CDGC porting: is someone still working on it?

[via Digitalmars-d]

On Thursday, 4 September 2014 at 08:57:53 UTC, ketmar via 
Digitalmars-d wrote:
> and does somebody working (or planning to work) on precise GC, 
> for that matter?

There are many discussions on the topic in the forums with 
various ideas and things being in the works, but nothing coherent 
I think.

What I'd prefer is to have a non-stack GC where you have 
GC-pointers within a graph and only trace those, backed by 
compiler magic. That's the most critical use scenario in my 
opinion. And it could be made very fast if you use cluster the 
GC-pointers to the same cache lines.

Good luck! :-)