It is the year 2020: why should I use / learn D?

[NoMoreBugs]

On Thursday, 15 November 2018 at 22:29:56 UTC, Stanislav Blinov 
wrote:
>> Writing C++ code therefore becomes an exercise in navigating 
>> the obstacle course of an overly-complex and fragile 
>> language...
>
> Same will happen to D. Or rather, it already has.

++1

That's my impression of D too, lately. It's seeking stability - 
but at what cost?

C++, at least, is boldly going where..perhaps it should not go ;-)

.. but at least it's moving in a direction that is (speaking for 
myself) making programming in C++ at little less brittle (*if* 
you stick to particular features, and learn them well). Trying to 
learn all of C++ is just complete nonsense (as it is for almost 
any language). It would take decades just learning it all (and 
it's a constant moving target now, making it even more difficult 
- i.e Scott Meyers).... and nobody needs to use all of the 
language anyway.

D is no exception to this - it is also a rather complex language 
with far too many features that any single programmer would need 
in totality. Pick a subset, get good at using it. Preferably the 
subset that can best provide guarantees of software correctness 
;-)

As for the next 'paradigm', it won't be 'unbridled freedom', I 
guarantee that.

The programmers may certainly want that freedom (I certainly do), 
but the institutions/corporations who will be impacted by that 
'unbridled freedom', will want better guarantees around software 
correctness - not more freedom.

So in my opinion, the language that can best provide such 
guarantees (with consideration to other constraints that might 
apply), is the language that people will flock too.

D provides a lot in that area (which is what attracted me to it), 
but, it breaks awfully in other areas ( I'm thinking implicit 
conversions (so old school), no concept of private state *within* 
a module (what! really!), no appetite at all for addressing many 
issues, ...etc..etc).

C++ is the Bear. Poke it at your risk.