|
Boost : |
From: Carl Daniel (cpdaniel_at_[hidden])
Date: 2002-01-08 10:13:28
From: "David Abrahams" <david.abrahams_at_[hidden]>
> From: "Greg Colvin" <gcolvin_at_[hidden]>
> > Perhaps an in-memory BTree would be a good compromise?
>
> I haven't seen a BTree implementation in years, but I'm guessing it
> generates lots of code and has complicated iterators, too. I'm sure it's
> sometimes useful, but it doesn't have many of the attractive qualities of a
> sorted vector.
Having just written a BTree with an interator-based interface, I can confirm that it has complicated iterators and a lot
of code :) Within the implementation of the BTree, I used a sorted vector (my own version - sequence_map<>) to
implement the pages of the BTree.
Let's get one (or more) of these alternate associative container implementations into Boost!
Dave - what were the issues with the version you & John Potter worked on several months ago? Why was that development
dropped?
I think there's sufficient justification for both the direct and indirect solutions - sometimes iterator stability is
important. Sometimes best performance is what's required.
-cd
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk