What's wrong? That's EXACTLY what the Arch wiki recommends. If you install linux after Windows, then iirc you can't configure some stuff like you can't resize the ESP without moving all the partitions.
@@ProferkFirst time I installed Windows after Linux, Windows f'd up my Linux boot partition ... 2nd Time I installed Windows after Linux, Windows created files on my 2nd drive ... I always liked to install Linux after Windows, because that way I can create a separate ESP partition, so Windows and Linux don't have to share one ... I had it like this: /windowsboot, /windowshome /boot/efi, /arch I just set arch boot as default, updated grup and windows appeared in my Arch grub menu. iirc, that doesn't work on MBR, just on UEFI GPT.
I've ran destructive (write) bad sector scan on a 12TB drive for like 5 minutes at 3 AM, it was not even the drive that was failing. Good times, never fix a server before going to bed.
because bed just got much further away! I've had a couple of those kinds of issues, fortunately not that bad. I did have a bunch of work go away though, when a lightning strike hit next to the office, and scrambled the drive.
@mvandios tbf, linus problem was a dependency problem in the pop os repository, but yeah you should at least have a vague idea of what each command does, AND ALSO READ THE PROMPT, AND AVOID CONFERMING WHEN THE PROMPT IS LITERALLY "YES BREAK MY SYSTEM" lol
First f-up I did was wipe a 16 year old harddrive filled to the brim with nostalgia (was intending to move them to a brand new ext4 drive for safe keeping). NEVER trust /dev/sdX names after a reboot.
I feel your pain. Iol I always # or echo my dangerous commands first, then go back and recheck my info with Is or whatever, against my built command, then run it if still looks right. Works in scripts to: l use echo and set -x in dangerous parts of bash script and see what it does after vars are expanded before I run it for real. Super paranoid for a reason
this is why i always double check the contents of a device/partition when doing possibly destructive stuff with them in chroot (e.g, changing filesystems, moving data to another storage device)
My mistake, and it was kind of funny was I was writing an install script for a program I wrote in python. I got distracted while writing the script, and when I went to run it, changed the permissions for every file on my system to "read only". There is no happy way to come back from that lol. Funny you mentioned a form of this same thing on your video.
Did you reinstall your system? I believe you should be able to do that, without ending up in the horrifying grub prompt, until you still have the efi folder intact
My lesson in humility came when I wanted to install a newer version of Thunar. It required newer versions of some libraries in the repo, so I force installed the newer libraries with deb packages. What an interesting time I had afterwards.
Reminded me when I had RedHat and Mandriva distros on CDs(it was cheaper to buy 4-6 cd distro). Both use rpm I tried to install some packages from mandriva to RedHat to have more software, and this is when I learned glibc is not fan of compatibility, libraries got intermixed and everything became unstable and unusable
You don't have to do that. You use the LD_LIBRARY_PATH variable. Then you put the libraries a binary needs off your system library path. What you do is make a shell script wrapper that exports that variable and then calls the program. So that automates everything for you.
@@littlepeon what I said in my previous comment already does that and it doesn't need to add anything. That's just how the loader works. There's a path it searches by default but in special cases you can direct it to load libraries from somewhere else. You just need to know how. I don't remember how I found out. But since I did I've been doing it for decades now. Dealing with linkage issues is pretty common. It's basic and simple stuff. But I guess when you don't know then it's complicated. Once people are aware of it I think it gets simple then. Everyone man ld.so for an explanation of how programs run. You should know this. Running a program through strace is pretty eye opening too. So also do that. See how the gears turn. Know thy OS.
This video is missing some tips to avoid these problems. - Use an alias for rm -rf to go into interactive mode. or better yet use trash-cli instead.(alias rm='rm -i') - Use /dev/disk/by-id instead of /dev/sda. It is much harder to make a mistake that way. - Backup and backup. And if that does not help, just backup some more.
I don't do backups. Had a disk spontaneously fail on me, and I still don't do backups. Why? Because it's always been easier to move on from scratch than restore from backup. What I do instead is replication of important data to multiple locations. Closest I get to a full back up is making a tarball of my home dir, and /etc before I reinstall. But only for future reference or to grab are file l forgot was important. These tarballs sit around for years until I need space and are deleted. Very rarely extract anything from them.
@@djyottaI call anything that makes sure you have more than one copy of the important data a backup. Though it may not be technically true, I'd call what you do, "backing up your data."
I also thought about doing the alias on rm -rf but it's dangerous. You then go and type it in a different machine, used to it being interactive and... You are in for some fun.
Now that I think about it, I suggest that you alias rm -i to "remove" or something similar, so that when you move to another machine the command will not be found and you won't mess up anything. Only downside is that you might forget that removing is done with rm but that looks difficult to forget when you have been typing it for years
I did a "chmod -R .." once. DON"T DO THAT. Worse< I misconfigured logrotate on a production system and deleted the live MySQL files. I was able to reattach the inodes to filenames again, and that was unbelievably tricky.
I've been a Unix/Linux admin for about 30 years. The best habit I developed was testing any recursive command with 'ls' first. That way I could check the output and make sure the 'rm' or 'chmod' command, etc. was going run on what I expected it to. I still managed to destroy a few systems, but this saved me many times.
For the formatting drive one, I actually even did that on windows, much less Linux. Long story short, disk part wasn’t working, instead of double checking why, I downloaded a data nuker, and ran it, long story short I was selecting the wrong drive, my C drive. That’s the story of how I started daily driving Linux for awhile.
I've gotten in trouble with symlinks before. For example, I wanted to create a single directory that contained only symlinks to my "important folders" that I wanted to backup. Nice and organized, one stop shop for everything important... I thought. I put this "one stop shop" in my home folder. Which also needed to be backed up. I used rsync -aL to follow all the symlinks and backup the data. This, of course, created an inifinite recursive loop of backups... /home/usopenplayer/drives/ /home/usopenplayer/drives/home/usopenplayer/drives/ /home/usopenplayer/drives/home/usopenplayer/drives/home/usopenplayer/drives/ I got very confused when my 8TB drive filled up with what I thought was 1.5 TB of data.
I did the rm -rf on home once before. I use zoxide, and didn't notice that it had "corrected" my previous cd command. I did it with verbose flag, and I found myself suddenly watching the terminal text flying by, telling me in-detail specifically how it was destroying my soul. I hit it with a interrupt before it finished, but the damage was done. I just ended up reinstalling my whole system, was easier than trying to figure out the mess I made. Luckily most anything critically important is always backed up on other drives and/or git.
Yup! I've had multi drive systems for 20+ years, usually one live Windows install (mainly for games these days, in fact 99% for games only !), plus a live/main Linux distro, and one or two 'trying these distros'.. drives..(or old distros when I've changed...) despite knowing the disk makes and types, I still find it more relaxing to just unplug the others first, it only take a few seconds!
I did it at some point which resulted in unbootable system because hda, hdb, hdc all get messed up. These days it's harder to mess up this way as /etc/fstab uses guids instead of /dev/sda
Destroying your optical backup disk on a disruptive hardware. Then, your system administrator repeating the same for your second optical backup disk. Then, taking your original hard disk with you to a new working place. And, finally, an unknown computer nerd there loaning your hard disk from your unlocked drawer without a permission during the weekend and accidentally removing all its contents. ... 20 years of your research etc. work destroyd permanently! This is a true story.
That's the right message at the end, create backups. I'm careful enough to always double check before running rm or dd. But I've been burned a couple of times by badly designed softwares that ended removing some of my data. Just yesterday I read that a theme from KDE store nuking user home directory. You won't know when or for which these would happen to you.
@@shreyasas3128 yep, and using the same tool your DE uses means less dependencies and more consistent behaviours, I want recommending anything tho, just throwing an information out there
With these dangerous commands like rm and dd, I always write out everything after the command itself first. Then I double and triple check that I have that part correct, then I actually type rm or dd and run the command.
My lessons: 1) Never use the -y option with a package manager on the command line on the first run. (Review what it would do, then either answer YES or re-run from command line history and add -y.) 2) In a script, set the -u option, which abends the script if it tries to expand an uninitialized variable. Many erroneous rm's (and other bugs) can be prevented due to typos in variable names. 3) Never add a random repo to your package manager that you find via a web search or cheatgpt or stackoverflow. And why doesn't apt have a command line option to enable or disable a configured repo, like yum does? Why should an end user, even one with root access, have to temporarily uncomment an entry in a repo list to install a package from a repo that's used only for unusual situations?
I've certainly deleted my entire system, back in my early days of using Linux. Downloaded a new DE to try, but wanted to delete it after realising I didn't like it. During the deletion process I quickly realised it was deleting things I wasn't expecting. Fortunately, however, I was staying with my dad at the time and he helped me save my data! 🙂
I once did a recursive chown on an OS directory by accident. Thankfully I had used the verbose flag, which shows the old user and group in the output. Was then a matter of copying the terminal output to a file and then writing a script to use that output to set each file back to it's original user and group. Amazingly the system didn't crash. 😅
For me, it was letting my laptop repeatedly get stuck on a TTY while it was supposed to be in sleep mode inside of a well insulated laptop carrying bag. Needless to say, it got VERY toasty.
I learned that one the hard way. I was working on the ISP firewall, and went to delete the old rules and instead deleted the active rules. I had to drive to town, call the NOC to explain why I needed to get in during the middle of the night. Fortunately I had a backup.
@javabeanz8549 Yeah, I was also a victim of my carelessness too. I was deleting old files and forgot to check my current directory. When I realized what I did, I used testdisk right away to recover the files. Boy, I'm glad I knew about testdisk.
@javabeanz8549 I know the feeling when you accidentally delete files, especially the crucial ones. That why, backups are so important to do. Happy to know that you got your files back.
I've never done sudo rm -rf, but I did do a sudo rm * once. Or rather, that's what I set out to do. What I actually did was sudo rm * (note that space between * and ). That cooperatively wiped out a Python file I'd been working on for work for about two weeks. I wanted to go throw up. Fortunately, I had just been editing the file and I remembered enough about what I'd just done to do a raw disk sector search with grep, and I had it return the entire sector.. That contained enough of my file to allow me to recover in a few minutes. So in the end it was all ok, but wow - trust me, it's NOT a good feeling. And it's impossible to lay the blame on anyone else.
DT mistakes funny you mentioned chown and chmod using the recursive flag commands. Two weeks ago I was playing around with an old drive that I use to created a XFS file system for file storage. Turn out that XFS drive was bad. I was tired after spending three days doing backups between three computers and I used those commands you referenced to reset file permissions on a directory which mess a directory on the /root. I was lucky and able to boot using the backup kernel off grub to correct my errors and it work. All was well except for the HP printer did NOT work gave some strange commands on the filter. After three hours of research, below are some come commands I used today to set everything straight on this box to once again print !
With the coreutils which I think rm is included in the package, I compiled it vanilla without the --no-preserve-root patch. Yes, I use LFS. Yes, in an effort to delete everything in another partition except the boot directory, I instead wiped my entire LFS system which I poured 40 hours into along with a year of user data. I was very sad that day, it's an unparalleled feeling for all the wrong reasons.
Booting into a btrfs snapshot without noticing for several weeks. The snapshot was writable so I didn’t notice. Until timeshift cleaned it up so it was basically deleting the system while I was running it. Was easy to rectify since I still had tons and tons of snapshot. But I first had to figure out what the hell is happening
@@user-mf6cu5xj3v I've never understood what that actually means. Do you press colon and then q, do you hold down colon while pressing q, or does ":" stand for some control key? It's anyone's guess. I like nano, much more intuitive - Ctrl+X, then follow the prompts if you need to save and you're done. Only way I've ever successfully closed vim is by literally closing the terminal window haha.
one of my more recent stupid mistakes was at work i put mx on this computer a year ago and i wasnt the only one who used it so it had repos from bookworm and buster. i assumed it used bookworm as the main repos so i deleted all the buster repos and gpg keys for buster. it was an old version of mx that still used debian 11 repos.
I was backing up my entire system onto an external SSD but I forgot to exclude the media directory, so the system began copying the entire SSD in a folder on the SSD. Right after that, I wanted to remove that entire directory that had the copy of all my backups with the rm rf command but somehow I nearly deleted everything on the SSD. Thankfully I noticed the mistake in time and stopped the command. It only removed some useless files in one of the backup folders, so I was able to recover them from the accidental dublicate that I made earlier before that.
1) Installed the bootloader in the wrong place, so after installation the machine booted into Windows. 2) Deleted a Windows partition while doing advanced disk partitioning 3) Overwritten stuff cp -rf 4) Overwrote a system file with stuff while editing in vi and broke stuffs 5) Deleted important stuff by using rm -rf.
11:38 the main reason is that the packages' names differ, Ubuntu has its own way to name the packages. For example, the chromium package is called `chromium` on Debian, while on Ubuntu it's `chromium-browser`
One of my colleagues did this infamous movement by "rm -rf /*" with root user. Not on Linux but on AIX server in enterprise environment. He was the teacher for new server admin guys on that day and shown them what should not do in any cases. So wrote this command in root session and hit the enter automatically instead of ctrl+c. When he realized this mistake he had the chance to show how to restore the filesystem of an AIX from backup when necessary.
When I swapped from Windows to Debian Slink, my "Linux friend" who helped me out explained /dev/null, in a different context, as basically the Trash bin in Windows. So...I may have moved files there, to remove them. As root. Which meant I replaced the null device with a file owned by root. Which caused all SORTS of havoc with daemons trying to write there.
I did a recursive rm on my boot directory once. I knew I had the boot directory backed up in a BTRFS subvolume (not a separate boot partition at the time, which was a confusingly bad idea), but I did not consider that rm would reach down into the EFI directory, crossing partition boundaries, and wipe the EFI partition. I had to learn how to chroot into the system and do a grub-install. Recursive chmod might not be fatal if you can reboot (maybe by forcibly powering off if necessary) and then boot into a backup BTRFS subvolume from GRUB. I have not tested it, but that should probably work. Regarding Debian and Ubuntu, Ubuntu is downstream of Debian Sid (i.e., Unstable), much like how Debian Stable is downstream of Debian Sid. Essentially, Ubuntu and Debian are each different ways and cadences of freezing Debian Sid. There is obviously more to it than that, but understanding this relationship between them reveals how adding repos from one to the other is a risky idea. (I should mention that a lot of Canonical developers contribute to Debian, so even though it appears that Debian has total control over Sid and that Debian is a complete single-project-controlled pipeline, Ubuntu has a lot of influence in it as well.)
Reformatting my boot partition. Look, I can explain. I thought it was from an old install. This was back when I used Arch (I'm a fedora fanboy now) and luckily I had a friend willing to let me borrow their laptop to flash the archiso onto a drive, and managed to reinstall the kernel and grub. Good thing I followed the manual install instructions, or I might not have known how to do that.
when it comes to rm -rf, I have trained myself to *never* type the -rf until the end of the command (after the path.) That way, you will not accidentally delete the wrong directory with a stray finger hitting Enter Key.
I made a backup-script with rsync, using variables in it not mentioning my windowmanager might not have these variables if I create a button for it. In bash, an empty variable is an empty string. Using this button (shutdown including backup) wanted in my case to delete all files beginning from root owned by user including the backup itself. This was a really nightmare! Another case is thinking about lightning flashes which might destroy everything connected in anyway. I think it's very important to have backups not connected by wire in any form.
I forgot the usermod -a -G "-a" option when i wanted to add a group to my user and lost the sudo group :-) on ubuntu it was an easy fix so, rescue boot and mounting root filesystem and adding the group manually.
Damn that rm command really fucks me up when i was starting on linux. I remember around 2011 i was asking on ubuntu forum how to install wifi driver since mine was not detected automatically. And some dude posted a sudo rm with an * and got a hundred + vote instantly. And me as a newbie run it and i never look back at linux for almost 5 years.
Just FYI: updating to testing works quite well as of past 2 years on debian, the only thing that can break is sometimes drivers but now you can sometimes make more things in your hardware just work(tm) by using testing distro
I have made my computer non-bootable a couple of times. I upgraded my kernel on Gentoo linux. I was tired and forgot to run the complete set of comands to update, reinstall, and configure Grub. I only updated it. Restarted and went to bed. Woke up to a grub prompt displaying errors. Wasnt a hard fix. But makes you sick to your stomach after you dedicate 14 hours into your system with compiles.
dumbest thing i did was go nearly 3 years on my arch install without backups or even snapshots. i recently was able to spare some money for an external HDD and copied my btrfs subvolumes over to there and compressed them. same install is still kicking today but i feel like this outcome is probably the least likely and i got lucky multiple times along the way
The biggest "don't" of this video is the one that's implied. Don't find yourself without tested, complete, bare metal, external, daily backups of your system. All of these mistakes are quickly fixed. "All of my important files are backed up" is a trap. It requires perfect accuracy in initially identifying which files are important, perfect accuracy in maintaining that list over time, and even if you managed both of those makes recovery tedious and time consuming.
Uploading the latest F3OS right now. I've learned to pretty much not even fool with chmod, chgrp, and chown. As for doing so as root, I've set root's bashrc to display a red and yellow prompt to remind users that they're running as root, which can be dangerous, while regular users have Debian's standard green and blue prompt.
After using Linux for a few years now, sometimes things just happen. And while it might be nice to understand what you've done to stuff your system, sometimes it's just quicker and easier to blow it away and start again. And you're back up and running in 30min or less. My biggest Linux mistake? Doing any updates to Manjaro 😂🤣
Man, when I used to run Manjaro I got my system destroyed by updates so many times....now I have been running Endeavour for a couple years, and it hasn't happened once...I was so scared to update in the beginning 😂
@@Syl4r93I know what you mean. I tried all sorts of ways to update, always had problems afterwards and generally needed to reinstall. After trying out a few Debian distros I ended up on Fedora and much happier. I spend more time using it than fixing things I shouldn't have to.
I have dual backup encrypted drives of 1TB each. These contain disk images, the ultimate time saving backup. Having disk images makes the system immune to RANSOMWARE.
back in about 2007-8, Gentoo (yes, I _ran_ Gentoo) changed their behavior to begin wiping /tmp on every boot. I blindly "emerge"''d world, and didn't notice, and had been using /tmp as "My Documents" for several months..no backup...lost a ton of data after that "change". _sigh_
Did the same this week, wiped out my home with rm rf... Following that I nerfed rm and improved my backup methods and will work on permissions and ownership protections...
Regarding recursive chmod, you can sometimes recover system files with getfacl on another system with the same OS and setfacl. Depending on how difficult backing up your data is this can save your bacon, or at least get you functional enough to do data recovery before reinstall
I never ran dd wrongly in ~30 years of linuxing. But that's mostly because I distrust myself and always do fdisk first to make sure I aim dd where I intend. Then I ctrl-c and change fdisk to dd. Always. No exception. Even if I write several sd cards in row. Now other stuff, like incorrect recursion and deleting packages I didn't mean to... Yes. So much yes. I think I uninstalled DE once, only "default" x booted (the one with static background) and even afetr installing DE it didn't setup itself correctly. Partial upgrades becamse less threat now when flatpak exist. So until flathub fails, it becamse much better. I can't imagine what a pooshow will happen once flathub goes offline for reasons that nothing is permanent (remember ftp cdrom?). Also messing up with grub. You either got lucky to never mess it up or forgot it or I skipped it. So many times I've installed grub where it didn't belong. I didn't mistype anything, I just chose incorrect drive. So once I removed drive that I thought was "secondary", my whole system no longer worked.
I installed Arch, KDE and the graphics drivers, why am I unable to get to a GUI and keep being dropped to a TTY? Hours later, 'did I install the display manager?' And then it clicked!
With repositories it's not that easy. For example I am on Ubuntu. But I use the Python repository and the Haskell repository. Other languages such as R, Common Lisp or Rust can also have their own repositories. And sometimes these repositories can have conflicts with the OS repositories. So if you use some of these programming languages, it's almost impossible to not use other repositories.
Even if it's not sudo rm -rf, you always have to be extremely careful about how you use rm. I try to write that command with as much specificity as possible, because like every other Linux user in existence, I have messed myself up in the past by recklessly throwing around rm.
Switching Debian to the testing repo is also bad. If you read the claims, you might think it's about as stable as Arch, but you're wrong. On Arch, we get a package broken by an update. It typically takes about a week before another update fixes it. On Debian Testing, I always got worse things happening than that.
My biggest mistake was reinstalling docker when I already had it then messing up a system service that was running... oops. I've also ofc done the classic chmod everything too
I almost caught a heart attack after formatting a drive I thought wasn't the right one for a second. I have 5 drives connected with only one of them meant for Linux. I didn't know that the drive names (sdb, sdc, sdd etc) can CHANGE after rebooting a live iso.
My first week on linux (elementary at the time) I CHROOT my entire home folder. I wasn't allowed to get into any folder in the system. Took me ages to fix it. After 3 years i am still reluctant change permission on any file or folder
Ah yes, I remember hard deleting everything in /var once by accident. That was interesting because the system didn't break right away, as the stuff in /var is meant to be temporary, but you could tell something was off.
When I compile from source I change the install prefix or just run it in place. Don't mess around outside your home directory. You don't need to. Leave things that don't belong to you alone. Don't taint the system.
How do people screw up with rm -rf? Do they also accidentally Shift-DEL folders on Windows? Careless with tab-completion? Not thinking what you're doing? When I used to use Linux a lot I never even used sudo, I was at all times logged in as root in one console and terminal window, I did stuff routinely as the root user and I never pissed on my pants because of it. The command prompt was different as a root user which was an immediately obvious hint to exercise extra caution.
Maybe I just never made those mistakes because I learned what did what first. Even when I was learning this stuff back in the 90s, there were things I didn't have to do because I understood the function.
I purged python and broke my desktop environment. I wrote a video generating pythkn script that sometimes froze my PC when it ran out of memory and I had to hard restart the machine. It corrupted my UEFI and I could only read but not write my drives. I needed to reflash the firmware, took me a week and a new SSD to debug. I used -r or -R on a zip command on a mac and made an empty archive and then sent it to my boss. I still dont know which one is right on linux, I dont trust myself. I changed my python version with ln links and my bluetooth GUI broke and I had to connect through CLI for 4 days until I found out (xfce)
They say that you learn from mistakes, but still the main thing is not to lose concentration, sometimes one space in the path to the file is enough to install the system again. And no one is immune from this.
Yes, putting rm -rf foldername/ * (still not sure why) was probably the dumbest thing I did in a directory full of databases when I was only working for a year or so. I noticed that it took too long when it had removed all databases from a to c before I cancelled it. We obviously had backups, but it wasn't fun asking someone else to put those back after I did something so dumb. I generally only put rm -rf * when I am IN the directory that needs to be cleared nowdays, I will do an rmdir after that the extra command is not worth the risk. Also, I had to fix a removed desktop environment at work for someone else. They had no clue how to fix it and claimed they didn't do anything, but either the desktop environment or x was completely gone if I remember correctly. I also killed my pacman in a VM once by updating with pacman -Sy (I think) and I don't think I knew how to recover it back then and just either reinstalled or put a backup of the VM back.
Of course `rm -rf /` was the classic one, specially when using root user as the main one!! About 30 years ago, using BitchX (IRC client) with root user, I ran /decode XXXXX, instead of /echo decode XXXXX, and the XXX in action was `rm -rf /` When I saw the HD light blinking, I imediately shut it off. A minor "kernel panic" but /home was safe! 😀
One time I was installing a custom auth module in /etc/pam.d and completely broke my system's ability to verify passwords. That was a tricky one to troubleshoot.
#HeyDT! I have a question: I see that there's a lot of recommendations about what would be a beginner's Linux distro. But there's nothing more comprehensive for more intermediate to advanced users. I mean, after you go over that "first distro", after you learn and pass the basics and get confortable with Linux, including a bit of CLI, what should a person, assuming the learning curve is no more of a concern, what should a person look to for the "advanced" distro to "settle" in ? When ease of use is no longer a need, what would be a list of distro that someone should check further, possibly finding the BEST distro for them, and why those distro ? I feel like a more comprehensive guide on that is missing. I use Gentoo -btw- fyi!
I do have one ubuntu package installed on my debian machine. And that's the manufacturer's software controller for my laptop's fans. Thankfully it works fine. I downloaded the deb directly and installed it because there is no way I am connecting my Debian machine to the Ubuntu repos.
Only real dumb thing I've done was to format my /home drive when I reinstalled the system after a disc failure, I had backup, but took way more time than expected 😊
Resizing ext4 partitions, while resizing a ntfs partition at the same time. Don't do it. I had use data recovery software to get some of thecsata back since the ntfs partition table was completely nuked.
Recently replaced my fstab with a customized one that had some mounted smb shares. Everything worked until I rebooted and realized this had UUIDs from another machine's drives for / and of course this machine would not boot anymore
Two things no one addresses or seems to know a fix for that are gonna end my switch to Linux about a month in. Every time I open a web page there is a 2-3 second delay before the page loads. Every browser every distro. Seems to be a known issue. Steam downloads at roughly 1/4 the speeds I get on Windows. Also a known issue. Any time I scroll and accidentally cross the top of the window instead of stopping scrolling it starts scrolling tabs messing up everything (KDE). Also a known issue. NONE of these have I been able to find a solution for. Seems everyone has the problems but no one has a fix.
I accidentally copied bash code into the CLI on a production NDB master node on primary/active "side" during business hours with PuTTY. Mild mistake, but still made my heart skip a beat.
Except for rm recursively, I used every stupid way imaginable to kill my system repeatedly. Thank god for backups, backups and backups. A lesson learned early.
2012 sudo pacman -Syu installed an icompatible libc and bricked my archlinux. apparently they made a change which required package signing keys to be installed on the system before updating and I missed the announcement.
