Path: utzoo!news-server.csri.toronto.edu!bonnie.concordia.ca!ccu.umanitoba.ca!herald.usask.ca!alberta!aunro!aupair.cs.athabascau.ca!atha!decwrl!elroy.jpl.nasa.gov!usc!samsung!munnari.oz.au!goanna!ok
From: ok@goanna.cs.rmit.oz.au (Richard A. O'Keefe)
Newsgroups: comp.lang.c
Subject: Re: Log Library - How is it done in the library code?
Keywords: log, library, series expansion
Message-ID: <4962@goanna.cs.rmit.oz.au>
Date: 14 Mar 91 00:17:35 GMT
References: <1991Mar11.022141.12068@m.cs.uiuc.edu> <15438@smoke.brl.mil> <1991Mar12.211047.826@m.cs.uiuc.edu>
Organization: Comp Sci, RMIT, Melbourne, Australia
Lines: 16

In article <1991Mar12.211047.826@m.cs.uiuc.edu>,
joshi@m.cs.uiuc.edu (Anil Joshi) writes:
> After going through them I figured out that my initial posting as well as
> the followup were not clear enough as to why I want roll my own log.
 
> 1. It is a new algorithm to solve LP problems that needs the log to be
> evaluated but allows for some lee-way.  The convergence rate of the 
> algorithm does not change[sic] on the accuracy of the log function.

In that case, why not just use 0?  If neither the correctness of the results
nor the rate of convergence depends on the accuracy of the logarithm, why
compute it at all?  It sounds as though frexp() would have done everything
you need (or logb() if you have it).

-- 
The purpose of advertising is to destroy the freedom of the market.