Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Suicide Linux (2009)

qntm.org

Post reply on HN
Print view View original

51–60 of 183 posts

Re: Suicide Linux (2009)

#51

Post by kashyapc »

dkarl wrote: post #27

> Perhaps rm -rf / should be replaced with something with more verbose flags set. That way, when you run a bad command, you are told immediately that things are being deleted This gives me an idea: Hell Linux. In Hell Linux, the operating system does its best to pretend that nothing is wrong, so it takes you as long as possible to realize that you've made a mistake. Commands ignore all unrecognized flags. In shell sc…

One of the "nice side effects" of your shell design is that one can do away with thinking about POSIX requirements / portability, among other "potential goodies". ;-)

    - - -
But yes, if Hell Linux exists, it'd make a great candidate for a future QEMU Advent Calendar[1] submission.

[1] https://www.qemu-advent-calendar.org/2018/#about

Re: Suicide Linux (2009)

#52

Post by 2is10 »

dkarl wrote: post #27

> Perhaps rm -rf / should be replaced with something with more verbose flags set. That way, when you run a bad command, you are told immediately that things are being deleted This gives me an idea: Hell Linux. In Hell Linux, the operating system does its best to pretend that nothing is wrong, so it takes you as long as possible to realize that you've made a mistake. Commands ignore all unrecognized flags. In shell sc…

> the operating system does its best to pretend that nothing is wrong

Call it "Trump Linux"! :)

Re: Suicide Linux (2009)

#54

Post by HashingtheCode »

segfaultbuserr wrote: post #37
HashingtheCode wrote: post #29

This could be turned into a really great security tool, akin to a deadman switch, which upon activation and incorrect use, deletes the contents of the users home directory. Example. In Mate Desktop. the default terminal is Mate Terminal. Upon installation of Suicide Linux, Mate Terminal could be modified to delete the users home, but provide no warning that it is running. For the user, they could use another terminal…

In this case, just "rm" everything is far from enough, it needs to be a secure deletion. And the best way to implement secure deletion is using full disk encryption, this way, instead of wiping the entire disk, you just need to wipe the header, which contains its master key. head -c 100000000 /dev/zero > /dev/sdX1; sync An even better solution is interfacing the motherboard with a small microcontroller and storing th…

Is it possible to find the sourcecode?

Edit: found it

https://github.com/tiagoad/suicide-linux

Re: Suicide Linux (2009)

#55

Post by remram »

A problem with most commands is that they usually don't need extremely-unfortunate typos or russian-roulette-enabled shells to turn into something wrong: they are usually wrong in passing, while you're typing them. I always wondered if something could be done against the effects of an accidental premature return. Some of us have cats...

`sudo rm -rf /tmp/workdir` starts with `sudo rm -rf /`

`git reset libs` starts with `git reset` which will wipe your staging area

`docker-compose rm -f db` starts with `docker-compose rm -f` which will remove all stopped containers (and their logs)

Most commands just tend to do something very broad with no arguments, and the fact that paths start with bigger folders and go down to files doesn't help either. No GUI tool would ask you to confirm before you type in the path, the way shell commands tend to.

Re: Suicide Linux (2009)

#56

Post by offtop5 »

Would be fun to create a reality TV show. Have new grads come in to work on a one month project.

The only os they can use is this. Who ever doesn't mentally break after 30 days gets 100k divided between the other survivors.

Also have a troll who randomly blocks various support sites like stack overflow.

Re: Suicide Linux (2009)

#57

Post by steffan »

segfaultbuserr wrote: post #17
megiddo wrote: post #13

This is like that time, last week, when I was having issues with do-release-upgrade and someone helpfully suggested to run sudo apt-get purge python* && sudo apt-get autoclean && sudo apt-get install python* Which I did without paying attention to what I was doing and it turns out that this is one of the worst things you could do ever on a Debian-based distro. Damn you, abu-ahmed al-khatiri.

Also on Gentoo. Once you've done it, your package manager is gone... At least you'll be warned when you attempt to do it, and often there will be two or more Python installations (2.7, 3.x), and Portage will run on any of them, so incidents like these are actually rather rare.

This reminds me of fun times compiling and upgrading libc on a running system. It's challenging, but if you plan it out, it's not bad. But if you miss a step...better hope you have some statically compiled tools somewhere that can help you out.

Re: Suicide Linux (2009)

#58

Post by t0astbread »

segfaultbuserr wrote: post #37
HashingtheCode wrote: post #29

This could be turned into a really great security tool, akin to a deadman switch, which upon activation and incorrect use, deletes the contents of the users home directory. Example. In Mate Desktop. the default terminal is Mate Terminal. Upon installation of Suicide Linux, Mate Terminal could be modified to delete the users home, but provide no warning that it is running. For the user, they could use another terminal…

In this case, just "rm" everything is far from enough, it needs to be a secure deletion. And the best way to implement secure deletion is using full disk encryption, this way, instead of wiping the entire disk, you just need to wipe the header, which contains its master key. head -c 100000000 /dev/zero > /dev/sdX1; sync An even better solution is interfacing the motherboard with a small microcontroller and storing th…

What's wrong with using FDE and entering the key manually on startup? You could then just shut down when you detect suspicious behavior and all is well or am I missing something?

Re: Suicide Linux (2009)

#59

Post by cobbzilla »

remram wrote: post #55

A problem with most commands is that they usually don't need extremely-unfortunate typos or russian-roulette-enabled shells to turn into something wrong: they are usually wrong in passing , while you're typing them. I always wondered if something could be done against the effects of an accidental premature return. Some of us have cats... `sudo rm -rf /tmp/workdir` starts with `sudo rm -rf /` `git reset libs` starts w…

My favorite is:

    rm -rf ${DIR_NAME}/*
Then you remember that DIR_NAME is undefined...

Re: Suicide Linux (2009)

#60

Post by jmnicolas »

dkarl wrote: post #27

> Perhaps rm -rf / should be replaced with something with more verbose flags set. That way, when you run a bad command, you are told immediately that things are being deleted This gives me an idea: Hell Linux. In Hell Linux, the operating system does its best to pretend that nothing is wrong, so it takes you as long as possible to realize that you've made a mistake. Commands ignore all unrecognized flags. In shell sc…

> the operating system does its best to pretend that nothing is wrong, so it takes you as long as possible to realize that you've made a mistake.

You're describing my (modest) experience with Javascript: it fails silently.

I was describing it this way to a colleague: Javascript is a Japanese girlfriend, too polite to show you your mistakes so you have a hard time to know what you did wrong.

I want a rigorous German girlfriend that won't let pass any mistake or even an Italian one that will burn my house because I forgot to call her by her right name!

Post reply on HN