do you use bs=2M or 1M for dd reformatting usb flash sticks?

polyex, I think both/either of 2M or 1M are both larger than the NAND flash stripe size and also a multiple. Meaning that either is okay.

I most often use 1M when I am writing NAND flash these days.

hi

Wait so FreeBSD 4.4 changed the functionality of `install` to `cp` not `mv`? =(

hi. yes.

FreeBSD 13 removed CU-SeeMe, this thing right? - en.wikipedia.org/wiki/CU-SeeMe

Title: CU-SeeMe - Wikipedia

haven't heard of that software in like 15 years

How else will I make my point-to-point 4-bit grayscale videocalls?

not as much of a need these days

I'm wondering if any of you folks have an example of a pf rule that would block traffic rom a jail to the rest of the LAN while allowing it to still reach the internet (e.g. any traffic from the bridge interface)

Hi all, is there a "smarter" `pkg clean` that would remove all but the N last packages? (a bit à la man.archlinux.org/man/paccache.8)

SamuuelMa: mv is cp + rm

err, wait - that's if it's crosses filesystem boundaries.

When a setup.cfg contains, coverage[toml]>=6.5.0, is this the right DEPENDS line? ${PYTHON_PKGNAMEPREFIX}coverage>6.5.0:devel/py-coverage@${PY_FLAVOR}

hello bsd folks. I have a pf problem if anyone would care to indulge?

just ask

Hrm. Anyone else have python39 installed via ports and getting hundreds (maybe thousands?) of checksum mismatches in their periodic daily reports?

I'm seeing this on multiple FBSD 13.1 systems.

paulf, it's double nat. my pf drops everything by default and no rules want to play nice with the port I'm trying to allow through

the upstream router works just fine, just not my router :(

ek, I had something happen a week or two ago that sounds like what you describe, although for me it was using pkg. Solved with "pkg update -f" followed by "pkg upgrade", IIRC.

V_PauAmma_V: Hrm. I tried rebuilding all the ports that are reporting the bad checksums. Everything rebuilt fine. However, immediately running another "pkg check" there are still a bunch of mismatched checksums. Not sure what's going on.

Long shot maybe, but do you also have another python 3.x installed?

V_PauAmma_V: it's the bytecode stuff that got reverted because someone got overzealous

ek: too

V_PauAmma_V: Naw. Just the python39 installed at the moment.

and this is why i'm not committing anything to the repositories for the foreseeable future

vishwin: Related to the github fiasco?

no

vishwin: Ah. I wasn't aware of such an issue.

the checksum mismatches with CPython bytecode files are from python packages still using direct setup.py invocations packaging bytecode, but then the (now reverted) trigger compiling them after install

they are harmless

i would direct you to the UPDATING entry but that got swept up as well

combined with nobody, including docs people, actually reading their own shit, is why we can't have nice things...ever...

vishwin: Oh! That makes perfect sense. Exactly what I'm seeing. Thanks!

yeah CPython bytecode isn't meant to be packaged, they're always supposed to be compiled on the target if desired

The periodic daily reports are so large that sendmail hadsa problem delivering them.

s/hadsa/has a/

i wanted to bundle the bytecode compilation for those cases into an overhaul of that "subsystem" to reduce PORTREVISION, but that obviously got derailed

s/PORTREVISION/PORTREVISION bumps/

Am I right when I say there is no way to change vt resolution without a KMS driver?

It's not really dependent on the KMS driver.

If you have UEFI with GOP and set it via the loader, the mode shouldn't change once you've set it.

debdrup there is no such rule.

prahou that is true.

Maybe I've just been lucky then.-

no UEFI here

thanks tsoome

prahou: if you do n't have UEFI, vidcontrol(1) can change sc(4)

but I don't want sc

I need unicode fonts

the problem is, some systems limit screen resolution, and you may want to get better ...

just making sure I didn't miss anything, thanks folks

prahou: where've you found a system produced in the last decade that doesn't have UEFI? :P

for example, my system (supermicro X10SAE) does only set 800x600, but the card can do up to 1600x1400. And BIOS mode goes happily with it. 800x600 is rather ugly on this 4k monitor:D

debdrup: who said anything about the last decade? ;)

vt has mechanism to have KMS backed gfx driver doing the draw

I have not checked what is needed to get the actual resolution change happening.

It works well here with the KMS driver (the resolution changes anyway), but it borks acpi suspend, which I couldn't resolve, and I tried.

prahou: really? What breaks ACPI S3 here is the TPM-

debdrup: it's a trial and error endeavor. but I can safely say that radeonkms is at least to some extent a culprit

I don't even know where to ask about this, so I'll try here first.

I'm trying to setup two OPNsense VMs (FreeBSD 13) and on one machine (Ubuntu 22.04) it works fine but on another (Ubuntu Studio 22.04) network access is not working.

I have pretty much identical network setups on both hosts (netplan, both laid out the same)

On the machine that it's not working on I have 4 other VMs (one windows, various linux) and they all work just fine. What could possibly be going on with FreeBSD that it doesn't want to work on this system?

Hi all, anyone have experience with FBSD 13.1 and Samba - Windows 11 22H2 clients - and also, is Samba 4.16 not suitable should I revert to 4.13?

roy2098: Yes 22H2 here.

Hi skered - which Samba version is working, can you provide any info on what you did to get network shares to work?

roy2098: 4.16 and the samething I did with 4.13 and all versions before that.

[name] path = ...

Nothing comples

complex even

did you make any changes to local security policies?

Just security = user

skered: also, which version FBSD?

and map to guest = Bad User

13.1 amd64

skered that's my environment - only twist is Samba is virtualized in one case under bhyve and another case under VirtualBox

skered: and that brings another issue - I wonder if there's a bug in the igb interface in FBSD 13.1

skered: i've already got too many variables

skered: what set this off is need to upgrade a working Samba 4.4 I think it was under FBSD 10.2 - both guest and hypervisor

maybe. But my issue with VMs network interfaces are mainly speed.

But then I always go paravirt for all of them

skered: I upgraded first the hypervisor(s) no problem

skered: then the fbsd guests

skered on one of the hypervisors, a previously older VirtualBox fbsd guest would not network

Are your issues with direct connections or you expecthing them to show up automagicly in Network?

skered: net use, File Explorer with ip address, nothing works now

skered: ironically, it does work fine under a FBSD 13.0 release (both hypervisor and guest)

skered - paravirt didn't help at all

skered - apologies I made an error - should be "on one of the hypervisors, a previously older VirtualBox Windows 10 guest would not network

skered - and you may be right about firewall/NAT . These are ubiquiti EdgeRouters and switches and maybe the addresses are stale or something

skered but you are telling me that you have working FBSD v 13.1, Samba 4.16 working with Win11 boxes with latest 22H2 update

skered so that means i have more digging to do

Yeah win11 (all versions), win10, macOS (with fruit/timemachine), and Linux.

skered: did you compile Samba or get from pkg? And any special switches/configs you set?

FreeBSD repos.

skered: not sure what you mean by that exactly...

pkg.

ok. so nothing set other than defaults

Yeah

btw are you using any Samba as domain controller?

Nope

I can only assume this is some type of information leakage attack? Should I unplug my Internet connection?

skered - are you kidding?

skered - why would i attack you?

Aren't we all?

hahaha

look, i appreciate your help; i have to head out now but if it's ok i'll look for you again. Thanks again!

Hi, are there any tools I could use to inspect wired memory?

is there any word when drm-511-kmod will drop? the PRs for 5.11 and 5.12 were merged months ago.