Why is D unpopular?

[Ola Fosheim Grøstad]

On Wednesday, 3 November 2021 at 06:08:57 UTC, harakim wrote:
> really like uncertainty. As a human developer who works with 
> developers, I have noticed that human developers aren't super 
> interested in sinking a bunch of time into something that will 
> not provide lasting value. They look at the long-term/Big-Oh

It is true that developers look for safe choices when picking 
languages in professional contexts, that basically always mean 
either a language that is already popular or a very narrow 
language that is well supported.

As such, small generic niche languages will always look unsafe 
and developers with (sound) risk aversion will avoid them.

So, we cannot extrapolate too much from what choices people make 
in professional software development contexts. We should expect 
that the majority of uptake comes from individuals that have a 
lot of freedom and can take higher risk.

But the critical adoption is not with average developers. Average 
developers are essentially consumers that can motivate people to 
write tutorials and so on (basically an audience), but they do 
not increase the quality of the eco system. The critical adoption 
is related to those that are highly skilled programmers that also 
have programming as their primary hobby. The ones that are 
willing to both work all day with programming using a cookie 
cutter language and in addition want to spend their evenings and 
weekends programming a niche language (+ those that have a lot of 
freedom in their daytime). Those are the ones that can drive 
early adoption and build a solid eco system for small languages.

D's major problem is not that it does not have sufficient numbers 
of "consuming programmers". I think it does. The major problem is 
that it does not have enough of those hardcore hobbyists. It has 
not been able to retain enough of them over time.

That is where the vision is very important. And it would also 
help if the compiler had a better architecture, and some semantic 
cleanup to bring it more in line with the ideals of computer 
science (as the most skilled programmers will know what the 
semantics ought to be, and that can also be a turn off).

Comparing to languages like Go, C# etc does not make any sense, 
because those languages have reached critical mass. They have so 
many users that they statistically also have a large number of 
hardcore programmers (even if that percentage is very low).

A small language needs a higher ratio of hardcore vs average 
programmers than a language that already has critical mass.