If the GC issue can not be tackled and even with the recent
communication blogs, it still keeps showing up. Is it maybe not
better to focus the marketing features that other developers (
none C++ ) may see as advantages and slow draw then in? High
performance web development package for instance. Vibe.d does
not count because has not been seen giving any descent public
performance in years ( like techempower or other sites ).
At the moment, for very demanding high-performance web
applications, the main choices are :
- Dart
- Go
- C/C++
- Java/Kotlin
This is somewhat confirmed by the following benchmark (despite
its well known flaws) :
https://www.techempower.com/benchmarks/#section=data-r14&hw=ph&test=query
I think that D, with its fiber system and extreme execution
speeds, should already have everything needed to compete on the
same grounds as Dart and Go.
But only if :
- the fibers automatically uses all cores if allowed to do so;
- the GC problems is designed so that the server never hangs for
more than a few milliseconds IN THE WORST CASE (progressive GC)
- dedicated HIGH PERFORMANCE web server (cfr Go/Dart) and
database access (MySQL, PostgreSQL, Aerospike) libraries are
available, preferably as a standard library (like in Dart and Go).
And for the third point, I really mean high performance,
something that make D's implementation of the (silly) benchmarks
immediately go to the top of the chart.
If all that is already available, perfect :)
Then we just need to add the D version of the benchmarks, and the
marketing is done.
And moreover I'd be delighted to start using D instead of Go for
my next web server developments.
