[Issue 13838] New: @safe by default

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=13838

          Issue ID: 13838
           Summary: @safe by default
           Product: D
           Version: D2
          Hardware: x86
                OS: Windows
            Status: NEW
          Severity: enhancement
          Priority: P1
         Component: DMD
          Assignee: nobody at puremagic.com
          Reporter: bearophile_hugs at eml.cc

Languages like Rust show us how important memory safety is today, and the
planned DIP69 works for @safe code, this means now in D it becomes more
important to use @safe functions in most cases.

There are still some cases where you can't use @safe even if they should be
safe, some of them are shown here, but Phobos/druntime is getting better, and
they will decrease:


void main() @safe {
    import std.stdio, std.algorithm, std.bigint, std.typecons, std.array;
    [1, 2].sort!("a < b", SwapStrategy.stable);
    auto r = [1, 2].sort().release;
    writeln;
    BigInt a;
    a = a + a;
    alias Foo = Tuple!int;
    Foo[] data;
    data.remove!(x => x == Foo());
    int[] b;
    auto c = b.capacity;
    b.schwartzSort!(x => x);
    const r2 = cartesianProduct([1], [1]).array;
    [Typedef!int(1)].array;
}


So perhaps it's a good idea to have @safe functions by default. This is how it
could be done:
Step 1) Introduce a "-safe" compiler switch that gives a warning where a
function unmarked with @system/@trusted calls a @system/@trusted function or
performs memory-unsafe operations. This will help D developers improve Phobos.
Step 2) The functions defined above generate a warning if the -wi/-w switches
are used (even if no -safe switch is used).
Step 3) The functions defined above generate a deprecation message (even if no
-safe switch is used). The compiler -safe switch is still allowed, but it's not
documented in the options help message of the compiler.
Step 4) The functions defined above give an error, and the -safe switch is
removed from the compiler (it's not recognized any more).

See also Issue 12941

--