Always false float comparisons

[tsbockman via Digitalmars-d]

On Monday, 9 May 2016 at 18:37:10 UTC, Xinok wrote:
> (1) Yes, emit a warning for this case.
> (2) Generalize it to all variables, like Nordlöw suggested.
> (3) Generalize it to all comparisons as well, including < and > 
> .
> (4) While we're at it, let's also emit a warning when comparing 
> signed and unsigned types.
> (5) Dare I say it... warn against implicit conversions of 
> double to float.
> (6) The same applies to "real" as well.
>
> All of these scenarios are capable of producing "incorrect" 
> results, are a source of discrete bugs (often corner cases that 
> we failed to consider and test), and can be hard to detect. 
> It's about time we stopped being stubborn and flagged these 
> things as warnings. Even if they require a special compiler 
> flag and are disabled by default, that's better than nothing.

(1) is good, because the code in question is always wrong.

(2) is a logical extension, in those cases where constant folding 
and VRP can prove that the code is always wrong.

(3) Makes no sense though; inequalities with mixed floating-point 
types are perfectly safe. (Well, as safe as any floating-point 
code can be, anyway.)

(4) is already planned; it's just taking *a lot* longer than 
anticipated to actually implement it:
     https://issues.dlang.org/show_bug.cgi?id=259
     https://github.com/dlang/dmd/pull/1913
     https://github.com/dlang/dmd/pull/5229