In a world where translators are still being asked to translate websites starting from snippets of text pasted into a Word document, PDFs of screens and even printouts of all the pages of a website, we can only hope that someday all content will be created in a structured environment like DITA XML for example.
I've created several text-rich, multilingual websites using DITA and the Open Toolkit and recently wrote an article about this for the US magazine Multilingual Computing.