It's amusing to me how every time my PPP reconnects (every 24h) I get an extra digit to my nickname.

PPPoE?

dautor851805, What IRC client are you using?

epicII, irssi

Usually irssi adds underscores to names until they become unique.

Before I switched to using certfp authentication I have seen problems with quick disconnects and reconnects leaving a "ghost nick" behind. I expect that is the problem with the nick adding numbers. superuser.com/questions/474001/irc-…ghost-and-get-back-my-original-nick

Title: IRC ghost - how do I unghost and get back my original nick? - Super User

Hello

How can I download opensmtpd <6.8 for armv7

?

>6.8*

Hello

Any information since my connection dropped?

for last hour

no

V-T60: i would image the tarball for amd64 is the sam as for armv7

oh neat so rtw88 just gave me a kernel panic when I went to install a package

also doesn't seem like the default shell has a chance to write out to its histfile when I reboot through the reboot command, hrm

actually, looking at the backtrace, it seems like it's crashing in some drm stuff, specifically skl_compute_wm. I'm guessing this is related to the whole thing where I have to restrict linuxkpi to a 32-bit address space for rtw88

rtw88 seems practically unusable anyway, can't seem to maintain a connection. bummer

wifibox it is, I suppose

rtprio: executable file? is the same?

how on planet is that possible, the machine code is totally different

seems the kmod-drm spits out some warnings about atomics or mutexes prior to the panic. suppose I should open up a bug report about this

I might also try the Linux 5.10 drivers

you would have to build it from source

V-T60:

V-T60: the source files are the same

Can I get some zfs help here? I have a single disk in a pool that I would like to make in a mirror with 2 larger disks. How would I go on about this?

CrtxReavr: yeah PPPoE; in Croatia every DSL user gets disconnected once every 24h and then gets a new address

rwp: I'm using thelounge. You think it could have something to do with that renaming?

So I attached one of the big disks to make a mirror and it is now resilvering

I have made sure that autoexpand=on

now I have to take out either the small disk or the big disk and add the second big disk. Which disk should I resilver from the small or the bing?

now I have to take out either the small disk or the big disk and add the second big disk. Which disk should I resilver from the small or the big?

I would remove the small one and after resilvering end up with 2 big ones in mirror

yes no the question is do I resilver form the small disk or from the big one?

is there a difference?

I'm saying I would do it from the big one. I don't know if there is difference - it just seems like less steps.

thanks for the reply.

satrapes: if you remove the large disk and replace it with another large disk, you'll just end up with an array with a small disk and a large disk... which is what you already have. if you want to end up with an array containing the two large disks, you need to add them both to the array. you could do that either before or after removing the smaller disk, but doing it after is safer

lw yes I know. I am asking if it is best to resilver from small disk or big disk. Then I will remove the small disk and end up with the 2 large disks

what do you mean by "take out" the disk? if you remove the disk from the array, you'll have to resilver again when you re-attach it

Ok I see. Misunderstanding on my part I thought that I can physically remove it (without zpool operations) and then when I plug it back in it will know that it is a mirror of whatever drive it was resilvered from

if you don't remove it from the pool that might work, but i'm not sure if it'll be happy with the vdev layout changing after the disk was removed... i would probably not do it that way, better to replace the disks to end up with the final configuration you want

lw thanks

has anyone successfully installed from a current snapshot iso recently? i'm getting a generic (and unhelpful) "Failed to extract /usr/freebsd-dist/..." using FreeBSD-15.0-CURRENT-amd64-20240201-2c191ba6b0b5-267921-disc1.iso

voy4g3r2: didn't you run into something like this ^?

Try with another snapshot is usually the first advice.

i'll just start from 14.0 now but i might have another look at this later

I am resilvering the second large disk. Curious to see if the size is going to be increased

Snapshots are little more than CI artifacts, if memory serves - but there was talk of a new -CURRENT stabilization week at the last conference, so maybe something'll make it a bit easier to use the head of the tree.

satrapes: it depends on whether autoexpand is set on the pool, i can't remember the default value. if not, you can use zpool online -e to force it to expand the disks

autoexpand is on I have quadruple checked

hm, every time i generate a password hash with slappasswd -c '$6$', using the same plaintext password, i get the same hash. obviously doing something wrong here, i guess it wants me to provide the salt?

dautor851805, even if you're actitively passing traffic?

lw do I need to restart after zpool online -e?

satrapes: no, should be enough to just run it on both disks and the new space will appear

CrtxReavr yeah :( it's been like that forever... for at least 15 years

I'm guessing some bbras for some technical reason did it a long time ago and they just assumed they have to continue doing it

That's why I do it manually once in a while in the middle of the night. That way it keeps recinnecting me in the middle of every night so it is not as annoying.

hmm zpool online -e did not work. I should probably say at this point that I am on proxmox so maybe not ideal

drives have defintely been resilvered with autoexpand on

satanist: are you using whole disks, not manually partitioned?

satanist: what's the `zpool get autoexpand <pool> ` set to?

oh they quit (i have join/parts turned off so i didn't notice)

so do i, lol

damn I have to go now but thanks for all the help I will ask again later

dautor851805, do you get an actual public v4 IP on your "router?"

I do. (I'm using fbsd as my router :) )

So not an IP in 100.64.0.0/10?

yeah, I'm not behind a CGN

i think i'm going to submit a base utility that can generate crypt hashes because this isn't the first time i've run into this problem, it seems like a useful thing to have

lw, please share you solution - I've often wondered how to manually generate them.

CrtxReavr: my temporary solution is this: le-fay.org/tmp/30d/P0zPqx.txt but it only works if you have openldap-server installed plus it's kind of annoying

(and it requires you to know how much salt each algorithm requires, which isn't even documented in crypt(3) from what i can see)

lw, what's the pwgen command actually doing there?

creating the salt

slappasswd doesn't do that so you have to add it by hand, otherwise you get a password with an empty salt (and i've just realised i used some of these saltless passwords in my ldap directory, so i need to go fix that)

openssl can do it, iirc

aha, openssl passwd -6. thanks

what am i missing with my NFS config here? le-fay.org/tmp/30d/wIh6FV.txt - i've tried putting the sec=sys export both above and below the krb5 export and it made no difference, i did restart mountd

lw, which package included pwgen?

CrtxReavr: sysutils/pwgen

so i tried my nfs mount using sec=krb5p,syskrb5,allgssname but that gives permission denied trying to access the mount without a tgt... i thought the point of allgssname was to allow that

sec=krb5p,gssname=host,allgssname doesn't allow mounting at all

oh, you have to start gssd for that... right. all working now

lw, little cleaner & slightly documented: bpa.st/RFBRU

Title: View paste RFBRU

CrtxReavr: as rtprio pointed out, it's much easier to just use 'openssl passwd -6' :-)

<Homer> BORING! </Homer>

lw: yes i got something like that before.. checksums were epic fails

i was informed it was a propogation issue

files made.. checksums not totally propogated, let me find the mail about it, give me a second

voy4g3r2: i saw the thread but honestly the explanation didn't make any sense

lw, better?: bpa.st/GSNY6

Title: View paste GSNY6

i'm using disc1 (and i think you were too) which doesn't need network access to fetch the packages, so it can't be a propagation issue

lw: agree

i used bootonly AND disc1

both gave same result: lists.freebsd.org/archives/freebsd-current/2023-December/005207.html

Title: Checksum Error on installer (3 iso images in CURRENT)

i left it alone for a few weeks and when i went down my bhyve rabbithole, found the 20240124 worked

let me see what iso i used

voy4g3r2: don't worry too much, i'll just upgrade 14.0 to 15.0 for this vm

i would like to find out what's going on here though

FreeBSD-15.0-CURRENT-amd64-20240118-7a4d1d1df0b2-267640-disc1.iso

FreeBSD-15.0-CURRENT-amd64-20240125-6ec8bf9f3d17-267775-disc1.iso

thes two worked

i didnt go down the route of root cause, i read the response and like.. that's nice i am going to see if works itself out :)

lw: the iso you downloaded, how "old" is it..? a few hours ?

i take it you are using bhyve and not virtualbox?

voy4g3r2: FreeBSD-15.0-CURRENT-amd64-20240201-2c191ba6b0b5-267921-disc1.iso so from yesterday

and yes, bhyve

dautor851805, In the US I was on CenturyLink and they would at random times disconnect me and reconnect me to another subnet as well. Usually before a disconnect there would be network problems and afterward it would be like a fresh reboot and everything would be much better. I always imagined that they were re-balancing load or shifting to different equipment.

lw: that is so odd.. it works then it doesn't and then it works.. there has to be SOMETHING but i do not have the brain capacity to even know where to begin

one day i will remember that 'service X enable' is a thing instead of doing it via sysrc

antranigv, you know your getting repeatedly kicked on efnet / freebsd

washingtonpost.com/obituaries/2024/…ork-time-protocol-internet-obituary :(

Freaky: I'm seeing some commits out of order here, freshbsd.org/freebsd/src?author[]=Lexi+Winter (the two noexcept ones) how's that happen

Title: FreeBSD / src - FreshBSD

lw: it's so good because: typos

meena: did i make a typo

I don't knew if you did. i would have

lw: openssl passwd -salt mysalt

rtprio: but the whole idea is not to have to specify the salt by hand!

lw: openssl passwd -salt `openssl rand -hex 32`

rtprio: openssl passwd -6 seems to generate random salt itself

meena: what what's so good, rebase? are we having this conversation across two different networks :-)

well, now you don't need slappasswd

rtprio: exactly

rtprio: that's why i said thanks earlier :-)

this method is much better

sorry was detached and didn't see the scrollback

although i think this is a bug in something because crypt(3) should probably not accept short salts, it's a potential accidental security issue

hmm, it's rather disappointing that ulrich drepper put lib/libcrypt/crypt-sha512.c in the public domain and freebsd went and put a copyright notice on it

lw: "public domain" doesn't mean the same thing in every country, that's why licenses are there

dstolfa: i know, adding a copyright notice doesn't fix that though, because if the public domain declaration isn't valid, neither is the copyright notice

i'm not so sure about that, but that's legalese. there was probably a good reason for doing it

not a lawyer :P

ianal either but the only reason freebsd is allowed to copyright is because it was PD... so if it's not PD... it would under some default all rights reserved license and you can't relicense it

(aiui, this is at least partly why freebsd doesn't allow PD code in the first place)

right. but consider that the licensing was done where PD is a thing, and now there is a properly licensed file distributed with FreeBSD in those places where PD isn't really a thing

however, copyright might be a thing in those places

so now you can take freebsd, make a product and legally sell it without worrying about PD semantics

dstolfa: but can you? let's say i live in Ipland, and in Ipland you can't put things in the public domain. so the crypt-sha512.c was never PD in Ipland. so freebsd's claim to have relicensed it also can't be valid in Ipland

i am not convinced the fact that it was relicensed by someone who lives in the US (or whatever) affects this

i dunno, this is lawyer talk

if you want to deal with legalese, you're more than welcome to but i won't even pretend to understand how it works :D

i'm mostly just sad that this is even a problem. fuck copyright, etc

the sad thing is that copyright laws got to a point where someone who just wants to release things to the public without ever wanting to deal with it can't really do so without a dozen gotchas

on non-copyright related things, crypt-sha512 apparently accepts short salts deliberately but i wonder if rejecting a zero-length salt is still reasonable

on the basis that this is almost certain user error

this would break existing hashes though :-/

oh ffs, traceroute is contrib

but i hasn't been updated from vendor since 2002 so i'm just going to declare it's not contrib anymore and move it to src

we can do that, right? it's under a BSD license

(3-clause)

👍

oh some of it's 4-clause

docs.freebsd.org/en/articles/license-guide says this is not acceptable

Title: FreeBSD Licensing Policy | FreeBSD Documentation Portal

daemon I am connected to EFNet???

I had no idea

:D

lw: I mean, how much is actually there from the PD implementation? It says 'based on,' so I assume if they licensed it differently it's substantially different enough to be valid

kevans: it's not about whether the freebsd version is different enough to claim copyright on (i'm sure it is) but whether the original public domain dedication is valid to begin with

lw, "oh some of it's 4-clause" The 4-clause license has had the advertising clause 3 rescinded in 1999. If it is an original 4-clause license then it has been officially changed to a 3-clause license. freebsd.org/copyright/license

Title: The 4.4BSD Copyright | The FreeBSD Project

that doesn't invalid all 4-clause, just those issued by UCB

s/invalid/invalidate/

Right. Though I did say "If it is an original 4-clause license" which implies that it was a UCB license.

In which case UCB the license holder has stated that clause 3 "is hereby deleted in its entirety".

I don't see how you make that jump to implying UCB, but I'm really not interested to probe any further

s/interested to/interested enough to/

But that is a good caution. If it is some other entity that is the license holder and they copied the 4-clause license then that rescind order does not apply.

(sorry, not trying to be a dick, just tired)

UCB is the entity that created the 4-clause license. So original would be UCB since they originated it. If another entity copied the 4-clause license then that's no longer the original as I read it since they copied it.

kevans, It's okay! No worries. Sleep is a good restorative. :-)

antranigv, 15:20, * antranigv (~antranigv⊙3272) has joined

if thats you ;)

antranigv, dpaste.org/0FGyf

Title: dpaste/0FGyf (Plain Text)

but yeah that has happened a few times over the last week or so :P

daemon that IS me!

my ISP is going crazy

we are expecting a possible war in the coming days

antranigv, that is not an isp problem that is the znc process quitting

the ISPs are doing their best to make sure things are alright

can tell with the quit message and version etc :P

that znc process is doing /quit

daemon yes, I have a daemon process that checks my ISPs internal stuff, and based on that does a restart. looks like it went nuts with all the changes they are doing

but it got stable as of... well looks like it's not

daemon lol just noticed your name is daemon. isn't that bothering? :D

ah :) anyhow I use znc too so know that its easy to setup server and forget about it so I though would notify you about it :)

it does get a little highlight happy

daemon thank you <3

np ;)

daemon please notify kastein that I disabled my ZNC->EFNet, I will rejoin when connections are back to normal. If they can unkick me please

oh wait, that's a kick, not a ban, ok ok

indeed no ban yet, from my log you have been kicked about 3 times

Title: dpaste/HdK87 (Plain Text)

oops sorry picked up an accidental one for supercat

hello there

I have a problem with bhyve

I don't know if this is the place to ask

can't start any vm because bhyve exited with status 71

fold, Either here or in #bhyve would be good. Simply ask and if anyone who is here knows anything they will say.

antranigv, 21:08, <daimon> trench, kastein, cpet: antranigv is aware of the issue he is working on it

they are cool with it

I don't know what that would mean, and it is not one of the documented exit codes. If no one else knows anything then ask in #bhyve and ask again in an hour or so when different people are here.

daemon thank you sooooo much

np

thanks rwp

Good luck!

hello, I am sorry for posting here but I figure that freebsd folk may be able to help me with zfs. I am using zfs on proxmox and I wanted to increase the size of my mirror by replacing each disk with a larger one etc. I was following the instructions here pve.proxmox.com/wiki/ZFS_on_Linux#_zfs_administration in Changing a failed bootable

device. I set autoexpand to on.

Title: ZFS on Linux - Proxmox VE

Yes I know, sorry left because my kid woke up. I am only here because you guys have a lot more experience on zfs. Since I am on proxmox (debian) I have to run the sgdisk commands which essentially create 3 partitions a) 1007K withoug a label b) ESP partition 512M vfat FAT32 c) the zfs_member. When I put this on gparted after resilvering both drives

I can see an unallocate drive equal to the size of large disk minus small disk.

I know it is off-topic in that it is linux but if you have any ideas on the zfs part that would be awesome. Thanks

forgot tosay that autoexpand=on and that I also run zpool online -e pool drives

meena: they're ordered by commit date, and they're identical