I spent the other reading a PhD thesis about time-based releases: http://www.cyrius.com/publications/michlmayr-phd.html All the projects involved seemed to prefer time-based releases, and there was much criticism of releases simply being endlessly dragged out and end-users being hurt by the delays and unclear forecasts. Let us say I was more than a little reminded of Xmonad and #xmonad. The last major release of xmonad was back when George W. Bush was president. There are no outstanding patches in http://darcswatch.nomeata.de/repo_http:__code.haskell.org_XMonadContrib.html or http://darcswatch.nomeata.de/repo_http:__code.haskell.org_xmonad.html Since then, awesome has made something like 10 releases; even dwm has made releases. There are plenty of bugs left in Xmonad, but I doubt many will be squashed given another few months or years. Let's just make the 1.0 release already. -- gwern http://www.gwern.net