That is, the file name will be splitted into words and the search term will be matched such that the search term matches start-of-words. Word boundary could be between any letter and non-letter character, and also on case change (I use this with Java files too, which typically contain files named like this TypicalJavaFile.java). Currently, I have to type "nl*h*m", which is unpractical (I turned off matching from the beginning).
This idea comes from IntelliJ Idea IDE and it quickly becomes addictive. There will be one more reason why I can't switch to Linux, because Total Commander is not there .
