Why lack of good IDE doesn't peek your attention

[Gru via Digitalmars-d-ide]

Hi, I have also been a long time lurker of the D language. D 
itself is very fine. I come from C# background for the most part, 
while having worked in other C-based languages like Java and C. I 
also used Visual Studio for years.

A couple of months ago, I tried setting up an IDE.
VisualD, MonoD, This thing with IntelliJ, and VSCode plugins - 
none of them worked out of the box.

The way I tried to install it is like a normal user. I would try 
installation while following the steps and if it doesn't work I 
quit. Yes, I could probably troubleshoot the installation but 
that's the core problem: if the installation is broken you will 
simply lose the confidence in the product as a person evaluating 
the product. While the OSS community "likes" these kinds of 
installations, they are a death for the interest of general 
public.

As someone coming into D from the outside, I cannot state how big 
of a problem this is for user adoption. Go is the proof that 
language is not sold by its features, it is sold by the 
perception for the users. What's more important than being good 
is making a good impression and that is basics of sales.

Also, you may think that a good IDE is not important. Let's put 
it this way, from an a lazy user point of view, if we wanted to 
skimp on productivity, why use a high-level language after all? 
Having a language in the family of C++/C#/Java without the good 
tooling support is ludicrous. If the code completion does not 
work, forget it. People spend enough time troubleshooting their 
own software. Imagine a compiler that works 100%, but only 90% of 
the time. This simply does not cut it. The relationship of the 
language and the tooling is very symbiotic. Even in C#, you know 
how in LINQ you type 'from' before 'select' even though you'd 
expect it to be the other way around? Well, one of the reasons is 
so they could offer better tooling support this way, and that's 
an example of how the language can be influenced by the IDE. At 
the end of the day, it all comes down to workflows. Either we are 
doing stuff (such as debugging) faster and better or we are not.

This is the Really important stuff, the end-to-end thinking which 
gets the job done and never focuses on the details at the expense 
of something more important down the pipeline.

I believe D can be a popular language, and it can build on the 
C++ popularity, but it needs to have really reliable tooling. In 
C#, language features and VS features go hand in hand. A language 
service for an IDE needs to work extremely reliably, just like 
any system.