I have created a very straightforward PHP & MySQL application.  I am trying to
render it offline so it can be more portable.  WinHTTrack goes a good job of
making an offline copy but randomly links will be hard links to the internet
address of a particular resource.

The really strange part (to me) is that these links are always part of a loop
in the PHP code where several links are generated at once.  (An example would
be where there's a list of navigational links across the top or side of a
page.)  Since they are created in a loop, they are all almost identical in
formatting.  There should be no way for HTTrack to differentiate between them
and randomly choose to drop certain ones.

Some times the links are media, sometimes they are text documents or other php
pages.

Each time I run the HTTrack program, different links are incorrect.  There is
no consistency!  I know there shouldn't be such a random behavior in a
computer program!  Could it be a time out issue in communications with the
server?