Hi David,

On Fri, May 02, 2014 at 09:11:55PM -0700, David White wrote:
> So a couple of things:
> 
> When I implemented the RNG I was aware of the potential problems with it. I
> maintain that these problems apply to using an RNG for cryptography or
> other security. I think this is a very different use case to using it for a
> game. I maintain that the current implementation is just fine for a game.

I agree, I mainly like to switch for the larger range as is sometimes
needed for WML.

> The C++11 standard includes RNG features. Is there any reason we can't use
> this?

I advised to pick a algorithm from boost random that is also included in
C++11.

> I think moving Wesnoth toward using C++11 in general would be nice, as
> all major compiler vendors are rushing to implement C++11.

In my opinion C++11 compiler support is not that great yet :-( If we
want to move Wesnoth to C++11, we should IMO also start to rewrite
existing code to C++11, for example enabling move support can change a
parts of the class hierarchy. I have no idea how unfamiliar C++11 will
look to our development team. (I have been using it for several toy
projects and feel comfortable to use it, but C++11 code can look rather
different from C++98 code.)

We have a Google Summer of Code project to rewrite the addon server.
This project will use C++11 and should use C++11 idoms. Trademark, the
mentor for this project is familiar with C++11. Since the addon server
has no real ties with the rest of the code and it is the ideal place
to have a look at C++11. So I'd like to evaluate C++11 after the summer.
(I think we should/will move at some point, I just don't know when.)


-- 
Regards,
Mark de Wever aka Mordante/SkeletonCrew

_______________________________________________
Wesnoth-dev mailing list
[email protected]
https://mail.gna.org/listinfo/wesnoth-dev

Reply via email to