On 2025-08-19 16:16, Ron via Talk wrote:

William Park via Talk wrote on 2025-08-17 21:51:

It's arithematic expression, so an easy way to remember:
     number 0      --> returns 1 ("fail")
     number non-0 --> returns 0 ("success")

You're right, once one knows about it it's easy to remember.

But it's highly counter-intuitive (*who* here knew about this behaviour
before this thread?).


And bash is littered with such weirdness.

This is actually not a bash-ism.

Early Unix (at least V7) used 0 as a successful program completion code.
Because numbers other than 0 were used to designate the kind of error.

The most logical would be to make true=1=sucess and false=0=failure but there is no way in the 8bit return code of V7 to specify what the failure was.

This specific problem is older than Bash.
Bash just inherited a weird answer.

BSD carried it on with and then Linux.
Take a look at /usr/include/sysexits.h

-- 
Alvin Starr                   ||   land:  (647)478-6285
Netvel Inc.                   ||   home:  (905)513-7688
alvin@netvel.net              ||