Componentizing D's garbage collector

[H. S. Teoh]

On Fri, Jan 10, 2014 at 08:53:44PM -0800, Andrei Alexandrescu wrote:
> On 1/10/14 8:26 PM, Timon Gehr wrote:
> >On 01/10/2014 09:03 AM, Andrei Alexandrescu wrote:
> >>
> >>* There will be no possibility to change the type of certain objects
> >>once allocated. An allocation for an immutable object cannot e.g.
> >>make it later a mutable one. (This is an essential issue with the
> >>current allocator, I think.)
> >
> >I assume you are aware that there is implicit casting to immutable
> >upon return from a strongly pure function. What about it?
> 
> I don't know. Need to think about it. Maybe it's a wrong design
> decision (or maybe separate heaps are wrong).
[...]

I dunno, separate heaps, esp. for immutable vs. mutable, seems like a
very powerful GC optimization opportunity. Missing it seems like such a
pity, esp. given the price we pay for immutability (transitivity, actual
immutability guarantees unlike C/C++ const, etc.).


T

-- 
If you want to solve a problem, you need to address its root cause, not
just its symptoms. Otherwise it's like treating cancer with Tylenol...