DIP 1019--Named Arguments Lite--Final Review

[nkm1]

On Thursday, 29 August 2019 at 21:53:36 UTC, Olivier FAURE wrote:
> On Wednesday, 28 August 2019 at 09:37:13 UTC, nkm1 wrote:
> I don't think it's obvious.

I guess it depends on whether you see functions parameters as a 
kind of struct (or both structs and parameters as kinds of 
tuple). Too me, they look almost indistinguishable. In any case, 
D already has certain rules for initializing structs and arrays, 
and the same rules _can_ be applied to parameters, so, using a 
variant of Occam's razor here...

> There's trade-offs to consider, that people have spent pages 
> debating.

That's what I'm saying, no? The DIP should discuss these 
trade-offs (but it doesn't).
Maybe D doesn't even need named parameters. Perhaps C-style 
struct literals with designated initializers would be perfectly 
sufficient (and more useful in general, and that also would 
adress Jonathan Davis' concerns about rewriting Phobos). That all 
should be discussed in the DIP.