• 0 Posts
  • 70 Comments
Joined 5 months ago
cake
Cake day: April 16th, 2026

help-circle

  • No, that is required for rm -rf /. But the command in question was rm -rf /*, note the asterisk. The * will be expanded by the shell, so what rm sees when it evaluates argv is not /, but [/bin /usr/ /lib ...].

    You can test this yourself. Run: rm --recursive --interactive /. It will abort, saying:

    rm: it is dangerous to operate recursively on ‘/’

    rm: use --no-preserve-root to override this failsafe

    Then, run set -x in your shell. That way your shell will print the command that is actually executed. Finally, run:

    rm --recursive --interactive /*.

    In my case, the output is:

    $ rm --recursive --interactive /*
    + rm --recursive --interactive /bin /boot /dev /efi /etc /home /lib /lib64 /media /mnt /opt /proc /root /run /sbin /srv /sys /tmp /usr /var
    rm: remove symbolic link '/bin'? 
    

    The line starting with a + is from the shell and shows which command was actually executed. After that, you can see rm will now happily start deleting files.



  • Any engineer fighting the use of LLM’s will ultimately find that their peers are achieving more in a shorter time then they are.

    How do you measure that?

    I am asking because I tried all the AI coding tools (copilot, cursor, claude code, etc.), even the paid versions, to find the utopia that doubles my productivity where lots of people are talking about. But everytime I find they take more time to generate code than me typing it, they rarely get things right the first time, they produce way more code than needed (which is much more to reason about) and all in all it takes longer to achieve something using “AI” instead of doing it on my own.

    The worst results are the ones that seem to be correct, but can only be found incorrect on closer inspection. For instances, a part of my job is to write data collection applications for sensors; those sensors come with a datasheet (usually 100 to 200 pages) that explain how to “talk” to the sensor. Sounds perfect for an LLM - throw the datasheet in, get a small C application out. But either I get something that does not even compile, or something that seems to collect data, but configures the sensor completely wrong. Debugging that takes more time than just working through the datasheet myself and program it myself.















  • Indeed I am always disguested when someone says “send a mail”.

    Maybe this is more common when English is not your native language; e.g. in Germany, “mail” is “Brief”, and “email” is just “E-Mail” or more short “mail”, because no one says “elektronischer Brief”.

    Interesting to see his take on “e-mail” vs “email”.

    So that might just be the origin of this language butchering you observed; as “email” means something different in various languages, lots of people who don’t speak English as their first language might have chosen to use “mail” for electronic mail. That is just speculation of course.