Do I have any control over torder in which objects are destructed? Is this standard? I am asking because I am thinking about how I might handle fatal errors in an application and thought about using RAII to make some things are destroyed. But sometimes, some objects might be dependant on another objects existance to be destroyed correctly (example, a device context object would need an hwnd to exist). Is it even appropriate to make RAII self-cleanup thingies for hwnd's, dc's, rc's, etc.?
Perhaps stuff you've already encountered, but the brief results of my skimming:
http://parashift.com/c++-faq-lite/dtors.html#faq-11.2
http://www.parashift.com/c++-faq-lit...html#faq-17.12
http://www.parashift.com/c++-faq-lit....html#faq-6.18
7. It is easier to write an incorrect program than understand a correct one.
40. There are two ways to write error-free programs; only the third one works.*
Well, RAII is exactly one answer (when used correctly of course) to problems derived from object destruction. The key notion behind the type of control you want is that of "object ownership" and under it, yes. You will have control on how and when an object is destroyed. Usually by relaying that control to the implementation part of your class.
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
