Currently IS practice abounds with application innovations, e.g. online auctions, blogs, wikis, chat, user spaces, multi-player games and reputation ratings, while over the same time few new IS academic theories have taken hold. To the practitioners who innovate, journals often seem out of date, over-rigorous and increasingly irrelevant. As IS practice innovates, IS theory seems to be if anything becoming more risk averse. If IS journals assume research rigor is research excellence, they will fall behind the cutting edge of IS progress. This would be unfortunate, as advances need to be driven by theory as well as practice. The key is recognition that good research involves two dimensions not one, namely rigor and relevance. These correspond to avoiding errors of commission (lack of rigor), and avoiding errors of omission (lack of relevance). These two dimensions suggest a selective but open IS electronic archive could increase both relevance (by electronically publishing all submissions) and rigor (by electronically supporting expert and Wiki style reviews). Since the IS community designs and creates online systems, we should be able to improve the highly successful Los Alamos physics preprint archive, and lead the way in electronic knowledge exchange systems.