Do we need Mat, Vec, TMmat, Diag, Sym and other matrix types?

[9il]

On Thursday, 15 March 2018 at 14:13:25 UTC, jmh530 wrote:
> On Thursday, 15 March 2018 at 12:49:22 UTC, jmh530 wrote:
>> [snip]
>>
>> It looks like it should expand the alias earlier. No problem 
>> with auto foo (T)(S!(1, T) v) {};
>
> Also, this issue also shows up in mir.ndslice.traits. I had to 
> do the equivalent of isV below. It doesn't work to do the 
> alternate version. However, given that you have the traits, 
> then you can use them in a template constraint. So you have to 
> repeat yourself in the trait once, rather than bunches of times 
> in each function that calls them.
>
> enum bool isV(T) = is(T : S!(1, U), U);
> enum bool isV_alternate(T) = is(T : V!(U), U);

This does not help in practice because ndslice has an alias this 
primitive to be implicitly convertible to const version.

For example for array one can write:

auto foo(T)(const(T)[] ar) {}

And this would work with double[] and immutable(double)[] as 
well. The same true for ndslice:

auto foo(T)(Slice!(Contiguous, [1], const(T)*)[] ar) {}

It is very important to do not create additional institutions for 
numeric code because usually it is quite heavy.

I don't know DMD internals. How the  aliasing issue can be solved?

Best regards,
Ilya