almost 14, nov 14, fbsd 14. will they release at 14h14m14s ?

Lol

woop 14.0-RELEASE soon

woop that means plenty of work

paste.purplehat.org/view/b55b1905 <--- anyone that can answer why portmaster says like this?

Title: portmaster - PHO Paste

Can FreBSD run Android Apps ?

paste.purplehat.org/view/b55b1905 <--- anyone that can answer why portmaster says like this?

Title: portmaster - PHO Paste

hm, not sure what portmaster uses to decide which are important packages and which are only there because they are build dependencies, are those included in portmaster --list-origins ?

nimaje: it seems all those packages that needs to update is in --list-origins to

openssl was not i noticed now

py39 openssl harfbuzz was not listed with portmaster --list-origins

had simular on another box to update but it worked flawless on that one

Hi

Good afternoon

n30: sorry no idea, maybe portmaster --update-if-newer '*' considers everything you have on your system, but you should consider switching to something doing isolated builds like synth or poudriere as your system isn't changed while building, so you don't have a partial updated system and you don't have the build dependencies installed on your system, only in the jails used to build the packages

Hi!

Hi

hi

I have this device (it is from OpenBSD host):

Ugen0 at uhub0 port 3 "Qualcomm Incorporated Qualcomm HS-USB" rev 2.00/0.01 addr 2

It is wwan interface, I inserted sim-card

But OpenBSD doesn't recorgnise it as an internet interface

So, where can I find info on what hardware is supported on FreeBSD?

Or better yet, can I get a live FreeBSD image to check? So I can check from FreeBSD host without having to install and delete my current installation

It's a USB device, right?

You could try installing FreeBSD in a virtualbox and attach the device and see if it works?

It is integrated

I think it is connected to internal USB on MoBo instead of connected to PCIe

(It is PCIe socket but I think it goes through USB bus lanes and is just PCIe to USB adapter so to speak, then connected to USB internals)

How can I get audio through hdmi? AMD APU(iGPU)

figure out which sound device it is, eg via cat /dev/sndstat and then change hw.snd.default_unit accordingly via sysctl

nimaje: Doesn't work?

Oh wait

That was setting default unit, not changing

But thanks! I can work from herr

Wait... Hmm...

I restarted mixer service but still no sound

I mean, sound is still on laptop speaker and not through HDMI

output of cat /dev/sndstat

I did: sysctl hw.snd.default_unit=0

still, device is the laptop speakers

hm, do you hear any sound? can you try cat /dev/random > /dev/dsp0 and the same for 1 and 2? (just terminate it after a few seconds via ctrl+c)

That gave sound on correct device

nimaje

I tried restarting firefox as well, incase firefox was "cached" on wrong device

Wait, let me try to restart x

so dsp0 is the correct device and works? hm, no idea then

Restarting x did not help

ok so that command needs a restart of another service to apply

audio/pulseaudio according to wiki, or pacmd and then input set-default-sink 0

To change on the fly

The command you gave needs reboot (and I guess persists between boots, obviously, lmao)

Hello everyone. With the 14-RELEASE just arround the cornor I would like to have your input on upgrading. More specific on when to upgrade. Is it considered best practice for production systems to wait for a .1 release?

thorre, I don't know offhand how applicable that advice was beyond my own circumstances and those specific versions, but someone (I think kevans) steered me toward 12.3 instead of 13.0 (which came close to each other) when I asked which to upgrade my daily driver (and single FreeBSD computer) to.

I feel that if one is asking that they should avoid the .0 releases. Especially if it is their one single main computer that must work for them.

On the other hand if you have a dozen machines all doing things and have been running it for a while then you definitely should install the .0 on a system and test it out and report bugs.

Thank you for your feedback. I will not jump the gun next week when 14 is released but give it a little time before I upgrade my main machine.

In any case, I strongly recommend testing an upgrade on a non-critical machine before moving on to production machines (on Fridays).

/set irc.look.smart_filter on

oops

lmao

:P

my look smart filter would probably just translate to /quit, so i stop embarrassing myself

zyxer: as stuff dynamically set via sysctl doesn't survive a reboot, it of course doesn't need a reboot, but no idea about pulseaudio, I just assumed you used sound directly and not pulseaudio on top, no idea if it notices when you change the default audio device that way

Fi phy pho plum. I smell a bunch of ... Crap I can't rhyme that without being mean. Oh well. What's new?

hi

i think i have found a bug in bash but i am not sure. Maybe it is bug in my script

but I have a problem with a weird behaviour or bash

may I share my script with you?

maybe first I check in #bash...

put the script on a paste site

58 - commended export = works fine

line 58 - uncommended export = eval in 59 runs 3 times and my desktop hungs

i fixed regex =~ ^\?

btw :)

sorry - not 58 but 59 (export line)

is that eval supposed to be re-executing this script?

no

i dont know how to push "value" with quotes but visible for child process as without them

because "bash run $string" where string is (--x "hello") results with that, the child script has x as "hello" with quotes

that's correct, yes

anyway, without eval i had the same issue

if you want to make it bash-specific, that's what arrays are for

will change it but first must not hung my desktop :|

what is ${arg_plugin}.sh whatever that is doing with XMENU_STRING ?

arg_plugin is --plugin argument. It is for example (--plugin rofi)

so when it is rofi then XMENU_STRING is a command to run rofi with theme and theme-str

plugin does for example (ls -la | $XMENU_STRING)

and rofi shows up

otherwise any other xmenu like dmenu (plan)

so XMENU_STRING is supposed to be a valid command and args after word-splitting

word-splitting?

but yes it is entire commad for plugins to run to show menu

you know that when you write $var in a script, the value of var is split into words, yes?

i.e. if you do var="1 2 3" and then somecommand $var then the command gets three args, yes?

yes

and that word-splitting doesn't respect internal quotation inside the value of var

plugin use eval

i.e. if the value of var is 1 "2 3" 4, e.g. if you set it as var="1 \"2 3\" 4" then it splits into 4 words, not 3

plugin uses eval how exactly?

sorry right. Plugins use eval, so "ls -la | eval $XMENU_STRING"

will change it to array

ugh. eval $XMENU_STRING and not eval "$XMENU_STRING" ? the difference matters a LOT

it matters depending on what XMENU_STRING can hold

but as long as you don't know, quoting that is probably a good idea.

wait...

should be with ""

is without

put differently, unless you WANT word splitting to happen on a " ", you want to quote that thing, yea.

the problem is that eval $XMENU_STRING ends up word-splitting the value, then concatenating it again

dont understand

so it messes up things like internal spacing in what you think are quoted strings

TMLP. too-many-layers-problem.

as I see, rofi launches 3 times and hangs desktop

hmmm

i.e. "foo bar" is split into "foo and bar" and then concatenated into "foo bar"

yeah, pretty much any use of eval in shell is a Big Red Flag

any use of word-splitting is a red flag, too

Guys when i put in a ssd in this laptop how do i get my system to reognize it ? i know the bios must recgnoize it. after that do i jsut format it with the freebsd installer and i should be fine or is there somethign else i need to do ?

RhodiumToad: it depends, if you really really know what you're doing, and you control the input, and you can predict the input, in THAT one case I might not see a problem.

jb1277976: as an additional disk? what do you want to use it for? or replacing the system disk?

armin: anyone who thinks they know what they're doing when it comes to shell is probably wrong

RhodiumToad: replacing the system disk

RhodiumToad: true.

jb1277976: as long as the bios sees it, then the freebsd installer should be able to handle the rest

Thanks

armin: for example, the only reason why the whole of /etc/rc* doesn't explode and die is that key paths like /etc are known not to contain whitespace or special characters

if I had infinite time I'd build a system where paths.h was modified to rename /etc to "/System Configuration" or some such, and document all the explosions

RhodiumToad: A large part of shell scripting is that you need to make assumptions. If the assumption you make is known to work in almost any damn case you can think of, why bother? Shell scripting is known as a "glue language" thing and I think that fits it quite well. You can use it to do almost anything you can think of, but you also get all possibilities to shoot yourself in the foot.

RhodiumToad: Of course, you *WILL* eventually shoot yourself in the foot while trying to master shell languages, no matter the interpreter, but that doesn't mean that interactive shell-use or shell-scripting is bad per-se, it's the other way around.

RhodiumToad: We need some kind of interface to speak to a computer and a shell is just a very perfect one.

RhodiumToad: for example you wouldn't check if a directory exists if you just mkdir -p'd it one line above...

RhodiumToad: but then again, someone evil might have created a file with that name before, what do you do? Check for all of these weird edge-cases? I don't think so.

RhodiumToad: I'd say many programmers will just live with that and assume that the user in SUCH a weird case knows what to do.

i have no idea how to run rofi now

eval hungs

variable - errors

bash is the pease of sh*

piece

bash is a horror, yes, that's why I love it.

i will try python wrapper

you can also just try to share your script (if you can share that data), I'll be happy to at least *try* to help you. <3

this issue would be the same in other shells, it's not likely to be bash-specific

That's correct.

but i can do system()

only for this one command

system(VALUE_FROM_ENV)

i have no idea why eval hungs it is non sense

doing system() from python or whatever is conceptually the same as eval "$var" in shell

it's not eval that is hanging, it's whatever you ended up invoking

have you used set -x to see what's actually being run?

garbage

this mode isn't helpful hre

of course it is helpful

Well set -x shows *exactly* what's actually being done, it could be.

it doesn't help me at all

but i will try -x to plugin

the script you pasted isn't the one with the issue.

$HOME/.xmenu/plugins/${arg_plugin}.sh <-- that's the one you need to be looking at

wait

this command is hanging

i copied it to xterm

and hungs

wtf

must check it

it is rofi fault

great

found a problem

you can'tchange rofi position by -theme-str

if you try rofi hungs

wrong - you must set location AND ancor

thanks for your time

i need something better than htop

can you suggest something guys?

tried "btop" or "bpytop"?

htop require double pressing F9 to kill

it is annoying

ctrl-c works for me

ah you mean to kill a process?

btop is to fancy

yes

when you move with arrows then you need press F9 twice

or maybe not

so htop has lags

not it is working well

i have no idea why

it didnt

ok thans for you time

again :)

must restart my computer

bye