Re: noverlay branch

[Stefan Monnier]

> I see you removed some of the null->parent trick just today.  I like
> that idea.  It is realtively easy to use actual NULL instead of a
> sentinel NULL in tree algorithms, and I think on modern processors this
> works out for the better.

The change I made only made it so the sentinel's fields are either
read-only or write-only, which means the sentinel is not used to
propagate information and is thus not a global state.

Whether using NULL is worse or better than using a sentinel, I don't
know, and I think the current code is fine in this respect.  If someone
wants to change it to use NULL, feel free, but it's not on my todo list.


        Stefan