Possible bug of "missing matching more than one candidate" in overload set

[user1234]

On Friday, 13 February 2026 at 15:12:16 UTC, monkyyy wrote:
> On Friday, 13 February 2026 at 14:56:43 UTC, user1234 wrote:
>> On Friday, 13 February 2026 at 14:54:44 UTC, monkyyy wrote:
>>> Template specialization rules can apply. I think its only a 
>>> bug if the order of the declarations effect behavior(which 
>>> they do often)
>>
>> No, order doesn't change which element of set is selected.
>
> Not my point, specialization overrides generic cases; an empty 
> list is less generic then a vartatic one.
>
> ```d
> import std;
> void f(int i:0)(){"0".writeln;}
> void f(int i:3)(){"3".writeln;}
> void f(int i)(){"?".writeln;}
> unittest{
>   static foreach(I;0..5){
>     f!I;
> }}
> ```
>
> Id rather make different rules based off "first match", its a 
> mess, but the rules are consistent.

ah... misunderstanding. I meant that if you declare the variadic 
`f` before the "no param one", the "no param one" is still 
selected, which illustrates that order of declaration is not part 
of the game.