> Actualy the file extention is shtml but 
> anyway.  It works fine for 99.9% of the files but there 
are 
> maybe 3 files on the whole website that come in 
incomplete 
> (litteraly if you look at the html raw it looks like it 
> just stoped downloading the page).

The only reason (if the problem does not generate any alert 
in hts-log.txt) I see is that the server did not give any 
Content-length, and broke the connection. Ensure that you 
are crawling the site using reasonnable settings (2 or 3 
simultaneous connection, maximum), and also ensure that you 
are using a recent (example: 3.22 or 3.23) release.