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

[Chris]

On Saturday, 13 April 2019 at 11:20:07 UTC, Andrei Alexandrescu 
wrote:
>
> Closely related to the dearth of great contributors is the 
> implication that we're blocking access of others to high-impact 
> and leadership roles. Let me state for the record that I am not 
> holding onto any seat, that I had to carry some roles simply by 
> default because nobody else would, and that there's many more 
> than two seats to start with, most of which are empty because 
> there's nobody able and willing to take them. The folks whom I 
> invited privately declined, thankfully save for Mike.
>
> Regarding some posters' discontent over poor performance of the 
> leadership/management. Speaking only of myself, let me state 
> for the record that I hold absolutely no illusion of my 
> performance as a leader in this community. Without being 
> dramatic I reckon I'm unfit for leadership. (Most certainly not 
> for lack of trying, but that doesn't matter.) An A leader 
> should get A+ people to follow. We don't have enough of those.

I said a few months ago that one cannot be leader and developer 
at the same time, because one will inevitably under-perform in 
one of the two departments, or in both. And unless you are 
supernaturally disciplined your personal preferences will creep 
in. The impression I have is that you and Walter are interested 
in one particular aspect that may be interesting and rewarding to 
work on and this aspect becomes top priority - just because you 
*can*. You are the leaders after all, nobody can tell you what to 
do. And the aspect you're interested in may change according to 
the latest discussions in CS, if Go or Rust has it, D needs it 
immediately. How often have I heard sentences like "This is the 
future of programming! This is the way to go!"? But this approach 
has created a solid mess. Half baked features. You use the actual 
D release as a playground instead of having a D Lab or D 
Experimental implementation. With something like that a lot of 
pitfalls could have been avoided and users wouldn't feel like 
guinea pigs.

Often less attractive features are not only neglected but 
regarded as beneath D as they belong to the realms of "mediocre 
coders who use Java, cough cough". But in the real world 
programmers need to get things done and competitors don't sleep. 
If a customer wants an app you cannot tell him / her that the 
language you're using is "too good" for that.

When I look at new languages, what are the features I look for? C 
interop, cross platform and *ease of use*. Many new languages are 
designed with all-round usefulness in mind and the devs try to 
make it as easy as possible for users. This is very important as 
it boosts productivity. Also, programmers just don't want to go 
through all the hassle of million compiler flags, complicated set 
ups etc. Not because they're stupid, but because it's the year 
2019 and what matters is the code and the product. Why would I 
use two stones to light a fire when I can just use a lighter?

One thing regarding the number of downloads. That doesn't tell 
you much. First, I often download languages out of curiosity, but 
do I use them? No. Second, I've often downloaded libs etc. in D 
but once they break, I have to stop using them. I have a 
graveyard on my machine. So the number of downloads does not 
necessarily tell you whether the language / library is actually 
being used.

It's too high a risk to use D in the real world. Imagine how a 
user feels when s/he reads on the forum that there will be yet 
another change that might break old code or that D / Phobos will 
now be rewritten. It's just mad.