How come there are so many bugs in XMaxima?

I think your question is worth careful consideration. Having done
thought about it for a while, I have some possible explanations.

1) There are different kinds of bugs. Bugs can be in the infrastructure
(for example, info not working) or in the actual mathematical algorithm.
There have been very few reports of Maxima bugs that result in
mathematically incorrect answers. When I was using Mathematica (years
ago) the bugs I saw in it sometimes resulted in actual mathematical
errors. The sorts of bugs we see right now in Maxima are often of the
form "function foo causes a lisp error." This type of bug is easier to
recognize and fix than the sort of bug that results in a few incorrect
steps in a large calculation.

2) I don't know what Macsyma looked like in 1968, but it had to be
fairly different from the Maxima we have today. Multiple dialects of
Lisp have been used in the mean time.

3) For a long time, Bill Schelter maintained Maxima more or less by
himself. I think we have a substantially larger number of people looking
at the code since he passed away.

4) Software really does rot. Many of the bugs that are being reported
recently are in functions/packages that haven't been exercised in years.
As more people exercise the code we should be able to clean out the
rotten bits.


--Jim

On Tue, 2002-03-19 at 13:33, synthespian wrote:
> Hi-
> 
> 	Please take no offense in my questioning this, I recognise
> people work hard on Maxima.
>         However, I read that Maxima was one of the most bug-free CAS
> amongst the rest, since it was made since 1968. Has there been a lot
> of code change?
> 	Are these bugs we see here from a recoding?
> 	
> 	Regs
> 	Henry
> 	synthespian@uol.com.br
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima