Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Non Professionals doing Engineering

[Subject Prev][Subject Next][Thread Prev][Thread Next]
On Mon, 25 Jan 1999, Dennis S. Wish PE wrote:
-- snip --
> The Computer Applications Committee of SEA of California is addressing the
> issue of control of computer programs. In this case it is an issue of
> submitting and accepting the reliability of a program. It was discussed that
> there may need to be a typical problem assigned to each category of design
> problem and any program which is submitted for this type of design should be
> able to submit calculation which proves that it can reliably solve the
> problem. This does not resolve potential bugs, but places the reliability of
> a program on a level platform so that those who create software in their own
> offices can assure the building official that the results are as accurate as
> the programs commercially sold. Commercial programs "generally" have better
> User's manuals and sample programs to support their programs abilities. This
> is not typically the case with proprietary software used by one or a few
> people.
> As we discussed in the past, the only assurance for reliability is the basic
> understanding of the materials and physics involved in solving the problem.
> Identifying a mistake is not easy unless the person reviewing it has solved
> enough problems (experience) to identify when a result appears to be out of
> order.
> Personally, I don't think that this needs to be done manually as a
> pre-requisite. The designer MUST have an understanding of the process, but
> with this understanding he or she should be able to evaluate the output - as
> would be expected of a plan check technician.
> I, too, am interested in the opinions of others on this issue.
> Dennis S. Wish PE
computer programs, even with those so called artifical intilligence,
are simply : tools.

the phrase in the defence of all engineers worth his salt is:

a fool with a tool, is still a fool.

I'm very much in favour of programs like MathCAD for this reason,
because they do not just present the resulting crunched numbers, but
include the complete sketch of the calcs in deriving the results.

This presents a way for anyone to independently verify the calcs/results
if he wants to.

  Andrew Goh