The easiest way to compete with Rust and cure D's GC reputation: switch to ARC.

[Dukc]

On Wednesday, 9 July 2025 at 17:10:26 UTC, WraithGlade wrote:
> In particular, it seems to me that by far the easiest way to 
> cure D of its GC-dependent reputation (which is one of the 
> biggest reasons why people avoid using D) is to switch the 
> whole standard language and library to use ARC based memory 
> management instead.

I don't think this is a good idea.

Because ARC suffers from the cyclic reference leaks that would 
have to be manually worked around, it'd mean that D programs 
would more or less have to be designed like Rust programs. With a 
tracing GC, we can almost always allocate in fire-and-forget 
style. While that'd still be the case most of the time with ARC, 
it is going to happen that we accidentally cause memory leaks 
with circular references, and then spend a lot of time figuring 
out the problem and then designing around it.

Second, it means all pointer assignments will have to be 
write-gated, to increment and decrement the reference counter. 
Not a big issue for most programs, but a minus nonetheless.

If it was just about the second issue, it might well be still 
worth it. Likewise, if the GC was mandatory and therefore made 
the most demanding real-time programming impossible, but it 
isn't. Considering how rarely you have to go totally without the 
GC, I think the first issue is far bigger a problem than the 
unpredictability of the tracing GC.

Then again, the last time I heard from Ali Weka.IO doesn't use 
the GC at all in their superperforming filesystem. Maybe they 
disagree with my assesment and if they do, I might well change my 
mind.