DIP 1019--Named Arguments Lite--Community Review Round 2

[aliak]

On Sunday, 9 June 2019 at 11:26:07 UTC, aliak wrote:
> On Sunday, 9 June 2019 at 11:15:39 UTC, aliak wrote:
>> [...]
>
> There're also some technical advantages/disadvantages to either 
> or:
>
> 1) searchability is out the window if you allow reordering
>   find "func(a:" will now not work
> 2) semantic overloading is out the window
>   lets(int doA, int thenB int thenC)
>   lets(int doA, int thenC, int thenB)
>
> I'm not sure what technical advantages reordering brings though.
>
> I also wonder what kind of overload resolution can be 
> accomplished with reorderable and unreorderable, and also what 
> kind of overload resolution is blocked with either of the two. 
> I believe that should also be taken in to account to make a 
> better decision here.
>
> Maybe the DIP can address that point as well?
>
> Cheers
> - Ali

Forgot another technical advantage: evaluation of arguments. With 
unreorderable it's obvious. With reorderable, as Walter pointed 
out, what't the order (parameter-wise or argument-wise).