Sweave is an implementation of the noweb standard for generating dynamic reports of R/S+ analyses in Latex.
Part of an ongoing groundswell of literate programming, i.e. the preparation of code for human readers. What appeals to me is the notion that the most succinct way of describing an analysis is to produce the code itself -- albeit a clean version -- along with pretty output. Important features are the dynamic, pipe-lined aspect of the process, and the latex format, which can easily be translated to html, xml, troff, ps, rtf, and (shock, horror) MSWord.doc formats, with greater or lesser success.
As I understand it, most such standards/ideas have originated in the effort to write documentation painlessly.
