Does anybody use syncoid on freebsd? Is there any way to use it without changing root's shell to sh?

mbuffer: fatal: unknown option "2"

syncoid just errors out constantly

i've also tried sh -c 'syncoid -r a@b:zroot/iocage zroot/iocage

' and it still errors

Does anybody use syncoid on freebsd? Is there any way to use it without changing root's shell to sh?

mbuffer: fatal: unknown option "2"

syncoid just errors out constantly

Anybody tried building the updated devel/nasm 2.16.01,1 in poudriere? I'm getting thousands of gmake processes, slowing down the entire system to a crawl.

not yet, sorry

sfox: root's shell is sh by default

fireglow: something seems amiss with that... why is nobody compiling it?

joko: you need to enable it to use RA

I can't untar my backup files

x srv/minecraft/crash-reports/crash-2019-06-20_06.42.40-server.txt: Cannot extract through symlink srv/minecraft/crash-reports/crash-2019-06-20_06.42.40-server.txt: No such file or directory

tar -C / -xvf /root/minecraft2.tar

what does that even mean?

sfox: man tar ;) -C destination location -x extract -v be verbose -f use file instead of stdin

rtprio: what would I need to enable? This is my current rc.conf line: ifconfig_epair0b_ipv6="inet6 auto_linklocal accept_rtadv autoconf"

yeah. pretty basic

why isn't it restore?

tar is archiver it does not restore files it extracts files from archives

so I can't just untar a tarball onto / with bsdtar?

sfox: yes you can

-C / right?

sfox: yes

that's what i did

tar -xvf file.tar -C /

I don't understand what i'm doing wrong

and it gives you error?

Check whether the symbolic link given exists in the extracted tarball, and whether the pathname it links to exists.

there is no symbolic link

/srv/minecraft is just a regular folder

what about crash-reports or crash-2019-06-20_06.42.40-server.txt

just a regular folder and plaintext file

I did managed to work around whatever the issue was with tar -C /srv/ -xvf /root/minecraft2.tar srv/minecraft and just moving the folder up a directory into place

fireglow: send an email to the ports maintainer.

But that is a really weird issue I'm curious how it happened

Do you have the exact command line used to create that tarball?

Which part of usr/src/release/ makes the tarfiles like base.tgz? It is a bit opaque.

libGL error: failed to authenticate magic 1

libGL error: failed to load driver: i96

any ideas on why that could be happening?

cracauer: it really is. what are you trying to do?

I want to make kernel.tgz and base.tgz from a 14-current build and untar that on a wimpy laptop.

and, like, make release doesn't do that?

It does?

I will try `make packageworld`

and with poudriere, I mostly create packages: alpha.pkgbase.live

Title: Unofficial FreeBSD pkgbase repository

cracauer: see release(7) and docs.freebsd.org/en/articles/freebsd-releng

Title: FreeBSD Release Engineering | FreeBSD Documentation Portal

OK, running now.

"The FreeBSD Subversion repository contains several hooks to perform sanity checks before any commit is actually committed to the tree." does the FreeBSD git repository also do that?

meena: yep

oh boy, had to remove mesa-dri and mesa-libs and install the previous version, totally broke Xorg. :(

what drivers are you using?

radeon-510

it's a mesa issue though, fixed in upstream, but not in pkgs it seems

on freebsd 13?

13.1, yeah bugs.freebsd.org/bugzilla/show_bug.cgi?id=268327

Title: 268327 – graphics/mesa-dri: 22.3.0 breaks direct rendering (radeon, SUMO, r600_dri)

found it via forums.freebsd.org/threads/xorg-server-segfault.87422 and backing down to mesa-dri/libs 21.3.8 makes X work again (yay for /var/cache/pkg holding onto old versions)

Title: Xorg server segfault | The FreeBSD Forums

anyone in here uses awesome wm?

12

In FreeBSD land there are no two people running the same desktop.

Except maybe a vt100.

Re: "no two people running the same desktop" -- Do you mean actual desktop environment and/or window manager; or what other auxiliary widgets people use? In case of "Wayland", there seems to be not much to choose

... for a window manager.

oh, wayland works on freebsd?

debdrup uses it, from the impression I got so far, exclusively with "sway"

# A comma could be in between "exclusively with" but works either way

I should try migrating to sway on my next reinstall tho.

i use wayland pretty exclusively on linux now, so it'd be nice not to regress to X on bsd if i ever use it as a desktop os

I do use wayland and sway, yes - but it's hard to get by completely with out Xwayland for the occational thing that doesn't support Wayland yet.

You don't need Xorg installed, though - only Xwayland.

Oh, and xorgproto according to `pkg info -x xorg` ;)

debdrup, Thank you for the amendment

No worries. :)

I have an SSD from the ZFS mirror my old desktop (FreeBSD 12... 3?) connected to my current system. I've created a raw disk in VirtualBox to point to it, and attached it to a VM. The VM asks for the GELI password, shows the FreeBSD boot menu, and hangs while booting.

My suspicion is that I need to disable some of the kernel modules such as for VirtualBox so they're not loaded in the guest OS, but I don't see /etc if I import the zpool on the host. What would be the right way to get to /boot and /etc on that zpool?

I could likely swap the SSD into another PC, boot from it, and edit from there, but it would be nice if there was a simple way to do it without having to move the disk back and forth.

import that pool with -R /somepath

Check.

I have it at /oldpcname

Is the problem of specifying GELI password?

check zfs list for the datasets you need, and mount those

There's no dataset for /etc.

and I don't see it under the mount point.

nor /boot.

Are those under different (unencrypted) partititions?

It only mounts tmp, usr, var, and zroot.

"/etc" would be in "/"; some portions of "/boot" also. All depends how the datasets were setup

It's the default setup from 12.x when choosing an encrypted filesystem.

Any idea which dataset /etc actually lives on? Is it zroot/ROOT/default?

Yes

OK, that was it. I thought it would mount with the others when I imported the pool, but it didn't.

Can you dual boot freebsd and openbsd?  I would think you just make a partition for freebsd, puts its bootloader into it, along with / everything in the rest of the partition, so you dividing that up so it hold all the partitions , files and folder.  Then do the same with openbsd.

The root dataset has "canmount=noauto" attriubute, at least on 1[34], which prevents the dataset being mounted via plain|automatic "zfs-mount"

Thanks. Sorry for the petty question.

Ah, OK.

It was just confusing to see the other dirs show up but not what I expected.

Try this: zfs list -o name,canmount,mounted,mountpoint

Yes, it has noauto.

s/attriubute/attribute/

/etc/rc machinery mounts everything under the active boot dataset even if a dataset has "canmount=noauto" attribute

Anyone here messing around with risc-v?

Not yet, what about you?

Exploring it for fun, as a side project.

I have been working to learn arm 32 bit and 64 assembly

and I have a pi  and a pine phone pro so I have been exporting much

oh and been looking at amd 64 assembly  too

arm basics are pretty straightforward, so that's good. I'm working with esp32.

I find , arm is way easyer , then when I look at the same code on amd64