Where will D sit in the web service space?

[via Digitalmars-d]

On Tuesday, 14 July 2015 at 15:17:13 UTC, Sebastiaan Koppe wrote:
> On Sunday, 12 July 2015 at 12:14:31 UTC, Ola Fosheim Grøstad 
> wrote:
>> What do you think about the future for D in the web service
>> space?
>
> What about this question: in 5 years from now what would be the 
> reason D failed?

It can't fail as long as Walter is having fun working on the 
compiler. :-)

> These come to my mind:
>
> Tooling
> Marketing
> Talent Pool (companies not willing to adopt)

I think neither, but what I have alluded to in this thread. A 
lack of decision making regarding picking an application domain 
that is large enough to sustain an ecosystem of libraries, and 
going 100% for honing the feature set towards that domain. 
"Possible" is not good enough.

I don't think Golang would have a chance to sustain adoption 
without being vocal about focusing on a low latency GC. Right now 
only Java/C# are typed alternatives. That makes Go pretty strong 
despite oddities.

I don't think Rust would stand a chance without being very clear 
on high level feature set, system level access, single threaded 
friendly with safe cross-thread-exchange linear typing memory 
model. It "sounds" clean. Only C++ is the perceived alternative. 
That's enough to get a surprising number of people to figure out 
how to deal with a rather challenging memory model. Not sure if 
Rust can sustain it, but hey, they might develop "best practice 
patterns" that makes it work...

I think the rest comes when you have the best feature set for a 
particular domain and a polished compiler/runtime. So yeah, maybe 
Game clients is the best bet, since you don't have to change the 
semantics too much (low latency GC and linear typing would take 
time to work in) and games benefits from C++/iOS interop. Indie 
games have low adoption threshold and could work as marketing.