"Walter Bright" <newshound1@digitalmars.com> wrote in message news:fjc8ls$7lq$1@digitalmars.com...
> Craig Black wrote:
>> Besides, given the scope of D 2.0, I can't quite fathom what features you would include in a D 3.0.
>
> It's far too early to make plans, but I expect 3.0 would be a push to support functional programming and other things for multicore programming.

I was under the impression that transactional memory would be included in 2.0.

Craig Black wrote:
> I was under the impression that transactional memory would be included in 2.0.


Too soon to tell.

Op Thu, 06 Dec 2007 12:58:31 -0800, schreef Walter Bright:

> Aggh! I should proofread my own posts!

Or you could try to borrow Guido's legendary time machine...  ;-)


-- 
JanC

On 12/8/07, Craig Black <craigblack2@cox.net> wrote:
> "Walter Bright" <newshound1@digitalmars.com> wrote in message news:fjc8ls$7lq$1@digitalmars.com...
> > Craig Black wrote:
> >> Besides, given the scope of D 2.0, I can't quite fathom what features you would include in a D 3.0.
> >
> > It's far too early to make plans, but I expect 3.0 would be a push to support functional programming and other things for multicore programming.
>
> I was under the impression that transactional memory would be included in 2.0.

It seems to me, D being a pragmatic language targeting realistic environments, STM wouldn't be a win with the number of cores most machines have these days (right?).

Plus, that area still seems to be seeing rapid evolution. It'd be pretty terrible if we ended up with some syntax in 2.0 for an experimental feature that could end up being the "const debate" of 3.0, simply by jumping the gun before the technology has stabilized enough to be generally useful.

>
>
>

David Wilson wrote:
> Plus, that area still seems to be seeing rapid evolution. It'd be
> pretty terrible if we ended up with some syntax in 2.0 for an
> experimental feature that could end up being the "const debate" of
> 3.0, simply by jumping the gun before the technology has stabilized
> enough to be generally useful.

IMO, the syntax Bartoz presented at the conference was fine.