contification paper

[Matthew Fluet]

> Since the writeup of the formal transformation is not done, I propose that we
> move the intuitive stuff about the transformation in 4.2 (i.e. the two bullet
> points) to the beginning of 4.3, and introduce the remainder of 4.3 as
> justifying why the bullet points are sufficient.

So, I introduced those bullets as a way to introduce safety.  But, if you
think it flows alright along the lines of:

An analysis is meant to guide the contification transformation.  However,
it should be clear that not all analyses will lead to sensible or correct
transformations.

then I think we can move those bullets.  And, as I read this, given the
examples already in the introduction, I think this is fine.  When I just
had the framework section, it felt a little out of the blue to define
safety, but I think the context is built up enough now.

I saw your todo note about Kelsey95.  I've got a copy of the paper and had
read it back when it first came up in the discussion with Reppy, but if
you know what you want to say there, go ahead and add it.