> but HTTrack could use a little bit more elegancy, 
> like Google for instance.

Google (and other search engines) can very easily spread 
hits by "mixing" the millions pages they have to fetch. For 
example, if a complete refresh is 1 week long, and if you 
have 7 pages in your website, search engines will fetch 1 
page per day on your site.

But HTTrack users would be a bit frustrated to wait 1 week 
for a 7-page website mirror, not to mention the need to 
keep the PC alive during the night (this is an extreme and 
louzy example, I admit)

Well, actually HTTrack does implements by default limits to 
avoid bandwidth overload from beginners:
- 25 KB/s maximum
- robots file and META enabled

The 25KB limit is the most important default feature - and 
it is generally the best solution to avoid overload.

Of course, users can override these limits and can abuse 
websites. But this is unfortunately beyond my control, and 
the reponsability of everyone. If someone is clobbering 
your bandwidth, the best way is to report the incident to 
the coresponding network postmaster that will take actions 
to stop the problem.