Has D failed? ( unpopular opinion but I think yes )

[Dukc]

On Friday, 12 April 2019 at 14:24:05 UTC, Chris wrote:
>
> Mind you, the first thing a lot of people ask is "Can I use it 
> on mobile and is it painless to do so?". This is the reality of 
> things. Kotlin devs realized that. Language adoption is not 
> _only_ about language features, it's about usefulness too. This 
> is why they simplified Scala.

I don't think the language features that have been worked on 
lately are because mobile/JS/ect would be considered less 
important. Quite the contrary for many of them, I give you 
examples:

-The idea behind -dip1008 is to eliminate reliance of the garbage 
collector when doing exception handling. If it, or something 
similar, will be succesful, that means you can use exceptions on 
precise real-time enviroments, that can't afford the GC. And you 
can do that without doing any special memory handling. For me, 
that's exactly what counts towards portability and usefulness on 
different platforms.

-Sebastiaan Koppe's excellent spasm library, that allows D code 
on browsers, uses stdx.allocator (essentially the same as 
std.experimental.allocator of Phobos, I'm not sure if they have 
ANY difference). Currently stdx.allocator won't "just work" in 
no-runtime enviroment, you need to do some hacks. But the point 
is, that work done on Phobos allocator module, that hasn't been 
around for long, is exactly what makes it easier to port D 
outside desktop enviroments.

-BetterC. I don't think I need to explain this one.

-Interfacing to C++ and Objective C. No need to maintain a 
separate C interface that always breaks when porting.