Subject: Re: speed results
From: Hubert Figuiere (hfiguiere@teaser.fr)
Date: Sat Apr 14 2001 - 01:52:25 CDT
On samedi, avril 14, 2001, at 04:17 , Joaquín Cuenca Abela wrote:
> In my tests, the algorithm to write a new word is ~O(n^2.5)
> Attached is a graph showing time(nb of words).
>
> I think that we should aim for 1.0 at (ideally) O(1), or at least
> O(log(n)).
> Worse than that is simply unacceptable.
Joaquin,
please fill a bug report for this so we can really plan it.
Sam, is it possible to add a "perf" keyword in bugzilla for this kind of
thing that would be used to tag performance issues ?
Hub
This archive was generated by hypermail 2b25 : Sat Apr 14 2001 - 01:52:28 CDT