Re: noverlay branch

[Matt Armstrong]

Gerd Möllmann <gerd.moellmann@gmail.com> writes:

> On 22-10-06 6:47 , Matt Armstrong wrote:
>
>> 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.
>
> Clang libstd++ uses NULL, BTW, and I already wondered a little bit why.

I believe GNU libstdc++ does not use sentinel nodes either.  I have yet
to see see sentinel nodes used in an optimized tree implementation.

I think in the context of this overlay work the performance difference
is not very significant, since the code is doing a lot of other stuff
while traversing the tree.