Hello,

If I do an update to a previously mirrored site, I 
noticed that there is no way to tell which non-HTML 
files have been re-downloaded and which have not. I 
have tried to hook into the 'save-file' callback, but 
for non-HTML files (PDF files for example) this 
function is called even when the file is not actually 
re-downloaded (because it hasn't changed). 

The function usercommand() (in htscore.c) which calls 
the 'save-file' callback is called for non-HTML files 
even if they are not re-downloaded.

The behaviour is different for HTML files. The 
usercommand() function (and thus the 'save-file' 
callback) is NOT called when an HTML file is not re-
downloaded. Is this different behaviour for HTML vs. 
non-HTML files done on purpose? 

The log messages also don't seem to help me detect 
this. Is there some other way I can detect when a non-
HTML file is not being re-downloaded? 

For my purposes, I would like to do an update, but 
know exactly which files have actually had to be re-
downloaded.

many thanks,
john