Re: [documentation] Confusion: vector::operator[] and exceptions

[Warren Young]

You might check out the current svn version of MySQL++ to see if this 
moves it in the right direction.  The move isn't complete.  I'm going to 
try adding BadIndex exception throwing from all operator[]'s and at() 
methods.

I might not be able to add it to all of them for compatibility reasons. 
  For instance, I already tried making mandatory the BadFieldName 
exception thrown by Row::operator[](const char*) when the field doesn't 
exist, but it turns out that SSQLS depends on being able to suppress 
these exceptions.  (It needs it for partial initialization of an SSQLS, 
a very helpful feature.)