Hi, Currently nutch isn't very friendly to windows users as it requires cygwin to run and there are a lot of issues with Hadoop 1.x branch, which nutch bundles with it, due to the "set tmp permission" issue.
What do you think about doing two things: 1. Move to Hadoop 2.4 to support windows/linux and the new map reduce api 2. Create bash scripts to run crawls with Relevant JIRA Issues: