Hardcore enterprise solutions, as well as other products, normally deal with large amount of text data. Those applications spend considerable time and memory to mess with Strings. It had been repeatedly shown that optimizing String usages will almost always give the immediate performance boosts. This is not to mention dodging OutOfMemoryErrors and the like.
In this talk, we will revisit the basic sins of working with Strings: gluttony of concatenation, wrath of substrings, greed of interning, pride of deduplication et cetera. We will also see the costs of sloth, believing JVM Almighty will do all the work for us.