On Jun 8, 2009, at 7:56 PM, Emil Dotchevski wrote:
On Mon, Jun 8, 2009 at 4:10 PM, David Bergman <David.Bergman@bergmangupta.com> wrote:
On Jun 8, 2009, at 6:58 PM, David Bergman wrote:
On Jun 8, 2009, at 6:40 PM, Emil Dotchevski wrote:
Specifically, what platforms that library runs on is not important.
It is important to the ideal of Boost, in my humble opinion (and interpretation of the stated goals of Boost.)
My response sounded harsh. What I think I meant is what is stopping someone from creating a Boost-compatible and Boostesque wrapper for DirectX? Why must it be part of the Boost libraries?
You can ask this question for any Boost library: why should it be part of Boost?
Most libraries are integral to Boost, and constitute aspects that together bring the developer to another level of abstraction and solution terminology. Boost is not a random set of code happening to be compilable with C/C++ and somehow having a potentially big audience. I do not welcome domain-specific or target-specific code with open arms.
Presumably the answer is "because many programmers (Boost users) would benefit from it."
No. The answer is - or should be - because programmers can benefit from FOO on any platform and most types of applications, and FOO is in harmony with the goal of Boost and the existing libraries. I am not even super-happy about quite specific math libraries entering Boost now and then - ending up with a bunch of them. /David