Thanks for the patches. I've fixed the warnings that make sense to fix. Unfortunately some of the warnings are not valid, and with GCC there is no way to turn off individual warnings, so it won't be possible to get Crypto++ to compile without warnings at -Wall.
On Tue, Nov 04, 2003 at 11:52:27AM -0600, Stephen torri wrote: > It seems a bit odd to do this at compile time. That is just my simple > understanding of the code so far. It says that you are accepting in a > const class T type by reference. You then make sure that the lenght of > it is only one character like an array. The assert is asserting that the first element of the string has size 1, in other words, that the string is not a string of wide characters.
