I really don't understand the reasoning for not removing as many
known sources of bugs as is reasonably possible *provided* that
doing so makes the situation incrementally better (rather than
worse or to no effect).
So will introducing non-nullable references make things worse or
have no practical effect?
There's also more to the equation than only reducing a potential
source of bugs, as it also eliminates the manual null checks that
programmers inevitably place in their code. More code always
means introducing more bugs, along with higher development and
maintenance costs.
I also know that allowing null references has a use (I use them),
but as was discussed before, we can have both options.
Non-nullable references look like a win-win to me.
--rt
