Thanks for asking anyway (:

@n0madcoder is that legal here?

This is the meme police. You are now under arrest

@unixmann the purpose of this chat is to stay FreeBSD related so people that want to share knowledge about it, or ask some technical questions can do it similar to irc. While I don't like to restrict every action out of that context, it's true that some users may like to keep this room tidy. Posting memes from time to time while related to FreeBSD is not a problem, unless it becomes a daily dose

@n0madcoder

This is the meme police. You are now under arrest

Is oracle JDK available in pkg? I don't see it in pkg search jdk but here says it's in pkg https://www.freebsd.org/java/

I mean I have PepperMint OS on my HDD

have you considered running it in a virtual machine instead ?

vbox

>uses Linux with root on ZFS with pkgsrc

>uses Linux with root on ZFS with pkgsrc

< uses devuan linux w/o systemd

>uses Linux with root on ZFS with pkgsrc

and break with kernel updates

< uses devuan linux w/o systemd

>uses KISS Linux

build with musl and clang

freebsd 14 current DIED can not recover block devices

that was with zfs

What did you tried to recover it?

i didn't

What did you tried to recover it?

once ZFS goes tits up, there is no point in trying to recover it anyways. thats why you use backups and stable branch for data

ZFS is considered rock stable.. what problem are you experiencing?

in other weirdness :: have 13 & ghost NO X-config, desktop works. 14 w/ X ~~~ exec mate-session kinda no selection of de w/ sddm

14 is expected to have issues tho (regardless of ZFS if you messed around with it)

once ZFS goes tits up, there is no point in trying to recover it anyways. thats why you use backups and stable branch for data

good to know

when is 14 expected to go stable ?

Maybe April next year. They usually add the date here at some point https://www.freebsd.org/releng/#schedule

good to know

it's basically because ZFS stores many snapshots that are supposed to guarantee data integrity. so when data is corrupted this means that all of those snapshots failed which is not supposed to happen at all. while it might be possible to recover from that the fact that ZFS has corruped data points to a larger issue...

it's basically because ZFS stores many snapshots that are supposed to guarantee data integrity. so when data is corrupted this means that all of those snapshots failed which is not supposed to happen at all. while it might be possible to recover from that the fact that ZFS has corruped data points to a larger issue...

NO! Snapshots are NOT supposed to guarantee data integrity

Snapshots are for your own ease of management, you can choose not to use them if you don’t like them. Data integrity is obtained by having multi-disk vdevs as base, and eventually raidz on top of these

so w/o raidz zfs is ~~pointless ?

~~somewhat

Not necessarily, it has a lot of cool features. Being a COW filesystem, data corruption is less likely in the first place

>uses Linux with root on ZFS with pkgsrc

oh gosh that sounds like hell

i cant imagine using thoose things on linux

I use Linux with ZFS :)

I use Linux with ZFS :)

is it nice?

It’s actually cool pulling FreeBSD jails from the server using zfs

is it nice?

Yes, quite. I use freeBSD as server and Linux on clients. Only downside on Linux is having to use unofficial kernel or DMKS but that’s about it. Love ZFS

K you manually partitioning that zfs drive?

NO! Snapshots are NOT supposed to guarantee data integrity

i used the word "snapshot" in a more colloquial manner than i should have. what i actually meant is the possibillity of rolling back transaction groups, because of the way ZFS keeps track of blocks. it stores a few versions of old states. which in essence are snapshots. but not the ones that a user would initiate, say before an uprade.

Yes, quite. I use freeBSD as server and Linux on clients. Only downside on Linux is having to use unofficial kernel or DMKS but that’s about it. Love ZFS

i love ZFS too. Makes thing easy

K you manually partitioning that zfs drive?

Yes

yeah ok

is your free space across mount points the same?

and break with kernel updates

zfs is unrelated to linux kernel as support is not native zfs support in linux is achieved by modules in initramfs

is your free space across mount points the same?

Drive on the notebook is: 550MB EFI All the other space ZFS Within ZFS I have datasets (home, system, containers etc…)

Yes, quite. I use freeBSD as server and Linux on clients. Only downside on Linux is having to use unofficial kernel or DMKS but that’s about it. Love ZFS

DKMS is crap..:(

I agree I don’t like it either

when the installer ask to use full drive and you say yeah, then the free space on all mount points is the same.

w/ zfs

I don’t use any installer, I do things manually

LOL how did ya get it on there in the first place?

Secondary laptop has 2 SSD, one 128GB and the other one 512 First disk has two partitions, EFI 550MB as always, and remaining space ZFS. Second disk is just ZFS. Cool thing is that the pool spans over both disks

LOL how did ya get it on there in the first place?

Arch Linux 🤷‍♂️

🙈

have you considered running it in a virtual machine instead ?

I will think about it. Im not so into VMs. Thanks

I'm working on an automated install iso. is there a way to access the ISO's/img's filesystem from bsdinstall? my installerconfig's script seems to be in a chroot under /mnt, but I'd like to access /usr/freebsd-dist in the image itself.

you can merely extract the txz archives from scripts

thats what the bsdinstall does

#!/bin/sh # # v1.1 # cd /tmp/zroot for i in base doc kernel lib32; do (xz -dc /usr/freebsd-dist/${i}.txz | tar xvf -) ; done zfs set readonly=on zroot/var/empty touch /tmp/zroot/etc/fstab echo zfs_load="YES" > /tmp/zroot/boot/loader.conf echo zfs_enable="YES" > /tmp/zroot/etc/rc.conf echo sshd_enable="YES" >> /tmp/zroot/etc/rc.conf

(for instance)

"iso img" contains only txz archives btw

so I would need to modify bsdinstall, or can I get this done within my installerconfig?

you dont need bsdinstall at all

you need to label the disk, create the pool, extract the stuff and create the configs

sieve I don't have dovecot-sieve package in freebsd

optionally chroot/install the packages

@drookie is your answer is for me?

nope.

you aren't asking for anything. just grieving.

(my condolences ! I suggest you install Linux and grieve there)

wait ,, I am asking a solution, I did all sieve configuration and now I don't see any sieve package

Uhm, dovecot... let me look.

I see it in website but, with pkg search sieve nothing no plugin for dovecot, I can't complie sieve script

so I suppose you did your configuration for nothing. :|

wait I search with pkg search dovecot I got this dovecot-pigeonhole-0.5.15 Sieve plugin for the Dovecot 'deliver' LDA and LMTP I think this is the right one.

done, YEEEEEEEEEEEEEPIEEEEEEEEEEEEEEEEEEEEEE

😎🎅

you dont need bsdinstall at all

gotcha, thanks. I had tunnel vision, I suppose.

gotcha, thanks. I had tunnel vision, I suppose.

actually there's an old FreeBSD wiki article about how do you label/partition the disk and how do you create the zfs datasets. pre-bsdinstall. still working though.

due to lack of manpower in FreeBSD they just automated it in bsdinstall and thats all it does actually

the only thing that can improve it - it's the be environments, but you can merely steal it from any actual FreeBSD installed

(can give you the script if you're lazy)