I noticed freebsd.org/community doesn't mention slack anymore, but the websute webapp working group is listed as slack channel #wg-www21. Does anybody know anything?

I've never used slack, but I want to speak with this group

concussious: the slack is invite-only afaik, kevans probably knows more

lw: perfect, thanks, I'll send a mail

some of the slack channels are proxied to #bsdcode (or maybe #bsdmips) so you can contact slack users that way, but i think that's just the lobby channel

cleaned my bong today

fyi

bro, that is so a #freebsd-social topic

this channel is for we

're tring to work over here :P

ah

but, im not a mod or something, idk how cool they'll be over there about that. I'm from ohio, where that's perfectly fine, but freebsd has an internation audience including places where the cops just merc you for that kinda stuff. i remember ~5 years ago things escallated to "openbsd is a drug dealer os" and it was quite stressful for a lot of people.

ya i guess the devil mascott makes me forget that sometimes it's not for serving porn and talking about drugs

and honestly, i don't think anyone was seriously using openbsd that way. that would be preposterous, openbsd is extremely rare and noticable on the network. security =/= anonymity

ya true

i bet more bankers use openbsd at home than dealers

^ me too lol

I worked out what is causing periodic to lock... the backup drive connect by USB has failed

s/connect/connected

and of course the warranty has expired, sigh

vortexx seems like that shouldn't lock. report bug?

if external, might just be the controller. Often if you "shuck" the drive, the bare drive plugged in internally will still work

vortexx: do you have a microcenter in your area?

it won't help this time, but microcenter branded flash of any time is lifetime policy, no paperwork or anything, you just bring it to them and they hand you another one, even they size you up when they dont make them that small anymore

polyex: I call a backup script via /etc/periodic/daily/999.local that backs up stuff to the drive and that's hanging since the zpool isn't available

concussious: not in the USA

it's a 14Tb WD Elements spinning rust drive

microcenter is from ohio, but they have lots of them elsewhere

I mean I'm not based in the USA

i see

I'm in central Europe

thanks anyway

np. I would if you can use "#!/bin/sh -e" to make it not hang?

I'm not sure -e works quite that way but idk

I will give shucking a go

ok I gotta go get some mulch before the store closes, thanks everyone!

SponiX: only problem with plugging in internally is this machine has all 4 drive slots used for the raid array. I'll have to boot with it offline to test. Doable all the same

ssh saper that's bizarre

ffs

I was forced to reboot my laptop and now xfce is completely unusable

all of my keybindings are gone and I can't just close a damn windo, or like... resize it, it seems

I just want my x and minimize buttons back man

a-ha, there we go

upgrade mishap

anyone ever install bsd on some super ancient hardware, just for the hell of it?

i always wonder how much pages an apache server could serve, if i ran it on an early 90s 68K 33MHz Mac

¯\_ (ツ) _/¯

you should try netBSD

i think i did on a 69k mac like 25 years ago but i didn't know what to do once it was installed and i went back to OS 8

GoSox: Depends on what you define as "super ancient", I think the oldest I've gone so far was IBM ThinkPad T43 and Apple PowerBook G4, installed OpenBSD on both of them.

The ThinkPad was reasonably fast, but booting it up or awakening it from sleep takes forever. The PowerBook is much slower, sleep mode doesn't work at all, and if you're using the Japanese keyboard layout, some keys that don't exist on the US keyboard layout will not get recognized.

Not only talking about the obvious keys like 変換, 非変換, かな, and 英数 (which is why our spacebars are so much smaller), other keys too like the 「|」 key that's the 3rd key after the 「0」 key, which is typically the backspace key on US layouts, and on the JP layouts it's 1 key more to the right.

So the backspace key on JP keyboards is the same size as all the other keys on the same row.

For this same reason I can't type "pkg_add" on it, because the location of where the 「_」 key is on JP keyboards doesn't exist on US keyboards.

I installed NetBSD and OpenBSD to a Compaq Evo N1000v and ran into several kernel bugs on NetBSD, but otherwise it was very responsive. OpenBSD on the other hand had no issues, but it was slow

I remember putting FreeBSD and OpenBSD onto my Titanium Powerbook G4 a couple years ago. All of them had problems with X11, Linux and FreeBSD had a black screen when the full radeon triver kicked in (but an external monitor worked fine), and OpenBSD had weird repeating patterns on the screen in X

I think NetBSD actually had working X11 on the machine, but it didn't cover the whole screen, only a portion of it

(and it kept randomly crashing)

I ended up putting MacOS 9 back onto it, it's a fun toy

that reminds me, I also installed NetBSD to the Wii and it works flawlessly. Linux on the wii had the issue that is was abandoned from the start, because no distro picked it up, there were only slightly modified versions created by the developers that didn't last long

OpenBSD was ok my Titanium G4, around 4 or 5 years ago

if i recall correctly

zwr: Do you actually do anything productive with NetBSD on the wii? or really just "because you can" demo :D

Dooshki: I have no X11 problems with OpenBSD on PowerBook G4, although I have the alluminum one, not the titanium one.

Oh, I'm actually just remembering, I had to disable the radeondrm driver for me not to lose the screen, so I was using a generic framebuffer driver

I guess if I was exceptionally bored one day, I could try debugging the black screen with modern Radeon driver issue. The thing is, my hands are already quite full of stuff to do

dtomato: I've been running programs on it and hoping to find a bug, no luck so far.

dtomato: I found a lot of bugs on the Evo but none on the Wii, I'm not sure why programs break on i386 but not on powerpc

I also use it to read NetBSD man pages

sorry, i missed the beginning, what about wii?

i am also running netbsd-wii

hmm, is there a restricted shell that comes with freebsd? nothing is jumping out at me but hoping i'm missing something

kevans: I was scratching my head whole day yesterday, but the results are 100% reproducible

zwr: there can my man reasons for this, some bugs just do not pop up on a different architecture. For example, on one architecture userland addresses may look like $800002a5 and on another $2a5. If you use a signed 32-bit integer strange things may happen on one platform, but not on the another.

very_sneaky: bash and ksh93 have restricted modes

saper: cheers. not installed by default unfortunately!

actually, that might not be right - i'm looking at pfsense not vanilla freebsd

Default is sh, csh and tcsh.

antranigv: I'm using samb 4.16

antranigv: and this is what I'm using for samba as a time machine. dan.langille.org/2024/01/06/creatin…tance-using-samba-freebsd-and-zfs-2

Title: Creating a Time Capsule instance using Samba, FreeBSD, and ZFS (latest) – Dan Langille's Other Diary

antranigv: I have another samba instance, but that's just for filesharing to macs.

saw 14.0 made pf do better scrubbing or something? but i thought i already had that working in 13. is the new "set reassemble yes" the same as the "scrub fragment reassemble" im using now?

anyone know what might cause a system stall on zio->io_cv? it's a new system, running 14.1-RELEASE

Hi!, does anyone know what happens with the host display when a VM uses GPU Passtrhough?. I mean, can the host and the vm run the same GPU at the same time?

martinrame that would be really cool. like if they could just share the gpu like how VMs share cpu

and ram

like obviously the vm host doesn't just go dark right?

polyex: I'm asking exactly that. Didn't test yet, my concern is that, does the host goes dark?

wcarson: running out of space, maybe?

hmm, no, only using like 100 GB out of 2TB

i see some CAM and Solaris errors on the console, but my little display is too small, so i setup syslog to a server and i'll see if i can capture the errors there

seems odd it would be a bad disk, it's a brand new nvme

wcarson: I have it all the time but I am running ZFS on a small, old laptop

wcarson: smartctl -a saying something on the drive? CAM errors are a problem, yes

wcarson: maybe you can dump dmesg to the place you can read better

i'm not sure how to do that

wcarson, Is this new system on the network? If so then you can do this: mesg | nc termbin.com 999

oh, i mean, i can get to it and run dmesg, but not when it crashes. i thought maybe there was some way to automatically send dmesg stuff over the network

Jun 16 17:39:16 <user.info> ports01 devd[1572]: Processing event '!system=nvme subsystem=controller type=RESET name="nvme0" reason="resetting controller"'

Jun 16 17:39:16 <user.crit> ports01 kernel: nvme0: Resetting controller due to a timeout and possible hot unplug.

Jun 16 17:39:16 <user.crit> ports01 kernel: nvme0: resetting controller

uhmm

There are some differences between what is logged to the syslog, what is written to the kernel ring buffer of dmesg, what is written to the system console.

Maybe you said earlier but I just joined back in, what is the type of machine you are running this one?

well that doesn't look good

it's a new minisforums nab9, core i9-12900hk, 2tb wd sn770 nvme, 64gb (2x32gb) ddr4 sodimm, 14.1-release

Agreed. Not good. Not good at all!

That does look like a very nice small form factor machine. Perhaps if it were me with this error I would boot a live-boot image from USB and then use it to run a read-write test across the NVMe. How is that NVMe attached? The specs I am looking at show only an M.2 SATA.

wcarson: bugs.freebsd.org/bugzilla/show_bug.cgi?id=262969

Title: 262969 – NVMe - Resetting controller due to a timeout and possible hot unplug

in 14's pf, does set reassemble yes no-df require random-id still?

rwp: seems to have m.2 2280 gen4 nvme & support for a 2.5" sata

armin: thanks, reading that

wcarson: Hmm, what sticks out is that the people having the same issue almost all report they have a WD 770, too

yeah, that's the conclusion i'm coming to

that... sucks.

i guess i can return it

thanks for the help everyone. i ordered a Crucial to see if that fares better

wcarson, The WD 770 is a very popular storage device though. I am using two of them in a Linux kernel system on a client machine with no problem. It might be that everyone with this problem is reporting a WD 770 because it is the most used device. In which case the finger of shame would be pointing to the FreeBSD kernel.

maybe, but it seems like in that openzfs thread that it's also a lot of linux users

but i admit i only skimmed that one

anyway, the crucial was like $50 cheaper and it's a no-hassle amazon return, so .. just wasting my time i guess, no harm there

wcarson: fascinating story someone measured the temperature of WD 770 rising before the crash

I think ZFS exercises storage media a lot. I have a feeling I kill mechanical hard drives with ZFS much faster than it was before with other filesystems.

I so far have not had ZFS kill any hard drives. However ZFS writes new blocks before freeing old blocks. I believe that causes more of the blocks on a hard drive to be used in a queue whereas other file systems might, maybe, reuse blocks faster and never write to the extra unused blocks. Maybe. I don't really know. It just seems like it might to me.

saper: yeah i saw that, crazy