> occasionally it happens, that a downloaded HTML file is 
> corrupt (eg. not complete). Unfortunately the corrupt 
file 
> is also in the cache (new.dat). So 'redownload' 
or 'update' 
> or 'continue' doesn't solve the problem, because HTTrack 
> thinks, that the cached file is identical to the server 
> version (that is my interpretation).
> You might guess my question: How can I repair the corrupt 
> files in the cache or how can I reinforce a redownload?
This isn't directly possible - but you can edit the hts-
cache/new.ndx file (which is a text file) using notepad or 
vi, search for the bogus html filename, and alter its name 
to something that won't be used (for example, replace 
the /bar/foo.html occurence by XXXXXXXXXXXXXX) ; and then 
save the file.

BTW, how bogus is the file? (missing data? currupted data?)