I am new with Hadoop. What are the specific reasons why Hadoop is so tightly bound with Linux, and the cluster it runs upon is homogeneous? I'm looking for really specific details that can tell me why Hadoop does not work well with windows, and if there are some libraries some specific scripts that are involved? My project is to deploy Hadoop without using Cygwin. I have already seen the article from Hayes Davis where he explained how to install Hadoop without Cygwin, but he said that there are some bugs. I might start from scratch to properly configure Hadoop on Windows, but if any one can explain what, specifically, are the reasons that Hadoop doesn't work well on windows that would be very helpful.