> > First I tried using the command line options on 
> WinHTTrack 
> 
> Ah, this is not possible (and this should be in the 
> documentation, yes, I know..), the GUI is not very 
clever, 
> and is just a frontend to the libhttrack.dll engine.
Ah, I understand.  Shame (it means we need two executables 
in the distro') but at least there's a logical reason...

> > I deliberately leave off the actual web address, hoping 
> > that HTTrack will take all relevant settings from 
> the .whtt file.  
> Not the .whtt file, but the doit.log file generated by 
the 
> engine AFTER you click 'Start' using the GUI.
I see.  This sounds like I should take that as a "yes, it's 
quite proper to run httrack without specifying a source and 
it should work as expected."

> > Unfortunately, it prints this message, and then says 
> Done! 
> > and stops.  There are over 900 links in detail.asp
> Ahah. But did you tell httrack that it had to crawl ALL 
> sites in this page?No: all the links (except the few in the "menu" block) 
are "local" documents.  The web page links to a set of 
scientific spec. and safety doc's for a range of products 
from a company often used by my colleagues.  The vast 
majority of these 900 links are of the form 
HREF="specs/x.pdf" or HREF="safety/x.pdf"

This is why I kind-a baffled...  If I run it from the GUI, 
it diligently displays all the URLs it's checking.  IIRC, 
the first page -- detail.asp which I believe is database 
generated -- is the only page to be updated (since it 
obviously can't have a revision date associated with it), 
but it takes a couple of minutes to run through all the 
other links to see that nothing else has changed.

I am using a proxy.  Without specifying any options, will 
the CLI HTTrack use the same proxy options as the GUI 
version?  I've only just thought of this as a potential 
issue -- Is this the obvious problem?  Am I being thick?  
Or does this also not matter?