opCmp / opEquals do not actually support partial orders

[Simen Kjærås]

On Thursday, 19 July 2018 at 10:14:34 UTC, Dominikus Dittes 
Scherkl wrote:
> On Wednesday, 18 July 2018 at 17:30:21 UTC, Jonathan M Davis 
> wrote:
>> On Tuesday, July 17, 2018 21:18:12 John Colvin via 
>> Digitalmars-d wrote:
>>> Just do what std.typecons.Proxy does and return float.nan for 
>>> the incomparable case.
>>
>> Since when is that legal? I thought that it was required for 
>> opCmp to return int. Certainly, the spec implies that it has 
>> to be int. The fact that the compiler allows it seems like a 
>> bug, though if Phobos is doing it, it wouldn't surprise me if 
>> Walter would choose to update the spec rather than fixing the 
>> compiler.
>
> It always worked with float as returntype (at least since I'm 
> using D
> - about 2.40 or so?), and it was necessary for the 
> (unfortunately
> deprecated) special operators <>, !<>, <>=, !<>=, ...

The oldest reference to opCmp returning NaN that I've found is 
from 2005, so it's not an entirely new thing:
https://forum.dlang.org/thread/dd3cea$26u7$1@digitaldaemon.com

--
   Simen