Hacker News Reader

Comment by mikorym | Hacker News Reader

Is there any more context to OP's post? I guess someone just noticed in the mailing list, but it's not clear whether maybe John Carmark does this with all code everywhere in the universe. And then probably still has some interesting coffee break conversations...

On a different note. I am something of a Ubuntu fan and try to be more involved there. But then again the xBSDs where x = {Free, Open, Net, ...} are one of those things that always prods the back of your mind. It's a bit like topics in mathematics that always catch my attention, only to be forsaken for more mundane tasks. Then again, mathematics is a bit like the art of making everything mundane.

View on HN

>Is there any more context to OP's post? I guess someone just noticed in the mailing list, but it's not clear whether maybe John Carmark does this with all code everywhere in the universe.

I remember a story about Carmack going off the grid with a laptop and OpenBSD for a couple of weeks, and him being delighted about their documentation and the ability to build things without access to the web.

Carmack is a hero in the C community, and OpenBSD is also all about C and classic Unix, so he might just feel at home there.

> whether maybe John Carmark does this with all code everywhere in the universe.

A lot of people rename things and add comments as part of understanding a new code base. I do and I'm sure I'm not the only one on HN.

I don't see any additional context but this does appear to be a new thing. Two commits credited to Carmack in the OpenBSD source and they are dated 5/16/2020 and 5/17/2020.

where are those commits? I didn't see anything credited to carmack on those dates. Must've missed them. Would be curious to see!

I found these two commits to the stc/sys/dev/ic/vga.c crediting Carmack

https://cvsweb.openbsd.org/src/sys/dev/ic/vga.c?rev=1.70&con... https://cvsweb.openbsd.org/src/sys/dev/ic/vga.c?rev=1.71&con...

To me this kind of stuff is fan junk news. I get that even hackers need celebrities, but front page of HN is a little extreme for a thing like this without context. If there was more context like, John C has re-written half of BSD, then we're talking - but this is very celebrity grocery list to me.

I'd bet everyone on this thread will be grabbing their cheerleading squad and their pom-poms everywhere if John Carmack visited them in person. They won't be able to contain themselves and start bursting into tears of joy.

'Extreme' maybe an understatement here but it seems like HNers here love seeing the second coming of King Midas when everything he touches turns into gold.

I’m sorry you feel this way, maybe if you could try contributing more positively to FOSS in some way and receive some accolades for yourself, you would not feel so spiteful about the attention Carmack receives.

I think you will find Carmack has contributed more to the hacker ecosystem than anyone else if you just take a very shallow peek.

I would suggest to start by reading the source code to Doom, it is surprisingly approachable and educational in several fields of math, geometry, data structures, and computer science. There are even books to help you understand it, heck, there are a dozen or more books about Id software and John Carmack, so there are many authors and readers who feel very differently from you, maybe you could consider the reason. There are many things in our field that would have happened eventually, no matter who did it, but much of Carmack’s contributions really would not have happened without him, he really is in a league of his own.

It’s really amazing to me that you feel so negatively towards people who have endearment for one of the hacker community’s greatest contributors. I can only guess that you might be very unhappy, for that I’m very sorry and I hope it turns around for you very soon.

I get what you're saying, but I think it's great for people to see that an old-school hard core programmer puts effort into improving documentation and names. I've worked with a lot of people who think programmers can and should be too good to invest time and thought into changes like that.

Also to see that he would be so tactful about asking if the changes are welcome, even though he is who he is and knows it. I like to think there are people in their twenties or teens or even younger browsing HN and subconsciously absorbing the fact that this is how a programming god behaves.

So many celebrities are celebrities for banal, superficial reasons.

Calling this a "very celebrity grocery list" story is a disservice.

Carmack's one of the most accomplished programmers in history in terms of innovation, skill, and longevity. His words and opinions also carry a fair bit of weight, even today. His contributions and technology choices matter in a way that a celebrity's grocery list do not.

If there's an injustice here, it's the fact that there are many Carmack-level coders out there, toiling away in relative obscurity, who don't have their contributions trumpeted in this way.

I would suggest to just install OpenBSD and try it (maybe in a VM?), one of the selling points is the simplicity of the setup process. It can be a one evening project.

I recently went and installed FreeBSD on my laptop. It has until now been a pretty good experiance. The main reason I went with FreeBSD over OpenBSD at the moment was that NVIDIA only supports FreeBSD. FreeBSD is also pretty simple (at least compared to linux) so it wins there as well.

By NVIDIA supporting FreeBSD you mean that half-assed thing with binary drivers allowing you to get a picture, but no CUDA?

Or is that old information which i should forget?

Sure, the factual parts of that are accurate. Your opinion colors some of the adjectives. OpenGL works well, for example. No CUDA, no NVENC/DEC.

Well it is a bit of a grey zone but I have gotten Vulkan to work and I think at least NVENC works as well. The solution is rather hacky though. It extracts libraries from the Linux driver and loads them when you run programs. A look at the code make it seems like cuda is getting loaded as well so it may work.

https://github.com/shkhln/nvshim

While i respectfully congratulate the authors of that shim for their ingenuity, that is no support.

Nothing drastically changed since Torvalds had sended his gratitude[0].

[0] https://youtu.be/MShbP3OpASA?t=2997 (time coded)

I concur.

Wayland support fiasco, changes in driver's EULA to deny the usage of consumer-grade GPUs in VMs are 2 things off the top of my head.

There was actually some period of time when the Tegra division (I think) contributed to Nouveau a bit (!) but that didn't really grow into anything good.

> changes in driver's EULA to deny the usage of consumer-grade GPUs in VMs

Which changes? Where? I'm looking at said EULA right now and I don't see anything related.

My bad, it was a silently added "feature" in Windows driver[0].

[0] https://wiki.archlinux.org/index.php/PCI_passthrough_via_OVM...

Nvshim is an amazing project, but it is a far cry from official Nvidia support for CUDA/NVENC/NVDEC.

(All Nvidia would have to do is build another binary; it's still chatting with the same driver blob internally so it's not really clear to me why they don't bother compiling it for FreeBSD.)

They could even do it with less churn, because of https://github.com/freebsd/freebsd/blob/master/sys/conf/opti... FreeBSD easily achieves binary compatibility with earlier versions. So no need to mess around with DKMS or their own thing. Or at least less so.

> FreeBSD easily achieves binary compatibility with earlier versions. So no need to mess around with DKMS

Binary compatibility is for userspace programs, not kernel modules.

GP's link is to the COMPAT_FREEBSD binary compatibility for earlier versions of FreeBSD, oddly. I don't think it (or linux emulation) has anything to do with Nvidia's needs re: CUDA/NVENC.

It means you have a module from some vendor, or something similar to the NVIDIA installer for linux, which technically is nothing else than a batch compiled makefile, spitting out a kernel-module made for version 1.2.3 while you are running 2.4.5. It worked. I don't know why it shouldn't now.

Edit: I'm not having any FreeBSD systems in use right now, so i don't know if they changed the defaults. What i remember is at least the last two versions were compiled into GENERIC, so you didn't even had to compile your own kernel when that range backwards was enough.

Similar thing in NetBSD, a loooong time ago, even their own driver for 3Com905 relied on the equivalent of that functionality. That caught me by surprise, because i used NetBSD like Gentoo, found some Compat foobarjurassicBSD, wondered why i should use that at all, and had no working net. Fun! :-)

That's not how the nvidia binary blob works in FreeBSD. Instead, there is some compiled-from-source-code OS portability glue, which is compiled against the target version of the FreeBSD kernel. Then the binary blob links the portability glue.

I believe something similar is done in Linux. The difference in experience may come from FreeBSD maintaining a stable KBI over a release version (e.g., stable/12), while Linux aggressively does not maintain any KBI stability; hence the need for DKMS/akmods.

This is unrelated to COMPAT_FREEBSD.

Sigh. Been typing too fast as usual. I know thats not the way the Nvidia-thing works, instead doing the glue-thing which interfaces to the binary blob like you mentioned. But it could (instead of that)! That's what i meant to say.

Ahem. Tell that to all the binary drivers for network cards and raid controllers which relied on that functionality, which i happened to use not all that long ago.

If anyone has any clout to get this to happen, John Carmack perhaps would.

Unfortunately, half-assed vendor support is still quite compelling compared to no vendor support.

It depends on your use-case, i guess. If my intent would be using FreeBSD as my daily-driver "desktop/laptop-OS" i'd avoid NVIDIA like the plague, becaue without CUDA their unique selling point does not exist, but the hassle of some "switcheroo-black-magic" possibly persists. Which i wouldn't have with an intel-gpu, also less power draw. Even AMD is often supported in a better way, be it opensourced, or relying on their binary blob.

> Even AMD is often supported in a better way, be it opensourced, or relying on their binary blob.

AMD binary blob? You mean that amdgpu-pro thing? That doesn't exist on FreeBSD.

Maybe not the latest and greatest "Pro", but otherwise?

https://wiki.freebsd.org/Graphics/AMD-GPU-Matrix

edit: maybe should have added https://www.x.org/wiki/RadeonFeature/#index5h2 too

A selling point in comparison to what? A "one evening project" installation doesn't seem to be all that amazing by any standard, in my opinion.

I think that the parent is saying that "trying OpenBSD" is a "one evening project". The installation should be complete within just a couple minutes.

I don't see how one could try OpenBSD, and derive any conclusions other than, "oh great, yet another Unix", in a single evening.

To get it working at all, you have to find the manual.

Once you find the manual, you’ll realize it’s well-documented and simple, and that simply using it has great educational value.

If those things don’t appeal to you, it’s probably not the OS for you. It takes about an evening to figure this out, in my experience.

But those things apply to Linux distributions too, e.g., Debian.

Having experienced both operating systems, Debian gets nowhere close. The Debian wiki is full of outdated articles.

OpenBSD's manual pages and online FAQ combined have never failed to answer a question I've had about the base system. It's really that good.

The ArchWiki is close behind in usefulness, but that's maintained by its community and article quality varies.

The Arch wiki is great. Reminds me of the old Gentoo wiki from way back when Gentoo was the new hotness. I'm not even an Arch user but documentation as to how things like Window Managers work is fantastic.

In my experience, the first step in trying out a new system is installing it, configuring everything (from network connectivity to various peripherals), installing your favorite software and so on. This gives a good idea on how to work with the system, how good the documentation is (as another poster already mentioned), how simple and consistent everything is.

Installing in a VM is not the best differentiation, since in many OSs these experience was streamlined and everything usually works out of the box, but the customization part is still a good example.

Of course in the end, you might get to a setup similar to other OSs - running a browser in a windows manager and a buch of other common tools (IDEs, editors, productivity apps), but the way you get there, and the amount of problems you encounter along the way is a major part of using an OS.

I think you can get a good impression in a few hours, and see how it's different. (For example, I had a similar experience with NixOS: In a few hours you see how radically different it is and learn how it fits you)

> maybe in a VM?

But that's the thing. The real benefits from OpenBSD come when it's the only system you need. Who cares how simple and elegant the guest system is, if you need to maintain a linux installation to host it?

A project on the scale of OpenBSD can't possibly support all the hardware that's out there. And the system is under active development, so there are constant regressions in hardware support. That's fine for John Carmack: he can fire up ed and fix a kernel panic, that's how he winds down after work. But the vision that attracts me to OpenBSD is a box in a closet that Google doesn't have the key to, which sits there for 6 months receiving my email and serving my homepage, takes 10 minutes to update, then sits there for the next 6 months; maybe once every 5 years I'd need to read a manpage and port my smtpd.conf to a new syntax. Hardware regressions are what stop that from happening.

The OpenBSD developers are getting what they want, and good luck to them. The shame is that a little bit of coordination, where they chose one model of ThinkPad every other year and committed to supporting it for a decade, would make the system so much more useful for so many other people.

> box in a closet

https://github.com/elad/openbsd-apu2

Mine is 4(?) years old, and has never crashed. I use it every day. If you want something newer, or have 10x the power budget:

https://flak.tedunangst.com/post/ryzen-build-for-openbsd

Yeah it generally works pretty happily with most hardware in modern X/T Thinkpads. Happily running it on my X270. It's missing the 802.11ac and Bluetooth support but I'll live. I bought a BT-W2 so that my Bluetooth headphones will work and OpenBSD handles it swimmingly.

There's very little that I can't do with it versus my Linux workstation, at least for my needs.

Thanks. I might try building one of those.

But the problem remains: eventually, the next upgrade will crash every time you boot it, and what will you do then?

> But that's the thing. The real benefits from OpenBSD come when it's the only system you need. Who cares how simple and elegant the guest system is, if you need to maintain a linux installation to host it?

I don't think he's suggesting that as the long term way to use it but as an introduction to the system.

> takes 10 min to update

If only that were true. And they do releases every six months. And you are not supposed to skip one, or you’re on your own.

BSDs are mundane and also stable, not like some rolling linux distros

The bsds are actually less stable than Linux.

Stable meaning changes, in this context, right? freeBSD changes ABI compatibility with every version.

If you mean that they have old software (and old==stable if you think Debian) then that’s because of BSD licensing vs GPL. GPL can take from BSD but the other way around is not possible.

Stability of uptime? Yeah, they’re rock solid.

For me, stability means uptime and infrequency of application crashes. Unfortunately, OpenBSD didn't meet that requirement on my (standard desktop) machine--it would frequently kernel panic on resume.

On the other hand, I found that OpenBSD has absolutely outstanding reliability and stability on hardware like my ThinkPad X220.

> The bsds are actually less stable than Linux.

> Stable meaning changes, in this context, right? freeBSD changes ABI compatibility with every version.

Every major version, right? From becoming -Stable to ending its support lifecycle, I'd think a FreeBSD version and its ABIs are supported as long as many Linux distro LTS releases.

The major difference being that if you have a statically compiled binary from linux 2.4 it will work on linux 5.5

The only difference will be any shared libraries and potentially the libc. But if we're speaking from a pure technical perspective, the libc is not linux.

Binaries from all the way back from FreeBSD 4.0 work on -CURRENT just fine.

And I seem to remember that FreeBSD can support linux binaries, too.

I know that they change (well, reserve the right to change) syscalls with every major version. But that's not a big deal, since only the base system is supposed to be using syscalls directly; everything else goes through libc.

So, are you saying that they break libc ABI?

NetBSD preserves backwards compatibility for its ABI.

How about stability of user interface?

More of what I learned using Linux 0.99 applies to present-day OpenBSD than present-day Linux, and I can’t think of any breaking changes specific to the Linux side that weren’t clear regressions (except apt and dpkg. Those rock.)

Apples and oranges. FreeBSD is providing the userland too so it’s not like the ABI change is going to break all the things it might in a Linux distro if the kernel broke ABI.

For me *BSD is like Plan9 or ZX Spectrum emulator … guilty pleasure with absolutely no practical use whatsoever.

You know, other than helping run the internet. They make great servers, so your viewpoint is a little confusing to me

Seems a little dismissive of BSD systems given how widely deployed the are in comparison to plan9 or zx spectrums. As far as I understand Netflix uses freebsd fairly extensively, and there are powerful tools like pfSense built on BSDs too, which are also used fairly extensively as far as i know.

I would describe that as having practical use.

I think there's a difference between using BSD to build an appliance and running BSD as a day-to-day system.

pfSense is a wonderful firewall appliance. FreeNAS is a wonderful storage appliance. If you need pf or native ZFS, a BSD is probably your best way to get it.

If you want to run a random headless server to mess around with, you are probably going to have a harder time with any of the BSDs than you would with even bleeding-edge Linux distros. If you want to run a graphical desktop with 3D acceleration and HD video you will almost certainly have a harder time than any Linux user.

---

There are, of course, people who are happily using BSD-based desktops right now and I am certainly not denying them anything, but even they'd have to admit they would have been able to get most of the same experience with a lot less effort on a modern Linux.

> If you want to run a graphical desktop with 3D acceleration and HD video you will almost certainly have a harder time than any Linux user.

With the exception of MacOS? Darwin is BSD.

Just because OSX has a very outdated BSD-derived userland does not make it a BSD. The OSX kernel a mach microkernel and shares no code from the BSD projects. This is a farcry from the *BSDs which can directly trace all of their code and history all the way back to 4.4BSD

> If you want to run a random headless server to mess around with, you are probably going to have a harder time with any of the BSDs than you would with even bleeding-edge Linux distros

Actually, I became a BSD user after just trying OpenBSD out of curiosity on my headless server after years of running Linux. The experience was so damn good.

> There are, of course, people who are happily using BSD-based desktops right now and I am certainly not denying them anything, but even they'd have to admit they would have been able to get most of the same experience with a lot less effort on a modern Linux.

It's the opposite for me. Modern Linux distros are a pain; I can't find one I really like. Whenever I ask people to recommend a distro that feels like OpenBSD, I get recommendations from people who obviously haven't used OpenBSD in two decades. Maybe they tried FreeBSD two decades ago, and then figured that some old Linux distro was kinda similar..

Frankly I don't think the Linux way of "ducktape a bunch of third party packages together" will ever produce a distro that comes close to being as pleasant as OpenBSD. At least they'd have to heavily patch things and diverge from upstream.

> guilty pleasure with absolutely no practical use whatsoever.

Are you sure you are talking about the right thing here?

I think you're confusing BSD with posting on HN.

Reply on news.ycombinator.com