/obsd/ - OpenBSD General

Ubuntu Gnome is better

The work on removing the kernel lock is making more progress! This time, they're removing the kernel lock during socket splicing:
marc.info/?l=openbsd-tech&m=156219064317315&w=2

Can you stop citing that stupid article? It's misleading clickbait.

why does openbsd uses ffs anyway?

Because it works

very nice pic OP, fukken saved

keep up the good threads and ignore the linux retards

Too bad I use OpenBSD GNOME.

>openbsd + openbox
perfection

Based and redpilled.

Like what? What uses does netbsd have?

>openbox
>not cwm
>on openbsd of all things

It supports more architectures than OpenBSD (even a fucking Dreamcast).

"supports"
yeah, try building a kernel on those platforms and see how many of them survive. every openbsd architecture is completely self-hosting.

Because it works for fuck's sake!

Help!
I can't install OpenBSD on my G4.
I get this error:
non-secure /pci@f2000000/mac-io@17/ata-4@1f000/disk@0:etc/boot.conf, will not proceed
booting /pci@f2000000/mac-io@17/ata-4@1f000/disk@0:/bsd.upgrade: /pci@f2000000/mac-io@17/ata-4@1f000/disk@0:/bsd.upgrade: Inappropriate file type or format
failed(12304). will try /bsd
Turning timeout off
What does it mean?
I found a similar report dating back from 2007!
openbsd-archive.7691.n7.nabble.com/Can-t-install-snapshots-2007-06-03-07-td208312.html
The difference is that I tried the original install iso. But then I tried the snapshot and didn't make a difference.
Any ideas?

Last thread:

Is anyone here excited about the upcoming GCC removal from the base system? That's going to be fucking awesome!
LLVM/Clang is the future.

Fuck GCC, I'm glad that this over-complicated non-modular decades-old crap is going.
Clang contains several cool stuff as well, like the static analyzer and the code refactorer, it truly feels like the compiler of the future.

total noob here, but how much openbsd knowledge is transferable to other unix-like systems in general?

It really depends on the "other Unix-like systems" that you mean.
If you mean another BSD or Solaris, then lots of knowledge is transferrable.
You still get the same UNIX shell tools (ls, grep, ps, cd, sh) in every Unix-like operating system, but they might have different parameters in rare cases (especially when you compare them to the GNU ones, since GNU likes to be contrarian a lot).
The init system should work similarly in all Unix-like systems (except Linux, where they mostly use some crappy non-standard init system called "systemd")
Overall, once you use a Unix-like operating system, it won't be hard to learn other Unix-like operating systems, since they all behave similarly.

Installed OpenBSD on my t420s and it’s heaven.

You tried to upgrade an existing install while there wasn’t one. Choose the correct installer option next time.

I haven't used macppc before but I don't think that's the problem.

I haven't used macppc before but I don't think that's the problem that he's facing.

They are all correct, except for the last two.

Also I don't know what do you mean with
> OpenBSD is single-threaded
Do you mean it doesn't do SMT (((hyperthreading_TM))) ? Because if that's the case they explained why they disabled it by default (of corse you can enable it with doas sysctl hw.smt=1 and you wont notice any difference except for specific case, in wich you wont use openbsd for other reasons

>They are all correct, except for the last two.
No they aren't all correct.
Here's his claims:
>OpenBSD does not have mandatory access controls
It has mandatory access controls, I responded to him yesterday () but he kept posting this despite being corrected, clearly says a lot about Linux shills.
>OpenBSD is single-threaded
False. Simply not true. Unless he's speaking about Hyperthreading, which can be enabled manually.
>OpenBSD has terrible virtualization
It has QEMU without acceleration or vmd with acceleration. vmd is not terrible by any means (unless you're looking for graphics without X forwarding, which is coming soon anyways). It even has QCOW2 support.
>OpenBSD is less based than Gentoo
The fuck is that even supposed to mean? Gentoo's portage was inspired directly from *BSD's ports system.
>OpenBSD’s best features are available elsewhere
This was also explained yesterday.

> has mandatory access controls, I responded to him yesterday
It doesn't have mac, pledge and unveil are not a mac mechanism.

>OpenBSD has terrible virtualization
It does by today standard. Don't be a mindless shill, please.

A lot, mostly. More so than GNU/Linux knowledge is transferable to others. This is because OpenBSD is more orthodox a Unix system. Of course, you'll always find differences.

>If you're using it in your home machine, then -current is most likely what you're looking for since it contains updated packages.
OP should kill himself.

Running on literal toasters

>It doesn't have mac, pledge and unveil are not a mac mechanism.
Unix's default permissions model, is a form of Discretionary Access Control, which is defined as:
>A means of restricting access to objects (e.g., files, data entities) based on the identity and need-to-know of subjects (e.g., users, processes) and/or groups to which the object belongs. The controls are discretionary in the sense that a subject with a certain access permission is capable of passing that permission (perhaps indirectly) on to any other subject (unless restrained by mandatory access control).
Source: csrc.nist.gov/glossary/term/discretionary-access-control
Mandatory Access Control, on the other hand, is defined as:
>An access control policy that is uniformly enforced across all subjects and objects within the boundary of an information system. A subject that has been granted access to information is constrained from doing any of the following: (i) passing the information to unauthorized subjects or objects; (ii) granting its privileges to other subjects; (iii) changing one or more security attributes on subjects, objects, the information system, or system components; (iv) choosing the security attributes to be associated with newly-created or modified objects; or (v) changing the rules governing access control. Organization-defined subjects may explicitly be granted organization-defined privileges (i.e., they are trusted subjects) such that they are not limited by some or all of the above constraints.
Source: csrc.nist.gov/glossary/term/mandatory-access-control
And Wikipedia says:
>mandatory access control (MAC) refers to a type of access control by which the operating system constrains the ability of a subject or initiator to access or generally perform some sort of operation on an object or target. In practice, a subject is usually a process or thread
By definition, OpenBSD's combination of pledge and unveil is a form of MAC.

That is a valid install option according to the official documentation.
openbsd.org/macppc.html
>If the system disk is shared with Mac OS, the ofwboot bootloader and the bsd.rd installation media can be copied to the first HFS or HFS+ partition. Then the installation procedure can be booted at the Open Firmware prompt, with
boot hd:,ofwboot bsd.rd
Supposedly, i should have been able to boot the installer and then indicate the directory of the bulk of the files (the thumb drive). I tried this route because there's no optical drive in the machine. As a last resort, I might get a Firewire DVD drive and try to boot the installer from there holding the C key at boot.

Nice. What DE/WM are you using?

>pledge and unveil
>has to literally modify the code and recompile the program
have fun getting that going on anything not written in C

btfo BSDshills

It works on any language that supports C bindings.

Unlike Loonix, where you literally have to write an extremely complex SELinux policy that can't even filter syscalls, OpenBSD's pledge (and unveil) can be simply implemented with one line of C code (or a C binding if you're using other compiled languages).
As for high-level languages that don't support C bindings, pledge is unnecessary because they're memory-safe anyways.
>inb4 AppArmor is easier than SELinux
Just like SELinux, AppArmor can't filter syscalls, and still requires a relatively complex policy in order to achieve anything useful with it.

Btw, you don't need to write anything yourself, the OpenBSD maintainers patch packages if they aren't pledged/unveiled upstream. Last time I used SELinux, I literally spent a few weeks learning it so that I can write some policies because most of my programs didn't work properly, even after writing the policies, I wasn't sure that I implemented them properly without leaving some hole somewhere because the whole concept is extremely complex, even Linus himself said that he disables it.

I was tired and made a mistake by calling pledge a syscall filter, it's actually more powerful than that:
man.openbsd.org/pledge

Most of the important packages have gotten pledge and unveil now, Firefox, Iridium, etc.

I wouldnt use Firefox with pledge
It does too many syscalls to everything to work properly
I remember a relevant openbsd mail thread about it too
Just use chromium or the like

This. Chromium is way more security-focused and it implements sandboxing better than Firefox.

Stop shitting this board and /fglt/ with your pro-BSD and anti-GNU spam. Yes, we know is YOU.

Attached: 1481032841105.jpg (1275x1200, 518K)

Linux is finished. GCC is finished. Emacs is cool though, and so is Icecat.

Hmmm... For some reason none of these mention BSD anywhere, yet Linux users (like you) try (and fail) to troll our threads all the time. It's almost like the exact opposite thing is happening.

Good thing you dont mention BSD much these days right? Stop spamming everywhere and fuck off.

>you
>implying that I'm the only one who talk about BSD on Jow Forums
This is a technology board, not a Linux board. You don't get to decide what gets posted here. So stop being a cultist child and deal with it.
You're an obvious troll, so I won't waste my time on replying any further.

Relevant discussion:
marc.info/?t=152475760400002&r=1&w=2

I really wish that OpenBSD adds ungoogled-chromium to ports, I hate Firefox and I hate Google and Iridium doesn't seem to be maintained upstream anymore.

Iridium is certainly maintained.

Their latest Windows builds on their official website dates back to 2019.04.73 (Chromium 73.0.3683.103):
iridiumbrowser.de/downloads/windows
The same release is the latest one in OpenBSD's ports as well.
It's unfortunate that Iridium died, it was a good project.

OP, how dare you disagree with the official documentation!
>New users should be running either -stable or -release.
openbsd.org/faq/faq5.html#Flavors

Attached: OpenBSD Puffy Apu.png (858x534, 82K)

>windows build
Kek, what’s this have to do with anything? So they haven’t had a release in 2 months, this is hardly a sign that a project is dead. They just released a new OS X build literally yesterday.

Not OP, but I'm the one who wrote this part of the post.
The reason I wrote that is because in order to get *package* updates, new users would have to either compile them from the ports tree or use a third-party repository like Mtier. (Do note though that the base system gets updated in -stable and -release through the syspatch(8) utility).
I didn't want to get into the hassle of instructing new users to add a third-party repository like Mtier or build their own packages, so I just instructed them to use -current. -current is very stable for home use and the OpenBSD developers do extensive tests before deploying code to -current, so rest assured, -current is suitable for home use.

iridiumbrowser.de/downloads/macos
Their MacOS and Linux builds are outdated as well.

Btw good job OP, you have been maintaining this thread really well. Thanks.

The most recent commit (not just release) was 2019.04.19.
git.iridiumbrowser.de/cgit.cgi/iridium-browser/log/

Did they all suddenly die or something?

I'm not privy to such information. Perhaps the CIA got them. Perhaps a rogue developer removed everyone's commit rights.

Can we contact the OpenBSD devs to add ungoogled-chromium and remove Iridium?

Adapt the iridium and chromium build scrips to port ungoogled-chromium to openbsd.

Recommend me a good WiFi usb stick please that works with OpenBaSeD.

Edimax EW-7811Un

Good. Another GNU trash out of the way.
Here's an example of their trash values: lwn.net/Articles/582241/
>clang vs free software
(they're making it seem as if Clang is a non-free compiler, in reality, it's freer than GCC and easier to understand as a result).
>The nonfree compilers that are now based on LLVM prove that I was right
(I'm curious about what "nonfree" compilers he's talking about).
>The existence of LLVM is a terrible setback for our community
>The only code that helps us and not our adversaries is copylefted code.
This guy is literally speaking like a cult leader, and the fact that he owns an organization that's called the (((Free))) Software Foundation is extremely concerning. He's literally saying that LLVM (an actually free compiler infrastructure) is a setback to his community (whatever that's supposed to mean). The president of the FSF is fighting Free software, how ironic.
It is only a matter of time before LLVM/Clang kills GCC, Stallman brought this on himself by deliberately instructing the developers of GCC to not make the code modular to "prevent
any parts of it from being used together with non-free software." (See: gcc.gnu.org/ml/gcc/2014-01/msg00209.html)
Even Rust chose LLVM instead of being a GCC target.

Not him, but are you sure that this card is supported? It doesn't seem to be listed here:
man.openbsd.org/?query=wireless&apropos=1
If it's listed, please mention its driver name.

It is kinda funny that ESR is the guy that started this, the same guy that knows nothing about writing C code:
catb.org/~esr/gif2png/NEWS
>Redirect segfault to a graceful exit. Tired of meaningless fuzzer bugs.
Yes. You read that correctly, he literally doesn't give a fuck about segmentation faults in his project.

urtwn

>They have been supported by Apple, the company which hates our freedom so much
Yes, Apple donated to LLVM, what's the problem?
It seems like Stallman forgot that the Linux foundation gets supported by hundreds to thousands of corporations, including freedom-hating corporations like Google.
His bias is so obvious and pathetic, he's really butthurt about LLVM being more successful than his trash, unmaintainable, spaghett-style compiler.

Computer security knower here. Consumers don't want security. It's inconvenient.

And...?

Thanks.

Eunuchs

Making a security-focused operating system seems like a weird thing.

Most of the security features in other operating systems (especially Windows and Linux) were made/implemented and enhanced by OpenBSD first.
And we don't really care about typical normal consumers, it's more of an operating system for tech/security enthusiasts and servers.

OpenBSD's not intended for general consumers; it's intended for OpenBSD devs and anyone else interested in it.

Hi bros! Void Linux user here (for about a year, after ~4 years of ubuntu/debian),
I'm currently learning C programming and some -serious- linux system administration
(aiming to finding work on that).
Since I want to migrate to OpenBSD on my main desktop, I want to ask for some experiences:
Do you have problems when you have to use Linux machines, if you aren't used to it?
Also is OpenBSD a good development environment for write portable code? What about shell scriting?
(I mean, how portable are ksh/sh scripts?)

Just a thought; I found this page:
blog.tintagel.pl/2017/06/09/openbsd-daily.html
It seems like a based goal to have and a great learning experience if is accomplished.
The commitment of OpenBSD's user base is what I like more. Is very unlikely to find
such passionated devs on the most boring open source project today :^)

>I'm currently learning C programming and some -serious- linux system administration
>(aiming to finding work on that).
Good for you, C programming gets you a lot of money if you work at the right place.
>Do you have problems when you have to use Linux machines, if you aren't used to it?
No, but that's because I used to use Linux before I switched.
>Also is OpenBSD a good development environment for write portable code? What about shell scriting?
OpenBSD is an excellent environment for portable code, unlike other platforms, they have a strict stand on following standards.
>(I mean, how portable are ksh/sh scripts?)
Very portable, more portable than Bash scripts.
>Just a thought; I found this page:
>blog.tintagel.pl/2017/06/09/openbsd-daily.html
>It seems like a based goal to have and a great learning experience if is accomplished.
Indeed it is, you'll learn a lot.

>The commitment of OpenBSD's user base is what I like more. Is very unlikely to find
>such passionated devs on the most boring open source project today :^)
I don't know about you but I find OpenBSD very interesting.

>inb4 GNUtards rushing to defend Stallman

Attached: 1483590349836.png (633x758, 34K)

tickless kernel when

>>OpenBSD has terrible virtualization
>It has QEMU without acceleration or vmd with acceleration. vmd is not terrible by any means (unless you're looking for graphics without X forwarding, which is coming soon anyways). It even has QCOW2 support.
OpenBSD _does_ have terrible virtualization but that’s on purpose. Virtualization is a huge security risk and the devs aren’t trying to kill their progress over trying to make something OpenBSD isn’t

we prefer iridium round these parts

>implying the purest form of love is not between a man and his boy
behold the (((catholics))) unironically running NSA code on their devices

>What about shell scriting?
(I mean, how portable are ksh/sh scripts?)
Object oriented ksh scripts aren’t portable but I’d probably know you by name if you seriously used those

No, it doesn't have terrible virtualization, and the OpenBSD devs are taking their time to implement it properly, which is why things like graphics are taking time.
But I agree with you in that virtualization can be a huge security risk
Many people (especially Linux users who come around here bragging about their shiny new distro that virtualizes everything and comparing it to OpenBSD) don't realize that virtualization is an extremely complicated subject, and many bugs have been discovered in the past where malware was able to execute code in the host operating system. Here's a Theo quote about this:
>x86 virtualization is about basically placing another nearly full kernel, full of new bugs, on top of a nasty x86 architecture which barely has correct page protection. Then running your operating system on the other side of this brand new pile of shit.
>You are absolutely deluded, if not stupid, if you think that a worldwide collection of software engineers who can't write operating systems or applications without security holes, can then turn around and suddenly write virtualization layers without security holes.
marc.info/?l=openbsd-misc&m=119318909016582

There isn't any work on this currently, but I'm pretty sure that they'll get around it eventually once they finish removing the kernel lock (they're working hard on this one).

Holy shit! The total amount of donations from small donors (i.e. donations made by individuals) have just reached Platinum ($50,000 to $100,000):
openbsdfoundation.org/contributors.html

I'm excited to see what they're going to fund this year, one of the biggest things that they funded last year was the development of AMDGPU/Vega support, and it worked out great!

As mentioned earlier, Iridium is a dead project.

It has the same virtualization capabilities of 9front wich is absurd considering the different amount of manpower.
Unveil it's a reimplementation of what Plan 9 does natively (and arguably better)

Quarterly releases aren’t even slow in the open source or proprietary software world.

So what will you say when they release another version soon? Track the release history yourself.