docs.bastillebsd.org/en/latest/chapters/subcommands/rdr.html might be helpful if I understand the question you're asking

Title: rdr — Bastille 0.10.20231125-beta documentation

hello anyone know if between yesterday and today there was been a huge amount of upgraded packages ?

voy4g3r2: also bastillebsd.org/blog/2021/01/13/bas…le-port-redirection-and-persistence maybe

Title: Bastille Port Redirection and Persistence | BastilleBSD

is 3 digit chmod as valid as 4? like chmod 755 vs chmod 0755

Okay so you read the man page and just wasn't sure. Okay. On FreeBSD there is no difference. AFAIK.

hello anyone know if between yesterday and today there was been a huge amount of upgraded packages ?

Are you worried about portability? GNU Coreutils chmod has some small difference with how it handles those cases. Because SysV like file systems like Linux file systems have a set-uid and set-guid directory permission bit. And BSD does not.

mystic, Yes. My desktop system upgraded about 300 of them today. My servers about 25-30 mostly perl modules there.

Did you experience any problems? Or just checking to see if today was normal?

rwp.. no problem, upgraded everything very well. But I suspect my pc was been hacked and then I'm becoming a bit paranoid

What symptoms have you seen that you worry might be indicating an intrusion?

it's pretty strange.. there are a lots of symptonm. One is that I see ads that is targeted on my private life. I mean things that I say in my home, or things that I do in my home. Other symptoms are that there are people that know what I say in private chat.. usually people don't believe me becasue it's very weird

I have often heard people give reports that they think their smart speaker system is listening to their conversations. I don't know if it is true or not. But this sounds less like an intrusion on your FreeBSD system than on advertising using tracking on your web browsing history. That latter we do know they do aggressively.

Let me strongly recommend installing the "uBlock Origin" plugin on your web browser as a security protection from malicious advertising malware.

If you are using Firefox also install the "Multi-Account Container" plugin and use it to compartmentalize your web browsing. Social media site in one container. Banking in a different container. And so on.

You might possibly want to install the "Privacy Badger" plugin from the EFF. I don't happen to be using it but friends do and comment positively on it.

That should for the most part avoid the tracking problem you are reporting. There are still other ways to track but that would help majorly with most of it.

rwp, I'm sorry but I'm not a newbie. I don't have a mic installed on my pc. And my phone is an old nokia without os and without internet. It's physically impossible that my voice is intercepted by my devices.

I am only responding to the information you provided. Unlike the spooks I am unable to listen to your thoughts and don't know anything more. But you did say, "things that I say in my home" and so I can only react to those things you say.

I had not responded to the "what I say in private chat" and was still contemplating that one. But if that was in a web browser then... maybe? I don't know.

I agree that it is very weird. Most of the hard to diagnose problems are weird and hard to diagnose. I wish luck and good fortune upon you. I must relocate. BBIAB.

rwp, yes I'm sorry if I haven't been very polite. So thanks for your answer but it's not a problem that I can solve here.

mystic: it doesn't even need to be your phone

Things like FreeBSD intrusion detection and hardening are good topics. I might talk about using mtree(1). But I must relocate. Later! :-)

ya i want to learn hardening more

all the "blue team" stuff

it can also be anyone who shares your ip address

rtprio: if you're talking about bugs, well I know. If you're talking about something else, please explain me.

rtprio: u mean a wifi ?

the example i'm aware of is; your wife* could have searched for toothpaste from, after mentioning it, from the home wifi

from an observer, talked about it, and now showing ads for it

but it wasn't from your device at all

unless you're of particular intrest to an adversarial nation state...

rtprio: I'm not married. My pc is used only by me. I don't use wifi. I have an openbsd router.

rtprio: yes could be, but it's illegal. I know who he is.. he just hate me and use gov infrstructure for personal use..

pkgbase looks really cool

the plan is to polish pkgbase through 14 and make it default in 15?

does the installer guided zfs let you choose what partition you want to install on?

alepzi: sounds plausible, yes. but we'll see

good good

really gotta be default in 15 then imo

alepzi: the first call for testing was in 2016

I answered it in 2019/2020 by offering a public repo

it's now defunct, because we have an official repo!

thx for helping it on

it's time for pkgbase

my 1 watt thunderbolt power virus is back...

but it's not freebsd-only anymore, it's also an issue on linux

in my xorg.conf i can fire up startx if i reference vesa. i am using radeon drivers prior to startx. How can i tell Xorg to use the a module pertaining to radeon?

what is modesetting driver? what is scfb?

dmr104, I am using the radeon driver and I set kld_list="radeonkms" in /etc/rc.conf in order to load that driver. Is that what you are asking?

the xorg.conf, not the rc.conf

I don't have an xorg.conf file bevause I am using all of the defaults and don't need one.

i seem to have gotten it working with modesetting driver. incidentally, do you use a display environment like lightdm?

I am one of those hippy folks that logs into the vt console and then starts my i3 window manager with xinit using a ~/.xinitrc file for configuration.

When I am setting up systems for others then I do usually install lightdm and use it as an X Display Manager for them to log in graphically. Most people prefer graphical logins.

I am using an AMD HD 5450 which is ancient by today's standards but it is driving dual 1920x1200 displays and works perfectly with the radeonkms driver..

I'll need to upgrade that card in order to drive any newer 2K or 4K displays. But it is quite good for my use.

can your system without a display environment switch between consoles returning to the window manager without encountering the "suspending AIGLX clients for VT switch" message?

What's wrong with that message, though?

hi i am failing; trying to install: /usr/ports/benchmarks/sysbench but it fails on deps of some sort how do i make it so that it also makes the deps too?

dmr104, Yes. No problems. I do it routinely. Is that a problem you are having? What graphics card do you have?

And I must relocate again.

mfisher: thank you, looks like an idea to try

it was odd that iocage and just basic jails.. just "worked" but bastille requires pf.conf which i am like..sure..

Um, what happened to the -6 switch for whois(1)?

Has FreeBSD whois ever had a -6 option? Isn't that a Mac whois option?

i do not see a reference to -6 on my mac

rwp, yes, it did.

And I figured out why the lack of it wasn't working.

My /etc/services file was hosed. . . the mergemaster for my upgrade from 12.4 to 13.2 didn't go well for some files.

Based on the man pages from man.freebsd.org, -6 was dropped in 7.3 and 8.0 releases (assuming 7.3 came out after 8.0 at the time)

yeah. . .. this all just started on this host (that I'm IRCing from) it wasn't working.

Basically, /etc/services was missing so the system didn't know what port whois.iana.org:whois was on.

is the use of the 3d processor on intel for X11 deprecated now

Ellenor, you try ports/x11-servers/xf86-video-i740 or xf86-video-intel ?

aren't both deeply deprecated?

Maybe it's time to switch Wayland.

I tried that once; I would later be convicted of indictable mopery.

interesting bastille makes a /usr/home/ setup for 14.0 but it should be /home

inb4 a political debate with X11 vs Wayland

that can be a scary rabbit hole kenrap

i made the mistake of watching some youtube videos from people like garyntech and robonuggie

now i get a bunch of "wayland vs x11" chaos on the linux land

a bunch of whining

tmux -> ssh -> tmux -> neovim == breakage of mappings, hardcore

kenrap, kinda

haha, do not get me started, i can tell when i am BSD too long, i am trying to do it in my terminal for mac

and like wtf.

now only if i can figure out how to get iTerm to show me which host i am on in the tabs.. but that is a whole other matter

Thinking about, X11 during it's very early stages kind of went through similar "breaking" stages akin to what people are already bitching about with Wayland. It just took a couple of decades for it to develop adhoc standards that people decided to follow. It was just unfortunate that serious adoption of Wayland development took over decade to happen.

I use screen, not tmux, but screen has a way to deal with being multiple screen sessions deep.

Basically, if you want to send a hotkey two sessions deep, you hit control-A twice.

Surely, tmux has thought of that?

CrtxReavr: quite possible, i was a long time screen user, then i "learned" about all the plugins and blah blah.. i got sucked in

man this sanoid/syncoid setup is humming along.. it is probably not liking how i am deleting and adding bastille jails but it is holding up

I've three decades of screen muscle memory.

Not switching to tmux now.

if it works, why change it

CrtxReavr, With default tmux, I believe you can Ctrl-B B. I've rebound my keys to Ctrl-A, and the B shortcut to A also.

which sends Ctrl-A into the terminal.

oh sweet.. thanks Ellenor

Now that I recall it, screen can also do serial device connections, which tmux lacks. So CrtxReavr can still benefit with screen for embedded development

For that, I'd use tip or cu, depending on OS.

Though. . . how many things have actual serial ports these days?

raspberry pi stuff, arduino, etc

Don't those just use USB?

oh wait, you said "actual"

anyway, I didn't mean to derial from tmux discussion

CrtxReavr, I am a newcomer having started with 12 so that's my excuse for not knowing about whois -6 here. :-)

I rebind the prefix character to Control-Z for both screen and tmux and both can pass it on by doubling it. I use job control all of the time outside of screen/tmux but in tmux I just start a new pane and keep it in the foreground in a pane. Using tmux is an alternate way obviating the need for job control.

Cut my teeth on 3.0.

I missed out on the earlier fun due to my own poor decision making skills but I am working the task of learning FreeBSD in depth now.

I am having problems connecting to my webmin. I have it installed and configured properly but chromium is saying localhost refused to connect

Can anyone give me any advice?

Nevermind I got it

is the year 2008 again?

poeple still use it; i didn't know it worked on freebsd

rwp: tmux is great

rwp: i used to update multiple systems (somtimes up to 8) with tmux synchronize-panes

(?)

Was kern.maxfiles increased significantly from 13.2 to 14.0? I see on a 13.2 server 517159 and on my 14.0 server 4189577

I'm looking into why editors/vscode keeps failing when build using poudriere where I have set max_files_vscode=4096 since the server is 14.0 with "too many open files"

Trying with max_files_vscode=8192 right now.

f451: what filesystem snapshotting helps with, is that you have a snapshot of exactly that moment, it could still be that applications are in the middle of writting something to the filesystem, but at least you can't get a mix of multiple states as if you do your backup by walking the filesystem, so you can't have that you backed up some application data, then the application writes new data and

metadata and then you backup the metadata, which would be a corrupted state for the application as it has new metadata but old data, still it could be in the middle of writing and not all applications can recover from that, but you would get in the same state if your system looses power in that moment

Always confuse the situation.

The "pkg" tools needs being always online for ANY funtionality. This is unfortunate, bad design!

...yes?

Not sure what you expected zagorkarabela

Say i install a package from it own cache it is refused because of its not being ONLINE.

it shouldn't need any network for pkg {add ,info,query,set,annotate,remove,autoremove,lock,unlock,which} and maybe others

Say I issued "pkg search" or "pkg info" or ...  it is ALWAYS refused because it is not ONLINE! This is bad design, unfortunate.

See  all other distros PACKAGE_MANAGER' s.

it is save db to local. AND is is done. All cmds works as expected!

May be i dont now because i am newbee. If it is the case please inform me to how to work with "pkg" tool.

pkg has a local db, pkg info should not try anything with the network

pkg install  always needs internet. Not lokup its own cache for whether pkgs is available there or not!

lw

Hi folks, is there any way to generate a pwn signal to a GPIO pin via gpioctl?

rogersm, what about pwm(8)?

oops

I had no idea it existed

I want to control the pwm fan via GPIO

well it needs pwmc(4)

rogersm, check whether you have a pwmc(4) driver for suitable hardware first

bit-banging PWM through a GPIO is extremely inefficient and inaccurate

I'm not too concerned. I just want the fan to spin to reduce cpu temp

is this a regular x64 type machine? why not just doing it in the BIOS/EFI ?

it's a raspberry pi

I feel like you'd get much better answers by just clearly stating your hardware :p

which RPi version?

rpi4

FreeBSD rpi-4 13.2-RELEASE-p3 FreeBSD 13.2-RELEASE-p3 GENERIC arm64

to prevent more layers of the xy-problem here: you can just slap on a fan without PWM

I don't have an rpi around right now but if you need/want PWM I'd suggest to first check whether you already have a suitable driver for that :)

thanks, I'll investigate the pwm route

FreeBSD 14.0-RELEASE is doing weird thing to my wifi. It's receives a DHCP lease, nameservers and search domain are set, as well as gateway, but when I do a ping to either a FQDN, hostname or IP address I get "the network is down". restart netif and it works. anyone got an idea what this might be?

Laptop was upgrades from 13.2 that didn't experience this issue ever.

weust, let me guess... iwlwifi? :D

yup

that is flacky on mine too :(

still under development

if you're lucky to have old enough hardware to be supported by iwm or iwn you could just fall back to that instead

I know. Just didn't expect to break from 13.2 to 14.0 like this

I'd recommend you to bother bz@ on the wireless mailing list :p

but first, check bugzilla for corresponding PRs

8265 / 8275 model. but iirc it's a 8265.

aight

before installing iwmbt-firmware it was far worse. Had to restart netif multiple times to get it working fully

is installing firmware a manual procedure?

I just load the driver and that's it

in my case I often seem to have to reboot

it associates, then a few seconds looses it and then it's stuck until I properly reboot

from what I can tell it's automatically, because I only need to restart netif once now

I remember having seen some messages in dmesg that indicate that loading firmware failed but oh well

time to compile llvm17 I guess

jbo: as regards bit-banging pwm inefficiency, do you mean that if one tries to write its own code to modulate the signal, the code will perform in an inefficient way?

rockyh, bitbanging means controlling the gpio state from userland. given that your typical 4-pin PWM fan requires a frequency range from 21 to 28 kHz you're gonna have a lot of software running to control a fan.

hence there are such things as dedicated PWM hardware: you give it a duty cycle value and it just keeps going without software overheard

yes, I was meaning from userland. Ok, so using (if available) pwmc(4) it's better because this operation is performed throurh a dedicated hardware

thanks!

Ellenor: i don't have a window manager installed at the moment, and it won't return to the display when i return to the console

weust: I have an 8265 here too, it doesn't seem to kick off dhclient properly here last time I checked it.

perhaps we should put a PR together on this

I run iwm still because suspend/resume just works, and thats what I need most in a laptop

dch: that seems to work alright here. I mean, I get a full lease and everything. but after that it's like traffic is blocked.

right, slightly different then

what would be the new way after iwm then right now?

weust: but your dhcp works at boot?

yeah, just fine

just use iwm and keep banging on the PR drum

before installing iwmbt-firmware it wasn't. couldn't even connect to the SSID, but since installing that is fine. super fast

hehe

oh i probably dont have the firmware

let me try this and compare notes

kinda busy with work too, so trying to have a quick peek at all the bugs under 273620 it depends on

alright

Only thing I noticed, is that my device ID doesn't match with the iwmbtfw man page.

reading a lot about crashes and kernel crashes, but I've never witnessed that before

kernel panics*

weust: I don't see this issue here, and now I'm loading the firmware, it connects automatically again (bonus)

my dmesg dpaste.org/gFo4U for comparison

Title: dpaste/gFo4U (Python)

it loads iwlwifi0-8265036.ucode

and my chip is 8265 REV 0x230 for comparison

and a crf-id 0xbadcafe ... suspicious :D

we need a oxgoodcafe

weust: this is definitely bad cafe, I got a panic in under 30s of iperf3

probably fixed on current, I'll install that later today

Here's a take so warm, it'll qualify as ice coffee: I definitely prefer a good café to a bad café.

:-)

in my dmesg I don't see any iwlwifi reference

debdrup: definitely a lukewarm café

weust: then maybe you're not using iwlwifi? I have an exclude for that

dch: I don't mind if the café is lukewarm, it's annoying when it's cold.

dch: I guess. but kldstat shows the module as loaded

weust: if it's not attaching, check the device IDs with pciconf

Wait, should it be attaching?

Vendor=0x8086 device=0x24fd

And should iwlwifi(4) attach to those?

tbh, I don't know. Iwas assuming that

Assuming is how you make an ass out of you and me. :)

The net.wlan.devices OID in sysctl(8) should tell you which device you have.

not on purpose..

weust: assume = ass u me ;)

iwm0

debdrup: aah :P

Well then, it's not iwlwifi(4) that's the problem.

pciconfig also shows iwm0@pci etc

..also, my T480s apparently has iwm(4) too.

Shows how much I use it :D

wiki.freebsd.org/WiFi/Iwlwifi is a bit weird in the FAQ.

Title: WiFi/Iwlwifi - FreeBSD Wiki

Q: I have an iwm(4) wireless card, can I try the new driver but fallback to iwm(4) for daily use for now?

A: Yes! See devmatch question above and Testing section below.

Q: I have an iwn(4) wireless card, can I try the new driver?

A: No. The dwm support of the iwlwifi driver is GPL-only code and thus not supported.

So can I use it or not?

weust: dev.iwm.0.debug OID in sysctl(8) and wiki.freebsd.org/WiFi/Debugging would be the next step.

Title: WiFi/Debugging - FreeBSD Wiki

OK

Do you _want_ to use iwlwifi(4)?

Before debugging, you need to decide what device you want to use.

well, right now I am not sure. I thought I was using iwlwifi, but iwm works apart from the weird network is down after setting everything up just fine on boot

Maybe rootcause the iwm(4) issue instead of switching drivers would be a better use of your time, as it seems like a regression has been introduced if it was working in 13.2?

yeah, that makes sense

first thing I want to figure out is if the iwmbt-firmware file is loaded, and which one. dmesg isn't showing that.

iwmbt is kinda useless here. you can specify a firmware path, but not list/show what is currently in use

weust: you can see what the `parent interface` is via `ifconfig wlan0` and the firmware I guess is only in dmesg /var/run/dmesg.boot

in my /etc/rc.conf I have both

devmatch_blocklist=if_iwlwifi

wlans_iwm0=wlan0

and when I switch to iwlwifi, I use instead:

devmatch_blocklist=if_iwm

wlans_iwlwifi0=wlan0

mkay

parent interface for wlan0 is iwm0. that makes sense. I have that same line in my /etc/rc.conf

only thing I see in /var/run/dmesg.boot is lines like: ACPI: \134_SB.WMI1.WQBA: etc etc

and that doesn't seem to be a file, I believe

wg devices in 14 have to be set up "by hand", right? ifconfig create, ifconfig inet, route, wg setconf?

the rc.d script from wireguard-tools works fine

you can also install wireguard-tools and use wg-quick, if you want to use a nearly 500 lines long bash script (that rc.d script of wireguard-tools uses it too)

Title: wireguard_lite.in « files « wireguard-tools « net - ports - FreeBSD ports tree

was looking at wireguard_lite just when mage responsed

there is also reviews.freebsd.org/D41318

Title: ⚙ D41318 Add the "missing" WireGuard rc.d script

dch: debdrup: comms/iwmbt-firmware is only for Bluetooth support for certain cards. I read it wrong

wait, when did that wireguard_lite rc.d script appeare?

thanks, so I wasn't overlooking something...

wireguard-tools-lite be it, then :)

with wireguard_enable and wireguard_interfaces in rc.conf it now makes better sense

i get garbled screen in gnome when it starts lockscreen

formely also the mouse would disappear, but logs showed that the package xdg-dbus-proxy was missing

i installed that and hope now that the issue with the mouse pointer is solved

but i still get the garbled screen, i have to sendkey ctrl-alt-f1 and back in order for it to restore to its original state, and when it restores i also see a stuttering input in gnome-terminal

at least 20-40 random chars appear there

Inline: got the drivers working last night, your tip about the pkg search function came in handy

aha good to know

dch: Gave iwlwifi a try. perhaps my 8265 is too old, but it utterly fails :-)

and works now ?

did you have to fiddle too much ?

weust shade :-( I am sure bz will want to know, he just asked me about mine

dch: yeah, will try to document everything I can and send it over

I used the /etc/rc.conf lines you put here.

is there a way to snoop on a socket file ? sort of like tcpdump but for sockets ?

in the worst case you can use dtrace, but maybe there is a better/easier method

anyone have an opinion on iso vs img formats for whatever tech work you do?

after i write a freebsd install iso to flash drive, any way to verify it was written cleanly? i checked its shasum against the CHECKSUM file but they didn't match for that's probably expected

You should be able to read it back from the flash drive and get the same shasum. The tricky part is reading the exact same size of image when the flash drive is larger than the image.

I would use dd where I can specify the size to read.

I assume you are wanting to verify the flash media as there are so many counterfeit flash drives now. It seems impossible to buy one now without getting a counterfeit one.

i just wanted to ensure a clean write, but wow that's really interesting. why is it so hard? i know amazon is a stinky garage sale but what about ordering directly from a reputable mfg?

I guess buying NAND Flash has become like buying drugs. "Know your dealer."

alepzi-, Earlier you asked "anyone have an opinion on iso vs img formats for whatever tech work you do?" and I did not understand that question. But now I can respond saying that for bare metal installs I always use ISO images written to USB NAND Flash storage.

me too!

.iso & .img are just part of the file name. . . 'file -s <that filename>' to learn more.

Also, remember that files are just arrays of bytes.

Trying to find out if smbus and pmbus is compatible. Im able to find a implementation for smbus but none for pmbus, are they perhaps the same thing but different names.

When I saw img I was thinking, is that asking about the .raw SD card images as described here: freebsd.org/releases/14.0R/announce

Title: FreeBSD 14.0-RELEASE Announcement | The FreeBSD Project

from freebsd repo i made an iso then i made an img. img was way bigger try it yourself

What did you do to make an img from an iso? I can think of several things that might mean.

Because if it is installing an iso and then making an img of the entire partition slice then yes an image will be MUCH larger than an iso.

no i made an iso from freebsd repo, then i made img from freebsd repo

using the standard releng build scripts

I have not yet used the SD card img images. I'll just say I assume they work fine. :-)

is there a backblaze annual report but for flash mem sticks?

They do have a report for SSDs and NVMe IIRC.

pajus claims he can hack debian

crack, not hack

I am hacking on FreeBSD, Debian, Ubuntu, Trisquel, Rocky, all day long.

First correct use today! \o/

OpenBSD is more stable because it has fewer hackers

error, missing one or more scripted disks! during an unattended bsdinstall. i assume i put the disks wrong in my installerconfig. export ZFSBOOT_DISKS="nda0 nda1" but that's wrong? fwiw they're 2 nvme drives

disks

crack, hack, hjesrk off, whatever

compromise

this makes him one of the most wanted people on the planet

drobban: pmbus is an extension of smbus

higher speed allowed, more registers standardized

(botyh are built on i2c)

babz: thanks for the info.

anyone who can hack debian should be eliminated

as rwp said, know your dealer

q

babz: are you accusing me of taking DRUGS? >:O

maybe the devil made you do it

lol

WHAT??

sounds like some fun conversation for #freebsd-social?

when i run git clone github.com/freebsd/freebsd-src --single-branch releng/13.2 --depth 1, it creates the dir tree releng/13.2/<repo here> but how can i have it just create freebsd-src/<repo here>?

Title: GitHub - freebsd/freebsd-src: The FreeBSD src tree publish-only repository. Experimenting with 'simple' pull requests....

alepzi-: --single-branch doesn't take an argument. you're looking for -b

--branch

or --single-branch --branch

any difference between just -b releng/13.2 and --single-branch -b releng/13.2?

seems like the former is cloning/pulling the whole repo and checking out a branch, whereas the latter isn't cloning the whole repo, only that one branch. but the two branches would be identical

i tried just -b version and git branch inside it only showed releng/13.2 branch, the active 1

why's that?

scoobybejesus: ^

what was your whole command? git pull -b releng/13.2 ?

no git clone

git clone ../freebsd-src -b releng/13.2 --depth 1

then i cd into freebsd-src, git branch shows 1 branch

someone else test it to verify?

what other branches did you expect to see? since you specified releng/13.2, that might be why there's only one

no i only wanted that 1

i had a short article on SR-IOV and was asked if i would expand it a bit. behold! :) markmcb.com/freebsd/vs_linux/sriov_is_first_class

Title: SR-IOV is a First Class FreeBSD Feature - Mark McBride

what are some fun things to do with FreeBSD

I've already formatted my terminal with colours

anything you would also do with linux, more or less

Here's one you can't do with linux, change your console font to one that includes east asian characters

I'm currently typing this from a freebsd console, no desktop environment whatsoever, and this console displays traditional chinese characters.

neat

In addition, we have a mouse in our console, nobody else has that.

yay!

doesn't linux have gpm, or whatever it's called?

(for mouse in the console)

I'm not familiar

I just looked it up, gpm seems like a cool add on. we have it as part of the design.

AumShivaya: FreeBSD has a utility for managing SCSI Enclosure Services (SES) devices that I've not found an equivalent for in Linux: sesutil. Quite hand if you have a lot of disks.

*handy

nope :)

Did you install with ZFS? If so, check out boot environments.

Title: bectl

boot environments? let me see. No, however, I did install with UFS

Linux does routinely have gpm. However in FreeBSD I now set moused_nondefault_enable="NO" in /etc/rc.conf to AVOID having a mouse in the vt console. Because I am an example that you can't make everyone happy! :-)

Linux does not have Boot Environments. That's definitely a Root on ZFS benefit that really cool.

jbo: i'm testporting mediaelch@all now, should have a new patch today (although poudriere wants to rebuild half the world first)