Andrei writes "The Case for D"

[bearophile]

grauzone:
> Just that LDC (why not mention it directly?) is starting to become more 
> stable than DMD, while the .NET port is in an early alpha stage at best.

I agree with you, mentioning LDC name there is better. This part of the article is bad.
And LDC is getting quite stable, I'm using it on Linux with good results.
I think the main missing parts of LDC (but please correct me if I am wrong) is that is lacks still: exception on Windows, built-in profiler, built-in code coverage. Once those three things are in place then LDC can replace DMD in most situations (I don't know much about the code coverage in LDC, I'll ask on IRC).


> Now isn't that quite implementation specific. Maybe you should also 
> mention that dmd is not only fast, but also contains lots of bugs that 
> become issues in real life.

I think the author of the article is trying to show only how D2 can be good in theory, its good sides, and not its "current" real-world faults. I don't agree with this choice, I prefer my texts to be more realistic and to list downsides/faults too. I try to be more balanced.

Bye,
bearophile