Larry Benson's view that papers should not need Supplementary Information (Nature 449, 24; 2007) has elicited lively debate on Nautilus (see http://tinyurl.com/2lfkx4).

Rather than being a way to place important-but-not-essential data at the readers' disposal — as originally envisaged — Supplementary Information (SI) has proliferated: ten pages are not unusual, and Nature's average is about five, writes Benson.

“Maybe the Editor(s) need to clamp down on the SI abuse?”, opines Richard Grant in response. Massimo Sandal, however, disagrees, saying that it is not the raw information that is the main problem, but that SI is “usually not as carefully crafted as the paper itself”. He concedes that he has read SI that was “wonderfully made and useful”, citing as an example a paper by P. W. K. Rothemund (Nature 440, 297–302; 2006).

Matthieu Vermeren sums up the mood, saying SI was initially “a great way to link a paper with files that are difficult to print such as movies”. But, he adds, it now contains “vast amounts of data that could either be part of the paper or not shown”.