>http://www.itworld.com/article/2901453/no-its-not-always-quicker-to-do-things-in-memory.html

Sorry Al, no Eye-Opener but only stupid pet trick.

Esp. telling is that the authors of the origina paper do ***not*** list the results of the mutable string types in Java, while going the extra mile with listing the times of the [lucky] benchmarking of an implemetation detail between Linux and Windows for copying on top of stack.

We even had a similar discussion UT about typical timing observed when creating long strings (size of multi-page reporst for instance). In languages like vfp it ***might*** be open to discussion, but for Java IMO they are just benchmarking programmer/rookie mistake, as the typical article on string vs. StringBuffer/StringBuilder will show.

While the authors in the conclusion list better education for programmers, the headline is meant to evoke further mentioning in print glossing over that not-so-PC finding ;-))