After a long hassle with errors that wouldn't stay fixed because my ISP had
moved the target website in the middle of an attempted HTTrack copy project,
the propagation of the target website's IP address finally went to completion,
and I was left with only one error to fix (out of 5,000 links).

However, when I fixed that error and saved the file both to my hard drive and
to the actual target website (my own, of course) the error came right back.

Another look at the offending link proved that I had indeed fixed the error,
and this was evident both in the file I had saved to my hard drive and also in
the source file retrieved from the online website.

My best guess is that HTTrack either looked at a cached version of the
original source file (still retaining the error) or looked at a truncated
version of the filename (leaving the erroneous portion of the link
unexamined).

If it's a cached version of the file, where might that cache reside so that I
may clear it before trying to update the website capture ? I do clear the
browser's cache every time I run HTTrack.