Reply to thread

Message: [QUOTE="zr, post: 3708713"] Point Taken. As mentioned in the OP, i used the example from Marshal Cline's FAQ, and wanted to make minimal changes to make things work. Besides, i ran this test on WIN32, so the pointer size and alignment were good enough for it run with no crash. This is great feedback. Thanks. Whoops. I totally misunderstood that. Thanks for pointing that out. But the whole point of this exercise was to define a smart delete operator such that the client code does not need to associate the deleted pointer with the memory pool that allocated it. What about using a try-catch block in Bomb::Bomb() ? [/QUOTE]