[GAP Forum] galois group

[Marco Costantini]

Dear all,
Gap is an open source system, so each user can look at the code of each 
function, to check both whether it is correct, and whether it does what the 
user needs. (Yes, this requires time, but also reading and full understanding 
a published paper requires time.)

Of course bugs in Gap happen, however, when bugs are reported, they are 
usually fixed within a reasonable time. On the other hand, usually "bugs" in 
published papers are not fixed, and it has been reported that one refereed 
paper out of four contains at least a "bug".

So I would say that using a result from Gap may be at least as safe as using a 
result from a published paper.

Best regards,
Marco Costantini


On Tuesday 25 April 2006 17:58, Alexander Hulpke wrote:
>
> Marta Aseda wrote:
> > I am wondering if galois group computation using gap is useable as a
> > mathematical result.
>
> In general, unless you explicitly turn off verifications or use
> functions which deliberately only return proibabilistic results (such
> as `ProbabilityShapes' for Galois groups), any result obtained with
> (documented functions of) GAP is proven correct. (There is of course
> always the philosophical problem of human error in  implementation or
> use, however the same problems are as well in published papers.) The
> aim is that one could use such a result in the same way as a result
> from a published paper.