Const-correctness and uniqueness. Again.
deadalnix via Digitalmars-d
digitalmars-d at puremagic.com
Wed Feb 11 10:50:15 PST 2015
On Wednesday, 11 February 2015 at 18:37:42 UTC, deadalnix wrote:
> On Wednesday, 11 February 2015 at 13:54:12 UTC, Dicebot wrote:
>> On Monday, 9 February 2015 at 23:40:31 UTC, deadalnix wrote:
>>> Someone is starting to see where I'm getting at when I'm
>>> pushing for owned...
>>
>> I don't think this specific case is a good justification for
>> your proposal - it is simple enough to accept a library based
>> solution. We need more accumulated anecdotal evidence.
>
> We do have. That is simply one more. On the top of my head:
> - Plug hole in the type system (implicit sharing).
> - Use the guarantee the type system is supposed to provide to
> get a better GC.
> - Construction of immutable objects.
> - make @nogc actually useful for something else than trivias
> (throw, defers usage of GC to the caller).
> - make std.concurency and std.parallelism safe end to end.
> - Be able to lock tree of objects.
> - Generating less garbage overall.
> - ...
Or we can add an easier hack for each of these. That is gonna
scale well.
More information about the Digitalmars-d
mailing list