Talk:Idempotence

Computer science

Idempotence is an extremely important computer science technique, in the design of transaction processing and database integrity. A number of the mathematical topics, created do have computer science, and since the concept is often taught in the "discrete mathematics" course delivered by computer science departments, I'd hate to do an artificial separation by disambiguation pages about different applications of the same concept. How do you think we should handle such topics? Howard C. Berkowitz 18:51, 23 December 2008 (UTC)

Well, I feel inclined to say "Just do it". Start off with sections "Idempotence in Mathematics" and "Idempotence in Computer Science", write what you want to write, add a sentence to the introduction and then take it from there. Let's see what happens? Richard Pinch 19:46, 23 December 2008 (UTC)

BTW, I feel that the case for separate articles depends on whether the meanings in the two arenas are actually incompatible or inconsistent. In this a reference here suggests that the two meanings are actually consistent. But then to me vi is a database query engine. Richard Pinch 19:56, 23 December 2008 (UTC)

Something like the following

Idempotence in Computer Science

In applications such as databases and transaction processing, idempotent operations are those for which the intended effect is that repeated application should have no effect, such as inserting a record in a file, an element in a set, or sending a message. Implementations must therefore be constructed in such a way that the intended effect is actually carried into practice. For example, messages might have unique sequence numbers with duplicates being discarded on receipt. A set might be implemented as a bit vector, and member insertion implemented by an idempotent mathematical operation such as inclusive or with a bit mask.

Richard Pinch 20:17, 23 December 2008 (UTC)