Talk:Error

I removed most of this entry. It was largely a combination of a less than fruitful attempt at a general-purpose definition of "error" and random trivia from a number of disciplines. Some particular comments:

An industry, insurance, has grown up to protect both victims and perpetrators from the baleful effects of error.

Removed this silly claim. Many kinds of insurance are designed not to protect you from error, but from "acts of God", such as fires and floods. Unless you mean to call God and/or "the Universe" a source of "error". Insurance protects you from many things that are not error, including chance and intentional malice.

Errors occur naturally, for example, an error in the replication of genetic material will result in a mutation, probably an unfavorable one.

Removed. Not NPOV. Whether errors occur "naturally" is entirely a matter of philosophical perspective. There is certainly a perspective from which anything natural is, by definition, free of error. In the case of genetics, you can only regard mutation as an "error" if you take as given that nature "wants" its replication processes to be exact replication processes.

Irrespectively of their inevitability, at law the fiction is maintained that an error is a wrong, negligence, a tort, for which anyone who suffers damage as the result thereof must be compensated by the perpetrator.

That's nice. Why can't this be in an article about law?

--Ryguasu 09:16 Nov 12, 2002 (UTC)

How about a section for Human Error? I would base it on the book of the same name written by James Reason. I may try it soon, but I don't currently have a copy of the book (an article on the book and its author would also be nice), but if anyone else wants to take a crack, give it a go. Spalding 16:01, Oct 10, 2004 (UTC)

Accidents

Accidents are often caused when a chain of errors occur that bypass safety countermeasures.

This applies most particularly in the fieds of transportation.

Syd1435 06:04, 2004 Nov 11 (UTC)

Software Engineering

I've never edited the wikipedia before and so don't know the protocols ... so when in doubt communicate (and be gentle with me if I have done wrong :-) The prior discussion described the term "fault" to be a bug. This is (a) wrong and (b) and ignores the critical distinction (widely recognized in availability engineering) between design elements that predispose a system to error (defects) and the conditions or events that exercise the defects (faults). The distinction is important, not merely for precision of communication, but has very practical value in that some things (that are technically referred to as defects) are nearly impossible to eliminate, and it is more practical to deal with the risk of failure by preventing the faults (which while less robust can be equally effective). For example, it is (in reliability parlance) a "defect" that human beings are so easily killed by bullets, so we attempt to reduce the likelihood of error (people being damaged by bullets) and failure (death or disability) by reducing the likelihood of fault-events (regulating fire-arms and wearing body armor).

There is another discussion that I only started ... but chose not to take deeper for fear of doing more harm than good. In hierarchical systems, we often use the term "error" to describe behavior that is entirely within specifications. When a degraded signal makes it impossible to recover data from a disk, the software in the controller detects this failure (through error detecting data encoding) and includes, in the request completion information, a description of the data read error. The disk controller is functioning exactly as specified ... which means that this does not (technically) qualify as an error. I mentioned that in hierarchical systems an error or failure at level N can turn into a fault at the next level up. I did not (for fear of muddying the waters) say that such faults are also usually referred to as errors.

MarkKampe 17:50, 5 November 2006 (UTC)[reply]

Headings should not be links

Don't all these headings violate this guideline? Shouldn't we move the links into the sections? Spalding 03:28, 24 February 2006 (UTC)[reply]

gaffe

I suggest merging gaffe into this article. --Dangherous 09:16, 14 April 2006 (UTC)[reply]

Whoever put the picture of the train wreck, I think that is a great illustration! Oops!
- I nearly soiled myself! Dorfl 17:55, 20 October 2006 (UTC)[reply]