There seems to be increased interest in OpenBSD given the recent CoC added to Linux and FreeBSD already having one.
In /fog/ we discuss all things OpenBSD. Questions, ideas, experiences related to OpenBSD belong here.
>Why OpenBSD? openbsd.org/goals.html Most notably: >"be as politics-free as possible; solutions should be decided on the basis of technical merit." >"Pay attention to security problems and fix them before anyone else does. (Try to be the #1 most secure operating system.)" openbsd.org/innovations.html
>How do I install OpenBSD? Until you are familiar with the system, be sure to try it in a VM first. openbsd.org/faq/faq4.html
>Tutorials? OpenBSD doesn't teach through tutorials. Think of the cliche old saying "Teach a man to fish..." Instead it opts to keep man pages of a very high standard. The man pages document things such as, system calls, general commands, library functions, the system manager's manual, and others. Unlike other systems, telling someone to read the manual isn't necessarily an insult because of the high quality of documentation. openbsd.org/faq/faq1.html#ManPages man.openbsd.org/
>What do I need to know coming from GNU/Linux 1. Services, basic system administration, etc. is different (ex: there isn't a big tool like systemd to handle a lot of things for you) 2. There are fewer tutorials online for how to do various things 3. Some standard commands like grep and make don't work the way you thought they would (they have been replaced with BSD licensed alternatives). GNU's make can still be used by running gmake. 4. There is no /proc/ virtual file system for viewing/setting properties of hardware, etc. This is a Linux specific thing. 5. Some commands just don't exist that you may use to understand your hardware layout. Things such as "lspci", "lsusb", ...
I hope this becomes an actual thing cause openbsd is pretty good
Michael Gray
I failed to install i3wm. I did everything, downloaded it, but it didn't wanted to run. I hope you keep these threads running so I will post error and probably maybe can help me...
Kayden Baker
JUST FUCKING PORT FSTRIM AND DMCRYPT
IT'S NOT THAT FUCKING HARD. I WANT TO USE YOUR FUCKING OPERATING SYSTEM JESUS
Easton Morales
you couldn't cd /usr/ports/X11/i3 and run make?
Samuel Watson
I'm thinking of building a NAS with OpenBSD and use sftp or sshfs. What do you guys think about the encryption overhead, is it noticeable?
Sebastian Morris
I will take a look on it when I will have time, I hope these threads will be posted here regularly.
Alexander Collins
i3 is available as a package, friends. Simply pkg_add i3 i3lock i3status. dwm is also available, if you prefer that.
James Parker
Just make sure that /etc/installurl exists and point to a valid mirror, then pkg_add i3 (pkg_info -Q i3 if you're unsure about tho package name) and edit your ~/.xsession (xenodm - the default x login - source that instead of ~/.xinitrc, never understood that desu)
Julian Roberts
I did that but then I tried to run it via exec/start i3, of course I couldn't because X was running so I tried to set it via xinitrc or xsession, honestly I fucked up something 100%. In other words it didn't worked straight after installation, when I tried to start it, it failed with some error which I now don't remember.
Maybe this is the part I screwed up. Thank you so far.
Nathaniel Evans
It has a cleaner, more consistent, more orthogonal, more secure design than Linux, but it's practically unusable as a desktop OS when it comes to driver availability and third-party-software conveniences.
Gabriel Powell
A few helpful things I learned as a newbie: 1) if after install you did not have internet when you first booted, run fw_update to automatically download any missing drivers for detected hardware. You can do this any time you have new hardware you think you may need the drivers for. 2) apply all available patches using syspatch 3) pkg_add -uvi to update all packages with verbose output and require confirmation 4) even if you had internet during install, unless you installed from http instead of a local disk, you won't have /etc/installurl set for installing packages. Be sure to read openbsd.org/faq/faq15.html
Evan Gomez
can't you achieve something like dmcrypt with softraid
Jace Jenkins
I have more than 20 TB of existing dmcrypt'd data that I'd like to be able to mount and use if I'm going to use openbsd
Hudson Morgan
My friend, this is a problem with every OS. CuckBSD uses its own geli, OpenBSD uses its softraid crypto policy, NetBSD i don't know what, DragonFlyBSD uses dm-crypt but it cannot mount ext4 AFAIK, and Linux has a bazillion of file system on top of dm-crypt. A cross-platform encryption and a non-joke fs (fat32 is a joke) would be indeed wonderful.
Samuel White
Looking through the mailing lists for openbsd, I get the impression that they don't like the design and quality of LUKS. There was a patch made for LUKS support in 2015, but it wasn't accepted.
Henry Turner
>just started learning linux >it gets cucked >now I have to learn openBSD >has even less software compatability
Ethan Lopez
>tfw own a OpenBSD shirt I hope I get interviewed at uni when I wear it so I can meme on that fat Arch dude
Since the apparent takeover of the Linux project by trannies, there has been a lot of talk about moving to OpenBSD. One of the criticisms of this OS is that its filesystem does nothing to protect against bitrot and data corruption in general. OpenBSD fans have responded to this by claiming that storage device makers are to be blamed for failures. Others have suggested that it is a result of 'bullshit writes' from large and bloated programs such as browsers. To be fair, I agree that modern browsers are shit, but I've been noticing this as a trend from OpenBSDfags on here. Shifting the blame from the OS to someone else. It's hard drive manufacturers, and if it's not them, it's browser devs. Pointing fingers doesn't solve problems. Actions do. What can hard drive manufacturers do to make their hardware failure-proof? Is that even possible with today's technology? No manufacturer has done it in the history of these computer components. What evidence makes you think they can do it now? What can browser developers do to fix their software? If they do not make their browsers as bloated as they are, 90% of the web will stop working, and that would prevent many people from doing what they want/need to do, since everything is done on the web. Perhaps there is room for a discussion on how the bloat got this bad and how to reverse it, but as it stands, the WWW won't be changing any time soon, and because of that, browsers can't change any time soon. So it is clear that regardless of who should be 'rightfully' responsible for the issue of bits being flipped, there is only one party that can do anything about it, and that is the OS developers. Considering the situation that is going on now, it is my hope that OpenBSD sees an influx of new devs that can bring it into the 21st century, although it will require a re-evaluation of how people think about the problems that face it.
>Probably crap as a server too, but I haven't used it like that. >Opinion discarded I was using openbsd before Linux got cucked. As for the others, openbsd doesn't discriminate about why someone is using their os. They don't care. The fact that you do says a lot about you. While I agree, I don't think it is likely to happen. What they have works so they'll probably keep it instead of increasing the size of the kernel. Remember that they have to audit all of the three code in their tree.
nice pasta, I haven't lost a file in ~10 years of OBSD use
the filesystem doesn't exactly impress me, but it's rock solid
DESU i switched my laptop over to debian because I like booting it fast, not rebooting to upgrade from snapshots, and the features that relatively poor security enables
my servers are all OBSD because debian is a bloated corpse to configure and manage, yet still doesn't include basic utils like ssh or tmux by default
>no non-web discord I'm sure ripcord could be ported
Sebastian Johnson
i've seen there is a openbsd version of gentoo how good is it?
Brody Allen
it's dead
i think the entire idea was pretty much porting portage to openbsd
Adrian Martinez
i hope maybe because of the linux kernel shitshow the idea would be brought back up again
Kayden Sanchez
I'm having dns problems. If i ping a site (like google.com), it says there are no addresses associated with the name. if i run host google.com it says not found 4(notimp). netstat -r takes a longass time to get past internet6. i have zero idea about networking so pls help
it's configured with dhcp. thought it would work. guess not. i have no idea where to go from there though.
Joseph Baker
this gentoo/openBSD would be a fucking GODSEND right now
William Watson
If your using dhcp, it should just be using the settings your dhcp server says to use for DNS. Are those settings wrong? You could try to tell openbsd override that setting and use a different DNS server (just use 1.1.1.1 by cloudflare to test) >openbsd.org/faq/faq6.html#DHCP
Robert Davis
I need to go to sleep so hopefully someone else can help.
If you've fucked up something, re-start clean. Detach your ethernet (I'm going to assume you're on ethernet), as root `ifconfig re0 delete` (replace re0 with your actual interface). Then attach your ethernet and `dhclient re0`(again, replace the interface name). If you haven't messed and fucked shit up with unbound you should get a ip and use your router dns. Otherwise setup /etc/resolve.conf to your likings. The next step is 'man hostname.if'
Julian Martin
Nah I just can't get it to work. Adding exec i3 in .xinitrc or .xsession won't even let me to log in. .xinitrc is full of some bullshit which I don't understand which is needed for running openBSD and which isn't. Something is blocking i3wm from running but I don't know what. When I try to execute it from console it tells me as always, which I know, that other wm is running.
Eli Ward
When is trim going to work?
Cameron Edwards
look at the bottom where it says "start some nice programs"
you should also have "exec startx" in the last line of your .profile
Justin Ortiz
>barely-functional it works better than any other dude
Chase Kelly
I followed this and it's working. openbsd.org/faq/faq11.html Maybe you forgot to enable the xenodm display manager? I only have "i3" in my ~/.xsession, no exec.
Luke Nguyen
never, its a useless feature that was only required for 1st gen ssds
William Reed
now it gives me fatal error no screens found?
Blake Edwards
I think you messed with too much stuff. You only need to rcctl enable xenodm, and put "i3" in ~/.xsession. I didn't touch .profile and don't have a ~/.xinitrc
Ayden Lopez
How is battery life on laptops?
Isaac Moore
It's possible. I will try xfce for now even though I'm used to i3. Shame I always fuck it up somehow, I will try fresh install later and what you did.
okay I got it working after all, thanks for patience with me and advice.
Logan Long
>Increased interest in OpenBSD naw
Luke Wood
Fuck you guys for suggesting this to me. Really.
>Installing via USB: it fails to mount the USB after booting >Installing via PXE: It boots but then can't get a lease from DHCP, can't do it with a static IP eiter >Trying to get basefiles from a USB, it says USB port is bad and just turns it off
If that's what non-SJW OS is I'm gonna transition and spread my legs for cultural enrichment today. Better dilate than deal with that shit.
Really cool OS with highly skilled dev working on it.
Don't know what's going on. Installed multiple times without troubles. You have to basically keep pressing enter 'till the end of the installer. If you want to install the sets from the USB instead of downloading them just answer "no" when it asks you if the USB is already mounted. And answer yes if it'll asks to to continue if the SHA256s are missing. Remember that, at any time, if you type ! and press enter the installer will drop you in a shell where you can inspect things :)
Ryder Long
Well I've been doing that, I'm not retarded. It just fails to use any of my USB ports or NICs on a hardware level, and I left with no options. No matter what USB device I'm trying to connect it just switches off the USB port, and after the network boot it can't be installed since I have no way of getting basefiles.
Charles Thomas
>no discord I'm sold.
>Just use raid if you're worried about bitrot. As far as I know, no soft or hard RAID controllers read more than one mirror to ensure consistency.
In my opinion, it's silly to expect black box software or hardware to not fuck up. The sane solution, OpenBSD or not, in my opinion, is to 1) checksum your data yourself, 2) have offline backups, 3) check your data and your backups yourself.
David Jenkins
Can you give specific errors (dmesg) and info about your hardware? Sounds like it might be a bug. I suppose you have tried a VM first and things worked well then?
Jeremiah Hughes
>As far as I know, no soft or hard RAID controllers read more than one mirror to ensure consistency. by which I meant RAID doesn't protect you from bitrot, just from device failure, if I wasn't being clear
Jaxson Jenkins
i don't want to go back to non-configurable binary packages
>hardware support Try it and see! >bugs You tell me!
Camden Watson
>not WINE-compatible sorry, not for me
Jose Green
>Raid doesn't protect from bitrot False, here's a good explanation from Reddit I found while searching (inb4 Reddit) >When anything is read or written, it's checked against the hash. So if bitrot occurs on a disk, when the relevant segment of data is read from again, it is compared for consistency across disks and hashes before being passed to the OS. If anything doesn't match up, the data is re-read from the disk and the most consistent data is returned.
>This also happens to be the reason why "bad" disks click back and forth when trying to read unrecoverable or damaged sectors. They re-seek out to the location from the head rest position (where it realigns), to try to find the data, and the drive keeps re-trying to read the sector, each time getting a CRC error, resulting in a re-try. High numbers of retries can also account for slowness in aging machines. This is why SMART monitors the relocated sectors count. More relocated sectors = more bitrot.
>Most of the time, a drive that's consistently used and checked (using even something as routine as checkdisk in windows), will be able to predict when a sector is becoming inconsistent, recover the data, relocate it, and flag the affected sector as bad. It usually happens completely silently, without the need for user intervention.
Continued next post
Gavin Young
>For RAID, they not only have the CRC that's validating each block on-disk, they have the data from the other drives, plus (in the case of RAID Z2 or RAID 6) two hashes with which to compare/detect, and if required, rebuild inconsistent data. This happens entirely transparently to the user, and it's the reason why "RAID 6 is slow". The controller is constantly hashing the data to ensure consistency at the cost of speed; it's a trade-off, you sacrifice some speed, and in return you get data consistency. In the case of RAID 1, the data is read off both disks, and compared, rather than hashed, which makes it "faster". in RAID 0, there is no hashing or comparison, which contributes to its speed, but at the cost of consistency (RAID 0 is no more consistent than a bare disk).
>This is just block-level consistency. Then you have file-system consistency on top of most RAIDs; RAID 6 just gives a raw device to the system. That raw device then gets a file system, which usually has it's own consistency mechanism.
>And everything is built this way, once you get past Authentication and Authorization, you look at all different methods of Accounting for the fact that the data has not changed. IP uses FCS or Frame Check Sequence, a hash of the data to ensure it arrived without being corrupted, and the FCS is on every layer of the protocol data units. Bad data is thrown out.
>DESU: even if the data, on-disk is inconsistent in a RAID 6, it will be picked up and fixed the next time you access the data, so I'm not sure why it would be relevant to spend time trying to check the data for any inconsistencies.
Wyatt Baker
While is true for FreeBSD's ZFS which in turn has a software RAID subsystem (this is what the discussion you found was about), no hardware RAID controller or straight soft RAID implementation I've ever come across had checksumming. Some RAID setups have parity but this really is a parity (i.e. to survive the loss of one disk) and not a checksum.
If you're relying on non-ZFS RAID to harden any critical data against bitrot, I would recommend that you start looking for a (real) solution ASAP.
Gabriel Rogers
Anyone know if Coffee Lake integrated graphics acceleration is supported? Specifically those chips use Intel Integrated Graphics 630 chips
Kayden Garcia
>If you're relying on non-ZFS RAID to harden any critical data against bitrot, I'm not. And I'd make sure it works before use is it in production
>rights and freedom guaranteed *as the FSF considers rights and freedoms* >amendments to the document re- its bureaucracy. >freedom from slavery >freedom from treachery Only literal, unironical commies believe that user All in all, it's different opinions. If you think the BSD license is enslaving you because companies *will take your code and fork it and make proprietary changes to it and not give it back* (which they do), simply don't use openBSD
Joshua Rogers
why the fuck are there so many goddamn trannies in free software. I've never even seen one irl but it looks like every single free software project is infested with these mental cases
Christopher Morgan
Not every tranny has an agenda. I don't like to look at their faces either, and don't even want to imagine what's between their legs, but I'm not about to sperg out over their life choices as long as they leave me in peace too
Isaac Adams
This. BSD devs know companies just take and modify the source. They like contributions and donations but in no way expect it.
