I understand that tar --list --file=x will list all files and folders. I am looking to just list the-top level data. Does anyone know how to do that? Alternatively, does anyone know how to list only the top-level files, but all folders including subfolders? Maybe with grep somehow? I'm after something that works on most nix flavors including MacOS.

Hello Unix/Linux community, looking for your help. I am running into an issue where I am trying to install Catalina 10.15.2 on a machine that has been wiped. The machine is in target disk mode and I am using the "Install macOS Catalina" app to image this mac. When I launch the app I see all of my mounted devices including the machine that is in TDM but when I select it as the disk where I want Catalina installed I get a message that states that: "You may not install to this volume because it is a Mac in target disk mode". Is it not possible to image a mac that is in TDM? BTW I have tried imaging the disk as APFS and Mac Os Extended (Journaled). Any help/guidance is appreciated. Thank you.

I'm trying to use Linux bash config on a mac. I don't use zsh only bash on my mac. I have problem with this variable:

# After each command, append to the history file and reread it
PROMPT_COMMAND="${PROMPT_COMMAND:+$PROMPT_COMMAND$'\n'}history -a; history -c; history -r"

I'm not sure what is the source of this code, but i use it on Linux to keep all history from different terminal tabs in sync. But on mac I got empty history. I'm not sure if this something that was from the start after I added this to my .bashrc or something that appeared just now. I didn't notice any problems before, but I use mac only for few weeks, and I may not use history at all. I have fzf setup, but I may not use CTRL+R at all. And I'm not sure if I used up arrow. This is something that I don't pay much attention to, I do this automatically.

Here on MacOS Catalina, when checking the memory usage of a process I see that the ps command shows a RSS value which is different from the memory usage shown in top:

$> ps e -o command,vsize,rss,%mem|grep "myapplication"|head -n 1
myapplication  4594896  51364  0.3

**RSS -> 51364**

top
PID    COMMAND       %CPU TIME    #TH             #WQ  #PORT MEM    
48106  myapplication 115.7        09:06.12 69/1   1    101   37M+

**MEM -> 37M** Why this difference? **UPDATE:** Another example with IntelliJ process:

top -pid 357

PID  COMMAND      %CPU TIME     #TH   #WQ  #POR MEM    PURG CMPRS  PGRP PPID STATE    BOOSTS        %CPU_ME %CPU_OTHRS UID       FAULTS    COW   MSGSENT  MSGRECV  SYSBSD
357  idea         2.6  03:16:46 112   1    925  4906M  0B   1583M  357  1    sleeping  0      0.00000 0.00000    281451937 28337096  54627 8404446+ 2733245+ 156093159+

Top shows **4906M**

ps aux

USER               PID  %CPU %MEM        VSZ    RSS   TT  STAT STARTED      TIME COMMAND

xxxxxxx            357   3.6 14.5 180050484 2430728   ??  S     1:44PM 196:48.70 /Applications/IntelliJ IDEA.app/Contents/MacOS/idea -psn_0_73746

ps shows RSS **2430728** (KB)

Updated sshd journal on both jumphost and jumptarget ---------- Full logs are too long to be pasted. uploaded to gist three files are logs that : Local -> Jumpper -> Debian by ssh -j Local -> Jumpper -> Debian by ssh Debian Local -> Jumpper by ssh jumpper Jumper -> Debian by ssh 10.10.10.3 FYI: During this time, the IP address was re-planned 10.10.1.3 -> 10.10.10.3, and a domain was added in the hosts. These shouldn't have any effect on the problem. ---------- A similar problem is [SSH ProxyJump on macOS Catalina is not working](https://superuser.com/q/1510997) . but the answer doesn't solve my case I'm configuring ssh jump with the following configures:

Host Jumper
	HostName domainName
	User username
	Port 1234

Host Debian
	HostName 10.10.1.3
	User username
	Port 22
	ProxyJump Jumper

The result of ssh Debian -vvv is

...
debug1: kex_exchange_identification: banner line 0: \033(B\033[mSSH-2.0-OpenSSH_9.2p1 Debian-2
kex_exchange_identification: banner line contains invalid characters
...

it shows OpenSSH, so I think the port should be correct ssh Jumper is fine, and manully ssh 10.10.1.3 on jumper is also fine. The local host is a Macos at OpenSSH_9.0p1 and the others are OpenSSH_9.2p2

I am using macOS 10.14.5, and I am trying to run a Windows exe file.  So I did brew install wine.  Then, using wine to run the program yields the following error:

0009:fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.Windows.Common-Controls" (6.0.0.0)

My wine can run some other exe programs. Because of this, I thought I needed to install mscomctl.ocx and comctl32.ocx into wine. I copied the files from my Windows 7 computer to ~/.wine/drive_c/windows/system32/ and added the corresponding overrides to the "Libraries" tab in winecfg. But I still got the same error. I tried another way of installing the two files.  I did brew install winetricks zenity, and ran sh winetricks dlls.  The GUI appeared, but it did nothing no matter what I clicked.  Also, whenever I clicked anything on the GUI, it outputs to the shell the same message

Gtk-WARNING **: Could not load a pixbuf from /org/gtk/libgtk/theme/Adwaita/assets/bullet-symbolic.svg.
This may indicate that pixbuf loaders or the mime database could not be found.

So how should I install OCX files to wine? "An OCX file is a Windows ActiveX control file."

I'm trying to ssh to a remote server. But after the ssh is successful the **connection freezes within 2-3 minutes and eventually drops**. On checking packet trace i observed my mac is randomly sending a **RST packet** to the remote server that is causing the connection to close. If this isn't already strange enough, i found that this happens if **open a file in vim** after establishing the ssh connection but if i ssh and leave the terminal as it is then the connection doesn't freeze. How can i debug this issue? How can i find why my mac is doing so? I tried to restart the ssh service on macbook but it didn't solve the issue If i do **hard reboot** of Mac then the issue is fixed PS: I have the following parameters in my ssh_config file

Host *
  ServerAliveInterval 60
  ServerAliveCountMax 30

This issue started after there was a network issue and all my ssh connections were abruptly dropped. After that when i'm trying to ssh i observer this behaviour

I'm on macOS and I set English as the locale except for LC_TIME:

~ $ locale
LANG="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_CTYPE="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="ja_JP.UTF-8"
LC_ALL=

Previously, the output of type command was (naturally) English:

$ type bash
bash is /opt/homebrew/bin/bash

However, after upgrading to bash 5.3, the output is now in Japanese:

~ $ bash --version
GNU bash, version 5.3.0(1)-release (aarch64-apple-darwin24.4.0)
Copyright (C) 2025 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 

This is free software; you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

~ $ type bash
bash は /opt/homebrew/bin/bash です

Why? --- By the way, I also have a very surprising result. type command seems NOT to be stateless. Specifically, it seems the command *remember*s the last environment in which it was previously executed:

~ $ echo $LANG
en_US.UTF-8

# The output is in Japanese.
~ $ type bash
bash は /opt/homebrew/bin/bash です

# When I override $LANG, the output is in English.
~ $ LANG= type bash
bash is /opt/homebrew/bin/bash

# As I used *simple command* in bash,
# the value of $LANG is still en_US.UTF-8.
~ $ echo $LANG
en_US.UTF-8

# Now, surprisingly, type outputs in English.
~ $ type bash
bash is /opt/homebrew/bin/bash

--- Current workaround is to use /usr/bin/type instead of the shell builtin type:

~ $ builtin type bash
bash は /opt/homebrew/bin/bash です

~ $ /usr/bin/type bash
bash is /opt/homebrew/bin/bash

--- Output of typeset -p | grep -e LC_ -e LANG (requested by a comment):

~ $ typeset -p | grep -e LC_ -e LANG
declare -x LANG="en_US.UTF-8"
declare -x LC_TERMINAL="iTerm2"
declare -x LC_TERMINAL_VERSION="3.5.14"
declare -x LC_TIME="ja_JP.UTF-8"

I’m trying to route all my PS4 traffic through a VPS-hosted VPN (using WireGuard), with the MacBook acting as the gateway. ✅ The Setup So Far * VPS (Ubuntu): Set up successfully using Algo VPN. WireGuard server running without issue. * Client (macOS, M1): Using laptop.conf from Algo. WireGuard connects and works fine on the MacBook. * Internet Sharing: macOS is set to share internet from Wi-Fi (en0) to Ethernet (en8). This is how I physically connect the Mac to the PS4. * PS4 gets an IP like 192.168.137.x (automatically via DHCP from the Mac). 📡 Interface Overview * utun4 is the WireGuard tunnel. * en0 is Wi-Fi (internet source). * en8 is the Ethernet adapter (PS4 is physically connected here). * macOS Internet Sharing assigns 192.168.137.1 to en8. 🔧 pf.conf (macOS NAT) I’ve enabled IP forwarding and configured NAT using /etc/pf.conf:

set skip on lo0
nat on utun4 from 192.168.137.0/24 to any -> (utun4)
pass in all
pass out all

Applied and loaded with sudo pfctl -ef /etc/pf.conf – no errors. Verified that pfctl -sr shows the rules correctly. 🔍 What Works * MacBook itself has full internet access via WireGuard ✅ * PS4 receives an IP from the Mac ✅ * PS4 test shows: * ✅ IP address assigned * ❌ Internet connection fails * ❌ PSN sign-in fails * ❌ NAT Type cannot be determined If I set DNS manually (like 8.8.8.8 or 192.168.137.1) on the PS4, it does resolve domains correctly (dig shows valid results when tested). 🚫 What I’ve Tried So Far * DNSMasq forwarding to 8.8.8.8 → no success * Tried different NAT statements: utun4, en0, etc. * Confirmed IP routing is enabled on macOS * Used tcpdump to check traffic – some packets go out, unsure if responses return * Reinstalled and reconfigured Algo/WireGuard multiple times * Flushed pf rules and re-applied minimal config * Disabled macOS firewall completely 🤯 Suspected Issue The PS4 gets an IP via DHCP but seems unable to properly route traffic back through the VPN tunnel, or maybe responses don’t get NAT’d correctly. Might also be a return path problem or DNS issue, but unclear why it’s isolated to the PS4. ❓Looking for Help With * Debugging pf/NAT config (especially macOS-specific issues) * Ideas for improving NAT/routing behavior on macOS * Alternative methods for routing all traffic from en8 through utun4 * Any known issues with macOS Internet Sharing + WireGuard + game consoles? 🙏 Thanks in Advance Any help or ideas would be appreciated! I’ve been deep in this for days and hit a wall. If someone has successfully routed console traffic via WireGuard on macOS, I’d love to hear how you did it. Let me know if I should post configs, logs, or test outputs.

*Please excuse the odd title. This is a strange one that I can't phrase in a single sentence.* My question: Why did the following happen? And, if possible, how can I change the IP that's used without changing the basic tech stack or how I connect? ----- What happened: I was just testing some code that is designed to behave differently based on which remote host it is talking to. The code identified my client IP as 185.125.190.39 which a reverse DNS lookup identified as aerodent.canonical.com. That was really odd because that's a *(well known)* public IP and definitely not one on my local machine. The the client was actually on the same physical box as the server. *This is not a coding error on my part...* I'm confident enough that the code I was testing is behaving correctly on this point. The IP was returned by python's inbuilt socket.getpeername() . The code was running inside docker which itself runs in a VM as part of docker desktop on MacOS. The container was started with a port forward (eg: docker run -p 8080:8080) and the client, run from the MacOS host machine, was just a simple curl http://localhost:8080) Now I know that the code was unlikely to see 127.0.0.1 or any form of localhost because the port forwarding meant that the connection was not coming to the container from it's own local host. The host machine is modeled as a remote. What really shocked me was the choice of alternate IP to use. It's neither in a link-local nor private CIDR range. Meaning there is a risk of collision with the outside world. So I'm left wondering why this IP, why did this happen? How can I change it to use a private IP in this context?

I need to create a shell script that checks for the presence of a file and if it doesn't exist, creates it and moves on to the next command, or just moves on to the next command. What I have doesn't do that. #!/bin/bash # Check for the file that gets created when the script successfully finishes. if [! -f /Scripts/file.txt] then : # Do nothing. Go to the next step? else mkdir /Scripts # file.txt will come at the end of the script fi # Next command (macOS preference setting) defaults write ... Return is line 5: [!: command not found mkdir: /Scripts: File exists No idea what to do. Every place a Google search brings me indicates something different.

Does anyone know how to get full screen mode working using QEMU on MacOS (Ventura or Monterey)? I tried entering full screen on my 5120 x 2880 resolution display, however Gnome Display Manager doesn't advertise a setting larger then 1620x... I'm wondering if this is something that needs to be set manually in X-Windows.

I have manually downloaded the latest python upgrade - 3.9.6, however when I check the version of python on my terminal (iMac - big Sur):

-iMac ~ % python --version
Python 3.8.8

However, when I try to upgrade my python using home-brew:

~ % brew upgrade python
Warning: python 3.9.5 already installed

I seem to not have the one that I downloaded - python 3.9.6. Is there a way to remove these other python languages i.e. 3.8.8, 3.9.5 and just keeping the latest python 3.9.6?

I'm running ArchLinux in a virtual machine in VMWare. I've *open-vm-tools* installed but I'm unable to copy from macOS host and paste into Hyprland. How can I achieve this?

I'm using the following command on my mac: $find . -name “*.java” -exec sed -i ’s/foo/bar/g’ {} \; and it seems to have no effect. I have two files in the directory that end in .java, which both have the foo text in them. Am I missing something? **EDIT : Results from request of comments** [aafghani-03:~/test amirafghani]$ find . -name "*.java" -exec sed -i 's/foo/bar/g' {} \; sed: 1: "./bar.java": invalid command code . sed: 1: "./foo.java": invalid command code .

There are several topics addressing similar issues, but none of the solutions provided there work for my environment, being: - macOS 11.2.3 (MacBookPro M1/Apple silicon) - GnuPG 2.2.27 (installed via Homebrew) - Zsh shell - GPG key with Auth-only subkey for SSH The *~/.gnupg/gpg-agent.conf* file:

pinentry-program /opt/homebrew/bin/pinentry-tty
enable-ssh-support

The *~/.gnupg/sshcontrol* file:

FADD8723...AE7ED (the keygrip retrieved with gpg -K --with-keygrip)

The *~/.zshrc* file:

unset SSH_AGENT_PID
export GPG_TTY="$(tty)"
export SSH_AUTH_SOCK=$(gpgconf --list-dirs agent-ssh-socket)
gpgconf --launch gpg-agent

I made sure *ssh-agent* is not running (also rebooted the system to be sure). But no keys are found when running:

ssh-add -L
The agent has no identities

What am i missing? EDIT: I should add that I also configured a few Debian 10 systems like this and it works just fine.

What is the current status for the Linux kernel's support for APFS or HFS+ (filesystems used by MacOS)? Does it support write, journaling etc? I saw https://unix.stackexchange.com/questions/481949/is-an-apple-file-system-apfs-driver-for-linux-available-or-in-progress but the responses there mostly deals with FUSE-solutions. I am, for performance reasons¹, interested in kernel based solutions. ¹ I am gonna run it on an embedded device with a slow CPU, low memory etc.

I have a usb disk and I have format it to ExFAT on Mac. Then I run the command to create a bootable windows on Mac: sudo dd if=Win11_24H2_EnglishInternational_x64.iso of=/dev/disk4 bs=8m status=progress The command finishes successfully but the disk format is changed to UDF (Universal Disk Format) which is not recognised by windows installation. How can I keep the disk format no change? I want it to be stay as ExFAT.

Is there an equivalent on MacOS to the Linux hostnamectl command?

I have downloaded the anaconda docker image from here and I am running it on MacOS. I have come across the following issue and I would highly appreciate your help! When I try running Ipython by the following command docker run -it pyimg ipython I get the error > docker: Error response from daemon: OCI runtime create failed: > container_linux.go:345: starting container process caused "exec: > \"ipython\": executable file not found in $PATH": unknown. This makes me think that the $PATH in docker does not entail the location of ipython. However, when I spin the container by first executing the bash and then by launching ipython, everything works fine! More precisely, docker run -it pyimg bash and then ipython works fine.. To make things even more confusing, docker run -it pyimg python works perfectly! I already checked inside the container and realized that both python as well as ipython are located in the same directory **/opt/conda/bin** How is it possible that this behavior is exhibited and how can I fix it? Thank you very much in advance for your help!