> How can get ALL a web site, even some pages without any 
> other link to other page.

This is not possible, if there are no links anywhere on a 
page that refers to this page. There is no way to "guess" 
the link names, the only solution for httrack is to fetch 
links embedded in pages to capture them.