At 8:14 AM -0700 7/20/96, Peter D. Junger wrote:

"Bill Olson (EDP)" writes:

: Reverse engineering is process of 'mimicking' the specifications of : another product by copying the 'abstract interface' of it. Example: : : I write a desktop application that greatly increases employee : productivity, and it sells like hotcakes. Another company decides that I : am gaining too much market share with my product and decides to reverse : engineer the product so that they can create a competing product. They : hire an engineer who takes the program and analyzes the input and output : with a detailed script of test patterns (heaven forbid he might even : decompile the program and snoop). By doing so, he now has a complete : product specification minus the implementation (i.e. how it works). He : then takes the product specification and gives it to another engineer : (actually it's done through 'clean' liaisons) who then creates a product : that does the exact same thing as mine--but with a different : implementation process. Because the product copies the specification and : not the implementation, it does not infringe on copyrights or patents.

Good explanation. But note that reverse engineering is not a way of getting around patent violations. It only works to protect oneself from copyright violations, since a reverse-engineered product is not (arguably) a copy of the original. It is also useful when the actual workings of the original, or the way the original is made, is a (trade) secret.

Important comment. Further, as I understand it if an implementation is obvious to one practiced in the art, one's ability to protect such an implementation is also limited. David