crap, i connected my local forgejo to push to github and it's somehow closed my open PRs... i guess the branch didn't exist

oh, if you push the branch again you can reopen the PR, okay

in rc(8), what is the difference between ${svc}_precmd and ${svc}_start_precmd? does the first one just run for any action?

Is there a simple way to copy config files from a dying server to a new one? Server only, no X.org, both have ssh accress

scp?

In the words of Homer Simpson ..... Doh

I'm reading up on it now

nimaje: your comment yesterday about setting svcj_options for ntpd prompted me to look into this in more detail and i think the issue is actually wider than just ntpd -- anything that tries to set $command_args in precmd is broken in svcj, including nfsd which supposedly supports svcj

anyone knows when the xorg-server 21.1.16 security update will arrive in quarterly branch?

polyduekes: i am rather surprised it hasn't already, you might want to mail manu@ and ask

the relevant commit is not marked with MFH: 2025Q1, which usually means there's no intention to backport it cgit.freebsd.org/ports/commit/?id=9…486db3008f77bebcc0887f8ecbc546fa007

ivy: thanks i wll try mailing

ah, hm

what does svcj mean

LXGHTNXNG: service jails, a new feature in 15.0 that allows rc(8) services to run inside a special type of jail

the idea being to reduce the amount of system access a specific service has (somewhat like apparmour or SELinux on Linux)

sounds interesting

hi, I am trying to write a rc for a custom service

a punk downvoted it -it was before asking here-, I wonder what is wrong with the question

no idea why it hangs rc, but isn't rserver_tcp_stop what rc.subr would do if you leave it out, as you have a pidfile? and I would set PATH with sh syntax, instead of using env and daemons --output-file instead of >>

can anyone audit my pf.conf script? the problem is it does not autoload after boot

i know it can be a matter of missing bracket

Have you tried pfctl -nf ?

wsky: do you get any output on the console during boot? there should be an error message printed if pf.conf isn't loaded

ivy i didn't this server is a vps

i mean i could

wsky: you can see console output in /var/log/console.log by following the instructions in /etc/syslog.conf and rebooting

jan0sch the script loads manually after boot, it just does not autoload

what if i just show the script here?

it's a little bit rude to ask us to test your pf.conf for you when you could easily do it yourself...

nacelle: awesome, good to hear

ivy i dont't know neither what's wrong with that script nor how is this rude :s

what did you already do to debug it? ideas to start debugging was already given here

nimaje: I tried with -o instead of redirects, still hangs

nimaje i've browsed the web

i know i needed brackets at some point and it worked and then, after upgrade to fbsd 14 it stopped working again

nimaje: it was adding REQUIRE: networking

uskerine_: you'd probably use the daemon output to log rather than redirecting like that

as daemon will accept signals for log rotation

Good night. Which pkg mirror is the fastest in Iran?

I am trying to install postgresql17 in a jail and the download ETA just for llvm19 is more than 1 hour

I can't achieve my goals with these speeds

sneakernet sounds like it would be faster

or use the package, which probably doesn't need llvm?

it does via the default LLVM option "Build with support for JIT-compiling expressions"

nimaje: is downloading them outside the jail any faster? there may not be much to be done

momken has problems downloading, but is already gone

quick question for developers in this group: is there any real reason to maintain the graphviz .dot files in src/share/misc/committers-*.dot?

these first appeared in v 7.0 and while they have been "kind of maintained" over time, they are "kind of a mess" right now.

is there any real value to these file? does anyone print them out on a plotter and have them hanging proudly in an office anywhere?

are they of any value?

I was looking at my HE.net /48 and /64, noticed that I was not using my /64 and feared I configured it wrong. Then I realized I assigned a /64 to each of my vlans and I'm using the /48 pretty much like I should.

dvl: your ISP doesn't provide you with native IPv6 ?

hello

okay team, I have not been able to figure this out. i ran out of space in my mirrored pool `tank` on my backup machine (I will run out tonight when syncoid runs), so I got new drives. I already re-silvered one disk. The issue is that `gpart show` doesn't show this disk and I want to label this drive. Do I need to remove this drive from the mirror, `gpart create` a scheme, `gpart add` a label, and then `zpool replace` and resilver?

It would be annoying to resilver all over again, but i'm thinking I messed up by not going a `gpart backup ada0 | gpart restore -F /dev/ada1` before adding this new disk to the mirror

When I offline the disk, i can do whatever I want to it. Then I put it back online and the changes are gone. I guess i need to do more than offline it. Fully remove it. gpart all the things. then add by the gpt label. and then it'll resilver all over again. yes?