Keep restarting an application in an interactive session after a short time

Question

So I'm aware that this is a somewhat convoluted usecase with probably not many practical applications, so unfortunately I wasn't able to find many helpful pointers so far.

Basically, I would like to have a terminal-only session that automatically runs a CLI application after login, and keeps restarting this application after a set amount of time (e.g. 1 minute) whenever it exits. For the duration where the application is exited, the shell should be regularly usable.

So here's my current setup (which works nicely so far):

• I have a terminal-only session (no GUI, linux mint in this case).
• After logging in, I automatically run an interactive CLI application (easy to do by adding it to .bashrc for example)
• After the user exits my CLI application, they are back in their regular bash terminal and able to run commands (as they should)

Now the tricky part which I was unable to figure out so far:

• How can I automatically restart the application in the user's terminal after a short time, while still keeping the terminal usable before that timeout is reached?
• We can assume that the user won't be doing anything in the bash that could break things (such as starting additional shells, exiting running processes or logging out). They have to be able to run 1-2 other commands that complete quickly, but not much else.

I tried accomplishing this by running a script that (in a loop) alternates between running the application and running the timeout command that should start a new bash, but was unable to get it right. It doesn't seem to be possible to start an interactive bash session using the timeout command. So this might be one possible way to do it (pointers on the correct syntax are welcome), but any other approach or idea is also perfectly fine.

Answer 1

I'm having the same thoughts here as Shelter Consider this pseudo-logic

answersRoomOneArray = {puzzleOneAnswer, puzzleTwoAnswer ... puzzleNAnswer}
# Room 1 Puzzles

puzzleOne () 
# Function to return a value
return puzzleOneAnswer

# Continue puzzles until Array is Full
...
if arrayCorrectRoomOne {} = answersRoomOne{} then
    unlockDoor ()
else
   punish ()
fi

You use the above format in each room terminal with a arrayCorrect stored somewhere in a directory the participants can't get to, in case they're terminal smart. Using this modular approach keeps you from needing a "controller script" without the need for a timer. If the script needs to restart clear the screen and restart it in the punish() function. If you only want so many attempts before they "die" add a counter to count how many times punish() was reached, and "kill" participants after X attempts.