When the program retrieves a page, like
the file generated will be
below the directory defined by the command directory. This way
pages coming from different machines will have a unique name, as well as
it is easy to guess the URL from the file name.
On the other hand many cases you want to retrieve pages only from one machine.
In such a case generating the directory com/mycom/www/80 is useless.
You can define
to generate the files under the directory defined by the command directory,
and forget the path generated from the machine name and the port number.
You can also define
to generate the full path in case you want to download pages from multiple sites.
This feature is going to be changed in future versions to provide a much higher flexibility for the user and possible being incompatible with the current version. TOC