I know that on Linux, a dir that is executable but not readable is accessible if you know a filename therewithin. The same behavior appears to apply on FreeBSD as I just tested it - is this standard behavior or do some UNIXes not do this?

It's how POSIX permissions work.

'r' lets you read a directory, but you don't need to do that if you know what's in it

AmyMalik, I'm pretty sure this is specified by Posix.

(Which mason already said.)

thanks

yuripv: net-snmp for the snmpwalk

pkg install seems to error out for me:

Creating user 'messagebus' with uid '556'.

pw: user 'messagebus' disappeared during update

pkg: PRE-INSTALL script failed

The user is present in /etc/passwd, no clue what it's complaining about

(I don't really think the package I want needs to pull in dbus as a dependency, but that's another matter)

root@daruma ~ ▶ grep '^messagebus:' /etc/passwd

messagebus:*:556:556:D-BUS Daemon User:/nonexistent:/usr/sbin/nologin

root@daruma ~ ▶ pw user show messagebus

pw: no such user `messagebus'

This doesn't seem right

do wlan devices not show their … "parent" device?

ccx: maybe you're missing a db update

ccx: did something go wrong during the pkg install?

oh, it's right there, PRE-INSTALL script failed

Probably. Is there an easy way to inspect/check the db file?

ccx: it's sqlite

I'm reading codeberg.org/FreeBSD/freebsd-ports/…src/branch/main/devel/dbus/Makefile trying to figure out where it's failing

Title: freebsd-ports/Makefile at main - freebsd-ports - Codeberg.org

the only remarkable thing about the port is the PORTEPOCH= 1

which means at one point, the version was downgraded

meena: Are you sure it's that and not db(3) - as in Berkeley DB?

ccx: are we talking about pkg's database? Then, Yes, I'm certain.

Ah, I was talking of pwd_mkdb which I assume is in play here

ccx: ah, yes, sorry

ccx: but you should be able to just open that with vipw

meena: Manually issuing `/usr/sbin/pwd_mkdb -p /etc/master.passwd` seems to have fix things.

ccx: with the messagebus user added, too?

Hi. Why does MPV pause after the videos ends if MPV has LOOP or SHUFFLE turned on?

Is it a bug or some setting? does this belong in #freebsd-social

PredatorONormies: the pause should be very minimal, IIRC, the loop setting takes a pause time

meena: Yeah, it was added to the plaintext files all right but somehow didn't propagate to the db? Though running `strings /etc/pwd.db | grep messagebus` shown that it there.

meena I meant pause as a pause, no a timeout paues

And it does not start playing again until I press the left arrow key aka to go backwards

if I press the O letter on keyboard to show OSD (or what it's called) - before the current video time there is a clock shown

a clock icon shown*

Want verbose or something?

PredatorONormies: Something like that has been happening to me for like a year or more, but I didn't find anything on internet nor do I know how to debug that.

And passing `--aid=no` to remove audio seems to fix it, but of course there's no audio.

Yeah

So something related to audio?

No idea, but it also happens when loading the next file, even if there's no shuffle.

I know because I have loaded directories of anime and this happened when loading next episode too.

well.. yeah for example mpv *mp4

oh, my, you have the same problem -- having to manually start each anime episode? lol

Clock usually means it's stuck loading next piece of source data. But it's quite possible it could be something else too.

Does it happen with all files? Could be malformed index or something if it's only a few.

Yeah - hence I asked if verbose output could be helpful

yes

all

would be worse if it's random or something, no? lol

Wait I forgot. I think it's only with video+audio, because I think there wasn't any problems with only audio.

oh

You might be correct

Some sort of syncronisation system then?

Hi, I am trying to n install freebsd on new amd64 hardware. I have 2 disks (one SSD one HDD). I am doing that from usb stick. (amd64-memstick. in graphical install I got an "unknown error" when I try to perfomr the install on both disks. Is there a way to debug this error ?

Always a bad thing when the error itself is unknown

Reminds me of Windows

yes indeed ...

I didn't check the image I am doing that

my configuration is an AMD64 5600G with 128G of RAM. the disk had an old freebsd install. maybe that is the cause

could be.. but I am surprised that there is an error handler in something like FreeBSD

Usually software written in C handles exit codes quite well (aka errors)

an error handler ?

But this just seems like some stupid brainless exception handler.. not that I've looked at installer/FreeBSD code

for examply - try excepts blocks in Python - that's an error handler

exception handler

same thing I think

ok yes

the screen seems new...

in the past it was a deep blue now it's a clear blue. Is there a new installer ?

No idea, the latest (I think) FreeBSD is my 1st

it seems the checksum isn't right

I am trying new download

no image is ok in fact :-/

with UFS it seems to work

0.o

yes... install has been completed with UFS right now.

Maybe my old disks was UFS. I don't know

:/

Try OpenBSD? XD

I am running FreeBSD on ZFS zroot, and all is fine in that regard. Then I have attached two other drives that is a ZFS mirror pool. Everytime I reboot the machine, I have to export the pool, because the dataset directories are empty, then import the pool manually and then it works. How can I troubleshoot why the pool doesn't get imported and mounted correctly at boot?

When I do a "zdb -CU /boot/zfs/zpool.cache" I can see that only the zroot is located in the cache.

I guess that's why the other pool doesn't get auto imported.

if you have to *export* it, it means it was imported, right? may be the fs you are looking for was not mounted?

does it have the mountpoint property set to noauto or legacy?

(and do you have zfs_enable="YES" in rc.conf?)

Yes, zfs_enabled="YES" is in rc.conf.

I have to export because the datasets are just empty directories, and only when I export and then import, the data is there.

i.e. files.

enabled or enable? there's a difference :)

That should be "enable", not "enabled", in case the typo was also in your ... beat me to it.

sorry, typo, it's enable and zroot is working.

Can it be because the mountpoint of the dataset is in a users home directory? The mountpoint is /home/foo/data, maybe it needs to be /usr/home/foo/data for it to work during boot?

Is it possible to parallelize compilation when building port ?

you can try make -jx but it doesn't work for many ports

Hi all. I'm currently trying to figure out where x11-themes/wallpapers-freebsd-kde pulls the files from. I see that in MASTER_SITES= it has the value LOCAL/avilla. But where is such path?

thanks sereg. I didn't think it was via make (I thought there was a configuration somewhere, since I will have always the same number of cpu)

perhaps there is somewhere but I do not know

I won't venture it

Chapter 6 of Porter's manual scared me.

Just realized that git was the versioning manager used in port tree. But I have an issue, I have installed port during the system install process and it is not a git repo. Can I remove /usr/ports without risk ?

MAKE_JOBS_NUMBER (documented in bsd.ports.mk)

err, bsd.port.mk, and yes, you can remove /usr/ports

thanks but I didn't ready to take the risk.

cool

ports building should be already parallelized unless explicitly disabled in port's Makefile

ah ok

is git the only way to upgrade port tree. I remember sometime ago there was cvs. (I am not a nostalgic of this tool, that's said)

yuripv, so the system determine the number of parallel jobs based on cpu number, right ?

Midjak: yes, github.com/freebsd/freebsd-ports/bl…dad77589258038/Mk/bsd.port.mk#L2049 (if not overriden by MAKE_JOBS_NUMBER)

Title: freebsd-ports/bsd.port.mk at 77c74c71009345aad77618eb6adad77589258038 · freebsd/freebsd-ports · GitHub

oh great thank you

Hello, i would like to try out freebsd but it seems my wifi card hasn't been supoprted for sometime. this is the latest i could find bugs.freebsd.org/bugzilla/show_bug.cgi?id=244261 my wifi card is AX201 can anyone verify this or something on the backend that i don't see

Title: 244261 – Add support for Intel AX200 (iwx) wireless devices

Is that an Intel WiFi card? It is "supported" via iwlwifi driver project on stable/13 & CURRENT

ax20*? you're better off with linux, which is where the driver gets it code from anywan

anyway*

Refer to wiki.freebsd.org/WiFi/Iwlwifi

Title: WiFi/Iwlwifi - FreeBSD Wiki

Alternatively there is net/wifibox port which uses a Linux VM for the iwlwifi driver. I use it on Framework laptop that has Intel AX210 card: github.com/pgj/freebsd-wifibox-port

Title: GitHub - pgj/freebsd-wifibox-port: FreeBSD port for Wifibox

idwer: I'm on Linux already

parv: This net/wifibox is it in the installer with FreeBSD?

Aren't we all, Linux is everywhere! on TVs, phones, routers, there's no escaping it.

morn

Mornin'

wez: o/~

jb1277976: its a port/package

koobs: You're heiling Hitler?

wez: sorry what?

wez: worth noting that people could misunderstand your 'joke'

koobs: you didn't include the ascii emoji's shoulder

wez: that assumes its not underarm stinky

.o/~

that has the shoulder

koobs: That means I have to get through the install first then install the port/binary?

jb1277976: there's affectively no difference, in the installer would pkg install a package

jb1277976: so just pkg install once youre done with the install

koobs: worthy note? you wrote the ascii text and that's what it look like, no joke about it

Ok, gotta find a Ethernet

careful with ascii emoji's that have a single arm raised with no indication of the direction being faced. that's why you add in the shoulder.

Hmm. All this time I took it as a representation of a note, the way I use it.

Please tell me parv isn't a bot?

That wave was clearly the left arm up anyway. :-) Left arm: o/ Right arm: \o Both arms: \o/

Only one way to find out. gonna get stable and do the install. i know before it formats the disk it will ask me to setup my wifi

jb1277976 yu can nomadbsd from a usb stick if youd like too

koobs: Nomadbsd looks nice. looking at the screenshots

koobs: I dont think it supports ax201 iwlwifi

ah, yes youll want current for that

ask lme (lars, nomadbsd author) (on #bhyve and #poudriere channels here on libera) for a CURRENT snapshot of NomadBSD :)

jb1277976: ^

Thanks going there now

Not sure who to report this to, but the plist search on Freshports is broken

Parse error: syntax error, unexpected identifier "count", expecting "," or ";" in /usr/local/www/freshports/classes/ports_by_pkg_plist.php on line 62

Ah there's a github

on a zfs array I've seen errors like this: da0: corrupt or invalid GPT detected

from geom.. but the pool still works..

should those be fixed?