Re: Mathematica goes Bad
- To: mathgroup at smc.vnet.net
- Subject: [mg59481] Re: Mathematica goes Bad
- From: "Harold" <Harold.Noffke at wpafb.af.mil>
- Date: Wed, 10 Aug 2005 02:58:11 -0400 (EDT)
- References: <dd9n09$iuh$1@smc.vnet.net>
- Sender: owner-wri-mathgroup at wolfram.com
Igor: It is well known in software engineering that any large system is never fully debugged. The maintenance process either introduces more bugs, or unexpectedly mucks up a larger process which uses code which is debugged or extended in some "in the small" obvious way. Dynamically bound code is very laborious to unravel when such maintenance artifacts suddenly appear. Statically bound code is _much_ easier on maintenance programmers trying to fix such problems. Does anyone know if Mathematica's "Solve" functions are dynamically bound or statically bound internally? If they are all dynamically bound, then unexpected results occurring after debugging or upgrading are, I'm afraid, the rule of the day. I've never seen a book or paper which lists more than vague pablums to keep such problems from happening. Inputs from anyone else on this subject would certainly be of interest to me. - Harold