, at least let me know whether i have anything to worry about with distributing my VM.

 

[drhowarddrfine]

a custom distro

The term "distro" was created to describe Linux created by packaging that kernel with third party software to create a complete operating system. FreeBSD is already a complete operating system. You are either using FreeBSD or you are not. Do not call any configuration of FreeBSD a "distro".

 

[Samuel Venable]

The term "distro" was created to describe Linux created by packaging that kernel with third party software to create a complete operating system. FreeBSD is already a complete operating system. You are either using FreeBSD or you are not. Do not call any configuration of FreeBSD a "distro".

That honestly has nothing to do with the point of this topic, so before it gets derailed any further I am going to ask things like this are kept to one's self.

Honestly, what are 10/10 people going to call GhostBSD, NomadBSD, or HelloSystem? A FreeBSD-based distro. You may be one of the odd few who disagrees, and it doesn't matter if you are right because again that's not what this topic is centered around.

Distro is short for distribution. I have absolutely no idea what the D in BSD stands for, and I don't want to know.

 

[drhowarddrfine]

You may be one of the odd few who disagrees, and it doesn't matter if you are right

Being right in terminology is everything and if you don't think the first thing that comes to mind when one says "distro" is Linux then you have another think coming.

The word "distribution" in FreeBSD relates to the original distributed copy of Unix that was passed around in the early days of BSD. Is that the same as "distro"? No it is not.

Get your terminology right to avoid confusion and you will go farther than hobby shop activities.

 

[Samuel Venable]

Being right in terminology is everything and if you don't think the first thing that comes to mind when one says "distro" is Linux then you have another think coming.

The word "distribution" in FreeBSD relates to the original distributed copy of Unix that was passed around in the early days of BSD. Is that the same as "distro"? No it is not.

Get your terminology right to avoid confusion and you will go farther than hobby shop activities.

Cool story bro.

 

[Beastie7]

The BSD license was drafted specifically to avoid these kinds of legal questions. It's copy-center. As long as you don't use the trademark "FreeBSD" in your marketing, you should be ok. Unless you want potential bureaucratic headaches in the future; I'd avoid any *GPL licenses if you intend on shipping this to the market.

I think you're getting off just showcasing FreeBSD on the desktop than trying to create another vertically integrated hardware/software platform. Once you factor in IHVs, product distribution, pricing, branding, attracting desktop developers (which FreeBSD has no ecosystem for), etc.; you'll likely give up.

Now, if FreeBSD (as a base system) had it's own Display Server/Compositor/Toolkit stack that the FreeBSD community owns; you may have a further head start.

 

[bookwormep]

I respect your right to ask these questions here; but, the answers you need would come from a licensed practicing attorney who has a specialty in intellectual property.

And I mean no disrespect to any of your posts; on the contrary, you have an excellent grasp of many topics and implementations for the FreeBSD operating systems and functionality.

 

[Samuel Venable]

The BSD license was drafted specifically to avoid these kinds of legal questions. It's copy-center. As long as you don't use the trademark "FreeBSD" in your marketing, you should be ok. Unless you want potential bureaucratic headaches in the future; I'd avoid any *GPL licenses if you intend on shipping this to the market.

I think you're getting off just showcasing FreeBSD on the desktop than trying to create another vertically integrated hardware/software platform. Once you factor in IHVs, product distribution, pricing, branding, attracting desktop developers (which FreeBSD has no ecosystem for), etc.; you'll likely give up.

Now, if FreeBSD (as a base system) had it's own Display Server/Compositor/Toolkit stack that the FreeBSD community owns; you may have a further head start.

I might have to rethink this stuff quite a bit then. Perhaps Lumina could be the base DE, but even then that uses Xorg, which what I read elsewhere has GPL deps. I haven't given up completely, I think the biggest factor is getting the GPL out of xorg, replacing that, and create a desktop theme that looks more polished than Lumina does by default. Thanks for the info, I'm not quite certain how Linux distro's manage to be sold built-in to certain computers and hardware without having the funds to pay for patents and other related issues.

I respect your right to ask these questions here; but, the answers you need would come from a licensed practicing attorney who has a specialty in intellectual property.

Of course, I don't consider replies to this topic free legal advice given by lawyers because I know it's not. I just wanted some brief feedback on this idea and I'm not really sure how much getting legal advice would cost or who'd be in my area that would be equipped to answer my concerns with more weight put into it, I still don't know where to begin, but I am making steps to figure this out the best I can.

 

[shkhln]

1) Do I need permission to use FreeBSD as the base system? Who do i contact about this if so, and how?

No permission necessary.

2) This is sort of a two in one question. I'd like to have GPL software and possibly even some software that is patented, such as ffmpeg just so it can have OBS studio and a basic video player built-in. Do i need to package the source code along with the OS itself, or is linking to the source where it is available sufficient in the case of a custom distro? (downloaded from the pkg package manager, not installed from some other weird or obscure method). Is all of this perfectly legal without having to pay for all the patent licenses?

Patents are not copyright, satisfying the license requirements doesn't necessarily allow you to do anything with patents. In this specific case, since FFmpeg doesn't actually own those patents, they can't grant you any rights even if they wanted to.

3) is there anything else not listed here i should be concerned about?

Well, there is the obvious question: who in their right mind would want a FreeBSD machine specifically for gaming?

 

[Samuel Venable]

No permission necessary.


Patents are not copyright, satisfying the license requirements doesn't necessarily allow you to do anything with patents. In this specific case, since FFmpeg doesn't actually own those patents they can't grant you any rights even if they wanted to.


Well, there is the obvious question: who in their right mind would want a FreeBSD machine specifically for gaming?

I appreciate this feedback. Also: I appreciate your honesty in that last comment lol I know of a lot people who think this idea is a bit odd and I can see why it viewed that way, especially when my original plans involved using GPL, even if that were viable, which It isn't, that's less reason to use any BSD as the base system. Several people have mentioned now (one in a private message) why GPL can't be used, so Lumina it is.

 

[GoNeFast_01]

Well, there is the obvious question: who in their right mind would want a FreeBSD machine specifically for gaming?

I think he meant the game developers of games, so they can use it to build with their game engine.

Having said that I think there could be a used case for FREEBSD Gaming Computer = Playstation isn't it BSD base ...

 

[Samuel Venable]

I think he meant the game developers of games, so they can use it to build with their game engine.

Having said that I think there could be a used case for FREEBSD Gaming Computer = Playstation isn't it BSD base ...

With some hacks you can probably get FreeBSD running on a PS3. There are still enough things Sony had to write on their own and stuff gutted/replaced so that might not be a good comparison but idk

 

[Beastie7]

I might have to rethink this stuff quite a bit then. Perhaps Lumina could be the base DE, but even then that uses Xorg, which what I read elsewhere has GPL deps. I haven't given up completely, I think the biggest factor is getting the GPL out of xorg, replacing that, and create a desktop theme that looks more polished than Lumina does by default. Thanks for the info, I'm not quite certain how Linux distro's manage to be sold built-in to certain computers and hardware without having the funds to pay for patents and other related issues.


Of course, I don't consider replies to this topic free legal advice given by lawyers because I know it's not. I just wanted some brief feedback on this idea and I'm not really sure how much getting legal advice would cost or who'd be in my area that would be equipped to answer my concerns with more weight put into it, I still don't know where to begin, but I am making steps to figure this out the best I can.

I'm sure Canonical and Redhat have their own legal teams who work with hardware vendors to sort out any potential legal problems. I haven't looked at X.org thoroughly but most of the libraries there are MIT licensed I believe. But X.org is just... bad. Usable, but bad. Also most (if not all) desktop software on FreeBSD is upstream Linux; shipping software with which you have no control of (at least a stack that's in base/BSD licensed), you're going to have a hard time ensuring quality assurance for people using it out of the box. Qt requires you to buy a license for commercial purposes, and the free version is GPL'd.. yuck.

Another hurdle is drivers.. and certifying that they work. That's a whole other bucket of tears right there.

 

[Samuel Venable]

I'm sure Canonical and Redhat have their own legal teams who work with hardware vendors to sort out any potential legal problems. I haven't looked at X.org thoroughly but most of the libraries there are MIT licensed I believe. But X.org is just... bad. Usable, but bad. Also most (if not all) desktop software on FreeBSD is upstream Linux; shipping software with which you have no control of (at least a stack that's in base/BSD licensed), you're going to have a hard time ensuring quality assurance for people using it out of the box. Qt requires you to buy a license for commercial purposes, and the free version is GPL'd.. yuck.

Another hurdle is drivers.. and certifying that they work. That's a whole other bucket of tears right there.

If I had the funds to financially support myself and entire dev team, what I'd ultimately love is to work full-time writing a permissively licensed toolkit and DE that looks nice and works well but targeted very specifically with the FreeBSD in mind. The problem with this, while I believe I have what it takes to be a somewhat decent of a help in such a project, there needs to be enough people of a like mind wanting such a thing to become a reality and (again) the funds. I don't know if any of this is realistic even in the long term, but that's what I'd really love to see happen even if it takes a lifetime to pull off (in all honesty). I have big dreams, but it has to start somewhere.

Sometimes it could just be a matter of getting lucky and find the right connections and have the right situations fall into place.

I'm tempted to update this topic and change the base discussion to see who's interested on starting such a thing. Although can't really say it would get anywhere.

 

[shkhln]

I think he meant the game developers of games, so they can use it to build with their game engine.

I can't imagine a competent developer buying a "custom gaming computer", honestly. Or having no software preferences.

 

[Beastie7]

If I had the funds to financially support myself and entire dev team,

You have an entire community at your disposal.

Honestly, your best bet would be to start contributing code to any desktop related topics or issues in the base system. Become a committer, then discuss ways to built upon what exists into a complete graphical system. Windows, macOS, et al. Learn how GPUs work, things like 2D/3D graphics, 2D drawing, color management, pixel manipulation, graphics drivers, typography/Iconography (which is hard as hell, btw), etc.

Unless the committers decide to ship a display server in base; FreeBSD in and of itself won't flourish on the desktop. And no, downstream projects like GhostBSD, NomadBSD, etc don't really count. It has to be in base.

 

[Samuel Venable]

I can't imagine a competent developer buying a "custom gaming computer", honestly. Or having no software preferences.

Some ideas can be a bit of a dud and I get that. If there's no audience for that, that may be great and all. I was outside of this VM project interested in investing in a Permissive Desktop similar to Lumina but without the Qt aspect.

So it sounds like I should take my VM down for the GPL and patent issues? I don't honestly know, but I highly doubt NomadBSD, etc paid for rights to distribute the H264 patented libraries. So I wonder if they are in any legal danger themselves.

What if I distribute a script that automates installation of all the GPL content? At least then they would be technically downloading and installing it on their own, and given the option, not mandatory, and the process made easy.

 

[GoNeFast_01]

I can't imagine a competent developer buying a "custom gaming computer", honestly. Or having no software preferences.

Ok sure let's assume a hypothetical....

If the competent developer had those preference and their pre-disposition of software were known ahead of shipping, let's give them an end product with all his tech in a compile BSD.

How does that future sound?

Question is there a market? Are there going to be pre-orders?

Anyways I want to point out that there are companies that do this... Look at LAMBDA.

https://lambdalabs.com/

 

[Samuel Venable]

You have an entire community at your disposal.

Honestly, your best bet would be to start contributing code to any desktop related topics or issues in the base system. Become a committer, then discuss ways to built upon what exists into a complete graphical system. Windows, macOS, et al. Learn how GPUs work, things like 2D/3D graphics, 2D drawing, color management, pixel manipulation, graphics drivers, typography/Iconography (which is hard as hell, btw), etc.

Unless the committers decide to ship a display server in base; FreeBSD in and of itself won't flourish on the desktop. And no, downstream projects like GhostBSD, NomadBSD, etc don't really count. It has to be in base.

One of my biggest concerns about developing with the official FreeBSD team as a committer, they don't seem to have much focus on the desktop side of things, so I don't see it going as far as a group of people who wanted to work on exclusively providing a polished and permissively licensed desktop environment and toolkit. FreeBSD's base is english only, DE's it can run don't have that limitation, although if I could pull off at least making an English-only OS as a starting point, that in itself would be a thing I'd feel pretty accomplished for pulling off with a team of any size. Not to mention I'm more well-aquanted with C++, it will be a while yet before I can be much help with an OS written in C.

 

[Beastie7]

Those are really non issues to be honest. The desktop team could form a design group or human interface guidelines committee where UI/UX can be discussed or something. On macOS; Helvetica is the default font type for all text in Aqua, for example. A simple language API can be included. It could be C or C++; both are in base. it doesn't really matter. Warner Losh wrote devd in C++, which is in base.

 

[richardtoohey2]

I can't imagine a competent developer buying a "custom gaming computer", honestly. Or having no software preferences.

Sorry to rain on your parade, OP, but I'd have to agree with this.

Why would anyone choose a FreeBSD-based computer for game development? I love the BSDs (well, technology in general) but can't see why you'd develop for the games market on an OS with such a small market share and not known for being a gaming platform (to put it mildly!) You'd be far better using Windows or a platform where you can develop for mobile devices.

I'd suggest some market research/polling to see if there's any demand.

Having said all that, when Google first emerged I said who would need anything to replace AltaVista? And how could Facebook ever succeed? Who would want or need that? And as for cellphones - there was no way they would ever catch on! So feel free to ignore my ramblings!

 

[reddy]

I can't imagine a competent developer buying a "custom gaming computer", honestly. Or having no software preferences.

Some ideas can be a bit of a dud and I get that. If there's no audience for that, that may be great and all. I was outside of this VM project interested in investing in a Permissive Desktop similar to Lumina but without the Qt aspect.

People like me could be interested, our situation is that we have plans to build certain game-like 3D simulation products in the future, but do not have any fix idea on the tools to use at the moment. I was therefore genuinely curious when I read "game development software that not many people know about" in the OP. Other people in a similar situation will probably feel the same way. So will people being frustrated with their current vendor / lock-in / approach and looking for alternatives.

That being said a question that comes to my mind is: why wouldn't we install the software ourselves on our FreeBSD workstations once we know the names of the tools being used? I believe that if you save us the hassle of finding out compatible hardware (video cards etc...) and build true game and video oriented boxes, this could really make sense.

Regarding the fact that there is no market for gamers running FreeBSD, let us remember that today there is this whole cloud-based gaming movement so gamers do not actually need to run games on their machines but frames of the game simply get streamed to their computer, just like when someone watch a video on YouTube. So you can here have a thin client, and play games. In addition to this there is also Web Assembly making cross-platform video games an option.

My advice would therefore be: do not market yourself as a distro with a bunch of pre-installed software. Market yourself as an opinionated workflow on how game development should be done. Then either build (or put together by gluing existing tools in a smart way) the first open-source solution allowing indie developers people to run their own stadia-like platform, so that they can use your framework knowing that the market will be able to purchase their game. Or make sure that your framework is able to export games to a Web Assembly binary out-of-the-box. Ideally do both, and begin by the easiest of these two. That way you can go all in on FreeBSD without worrying about porting your work to Linux, Mac & Cie, this will save you development effort

Just giving a freebsd-only framework, without holding the hand of people to get them to market in a cross-platform manner is a recipe for failure in my opinion. I think this may also be what drhowarddrfine was getting at, because here you'd be plainly saying that your stuff runs on FreeBSD period. And you'll see yourself as someone developing a FreeBSD package, and all your licensing concern can also go away (this is a different way to approach your questions). You can customize everything on FreeBSD anyway using post-install scripts, you can even set a custom bootsplash. This will reduce your maintenance burden so that you can focus your time on your package.

One last remark: to me, as a prospective customer, the minute your business starts to even remotely look like Apple, where you have a walled-garden and stuff only working on your hardware, or if I can smell that this is the direction you intend to take down the line, or if I perceive that there is a risk that you do so and my business could be disrupted I will not even bother touching your solution. My point is you need to keep in mind that your target audience will likely be sensitive to vendor lock-ins. So I'd suggest making all your software tooling open-source, and providing hardware boxes as a premium option for interested people.

Initially you can glue together the core game development framework, while ensuring that one of the two go-to-market solutions I mentioned are built-in, and you could then reserve the DE for the hardware boxes. Just my 2 cents.

 

[kpedersen]

Some random notes...

From what I can see, similar projects using Linux are just about scraping by. The more successful I have seen is:
https://puri.sm/

And from the looks of it, they are now focusing on shite phones instead. My guess is that they realized laptops / consumer hardware is not a viable way forward.

iXsystems and TrueNAS seems to be doing OK for the server market. Our company is also dabbling with Desktop as a Service (Daas) based on FreeBSD but honestly, it isn't really any better than a Linux offering, there are a lot of disadvantages (Linux unfortunately has proprietary software which is what the general public wants!). The main offering we provide is a lot of effort tailoring it to their requirements. Hardly an off-the-shelf product!

For gaming, you could consider specializing in build servers. Something like a Jenkins instance on FreeBSD that can then build to as many platforms as possible (i.e via Wine, Linux compat). Basically every platform will be necessary for many games companies *other* than FreeBSD annoyingly. The weak spot is Apple which is still fairly toxic to build for using proper infrastructure.
UE4 support on FreeBSD will be a massive bonus here. Unfortunately many smaller / indie software houses (i.e your target market) will also be using Unity which requires far too many hacks to be made to run.

 

[tingo]

OP. if you want any input from another relevant source (this community is a relevant source), why not contact the FreeBSD Foundation? Worst thing that happens is that they're not interested in answering your questions (but I don't think that is likely).

 

[Jose]

From what I can see, similar projects using Linux are just about scraping by. The more successful I have seen is:
https://puri.sm/

There's also System76

System76 - Linux Laptops, Desktops, and Servers

System76 computers empower the world's curious and capable makers of tomorrow

system76.com

And from the looks of it, they are now focusing on shite phones instead. My guess is that they realized laptops / consumer hardware is not a viable way forward.

I'm actually thinking about getting one of these phones. I made the mistake of ordering a Samsung phone. It came with a staggering amount of crapware you can't uninstall and that requires permissions that can't be revoked. Yes, I returned it. My vain hope is that Starlink will catch on, and that will mean death to all the institutional wireless carriers. Once we have ubiquitous wireless Internet, all you need is software to make calls and exchange text messages. A lot such software already exists.

There's also a chance now that truly open RISC-V hardware will become widely available. This could mean really cheap, truly open handheld wireless terminals (fine, call them phones.) The combination of these two things could mean the end of the SIM card nonsense or paying 25 cents per text message.

For gaming, you could consider specializing in build servers. Something like a Jenkins instance on FreeBSD that can then build to as many platforms as possible (i.e via Wine, Linux compat). Basically every platform will be necessary for many games companies *other* than FreeBSD annoyingly. The weak spot is Apple which is still fairly toxic to build for using proper infrastructure.
UE4 support on FreeBSD will be a massive bonus here. Unfortunately many smaller / indie software houses (i.e your target market) will also be using Unity which requires far too many hacks to be made to run.

Gentoo had a bootable CD that ran America's Army back in the early aughts. I think that would be the best approach for evangelizing Freebsd for games. A read-only memstick image that boots straight into a free game, or maybe even into Big Picture Steam.