www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Channel 9 interview with Scott Meyers, Herb Sutter, and yours truly

reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
Includes a long discussion about D:

http://channel9.msdn.com/posts/Scott-Meyers-Andrei-Alexandrescu-and-Herb-Sutter-C-and-Beyond


Andrei
Aug 24 2011
next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 8/24/2011 11:50 AM, Andrei Alexandrescu wrote:

 Includes a long discussion about D:

 http://channel9.msdn.com/posts/Scott-Meyers-Andrei-Alexandrescu-and-Herb-Sutter-C-and-Beyond




On reddit: 
http://www.reddit.com/r/programming/comments/jt7zm/scott_meyers_andrei_alexandrescu_and_herb_sutter/
Aug 24 2011
next sibling parent bearophile <bearophileHUGS lycos.com> writes:
Walter:


 http://www.reddit.com/r/programming/comments/jt7zm/scott_meyers_andrei_alexandrescu_and_herb_sutter/



The interview was nice. Sadly they don't talk about the "pure" attribute of D
:-)
Andrei needs more water.

From that Reddit thread:


I dislike C++ so much that I've been writing emulators in C#. It's quite nice
to program with, but the speed penalty is not minor for this sort of cpu-bound
work. I'm currently evaluating the feasibility of porting my emulation cores to
D (while leaving the client in C#).<



Even if the result of such evaluation of D will be "no", it will be quite
interesting to know why. Failed use cases too are interesting.

Bye,
bearophile
Aug 24 2011
prev sibling next sibling parent reply Andrej Mitrovic <andrej.mitrovich gmail.com> writes:
Cool interview. The D discussion was looooooong, I was expecting a
2-minute talk about D and then moving onto C++ business. Pretty cool!

Now if we could have an easy way to enforce the GC out of compilation
via some compile-time switch we could forever get rid of that GC
equals slow application argument. I find that "D is 90% of C++'s speed
and that's it" comment to be silly. I'm paraphrasing there, I can't
recall the exact words, but it was something to that effect.

Anyway, interesting times ahead with D and C++. Did they do any work
at all regarding the ABI for C++11? Like define it properly for
interoperability with other languages?
Aug 24 2011
next sibling parent Timon Gehr <timon.gehr gmx.ch> writes:
On 08/25/2011 04:13 AM, Andrej Mitrovic wrote:

 Cool interview. The D discussion was looooooong, I was expecting a
 2-minute talk about D and then moving onto C++ business. Pretty cool!

 Now if we could have an easy way to enforce the GC out of compilation
 via some compile-time switch we could forever get rid of that GC
 equals slow application argument. I find that "D is 90% of C++'s speed
 and that's it" comment to be silly. I'm paraphrasing there, I can't
 recall the exact words, but it was something to that effect.



Once the D GC is replaced by a sufficiently sophisticated one, I believe 
D with GC could quite easily beat C++ in benchmarks.


 Anyway, interesting times ahead with D and C++. Did they do any work
 at all regarding the ABI for C++11? Like define it properly for
 interoperability with other languages?
Aug 24 2011
prev sibling parent "Paulo Pinto" <pjmlp progtools.org> writes:
"Andrej Mitrovic" <andrej.mitrovich gmail.com> wrote in message 
news:mailman.2507.1314238444.14074.digitalmars-d puremagic.com...

 Anyway, interesting times ahead with D and C++. Did they do any work
 at all regarding the ABI for C++11? Like define it properly for
 interoperability with other languages?



This is not possible to do properly. For 64bit there is one ABI defined,
which was based on Intel's Itanium ABI, but it only concerns C++.

It is not possible to properly define a C++ ABI that can be used across
languages with multiple concepts. The only reason that C suceedes here
is because it is usually the OS ABI and as an high level assembler, most
languages can easily map C's concepts and they need anyway to make
use of the OS APIs.

With C++ is a different story, or with any other language that compiles
to native code, even D. You would need to think of ways how to expose
language specific concepts, or limit what can be made part of the ABI
like .Net does with the CLS specification.

--
Paulo
Aug 25 2011
prev sibling parent Jesse Phillips <jessekphillips+d gmail.com> writes:
On Thu, 25 Aug 2011 04:13:55 +0200, Andrej Mitrovic wrote:


 Cool interview. The D discussion was looooooong, I was expecting a
 2-minute talk about D and then moving onto C++ business. Pretty cool!
 
 I find that "D is 90% of C++'s speed and
 that's it" comment to be silly. I'm paraphrasing there, I can't recall
 the exact words, but it was something to that effect.



He was saying that D was 90% native, 10% virtual machine.
Aug 25 2011
prev sibling next sibling parent Andrej Mitrovic <andrej.mitrovich gmail.com> writes:
Still 15 minutes to go, 1.2 gig download! (the video they broadcast is
of poor quality so I'm fetching the high quality one..)
Aug 24 2011
prev sibling parent "Paulo Pinto" <pjmlp progtools.org> writes:
Cool, already know what to watch tonight's evening. :)

"Andrei Alexandrescu" <SeeWebsiteForEmail erdani.org> wrote in message 
news:j33h63$2opt$1 digitalmars.com...

 Includes a long discussion about D:

 http://channel9.msdn.com/posts/Scott-Meyers-Andrei-Alexandrescu-and-Herb-Sutter-C-and-Beyond


 Andrei
Aug 25 2011