Currently both XmlParser and XmlSlurper elide text that is only whitespace. Surely it is for the application and not the parser to determine what to do with whitespace that is not inside an opening or closing tag. Both classes allow for changing the behaviour programatically. I argue that the current default in not consistent with generally expected behaviour and so should be changed. Clearly this is a breaking change and so should not happen in the 1.8.x series. The 2.0.x series should take the opportunity to make this change.