Unintuitional behaviour of range(-1,1)

[hirschenberger]

When iterating by using range with signed variables, the iteration doesn't start at all because the int gets converted to uint and -1 is set to std::uint::max, that's unintuitional.
There should be a runtime error (start > end) or the use of uint should be enforced somehow.

[thestinger]

There should be a runtime error (start > end)

These are meant to be easily composed without failure. It's sensible to not produce any numbers if the start is larger than the end, rather than treating it as error. This is usually the desired behaviour.

When iterating by using range with signed variables, the iteration doesn't start at all because the int gets converted to uint and -1 is set to std::uint::max, that's unintuitional.

I think you're running into a different problem, and mistakenly attributing it to range. It does work with signed integers and I don't see a reason to lose this useful functionality.

fn main() {
    for i in range(-10, 10) {
        println!("{}", i)
    }
}

Output:

-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9

[thestinger]

The real issue you're running into is covered by #11273.

[hirschenberger]

Yes you're right, that's the real issue