Re: Red/black trees
Matthew Dillon <dillon@apollo.backplane.com> writes:
> :So what advantage do the red-black trees have, since the previous data
> :structure has been working fine for some time?
> :
> :Jonathon McKitrick
> :--
> :My other computer is your Windows box.
>
> The previous data structure was a doubly linked list with a bunch of
> hacks to try to make insertions go faster. The red-black tree should
> be just as fast, and not require any hacks. Plus we can simplify
> the clustering and fsyncing code, and make other optimizations.
is there are reason to not use skip lists ? now that we don't have to
rely on probabilistic measures anymore.
thanks
kind regards
anupam
>
> -Matt
> Matthew Dillon
> <dillon@backplane.com>
討論串 (同標題文章)
完整討論串 (本文為第 4 之 11 篇)：