Sovereign Tech Fund to Invest €686k in FreeBSD Infrastructure Modernization | Hacker News Reader

Sovereign Tech Fund to Invest €686k in FreeBSD Infrastructure Modernization

freebsdfoundation.org

158 pointsnesarkvechnep94 comments

Interesting, the first time I've heard of the https://www.sovereigntechfund.de/ .

They have funded a variety of tooling, including curl, ffmpeg, gnome, php, etc: https://www.sovereigntechfund.de/tech

My favorite part is they're funding making home directory encryption a first-tier integrated feature of GNOME[1]

[1] https://gitlab.gnome.org/Teams/STF/homed/-/issues/42

I've been personally most excited for this, simply because I've personally been looking at making my own image based system to daily drive and having an empty (and possibly read-only) /etc/ is kind of a prerequisite, which passwd and shadow kinda make difficult.

For anyone that doesn't know, homed allows using different mechanism as backing storage (e.g. directory, luks loopback, btrfs subvolume, separate drive and more). It stores user information usually store in passwd and shadow in a signed ~/.identity file allowing it to be entirely portable. Changes need to be done via homectl as it needs to be resigned by a system specific key (same for when a new home dir is added) and some changes are limited to admin only, while others can be done by the user themselves.

I am really surprised by all the negative comments in this thread about FreeBSD by people who seemingly use Linux and think this is an appropriate way to behave.

There are many reasons why someone might use BSD, and it has several specific advantages over Linux for some use cases. That aside, even in cases where Linux is "better", FreeBSD is more than sufficient. There is no downside and many upsides to having multiple free operating systems that are aligned around the same core standard (POSIX) with tooling for software portability.

Thank you Germany for helping fund open source.

You're surprised by negative comments on HN?

I am not. When I used Coherent OS, Linux was starting out and Linux people were spamming comp.os.coherent on USENET incessantly. The very few BSD people who posted were respectful of people on that group.

Seems over the years, times have not changed :(

And yes, a big thank you to Germany.

I'm not surprised, but I am a little disappointed. Some random tech fund gives a small amount of money to a project that can put it to good use, and even plans specific useful things with said money, and then some ninnies come along to complain about it? Why? Is it their money? Do they dislike FreeBSD benefiting from things? Are they jealous because their pet project didn't get any money in this particular instance?

I'm sure some money goes towards silly things and could be spent on more worthwhile things, but this doesn't strike me as a waste of money...

This seems like a strange use of money, surely? Nothing against FreeBSD, but Linux is where most of the focus already is. People using FreeBSD are doing so due to preference or legitimate need due to their software somehow specifically being BSD dependents.

Not that this isn't 'nice', but is that the goal here? I would have thought it would make more sense to put this money toward something underfunded but still useful for the government.

How do we know it's not useful to the German government? Without context, I would have assumed that "infrastructure, security, regulatory compliance, and developer experience" meant they were trying to secure FreeBSD's builds because they were using them and identified them as a weak point.

Then they are wasting taxpayer dollars and should switch to Linux.

686000 euros are not taxpayer dollars.

You have a point, and I agree that these days, for most people, Linux is where the action is. But it might not be desirable to end up in a future where Linux is the only game in town. Therefore it could be wise to support other FOSS operating systems, to keep some kind of competition.

Is this not an argument against vendor lock in? I'd argue that doesn't really apply to a FOSS project like Linux, and that BSD isn't sufficiently different to help diversity. If that was the goal, there are a ton of smaller OS projects that I think would make sense to fund. Like ReactOS.

huh?

https://en.wikipedia.org/wiki/List_of_products_based_on_Free...

https://freebsdfoundation.org/end-user-stories/

ex Junipers junOS, Sonys OS for PS3,4,5 is based on FreeBSD. Netflix and Netapp use it for their main stuff.

Lots of "under the hood" stuff is FreeBSD. Even I know this, and I'm a filthy Windows user.

I’d wager BSDs move more packets every day than any other major OS. Network infrastructure is often running BSD, so does Netflix on its cache servers, network appliances…

You're 'huh' is misplaced.

Some stuff having some BSD code in it is entirely beside the point.

You should be running a mix of FreeBSD anyway. Do not get yourself locked into one vendor solutions even if the vendor is open source.

The BSD license is more permissive than the GPL/Linux license, so one could argue source code investments made here can more readily be integrated into all systems, even closed source.

I don't see how diversity of platforms/OS could be anything but a good thing. FreeBSD has been making a lot of QoL changes over the last several years and is laid out much more like a cohesive system than any Linux distribution I've used. Not to mention it is still used elsewhere.

Putting all your eggs in one basket is usually a good way to end up hungry.

> I don't see how diversity of platforms/OS could be anything but a good thing.

I didn't say it wasn't a good thing, I said I don't think it giving €686k to FreeBSD is the best use of that money, even if diversity of systems is a goal. In that case, give it to ReactOS or something.

This seems like a strange use of money, surely? Nothing against ReactOS, but Linux is where most of the focus already is. People using ReactOS are doing so due to preference or legitimate need due to their software somehow specifically being ReactOS dependents.

Not that this isn't 'nice', but is that the goal here? I would have thought it would make more sense to put this money toward something underfunded but still useful for the government.

I'll rephrase since my last comment was flagged.

FreeBSD in this context is basically Linux-lite. It's not really adding diversity. Something like ReactOS is. That's why trying to take my words and use them against them here doesn't work. At all.

>This seems like a strange use of money

I do not :)

Linux and even more so, us, needs FreeBSD and the other BSDs. IMHO Linux, at the moment, seems to be heading to Corporate Domination and is slowly becoming a Windows type Clone. Plus with Wayland being Linux only and its largest funder is Red Hat, it seems to be trying to eliminate BSD.

The *BSDs right now is a option for people who want an OS to work the way they would like it to, not a way a Company like Red Hat/IBM wants it to be.

So great for FreeBSD, and the way the BSDs operate, developments will cross-pollinate the other BSDs.

> Wayland being Linux only and its largest funder is Red Hat, it seems to be trying to eliminate BSD.

https://docs.freebsd.org/en/books/handbook/wayland/

> The *BSDs right now is a option for people who want an OS to work the way they would like it to, not a way a Company like Red Hat/IBM wants it to be.

I mean, Linux is just the kernel. Any number of distros are further than the BSD's for how Red Hat and IBM would want it to be.

Of note: Hacker News, this very web site, runs FreeBSD as their server.

Is anyone aware of similar funds or structures from different countries?

Sovereign Tech Fund is German, but I did not find (albeit not looked too much) any in other European countries.

https://hn.algolia.com/?q=nlnet ?

NLnet is somewhat similar: https://nlnet.nl/

Nlnet gets funding from NGI https://ngi.eu/ and unfortunately there have been insider reports that EU commission gave top priority to AI and from 2025 cut all funding of NGI to free up money for the AI initiative. Of course this is pretty lame and terrible as lot of good independent projects got funding there.

For context, NLNet's NGI0 programs are EU-funded (for now [1]).

[1] https://pad.public.cat/lettre-NCP-NGI

Every country has some sort of quango that throws money at "weird" projects.

FreeBSD is not weird, it's a good and very reliable OS, I'm very happy this happens.

FreeBSD is excellent if your use-case is flinging bits as fast as possible, but it is kind of niche. Who else is going to fund them? Netflix?

>Who else is going to fund them? Netflix?

So you don't have to do the "research" yourself:

https://freebsdfoundation.org/our-donors/donors/

https://freebsdfoundation.org/our-donors/donors/?donationYea...

BTW: Those are just the monetary donations towards the foundation, additionally you have developers that are directly paid to work on the FreeBSD project (search for "Sponsored by"):

https://www.freebsd.org/releases/14.1R/relnotes/

https://www.freebsd.org/releases/13.3R/relnotes/

There's also a lot of FreeBSD code in Apple OSes.

There is! I wonder how much new *BSD code Apple has imported the past ~5 years or so. Does anyone know?

The code can be found here: https://github.com/apple-oss-distributions

One example I'm aware of is that they switched from GNU libiconv to FreeBSD libiconv in macOS 14.

Can't argue it's niche but it's far from uncommon. The BSD licensing allows usage in places allergic to the GPL so you see it (or don't) often used behind the scenes in lots of products.

...and I'm writing this comment on a Lenovo T450s running FreeBSD. Dang can probably verify the user agent of my POST, if he has nothing better to do (pretty sure he does).

The experience is not perfect (just _now_ I'm enjoying fighting with a deskhop (https://github.com/hrvach/deskhop) which isn't seen as a ums pointing device unless another usb mouse is also present, but that's the first problem in months (admittedly it's also the first change in as many months)).

Would be interested to hear more about BSD as a daily driver.

Is the slower wifi problematic? How about browsing the web, are the major bowsers (chromium or Firefox) supported? What about running the tools necessary for web development (or any other language that isn’t C)?

Chromium and Firefox are both there, yes. Sometimes one or 2 versions behind if you go for the packaged version, but I've compiled them both before quite painlessly. Even in the packages you may occasionally run into something not working - browsers are so big now and not all corner cases get tested for every possible OS. I've had tabs crashing, although not since a couple years.

The wifi is not really slow but disconnects with a certain frequency (at least 1 brief drop per day) which on the same hardware didn't happen on Windows or Linux. The slowness is not noticeable until you're trying to move gigabytes - watching youtube is unaffected for example.

Can't help with the web development info, I'm relatively ignorant there - more on the sysadmin side of things :) but a lot of Linux software can be run directly thanks to a compat layer (linuxulator) which gives you a convincing impersonation of a centos, debian or ubuntu - or you can actually run the distro themselves in a jail.

The biggest absence is Docker. Not as a concept (jails are actually better under many aspects, and native to FreeBSD), but because so much software is now shipped like that. There's progress to get them working properly, otherwise you run them in a Linux Bhyve (BSD hypervisor) VM.

I haven't tried running an LLM since none of my FreeBSD machines have hardware even remotely capable of that, would be nice to hear if anyone was successful.

For Germany that would be https://prototypefund.de/en/

> A quango or QUANGO is an organisation to which a government has devolved power, but which is still partly controlled and/or financed by government bodies.

A quasi NGO. First time I heard this term.

It's a British term; it basically means a somewhat independent authority nominally at arm's-length from the government, for example, regulatory agencies or national museums.

I am interested in the details, particularly how they plan to reduce the technical debt. Are there any additional resources available?

Presumably it means paying people to work on all the painful crap they’re not getting contributions for.

Most people who want to help out a project like FreeBSD will work on something that interests them; I have an hard time imagining that many people out there want to work on some libtool removal or cleaning up deep in the plumbing of buildworld or such.

Plenty of people are happy to be paid to do such tasks tho, and this pays for that. We all win :)

On the one hand, this sounds like a lot of money, on the other, it's ~765k USD. If you just go off the rough assumption that 1 engineer/100k USD/yr then it's either a team of 7 for a bit over a year or a single engineer for a decade or so. I don't know if that's enough to make a significant impact? I'd like to think so, but a project that big, once it gets funding, is sure to have some tech debt that needs paid before anything new, so, will new things come of this?

There are a few exceptions, but overall the BSDs have very well engineered and maintainable code. They don’t move as fast as Linux distros do partly due to less manpower and partly due to taking the time to do it the right way. There is much less churn and they don’t have the cruft of supporting obscure UNIX flavors that have faded into distant memory like GNU utilities do. Take a look at some FreeBSD code and their GNU and Linux equivalents and you’ll see what I’m talking about.

Interesting! Does that make it perform worse on some modern hardware due to having, say, worse scheduling for CPUs with multiple CCXs or worse graphics drivers? I'm obviously wandering a bit outside my knowledge here.

Yes FreeBSD does generally benchmark lower than Linux. The BSDs typically prefer correctness and maintainability to aggressive micro optimizations. There’s nothing wrong with either approach, but I can usually follow FreeBSDs code much easier than I can GNU coreutils or the Linux kernel for example.

The BSDs also have excellent documentation. Every utility, kernel module, and system configuration file have robust man pages.

Where are you getting this from? Most benchmarks I’ve seen put FreeBSD on par with linux.

https://www.phoronix.com/review/bsd-linux-threadripper-7980x...

Fascinating and I stand corrected!

Why FreeBSD? How did they decide that it's worth supporting? What return is expected?

I don't even know a reason to use FreeBSD these days apart of being a FreeBSD enthusiast.

pfsense is fairly widely deployed in a lot of 'secure' infrastructure.

>I don't even know a reason to use FreeBSD

That's funny because i don't even know a reason to use Linux these days.

And since there is the Linux-Foundation who invests whopping 5% of their money into the Linux Kernel, an opensource alternative is always good to have.

Edit: Correction it's not 5% but less then 3%

I’m a big fan of the BSDs, but if you’re talking about the desktop, Proton and Steam, WiFi support, and (for me) Niri. You can get Niri working, probably, with some effort, but it comes out of the box on Fedora and Arch. The case of Niri is representative, in my limited experience, of the general case. The dev tooling I use always supports Linux as a 1st class citizen, but rarely supports the BSDs.

Well there is proton and steam on FreeBSD and i can even play Elden Ring and X4 for 2 minutes.....then my machine resets because there is no way to adjust the AMD-GPU fans, so i made a bhyve Win11-VM with passthrough to the discrete GPU and Bluetooth (PS4 controller), and now i can play everything, even my beloved Gold-Box Games with the GBC.

>WiFi support

Yeah that's really a problem, but i just use my Laptop for work so ~6 MB/s is good enough for me, otherwise i would use FreeBSDWifibox it's a hack but hey we use Unixes right? ;)

>Niri is representative

See what Niri is for you, ZFS is for me, i could say that the ZFS-module is representative (on mainstream distros) for Linux and we are both right, FreeBSD and Linux are both not perfect.

BTW Niri is already in ports since ~9 month:

https://www.freshports.org/x11-wm/niri

>dev tooling I use always supports Linux as a 1st class citizen, but rarely supports the BSDs.

I don't know what "1st class citizen" means, if it works it works, i often use dev-tools where Mac and Windows seam to be the "1st class citizen" but if they work they work.

So yes, there are problems like the Overheating AMD-GPU and Wifi but apart from that i am super happy with FreeBSD, and most important everyone is happy with the OS/Distro choice one made.

For WiFi you should really try bhyve with pass through to an OpenBSD guest. I get wifi speeds comparable to what I got with Alpine(wifibox), but a much nicer environment. Not particularly difficult to setup or maintain.

> That's funny because i don't even know a reason to use Linux these days.

Can I run apps like Slack or Spotify on BSD? Will it have WiFi and Bluetooth drivers for my laptop? Not even talking about games.

Slack and Spotify if needed in a Linux Jail, yes to Games (Proton or Linuxlator or native), maybe to Wifi and BT.

Read my comment two steps down the line (bhyve-vm etc).

> Slack and Spotify if needed in a Linux Jail, yes

Ha, just after you last said:

> That's funny because i don't even know a reason to use Linux these days.

Well, there's one reason to use Linux, lol - to run all the apps on BSD that BSD doesn't support.

>Ha, just after you last said

I don't use a Slack Client but the Web one, and i don't use Spotify.

>Well, there's one reason to use Linux, lol

True, you have a point to compare two proprietary Blobs, oh BTW is Adobe and nearly ever AAA-Game supporting Linux or do you need Wine/Proton (that Windows not emulator) thing?

And to be clear a LinuxJail is actually NOT using the Linux Kernel but the GNU-Userland ;) so no Linux present, Wine is the ~same as a LinuxJail from your viewpoint...but without isolation ;)

Linux (Wine):

Windows -> Linux SystemCall-translator

FreeBSD (Linxulator):

Linux -> FreeBSD SystemCall-translator

With more or less your Words:

>Well, there's one reason to use Windows (Wine), lol - to run all the apps on Linux that Linux doesn't support.

To be short, double lol

> I don't use a Slack Client but the Web one, and i don't use Spotify.

The comment you were replying to was talking in general terms so I thought you were as well.

> And to be clear a LinuxJail is actually NOT using the Linux Kernel but the GNU-Userland

Huh, that means even less stuff will run. I thought it was more than just an emulation layer.

> To be short, double lol

Not sure what this means. Long and short of it is though the original commenter was right IMO. The BSDs are generally the OS of hobbyists and sometimes contrarians, and it doesn't make sense these days when there are better options in most situations. If often makes as much sense as running Hurd or P9.

>I thought it was more than just an emulation layer.

Gosh i give up.....

>and it doesn't make sense these days when there are better options in most situations

Since you know best what makes sense (by not knowing what FreeBSD even can and the difference between emulation and translation) perfect self-report.

> Gosh i give up.....

OK? Someone said what about apps that only run on Linux, you mentioned you run them in a linux jail.

I like the BSD, mainly Net, but I don't try to force them to be a desktop OS when they are not suited for that, and so don't try running Linux stuff under them.

Besides, even it's it's an emulation layer, you're still basically running the linux userland, and we both know the FreeBSD kernel isn't significantly different or better for most desktop uses.

> Since you know best what makes sense (by not knowing what FreeBSD even can and the difference between emulation and translation) perfect self-report.

Oh, I'm perfectly familiar with FreeBSD. I just don't try and use it as a desktop OS for the sake of being different. You might like it just fine, and maybe you're doing some tinkering with something BSD adjacent so it makes sense to run it. But objectively for most users and cases, it's the inferior option for desktop computing.

You're here trying to argue it's as good or better than Linux, if I haven't misunderstood your position, and I don't think that's something you can support at all. You can give up if you like, but it won't be because I thought linux support on FreeBSD was more than an emulation layer, it will be because you can't support your position.

How's 802.11ax and 802.11av support? Bluetooth work fine? Realtime audio multiplexing? What if you want to run a newer game and the FreeBSD version of Wine hasn't backported enough support yet?

There's so many apps and hardware that won't work well on FreeBSD, or won't work at all, because making it work isn't a priority of the project. That's just a fact.

it boost rate of R&D and innovation in EU so statistics look better. It reminds me a story told by my professor - years ago (70s) he implemented accounting system for some company and later noticed a serious bug which made it unusable. It wasn't a problem because system was never deployed - it was needed so the (state owned, because it was Poland) company could report deployment of innovation.

Very cynical, but wrong in this case. The point of the sovereign tech fund[0] is to throw some (relatively small in the scheme of things) money at open source infrastructure that everyone relies on, no matter if that infrastructure itself is German, or even European. Basically, as their graphics on their homepage allude to, aiming to improve https://xkcd.com/2347/

IMO this is exactly the common-sense investments that every country should be making in their own self interest, and is one of the things that Germany is actually doing right.

[0]: https://www.sovereigntechfund.de/mission

Obviously the likes of Juniper, NetApp, Sony, Netflix, &c., know something you don't.

Also, FreeBSD is fairly heavily used in Germany.

> Obviously the likes of Juniper, NetApp, Sony, Netflix, &c., know something you don't.

Or there was no advantage and it was just the preference of the person designing the architecture.

I'm sure all those companies decided to base their flagship products on a particular (implied inferior) system all because of one dude's preference

Junior Developers: There's this guy in the cellar, his beard is 1 metre long, he hates the Sun but he likes FreeBSD.

Management: He must be right if he has such a long beard, let's take FreeBSD

Marketing: At least he is not eating from he's feet or wants to retroactively abort someone, let's go for it.

It's not that far fetched.

Someone is going to be the lead architect. If BSD works as well as Linux but the guy in charge has a preference for BSD, that's enough for them to go BSD.

I am sure companies design their entire OS based on that...

Aren't NetApp and Juniper slowly switching to Linux too?

The license could be one

Some (former) apple users in my small sphere that now also run FreeBSD mentioned feeling happy about helping to ensure there will always be well-tested BSD/MIT licensed code Apple then can incorporate/use.

Wait, they do it for Apple Corp??

> I don't even know a reason to use FreeBSD these days apart of being a FreeBSD enthusiast.

In addition to the technical reasons other people have given, just running on a slightly "exotic" platform is often sufficient to avoid most of the common exploits. Not because there are inherently less vulnerabilities on your platform, but because it's less likely to be an explicit target.

FreeBSD has good ZFS support, and nice package update cycle (quarterly + ongoing security updates), halfway between rolling release and point release, coppled with the stable "core" part of the system.

Hardware support is shitty, even worse than OpenBSD, and it tells a lot. 12th-gen Intel integrated video driver is scheduled to appear only in September 2024.

hardware support being bad is a feature.

remember BSDs are chosen by companies to whom linux is license-toxic. if they distribute hardware, like apple, gpl is poison.

this is a way to use your taxes to get free (as in beer, primarily) software they can ship on their products. less other companies can easily use that code the better.

This is BS, as FreeBSD use linux kernel modules for half of its hardware and, OpenBSD with a lot smaller team has better support for various devices.

It's not BS. They port a lot of modules but based on the needs of the developers, and often out of date.

Half is definitely an exaggeration for LinuxKPI modules that cover WiFi and GPU drivers.

Still, most of radeon/amdgpu DRM driver is MIT-licensed and doesn't require abiding by the GPL (not sure about other drivers as I don't use them, but I assume licenses vary).

This isn't generosity. They probably need it for a bank or something.

> Sovereign Transportation Fund to invest €686,400 into Highway Infrastructure

This isn't generosity. They probably need it for a Hospital or something.

It's not generosity, it's investment in infrastructure.

who cares if its generosity or not... it furthers the OS platform. Provided the money has no strings about the license...

Motivation can be important. After Mozilla started taking Google's money they lost a lot of urgency about competing with Chrome to offer a different view of the internet.

In this case there don't seem to be any concerning dynamics and generosity or lack thereof is irrelevant. It'll be good for FreeBSD.

How could Firefox as a company afford to compete with Chrome or do anything without any revenue? Unfortunately Google's money is almost the only reason why Firefox is still a thing...

It's better that it's not funded out of generosity.

Now the OS projects can have two revenue streams: one funded by generosity, one by self-interest.

Thank god. Generosity really doesn't align with any of my values. Where's the growth in that.

and your point?

What is generosity? Does the word even apply to organizations?

I’m not sure that needs specifically pointing out because nowhere in the title nor article does it call this a donation. It’s clearly described as an investment.

The end result is the same though: everyone who uses FreeBSD benefits from this investment.

Reply on news.ycombinator.com